WEBVTT 00:00:00.000 --> 00:00:00.215 align:middle line:90% 00:00:00.215 --> 00:00:02.340 align:middle line:84% ISAAC SAMUEL KOHANE: So what we'll talk about today 00:00:02.340 --> 00:00:04.950 align:middle line:84% is a very practical genomic medicine, 00:00:04.950 --> 00:00:08.160 align:middle line:84% by which I mean not only stuff that we 00:00:08.160 --> 00:00:12.060 align:middle line:84% think is going to be imminent in two years 00:00:12.060 --> 00:00:16.650 align:middle line:84% but what it means today to do genomic medicine, nothing 00:00:16.650 --> 00:00:17.860 align:middle line:90% future, just today. 00:00:17.860 --> 00:00:20.550 align:middle line:84% And if you think that I have stepped across a line 00:00:20.550 --> 00:00:24.400 align:middle line:90% into the future, call me on it. 00:00:24.400 --> 00:00:27.570 align:middle line:84% So let's talk about different diffusions of knowledge. 00:00:27.570 --> 00:00:31.470 align:middle line:84% The genome draft, as you know, was much 00:00:31.470 --> 00:00:33.836 align:middle line:90% heralded with Clinton-- 00:00:33.836 --> 00:00:35.880 align:middle line:84% was Clinton still president in 2001? 00:00:35.880 --> 00:00:36.750 align:middle line:90% No. 00:00:36.750 --> 00:00:38.610 align:middle line:90% So before-- when was the draft? 00:00:38.610 --> 00:00:39.920 align:middle line:90% The first-- 00:00:39.920 --> 00:00:41.820 align:middle line:90% AUDIENCE: [INAUDIBLE]. 00:00:41.820 --> 00:00:46.980 align:middle line:84% ISAAC SAMUEL KOHANE: When was Venter and Collins-- 00:00:46.980 --> 00:00:49.440 align:middle line:84% and wasn't Clinton-- yeah, it was Clinton. 00:00:49.440 --> 00:00:50.580 align:middle line:90% AUDIENCE: It was 2000. 00:00:50.580 --> 00:00:51.955 align:middle line:84% ISAAC SAMUEL KOHANE: It was 2000. 00:00:51.955 --> 00:00:53.180 align:middle line:90% That's an error. 00:00:53.180 --> 00:00:54.180 align:middle line:90% So it was much heralded. 00:00:54.180 --> 00:00:59.040 align:middle line:84% And there was a lot of promise about this would solve problems 00:00:59.040 --> 00:01:01.260 align:middle line:84% in human disease and in medicine. 00:01:01.260 --> 00:01:03.120 align:middle line:90% And it made it sound-- 00:01:03.120 --> 00:01:05.790 align:middle line:84% the way it was described-- as fairly imminent. 00:01:05.790 --> 00:01:07.710 align:middle line:84% And it was going to transform medicine. 00:01:07.710 --> 00:01:10.920 align:middle line:84% And just as a measure of calibration, 00:01:10.920 --> 00:01:14.190 align:middle line:84% the development of penicillin by Chain and Florey in '41, 00:01:14.190 --> 00:01:18.460 align:middle line:84% was saving thousands of lives within months. 00:01:18.460 --> 00:01:22.920 align:middle line:84% And so has genomic medicine been successful by that measure? 00:01:22.920 --> 00:01:25.470 align:middle line:90% No. 00:01:25.470 --> 00:01:30.480 align:middle line:84% Do we think it's going to impact medicine in the future? 00:01:30.480 --> 00:01:31.350 align:middle line:90% Well, sure. 00:01:31.350 --> 00:01:34.410 align:middle line:84% I hope I've convinced you in the course of the class 00:01:34.410 --> 00:01:37.530 align:middle line:84% that it's going to be important in the way we diagnose 00:01:37.530 --> 00:01:39.840 align:middle line:84% our patients, in the way we manage patients, 00:01:39.840 --> 00:01:41.310 align:middle line:84% even the way we treat our patients, 00:01:41.310 --> 00:01:42.790 align:middle line:90% but clearly not by this measure. 00:01:42.790 --> 00:01:44.165 align:middle line:84% And I think we have to appreciate 00:01:44.165 --> 00:01:48.640 align:middle line:84% that probably timescales on the order of 10 to 20 years. 00:01:48.640 --> 00:01:52.230 align:middle line:84% And, in fact, one of my colleagues, one of my friends 00:01:52.230 --> 00:01:54.930 align:middle line:84% used to be a reporter for the New York Times 00:01:54.930 --> 00:01:57.210 align:middle line:84% when this was announced in 2000, asked me, 00:01:57.210 --> 00:02:01.560 align:middle line:84% so what's going to be the main impact of the human genome 00:02:01.560 --> 00:02:02.910 align:middle line:90% draft for the next 10 years? 00:02:02.910 --> 00:02:05.130 align:middle line:90% I said, a lot of bad news. 00:02:05.130 --> 00:02:06.630 align:middle line:84% By which I meant that we're probably 00:02:06.630 --> 00:02:08.340 align:middle line:84% going to be able to diagnose a lot of things 00:02:08.340 --> 00:02:10.173 align:middle line:84% that we had not been able to diagnose and do 00:02:10.173 --> 00:02:12.850 align:middle line:90% nothing about it. 00:02:12.850 --> 00:02:14.430 align:middle line:90% And that's probably why not. 00:02:14.430 --> 00:02:19.110 align:middle line:84% But there's another reason why not, which is, well, 00:02:19.110 --> 00:02:21.150 align:middle line:84% what other problems have there been 00:02:21.150 --> 00:02:26.190 align:middle line:84% in preventing the diffusion of genomic medicine? 00:02:26.190 --> 00:02:26.903 align:middle line:90% Any ideas? 00:02:26.903 --> 00:02:28.320 align:middle line:84% AUDIENCE: Massive amounts of data? 00:02:28.320 --> 00:02:28.497 align:middle line:90% 00:02:28.497 --> 00:02:29.580 align:middle line:90% ISAAC SAMUEL KOHANE: What? 00:02:29.580 --> 00:02:30.480 align:middle line:84% AUDIENCE: Massive amounts of data. 00:02:30.480 --> 00:02:32.340 align:middle line:84% ISAAC SAMUEL KOHANE: Massive amounts of data and-- 00:02:32.340 --> 00:02:33.900 align:middle line:84% AUDIENCE: Lack of trained personnel. 00:02:33.900 --> 00:02:36.108 align:middle line:84% ISAAC SAMUEL KOHANE: --and lack of trained personnel. 00:02:36.108 --> 00:02:40.140 align:middle line:84% Basically, the medical system does not have any idea 00:02:40.140 --> 00:02:41.470 align:middle line:90% how to deal with this. 00:02:41.470 --> 00:02:45.030 align:middle line:84% And just to give you some calibration around that. 00:02:45.030 --> 00:02:49.920 align:middle line:84% I gave a lecture a year ago to American Professors 00:02:49.920 --> 00:02:52.830 align:middle line:84% of Medical Genetics, not genomics but medical genetics. 00:02:52.830 --> 00:02:55.560 align:middle line:84% And I was telling them about the whole genomic revolution. 00:02:55.560 --> 00:02:58.142 align:middle line:84% And I think it's an odd thing for me 00:02:58.142 --> 00:02:59.850 align:middle line:84% to talk to professors of medical genetics 00:02:59.850 --> 00:03:01.530 align:middle line:90% and the national association. 00:03:01.530 --> 00:03:05.363 align:middle line:84% But they were polite and interested in my lecture. 00:03:05.363 --> 00:03:07.530 align:middle line:84% But when I asked them, is this what you see yourself 00:03:07.530 --> 00:03:10.080 align:middle line:90% doing for the next 20 years? 00:03:10.080 --> 00:03:10.853 align:middle line:90% Absolutely not. 00:03:10.853 --> 00:03:12.270 align:middle line:84% They wanted to restrict themselves 00:03:12.270 --> 00:03:17.340 align:middle line:84% to monogenic, strongly highly-penetrant diseases that 00:03:17.340 --> 00:03:18.200 align:middle line:90% were extremely rare. 00:03:18.200 --> 00:03:19.450 align:middle line:90% That's what they wanted to do. 00:03:19.450 --> 00:03:23.550 align:middle line:84% And I said, well, you share at least a substring 00:03:23.550 --> 00:03:24.632 align:middle line:90% with genomics. 00:03:24.632 --> 00:03:26.340 align:middle line:84% Don't you see it as a part of your field? 00:03:26.340 --> 00:03:27.630 align:middle line:90% No. 00:03:27.630 --> 00:03:32.190 align:middle line:84% And so the geneticists are not holding the torch. 00:03:32.190 --> 00:03:34.110 align:middle line:84% Right now, the medical students are not 00:03:34.110 --> 00:03:38.580 align:middle line:84% being taught this in any detail whatsoever. 00:03:38.580 --> 00:03:41.830 align:middle line:84% And so there's going to be a problem. 00:03:41.830 --> 00:03:45.330 align:middle line:84% So who is going to practice genomic medicine? 00:03:45.330 --> 00:03:48.450 align:middle line:84% There's a variety of possibilities. 00:03:48.450 --> 00:03:52.530 align:middle line:84% Shown in light blue, for some reason, 00:03:52.530 --> 00:03:54.600 align:middle line:90% is the medical geneticist. 00:03:54.600 --> 00:03:58.080 align:middle line:84% The person who typically has been doing-- 00:03:58.080 --> 00:04:01.500 align:middle line:84% sorry about that-- who has typically-- 00:04:01.500 --> 00:04:05.400 align:middle line:84% that will be the last light blue thing, I hope-- 00:04:05.400 --> 00:04:08.490 align:middle line:84% who has been doing medical diagnosis and counseling. 00:04:08.490 --> 00:04:11.430 align:middle line:84% But what I'm telling you is they do not see themselves 00:04:11.430 --> 00:04:17.160 align:middle line:84% as bringing into the for the entirety of the impact 00:04:17.160 --> 00:04:25.660 align:middle line:84% of genetics into the broad swath of diseases, 00:04:25.660 --> 00:04:28.710 align:middle line:84% whether they're primarily inherited 00:04:28.710 --> 00:04:34.170 align:middle line:84% or primarily modifying environmental influences. 00:04:34.170 --> 00:04:37.230 align:middle line:84% A logical place would be primary care. 00:04:37.230 --> 00:04:39.150 align:middle line:84% Because, after all-- as I hope I've 00:04:39.150 --> 00:04:41.880 align:middle line:84% convinced you-- that, indeed, part of what genomics gives 00:04:41.880 --> 00:04:45.270 align:middle line:84% you is the ability to prognosticate for the future. 00:04:45.270 --> 00:04:48.540 align:middle line:84% Part of preventive care should be 00:04:48.540 --> 00:04:51.630 align:middle line:90% genetic and genome-wide testing. 00:04:51.630 --> 00:04:54.365 align:middle line:84% So a pediatrician, internist, OBGYN 00:04:54.365 --> 00:04:55.740 align:middle line:84% would be a natural place for this 00:04:55.740 --> 00:05:00.110 align:middle line:84% to happen, and also for specific diseases. 00:05:00.110 --> 00:05:02.990 align:middle line:84% We've seen a lot of application of microarrays 00:05:02.990 --> 00:05:05.090 align:middle line:84% to cancer, for instance, so oncologists 00:05:05.090 --> 00:05:10.280 align:middle line:84% would be the natural person to order these tests. 00:05:10.280 --> 00:05:13.080 align:middle line:84% And a gastroenterologist-- we looked at Crohn's disease. 00:05:13.080 --> 00:05:16.310 align:middle line:84% And it would not be illogical for gastroenterologist 00:05:16.310 --> 00:05:19.020 align:middle line:90% to screen for risk factors. 00:05:19.020 --> 00:05:21.020 align:middle line:90% So what's, in fact, the case? 00:05:21.020 --> 00:05:22.940 align:middle line:84% So there was a study done just looking 00:05:22.940 --> 00:05:25.940 align:middle line:90% at cancer susceptibility tests. 00:05:25.940 --> 00:05:27.440 align:middle line:84% So it's a substantial study looking 00:05:27.440 --> 00:05:32.000 align:middle line:84% at over 1,200 physicians, 820 of which were in primary care. 00:05:32.000 --> 00:05:34.220 align:middle line:84% And in 12 months, approximately 30% 00:05:34.220 --> 00:05:40.610 align:middle line:84% ordered or referred genetic testing, 00:05:40.610 --> 00:05:43.550 align:middle line:84% looking for susceptibility, not looking 00:05:43.550 --> 00:05:48.000 align:middle line:84% for diagnosing a patient but for looking were they at risk. 00:05:48.000 --> 00:05:51.677 align:middle line:90% So that's kind of impressive. 00:05:51.677 --> 00:05:52.760 align:middle line:90% So a bit less [INAUDIBLE]. 00:05:52.760 --> 00:05:56.960 align:middle line:84% Only 7% of them directly did it, whether because they 00:05:56.960 --> 00:05:59.180 align:middle line:84% were uncomfortable or not knowledgeable enough 00:05:59.180 --> 00:06:00.920 align:middle line:90% to actually directly order it. 00:06:00.920 --> 00:06:05.630 align:middle line:84% But that's telling you that 30% of this random sample 00:06:05.630 --> 00:06:07.400 align:middle line:84% of physicians were actually ordering 00:06:07.400 --> 00:06:10.380 align:middle line:90% these susceptibility tests. 00:06:10.380 --> 00:06:16.340 align:middle line:84% What do you imagine are the factors affecting ordering? 00:06:16.340 --> 00:06:21.410 align:middle line:84% Let me give you some potential-- what were you saying? 00:06:21.410 --> 00:06:22.557 align:middle line:90% AUDIENCE: Cost. 00:06:22.557 --> 00:06:23.640 align:middle line:90% ISAAC SAMUEL KOHANE: Cost. 00:06:23.640 --> 00:06:27.460 align:middle line:90% Any other suggestions? 00:06:27.460 --> 00:06:29.400 align:middle line:84% AUDIENCE: Whether you can interpret the test. 00:06:29.400 --> 00:06:31.733 align:middle line:84% ISAAC SAMUEL KOHANE: Whether you can interpret the test. 00:06:31.733 --> 00:06:33.715 align:middle line:90% Any other suggestions? 00:06:33.715 --> 00:06:34.215 align:middle line:90% What? 00:06:34.215 --> 00:06:35.317 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:06:35.317 --> 00:06:37.650 align:middle line:84% ISAAC SAMUEL KOHANE: Which physicians were more likely-- 00:06:37.650 --> 00:06:38.190 align:middle line:90% no, no. 00:06:38.190 --> 00:06:39.390 align:middle line:84% What I meant was which physicians were 00:06:39.390 --> 00:06:40.515 align:middle line:90% more likely to order tests. 00:06:40.515 --> 00:06:43.470 align:middle line:84% I apologize for-- let me rephrase the question. 00:06:43.470 --> 00:06:45.177 align:middle line:90% What was it about them? 00:06:45.177 --> 00:06:47.760 align:middle line:84% What property of the physicians made them much more likely to. 00:06:47.760 --> 00:06:49.800 align:middle line:84% AUDIENCE: Probably research hospitals. 00:06:49.800 --> 00:06:51.467 align:middle line:84% ISAAC SAMUEL KOHANE: Research hospitals. 00:06:51.467 --> 00:06:54.990 align:middle line:90% 00:06:54.990 --> 00:06:57.910 align:middle line:84% AUDIENCE: And also probably in disease areas 00:06:57.910 --> 00:06:59.323 align:middle line:90% where there's a lot of-- 00:06:59.323 --> 00:07:01.240 align:middle line:84% ISAAC SAMUEL KOHANE: So places where they have 00:07:01.240 --> 00:07:03.910 align:middle line:90% a high burden of genetically-- 00:07:03.910 --> 00:07:04.980 align:middle line:90% disease. 00:07:04.980 --> 00:07:05.980 align:middle line:90% Well, here's the answer. 00:07:05.980 --> 00:07:10.405 align:middle line:84% The first and foremost was being in the Northeast. 00:07:10.405 --> 00:07:11.530 align:middle line:90% I don't know what it means. 00:07:11.530 --> 00:07:13.310 align:middle line:90% But that's, in fact-- 00:07:13.310 --> 00:07:14.570 align:middle line:90% it was a by far-- 00:07:14.570 --> 00:07:18.010 align:middle line:84% so there's plenty of wonderful tertiary care centers 00:07:18.010 --> 00:07:20.290 align:middle line:90% on the West Coast, I hear. 00:07:20.290 --> 00:07:23.860 align:middle line:84% But that was not a telling thing. 00:07:23.860 --> 00:07:26.550 align:middle line:84% Feeling confident that, as you mentioned, 00:07:26.550 --> 00:07:28.870 align:middle line:84% the presence of advertising materials 00:07:28.870 --> 00:07:32.410 align:middle line:84% and, most importantly, having the patient ask for it. 00:07:32.410 --> 00:07:34.690 align:middle line:84% The point here is that the reason there 00:07:34.690 --> 00:07:38.320 align:middle line:84% was such a high percentage, 30%, is not 00:07:38.320 --> 00:07:41.290 align:middle line:84% because doctors have been trained to order those tests. 00:07:41.290 --> 00:07:45.970 align:middle line:84% It's because the patients are reading lay literature. 00:07:45.970 --> 00:07:49.570 align:middle line:84% And if they have breast cancer in the family 00:07:49.570 --> 00:07:54.610 align:middle line:84% or have ovarian cancer, or the have colon cancer, 00:07:54.610 --> 00:07:58.300 align:middle line:84% they're asking the doctors to test them. 00:07:58.300 --> 00:08:01.720 align:middle line:90% That's the real insight. 00:08:01.720 --> 00:08:02.995 align:middle line:90% [SIDE CONVERSATION] 00:08:02.995 --> 00:08:13.940 align:middle line:90% 00:08:13.940 --> 00:08:15.080 align:middle line:90% Go ahead. 00:08:15.080 --> 00:08:17.690 align:middle line:84% AUDIENCE: When you say susceptibility test-- 00:08:17.690 --> 00:08:19.280 align:middle line:84% presumably some kind of multiplexed 00:08:19.280 --> 00:08:21.890 align:middle line:84% PCR looking at different [INAUDIBLE] 00:08:21.890 --> 00:08:26.470 align:middle line:84% or associated risks that actually [INAUDIBLE].. 00:08:26.470 --> 00:08:28.220 align:middle line:84% ISAAC SAMUEL KOHANE: It's a multiplex test 00:08:28.220 --> 00:08:31.010 align:middle line:90% not an array test. 00:08:31.010 --> 00:08:32.510 align:middle line:84% AUDIENCE: Question about Northeast-- 00:08:32.510 --> 00:08:35.502 align:middle line:84% is it just simply medical professional density? 00:08:35.502 --> 00:08:37.460 align:middle line:84% ISAAC SAMUEL KOHANE: No, this corrected for it. 00:08:37.460 --> 00:08:38.104 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:08:38.104 --> 00:08:38.179 align:middle line:90% ISAAC SAMUEL KOHANE: Yeah. 00:08:38.179 --> 00:08:40.182 align:middle line:84% AUDIENCE: The New York Times or [INAUDIBLE].. 00:08:40.182 --> 00:08:41.890 align:middle line:84% ISAAC SAMUEL KOHANE: It's something else. 00:08:41.890 --> 00:08:42.799 align:middle line:90% It's corrected. 00:08:42.799 --> 00:08:48.230 align:middle line:84% But it's something about the nature of the training 00:08:48.230 --> 00:08:51.620 align:middle line:84% or it may be patients in the Northeast. 00:08:51.620 --> 00:08:55.070 align:middle line:90% 00:08:55.070 --> 00:08:56.060 align:middle line:90% Let's put it this way. 00:08:56.060 --> 00:08:59.770 align:middle line:84% Maybe there's a lot of Ashkenazi, Jewish women worried 00:08:59.770 --> 00:09:01.520 align:middle line:84% about breast cancer knowing that they have 00:09:01.520 --> 00:09:05.500 align:middle line:90% this BRCA1, BRCA2 risk factor. 00:09:05.500 --> 00:09:08.030 align:middle line:84% The article did not actually elaborate on that. 00:09:08.030 --> 00:09:08.930 align:middle line:90% But that's my guess. 00:09:08.930 --> 00:09:10.763 align:middle line:84% I think the patients are the driving factor. 00:09:10.763 --> 00:09:13.160 align:middle line:84% And for those of us who are in medicine, 00:09:13.160 --> 00:09:14.750 align:middle line:84% I think we have to recognize this. 00:09:14.750 --> 00:09:16.850 align:middle line:84% And it's actually a sad fact that it's the patients that 00:09:16.850 --> 00:09:17.475 align:middle line:90% are driving it. 00:09:17.475 --> 00:09:21.830 align:middle line:84% But it's a problem because if one of the things that you 00:09:21.830 --> 00:09:24.800 align:middle line:84% bring to bear, as a doctor, is an appreciation 00:09:24.800 --> 00:09:29.510 align:middle line:84% of prior probability and what tests 00:09:29.510 --> 00:09:33.260 align:middle line:84% you want to do in order to avoid false positives. 00:09:33.260 --> 00:09:35.990 align:middle line:84% Because, remember, a lot of these polymorphisms 00:09:35.990 --> 00:09:41.360 align:middle line:84% or mutations may not actually be the causative element. 00:09:41.360 --> 00:09:43.860 align:middle line:84% They may be in linkage disequilibrium, for instance. 00:09:43.860 --> 00:09:47.030 align:middle line:90% And so it's not a 1 to 1. 00:09:47.030 --> 00:09:50.060 align:middle line:84% It's not if you have this, you're going to have cancer. 00:09:50.060 --> 00:09:51.333 align:middle line:90% It's a probabilistic measure. 00:09:51.333 --> 00:09:53.750 align:middle line:84% And so there is going to be a false positive rate with any 00:09:53.750 --> 00:09:55.235 align:middle line:90% of these measures. 00:09:55.235 --> 00:09:57.110 align:middle line:84% And if the doctor is really not knowledgeable 00:09:57.110 --> 00:10:01.010 align:middle line:84% about that interpretation, then they're going to find problems. 00:10:01.010 --> 00:10:04.970 align:middle line:84% Just as if doctors did the routine CT scans on everybody, 00:10:04.970 --> 00:10:08.310 align:middle line:84% you'd find on the order of, people my age, 00:10:08.310 --> 00:10:12.650 align:middle line:84% maybe 10% to 20% pituitary microadenomas, as we 00:10:12.650 --> 00:10:17.090 align:middle line:84% find when we do autopsies of car accidents in people my age. 00:10:17.090 --> 00:10:19.760 align:middle line:84% But they have no clinical meaning that we can tell. 00:10:19.760 --> 00:10:21.830 align:middle line:84% But if you do routine testing, you'll find that. 00:10:21.830 --> 00:10:25.040 align:middle line:84% And this is going to be quite a huge problem of false positives 00:10:25.040 --> 00:10:27.620 align:middle line:84% if we continue to have patient-driven demand 00:10:27.620 --> 00:10:32.690 align:middle line:84% for testing, because it should be done knowledgeably. 00:10:32.690 --> 00:10:37.520 align:middle line:84% AUDIENCE: Can that be met with if a doctor orders this 00:10:37.520 --> 00:10:41.720 align:middle line:84% or refers them to a specialists [INAUDIBLE].. 00:10:41.720 --> 00:10:45.490 align:middle line:84% When these tests come back, I'm assuming you don't-- 00:10:45.490 --> 00:10:47.240 align:middle line:90% this is probably wrong-- 00:10:47.240 --> 00:10:50.750 align:middle line:84% that they don't come back the way the normal blood test-- 00:10:50.750 --> 00:10:53.580 align:middle line:84% there's certain numbers and whatever the doctor's-- 00:10:53.580 --> 00:10:55.070 align:middle line:90% blood tests. 00:10:55.070 --> 00:10:58.077 align:middle line:84% ISAAC SAMUEL KOHANE: The short answer is theoretically not. 00:10:58.077 --> 00:10:59.125 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:10:59.125 --> 00:11:01.500 align:middle line:84% ISAAC SAMUEL KOHANE: No, there actually has to be person. 00:11:01.500 --> 00:11:02.820 align:middle line:90% We'll get to that. 00:11:02.820 --> 00:11:07.610 align:middle line:84% But even so, if you have an ascertainment bias because 00:11:07.610 --> 00:11:11.420 align:middle line:84% of who's doing it without actually letting 00:11:11.420 --> 00:11:17.210 align:middle line:84% the medical system try to define a appropriate measure of when 00:11:17.210 --> 00:11:22.160 align:middle line:84% you do a test, then you're going to run into those problems more 00:11:22.160 --> 00:11:24.470 align:middle line:90% than less. 00:11:24.470 --> 00:11:27.800 align:middle line:84% AUDIENCE: Oh, so you're suggesting that maybe-- 00:11:27.800 --> 00:11:29.710 align:middle line:84% so patients are asking for a particular test 00:11:29.710 --> 00:11:30.710 align:middle line:90% because they're worried. 00:11:30.710 --> 00:11:31.280 align:middle line:84% ISAAC SAMUEL KOHANE: They're worried. 00:11:31.280 --> 00:11:34.340 align:middle line:84% AUDIENCE: So you're suggesting that maybe that's a bad thing 00:11:34.340 --> 00:11:38.372 align:middle line:84% not because the doctor should be the ones to say, hey, 00:11:38.372 --> 00:11:39.830 align:middle line:84% you might want to think about this. 00:11:39.830 --> 00:11:42.530 align:middle line:84% But you're saying that maybe they-- 00:11:42.530 --> 00:11:45.230 align:middle line:84% ISAAC SAMUEL KOHANE: It's a self-selected group. 00:11:45.230 --> 00:11:48.800 align:middle line:84% And it may be that the people who are truly at risk 00:11:48.800 --> 00:11:50.960 align:middle line:90% are not getting tested. 00:11:50.960 --> 00:11:53.700 align:middle line:84% Basically, the doctor is not a decision maker here. 00:11:53.700 --> 00:11:56.810 align:middle line:84% And the whole role of a doctor is to decide, if anything, 00:11:56.810 --> 00:12:01.100 align:middle line:84% when is a test going to be cost effective and sufficiently 00:12:01.100 --> 00:12:02.400 align:middle line:90% useful for the patient. 00:12:02.400 --> 00:12:04.400 align:middle line:84% And if they disintermediate themselves and allow 00:12:04.400 --> 00:12:06.710 align:middle line:84% the patient to do that, then there's 00:12:06.710 --> 00:12:10.070 align:middle line:84% a risk that there's a group of people who will not be screened 00:12:10.070 --> 00:12:11.900 align:middle line:84% and a group of people who have been 00:12:11.900 --> 00:12:14.300 align:middle line:84% over-screened and, therefore, have an unacceptably 00:12:14.300 --> 00:12:15.465 align:middle line:90% high false positive rate. 00:12:15.465 --> 00:12:17.090 align:middle line:84% Because, basically, when people come up 00:12:17.090 --> 00:12:21.380 align:middle line:84% with calculations of sensitivity, specificity, 00:12:21.380 --> 00:12:24.440 align:middle line:84% they do it based on a certain population of patients, 00:12:24.440 --> 00:12:26.090 align:middle line:84% when a genetic counselor says, you 00:12:26.090 --> 00:12:30.980 align:middle line:84% have X, Y percent risk based on a general population. 00:12:30.980 --> 00:12:33.590 align:middle line:84% They don't based on the group of patients 00:12:33.590 --> 00:12:35.930 align:middle line:84% who are neurotic and worried, maybe appropriately, 00:12:35.930 --> 00:12:37.925 align:middle line:90% about their family history. 00:12:37.925 --> 00:12:40.550 align:middle line:84% That might be a different set of specificity and sensitivities. 00:12:40.550 --> 00:12:42.780 align:middle line:84% I'm not saying that patients are wrong to be worried. 00:12:42.780 --> 00:12:45.200 align:middle line:84% What I'm saying is the medical system is failing. 00:12:45.200 --> 00:12:46.790 align:middle line:84% They don't have educated doctors who 00:12:46.790 --> 00:12:50.390 align:middle line:84% can intermediate about what is the appropriate testing 00:12:50.390 --> 00:12:51.080 align:middle line:90% to be done. 00:12:51.080 --> 00:12:53.750 align:middle line:84% I mean, if a patient comes to me as an endocrinologist-- says, 00:12:53.750 --> 00:12:55.400 align:middle line:90% I want growth hormone testing. 00:12:55.400 --> 00:12:57.410 align:middle line:84% I not to do growth hormone testing when 00:12:57.410 --> 00:13:01.290 align:middle line:84% everybody asks because they have a significant false positive 00:13:01.290 --> 00:13:01.790 align:middle line:90% rate. 00:13:01.790 --> 00:13:04.310 align:middle line:84% And they'd end up treating a whole bunch of people 00:13:04.310 --> 00:13:05.900 align:middle line:84% with growth hormone for no reason. 00:13:05.900 --> 00:13:08.930 align:middle line:90% 00:13:08.930 --> 00:13:13.460 align:middle line:84% And that's the whole point of having an expert 00:13:13.460 --> 00:13:14.520 align:middle line:90% as part of their loop. 00:13:14.520 --> 00:13:16.700 align:middle line:84% And what I'm saying is this is telling us 00:13:16.700 --> 00:13:17.803 align:middle line:90% that there's a demand. 00:13:17.803 --> 00:13:19.220 align:middle line:84% And the medical system is failing. 00:13:19.220 --> 00:13:22.260 align:middle line:90% 00:13:22.260 --> 00:13:24.720 align:middle line:84% So here's the conventional view of genetic information 00:13:24.720 --> 00:13:25.220 align:middle line:90% management. 00:13:25.220 --> 00:13:27.695 align:middle line:90% It's worth going through it. 00:13:27.695 --> 00:13:31.490 align:middle line:84% You can think of disorders being single-gene or chromosomal, 00:13:31.490 --> 00:13:35.630 align:middle line:84% major gene or multifactorial, or these complex traits, complex 00:13:35.630 --> 00:13:36.860 align:middle line:90% multifactorial. 00:13:36.860 --> 00:13:39.800 align:middle line:84% And the conventional view of how this should be used 00:13:39.800 --> 00:13:42.170 align:middle line:84% is, in primary care, the point would 00:13:42.170 --> 00:13:44.270 align:middle line:84% be to recognize signs and symptoms, 00:13:44.270 --> 00:13:47.653 align:middle line:84% make referrals, support family long-term care. 00:13:47.653 --> 00:13:49.070 align:middle line:84% The primary care should appreciate 00:13:49.070 --> 00:13:50.420 align:middle line:84% the role of family history, arrange 00:13:50.420 --> 00:13:52.253 align:middle line:84% testing, referral to specialists, as needed, 00:13:52.253 --> 00:13:53.600 align:middle line:90% provide longitudinal care. 00:13:53.600 --> 00:13:55.580 align:middle line:90% And for complex traits-- 00:13:55.580 --> 00:13:58.070 align:middle line:84% use of genetic tests to guide prevention treatment. 00:13:58.070 --> 00:14:01.730 align:middle line:84% The specialist would manage specific problems. 00:14:01.730 --> 00:14:05.170 align:middle line:84% They would diagnose and manage system-specific problems. 00:14:05.170 --> 00:14:06.545 align:middle line:84% And they'd use also genomic tests 00:14:06.545 --> 00:14:08.848 align:middle line:90% to guide, prevent, and treat. 00:14:08.848 --> 00:14:10.640 align:middle line:84% And the medical geneticist-- the role there 00:14:10.640 --> 00:14:13.340 align:middle line:84% is counseling, longitudinal care, 00:14:13.340 --> 00:14:16.280 align:middle line:84% advice and interpretation of test results, 00:14:16.280 --> 00:14:21.080 align:middle line:84% and a reservoir of knowledge of handling of complex cases. 00:14:21.080 --> 00:14:23.750 align:middle line:84% But I just told you that medical geneticists 00:14:23.750 --> 00:14:27.245 align:middle line:90% have largely not met this role. 00:14:27.245 --> 00:14:29.830 align:middle line:90% 00:14:29.830 --> 00:14:31.870 align:middle line:90% So is there a different role? 00:14:31.870 --> 00:14:33.940 align:middle line:90% What to do. 00:14:33.940 --> 00:14:37.630 align:middle line:90% Is there a medical genomicist? 00:14:37.630 --> 00:14:38.860 align:middle line:90% What do you think? 00:14:38.860 --> 00:14:42.430 align:middle line:84% Is that a reasonable specialty or is that the wrong answer. 00:14:42.430 --> 00:14:45.205 align:middle line:90% 00:14:45.205 --> 00:14:49.981 align:middle line:84% AUDIENCE: Explain what a medical genomicist means. 00:14:49.981 --> 00:14:52.812 align:middle line:84% ISAAC SAMUEL KOHANE: If you wanted a good preventive care 00:14:52.812 --> 00:14:54.520 align:middle line:84% and you wanted someone who would give you 00:14:54.520 --> 00:14:57.700 align:middle line:84% appropriate counseling about either lifestyle 00:14:57.700 --> 00:15:03.850 align:middle line:84% changes or appropriate drugs that you need to consider, 00:15:03.850 --> 00:15:06.880 align:middle line:84% or a specific diagnostic test that you 00:15:06.880 --> 00:15:09.850 align:middle line:84% have to do for all possible genetically 00:15:09.850 --> 00:15:12.040 align:middle line:84% influenced diseases, what kind of person 00:15:12.040 --> 00:15:13.600 align:middle line:84% would you what you want to go to? 00:15:13.600 --> 00:15:17.100 align:middle line:90% 00:15:17.100 --> 00:15:17.625 align:middle line:90% Any ideas? 00:15:17.625 --> 00:15:21.420 align:middle line:90% 00:15:21.420 --> 00:15:23.800 align:middle line:84% AUDIENCE: Actually primary care is my first reaction. 00:15:23.800 --> 00:15:26.300 align:middle line:84% ISAAC SAMUEL KOHANE: So your first reaction is primary care. 00:15:26.300 --> 00:15:29.610 align:middle line:90% 00:15:29.610 --> 00:15:31.560 align:middle line:84% Do primary care practitioners currently 00:15:31.560 --> 00:15:33.090 align:middle line:90% know about any of this stuff? 00:15:33.090 --> 00:15:34.765 align:middle line:90% No. 00:15:34.765 --> 00:15:36.390 align:middle line:84% AUDIENCE: They [INAUDIBLE] with someone 00:15:36.390 --> 00:15:38.730 align:middle line:84% who would give them a very simple explanation. 00:15:38.730 --> 00:15:41.282 align:middle line:84% If someone has this test and [INAUDIBLE] what it means. 00:15:41.282 --> 00:15:43.740 align:middle line:84% And they're going to tell you what it means and then manage 00:15:43.740 --> 00:15:44.520 align:middle line:90% that through your whole-- 00:15:44.520 --> 00:15:46.853 align:middle line:84% ISAAC SAMUEL KOHANE: So let's do it slowly with feeling. 00:15:46.853 --> 00:15:49.740 align:middle line:84% Because you're now the czar of medicine. 00:15:49.740 --> 00:15:56.760 align:middle line:84% And you're trying to start a company in this area, 00:15:56.760 --> 00:15:58.950 align:middle line:90% in the United States, in 2004. 00:15:58.950 --> 00:16:01.810 align:middle line:84% Who's going to actually do this job? 00:16:01.810 --> 00:16:04.860 align:middle line:84% AUDIENCE: Well, I think the idea of going 00:16:04.860 --> 00:16:06.660 align:middle line:84% to any one, single person, whether it's 00:16:06.660 --> 00:16:10.795 align:middle line:84% specialist, genomicist, geneticist, or primary care-- 00:16:10.795 --> 00:16:12.045 align:middle line:90% I think is the wrong approach. 00:16:12.045 --> 00:16:15.120 align:middle line:90% I don't think any one person-- 00:16:15.120 --> 00:16:18.670 align:middle line:84% just in their specific roles can address-- 00:16:18.670 --> 00:16:22.560 align:middle line:84% especially as you move down that leftmost column 00:16:22.560 --> 00:16:25.810 align:middle line:90% to the very complex disease. 00:16:25.810 --> 00:16:26.760 align:middle line:90% I think that's the-- 00:16:26.760 --> 00:16:28.385 align:middle line:84% ISAAC SAMUEL KOHANE: So what do you do? 00:16:28.385 --> 00:16:30.340 align:middle line:90% 00:16:30.340 --> 00:16:31.855 align:middle line:84% We know that there's a large number 00:16:31.855 --> 00:16:34.230 align:middle line:84% of people who are going to die of complications of type 2 00:16:34.230 --> 00:16:34.890 align:middle line:90% diabetes. 00:16:34.890 --> 00:16:36.150 align:middle line:90% And it's exploding. 00:16:36.150 --> 00:16:39.675 align:middle line:84% We know there's a large cancer burden. 00:16:39.675 --> 00:16:42.660 align:middle line:84% We know there's a large cardiovascular disease burden. 00:16:42.660 --> 00:16:45.860 align:middle line:84% And we know that some component of all of these diseases 00:16:45.860 --> 00:16:47.770 align:middle line:90% is, in fact, down at the bottom. 00:16:47.770 --> 00:16:51.210 align:middle line:90% So what are we going do? 00:16:51.210 --> 00:16:55.680 align:middle line:84% AUDIENCE: So I think rather than having a medical genomicist, 00:16:55.680 --> 00:17:02.700 align:middle line:84% it would be nice to have a type of physician who organized 00:17:02.700 --> 00:17:06.450 align:middle line:84% with different levels of care into a seamless approach 00:17:06.450 --> 00:17:09.270 align:middle line:84% to be able to [INAUDIBLE] approach, 00:17:09.270 --> 00:17:11.040 align:middle line:84% to address the [INAUDIBLE] you have. 00:17:11.040 --> 00:17:15.405 align:middle line:90% 00:17:15.405 --> 00:17:19.589 align:middle line:84% Maybe the genomicist would be appropriate for homing 00:17:19.589 --> 00:17:22.950 align:middle line:84% the genetic aspects [INAUDIBLE] collaboration. 00:17:22.950 --> 00:17:27.390 align:middle line:84% You have somebody who is seeing a patient, outside 00:17:27.390 --> 00:17:33.180 align:middle line:84% of their regular PCP, who can integrate the latest 00:17:33.180 --> 00:17:36.090 align:middle line:84% technologies along with the traditional therapies 00:17:36.090 --> 00:17:37.920 align:middle line:84% and can communicate the process-- 00:17:37.920 --> 00:17:39.900 align:middle line:84% ISAAC SAMUEL KOHANE: So basically someone 00:17:39.900 --> 00:17:42.660 align:middle line:84% who's a generalist in medicine but a specialist 00:17:42.660 --> 00:17:48.060 align:middle line:84% about inherited diseases, and complex diseases, and also 00:17:48.060 --> 00:17:51.413 align:middle line:84% a team communicator, coordinator. 00:17:51.413 --> 00:17:52.080 align:middle line:90% AUDIENCE: Right. 00:17:52.080 --> 00:17:57.372 align:middle line:84% So for each patient, you've got one person who's 00:17:57.372 --> 00:17:59.580 align:middle line:84% not having to worry about the occasional sniffle here 00:17:59.580 --> 00:18:02.010 align:middle line:84% or there, like a PCP, but also is 00:18:02.010 --> 00:18:04.890 align:middle line:84% focused on a bigger picture than maybe the specialist 00:18:04.890 --> 00:18:06.100 align:middle line:90% or the geneticists. 00:18:06.100 --> 00:18:07.350 align:middle line:90% ISAAC SAMUEL KOHANE: Go ahead. 00:18:07.350 --> 00:18:09.475 align:middle line:84% AUDIENCE: I think it's unrealistic to have a team-- 00:18:09.475 --> 00:18:11.358 align:middle line:84% I wish I had a team of specialists around me 00:18:11.358 --> 00:18:12.900 align:middle line:84% all the time to do this stuff for me. 00:18:12.900 --> 00:18:15.960 align:middle line:84% I think the reality is you're going to have, 00:18:15.960 --> 00:18:17.700 align:middle line:84% at most, one person who knows anything 00:18:17.700 --> 00:18:20.670 align:middle line:84% about your health, any ongoing basis. 00:18:20.670 --> 00:18:22.770 align:middle line:84% And that person should be the counselor 00:18:22.770 --> 00:18:24.930 align:middle line:84% that let you know things about lifestyle, 00:18:24.930 --> 00:18:27.360 align:middle line:90% things about preventive care. 00:18:27.360 --> 00:18:29.650 align:middle line:84% And they should have genomic information 00:18:29.650 --> 00:18:32.850 align:middle line:84% to help give you advice about your particular lifestyle 00:18:32.850 --> 00:18:33.960 align:middle line:90% decisions you make. 00:18:33.960 --> 00:18:35.920 align:middle line:84% And that should be the primary care physician. 00:18:35.920 --> 00:18:37.950 align:middle line:84% So I think that it's going to be really 00:18:37.950 --> 00:18:40.440 align:middle line:84% hard to have a lot of people around, 00:18:40.440 --> 00:18:42.600 align:middle line:84% these specialists like genomicists giving you 00:18:42.600 --> 00:18:46.470 align:middle line:84% this preventive care type information. 00:18:46.470 --> 00:18:48.280 align:middle line:90% Because they don't that well. 00:18:48.280 --> 00:18:49.800 align:middle line:84% ISAAC SAMUEL KOHANE: So you say the primary care practitioner. 00:18:49.800 --> 00:18:50.502 align:middle line:90% Cecily. 00:18:50.502 --> 00:18:52.710 align:middle line:84% AUDIENCE: I mean, to be honest, more and more genomic 00:18:52.710 --> 00:18:53.525 align:middle line:90% tests come online. 00:18:53.525 --> 00:18:55.530 align:middle line:84% Like, look at cancer microarrays. 00:18:55.530 --> 00:18:58.380 align:middle line:84% You're just going to need people to know about the techniques 00:18:58.380 --> 00:18:59.730 align:middle line:90% and know about the methods. 00:18:59.730 --> 00:19:03.360 align:middle line:84% And you can help the physicians interpret it. 00:19:03.360 --> 00:19:05.633 align:middle line:84% I mean, there's going to be a role for them. 00:19:05.633 --> 00:19:07.050 align:middle line:84% We're seeing already there's going 00:19:07.050 --> 00:19:11.880 align:middle line:84% to be an explosion of SNPs, and pharmacogenomics, 00:19:11.880 --> 00:19:12.775 align:middle line:90% and all this stuff. 00:19:12.775 --> 00:19:14.400 align:middle line:84% And you're going to need to have people 00:19:14.400 --> 00:19:17.520 align:middle line:84% who know the techniques who can help the primary care-- 00:19:17.520 --> 00:19:21.662 align:middle line:90% 00:19:21.662 --> 00:19:24.120 align:middle line:84% ISAAC SAMUEL KOHANE: Absolutely true of what you both said. 00:19:24.120 --> 00:19:27.150 align:middle line:84% On one hand, it's unlikely we'll have swarms 00:19:27.150 --> 00:19:29.340 align:middle line:90% of people worrying about us. 00:19:29.340 --> 00:19:32.730 align:middle line:84% And yet, there is going to be this explosion. 00:19:32.730 --> 00:19:33.540 align:middle line:90% And no one person-- 00:19:33.540 --> 00:19:37.440 align:middle line:84% I mean, it's already pretty hard to be a good primary care 00:19:37.440 --> 00:19:41.520 align:middle line:84% practitioner with knowledge circa 1990. 00:19:41.520 --> 00:19:43.990 align:middle line:90% How are they going to do that? 00:19:43.990 --> 00:19:44.490 align:middle line:90% Any ideas? 00:19:44.490 --> 00:19:51.060 align:middle line:90% 00:19:51.060 --> 00:19:54.360 align:middle line:84% Operationalize what you were suggesting. 00:19:54.360 --> 00:19:57.510 align:middle line:84% AUDIENCE: You might have, in a hospital, a medical genomicist 00:19:57.510 --> 00:20:00.780 align:middle line:84% that these primary care physicians can call up 00:20:00.780 --> 00:20:04.890 align:middle line:84% and refer a patient to the help interpret things. 00:20:04.890 --> 00:20:07.170 align:middle line:84% ISAAC SAMUEL KOHANE: But what if it's primary care? 00:20:07.170 --> 00:20:09.930 align:middle line:84% I mean, we're all going to, unfortunately-- 00:20:09.930 --> 00:20:12.540 align:middle line:84% unless one of is a major messianic figure-- 00:20:12.540 --> 00:20:16.120 align:middle line:84% going to die from one disease or other. 00:20:16.120 --> 00:20:17.730 align:middle line:90% So we're at risk. 00:20:17.730 --> 00:20:20.170 align:middle line:84% And, hopefully, it will not be a bus hitting us. 00:20:20.170 --> 00:20:22.437 align:middle line:84% So it's going to be something that's going to come-- 00:20:22.437 --> 00:20:24.520 align:middle line:84% interaction between the environment and our genes. 00:20:24.520 --> 00:20:27.370 align:middle line:90% 00:20:27.370 --> 00:20:28.720 align:middle line:90% So that's for all of us. 00:20:28.720 --> 00:20:31.240 align:middle line:84% So, basically, we need that information, 00:20:31.240 --> 00:20:32.830 align:middle line:84% unless we don't want the information, 00:20:32.830 --> 00:20:35.320 align:middle line:90% for every one of us. 00:20:35.320 --> 00:20:39.540 align:middle line:84% It's not going to be, hey, I have an interesting problem. 00:20:39.540 --> 00:20:43.270 align:middle line:84% It's part of routine care, I think. 00:20:43.270 --> 00:20:44.560 align:middle line:90% Everybody needs to be-- 00:20:44.560 --> 00:20:48.710 align:middle line:84% AUDIENCE: There could be a department of genomics in any-- 00:20:48.710 --> 00:20:50.710 align:middle line:84% ISAAC SAMUEL KOHANE: That's the academic answer. 00:20:50.710 --> 00:20:52.720 align:middle line:84% I'm talking about the process in the field. 00:20:52.720 --> 00:20:55.870 align:middle line:84% Who's going to be the guy or girl who's 00:20:55.870 --> 00:20:59.230 align:middle line:84% going to be providing that knowledge, 00:20:59.230 --> 00:21:02.260 align:middle line:84% either as a primary care practitioner 00:21:02.260 --> 00:21:08.100 align:middle line:84% or to the primary care practitioner? 00:21:08.100 --> 00:21:11.190 align:middle line:90% So here's some ideas I had. 00:21:11.190 --> 00:21:13.950 align:middle line:84% Your ideas are-- by the way-- are as good as mine. 00:21:13.950 --> 00:21:16.440 align:middle line:84% And I want to tell you, this is actually the central-- 00:21:16.440 --> 00:21:18.420 align:middle line:84% I mean, this course is about genomic medicine. 00:21:18.420 --> 00:21:22.440 align:middle line:84% And I think this is a central conundrum. 00:21:22.440 --> 00:21:26.868 align:middle line:84% There is no plan right now how to bring all of this 00:21:26.868 --> 00:21:27.720 align:middle line:90% into the field. 00:21:27.720 --> 00:21:30.630 align:middle line:84% And I can tell you, in pediatric endocrinology, 00:21:30.630 --> 00:21:36.480 align:middle line:84% we're not even ordering the autosomal-dominant, 00:21:36.480 --> 00:21:38.760 align:middle line:84% well-understood, highly-penetrant single genes 00:21:38.760 --> 00:21:41.760 align:middle line:84% that we know about because of the educational problems, let 00:21:41.760 --> 00:21:47.770 align:middle line:84% alone all these other complex diseases. 00:21:47.770 --> 00:21:52.920 align:middle line:84% So any idea that you have could be of great interest 00:21:52.920 --> 00:21:56.760 align:middle line:84% to our system or to a company, if you could actually figure 00:21:56.760 --> 00:21:58.740 align:middle line:90% out how to do this right. 00:21:58.740 --> 00:22:01.200 align:middle line:84% AUDIENCE: Some of it is going to either market-driven-- 00:22:01.200 --> 00:22:01.620 align:middle line:84% ISAAC SAMUEL KOHANE: It's definitely 00:22:01.620 --> 00:22:02.580 align:middle line:90% going to be market-driven. 00:22:02.580 --> 00:22:04.538 align:middle line:84% AUDIENCE: --I mean, right now, [INAUDIBLE] done 00:22:04.538 --> 00:22:07.500 align:middle line:84% a lot of research in but there's a really good test and really 00:22:07.500 --> 00:22:11.280 align:middle line:84% good results come online and generate high-profile papers 00:22:11.280 --> 00:22:14.100 align:middle line:84% and get into some of the patients 00:22:14.100 --> 00:22:15.870 align:middle line:84% are going to start demanding these things. 00:22:15.870 --> 00:22:17.620 align:middle line:84% ISAAC SAMUEL KOHANE: But that's a problem. 00:22:17.620 --> 00:22:20.820 align:middle line:84% For instance, a prostate-specific antigen 00:22:20.820 --> 00:22:23.190 align:middle line:84% done wrongly gets a bunch of poor guys 00:22:23.190 --> 00:22:26.172 align:middle line:84% get their prostates removed for nothing. 00:22:26.172 --> 00:22:32.388 align:middle line:84% AUDIENCE: [INAUDIBLE] error too, 7% false-- 00:22:32.388 --> 00:22:33.930 align:middle line:84% ISAAC SAMUEL KOHANE: False positives. 00:22:33.930 --> 00:22:35.097 align:middle line:90% AUDIENCE: --false positives. 00:22:35.097 --> 00:22:37.830 align:middle line:84% ISAAC SAMUEL KOHANE: Yeah, and so that's the problem. 00:22:37.830 --> 00:22:40.800 align:middle line:84% And, as I explained to you in my first lecture, 00:22:40.800 --> 00:22:43.740 align:middle line:84% there's 7,000 articles just on appropriate use 00:22:43.740 --> 00:22:47.280 align:middle line:84% of the prostate-specific antigen. 00:22:47.280 --> 00:22:51.570 align:middle line:84% We have now 30,000 genes to worry about. 00:22:51.570 --> 00:22:54.060 align:middle line:90% And what are the right cutoffs. 00:22:54.060 --> 00:22:56.550 align:middle line:84% And let's be optimistic and say we can actually get 00:22:56.550 --> 00:22:58.170 align:middle line:90% the optimal answers for these. 00:22:58.170 --> 00:23:03.160 align:middle line:84% No one individual I can think can actually solve it. 00:23:03.160 --> 00:23:04.285 align:middle line:90% So you're absolutely right. 00:23:04.285 --> 00:23:05.327 align:middle line:90% It will be market-driven. 00:23:05.327 --> 00:23:06.270 align:middle line:90% That's certainly true. 00:23:06.270 --> 00:23:10.230 align:middle line:84% But the question is, how will the market solve that problem? 00:23:10.230 --> 00:23:12.510 align:middle line:84% And I'm claiming there are companies out there that 00:23:12.510 --> 00:23:13.320 align:middle line:90% will solve that. 00:23:13.320 --> 00:23:16.770 align:middle line:84% And either it'll be inspired by ideas 00:23:16.770 --> 00:23:18.960 align:middle line:84% similar to ones we're going to articulate today 00:23:18.960 --> 00:23:20.080 align:middle line:90% or by other ideas. 00:23:20.080 --> 00:23:22.680 align:middle line:84% So all I've been calling for is thought about this. 00:23:22.680 --> 00:23:25.560 align:middle line:84% So one was internet-enabled triage specialists. 00:23:25.560 --> 00:23:28.980 align:middle line:84% Basically, you feed your genome to, essentially, a service 00:23:28.980 --> 00:23:30.060 align:middle line:90% on the internet. 00:23:30.060 --> 00:23:34.290 align:middle line:84% And they take whatever the primary care physician says 00:23:34.290 --> 00:23:36.735 align:middle line:84% and they basically-- in India or somewhere else-- 00:23:36.735 --> 00:23:39.780 align:middle line:90% 00:23:39.780 --> 00:23:42.270 align:middle line:84% provide you with your risk profile and the next steps 00:23:42.270 --> 00:23:42.960 align:middle line:90% for you to take. 00:23:42.960 --> 00:23:44.970 align:middle line:90% That's one possibility. 00:23:44.970 --> 00:23:46.410 align:middle line:84% So you check each patient in door 00:23:46.410 --> 00:23:49.830 align:middle line:84% for inheritable genetic or epigenetic phenomenon. 00:23:49.830 --> 00:23:54.570 align:middle line:84% A parallel system-- that's, essentially, 00:23:54.570 --> 00:23:58.850 align:middle line:84% a fly on the wall to a mental interview, 00:23:58.850 --> 00:24:00.830 align:middle line:90% provide decision supports. 00:24:00.830 --> 00:24:05.390 align:middle line:84% And this could be a person or, more likely, 00:24:05.390 --> 00:24:08.330 align:middle line:90% a computer program. 00:24:08.330 --> 00:24:11.270 align:middle line:84% But it's going to don't have to be a damn good computer program 00:24:11.270 --> 00:24:12.360 align:middle line:90% to be able to do that. 00:24:12.360 --> 00:24:16.740 align:middle line:84% But I was involved in knowledge representation 00:24:16.740 --> 00:24:20.600 align:middle line:84% and knowledge-based systems in the 80s when I did my thesis. 00:24:20.600 --> 00:24:22.808 align:middle line:84% But there really wasn't any good motivation for it. 00:24:22.808 --> 00:24:24.350 align:middle line:84% Because, in the end, doctors know how 00:24:24.350 --> 00:24:28.520 align:middle line:90% to diagnose acid base problems. 00:24:28.520 --> 00:24:30.810 align:middle line:84% But doctors will never be able to do this. 00:24:30.810 --> 00:24:32.570 align:middle line:84% So this is a true motivation, I believe, 00:24:32.570 --> 00:24:33.920 align:middle line:90% for automated decision support. 00:24:33.920 --> 00:24:37.830 align:middle line:90% 00:24:37.830 --> 00:24:42.600 align:middle line:84% The alternative is just to redo the curriculum in a major way, 00:24:42.600 --> 00:24:47.910 align:middle line:84% and even so, to teach students how to use electronic resources 00:24:47.910 --> 00:24:51.030 align:middle line:90% in near real time. 00:24:51.030 --> 00:24:54.732 align:middle line:84% Because no one will be able to keep this in their head. 00:24:54.732 --> 00:24:56.190 align:middle line:84% AUDIENCE: For the last one, I think 00:24:56.190 --> 00:25:02.190 align:middle line:84% it would require such simplification of outputs 00:25:02.190 --> 00:25:06.270 align:middle line:84% in terms feeding a genome in and spitting out numbers, 00:25:06.270 --> 00:25:08.835 align:middle line:84% the way that [INAUDIBLE] and all these other things, 00:25:08.835 --> 00:25:09.960 align:middle line:90% because that's how-- right? 00:25:09.960 --> 00:25:10.560 align:middle line:90% I mean-- 00:25:10.560 --> 00:25:12.720 align:middle line:84% ISAAC SAMUEL KOHANE: But it would not be a number. 00:25:12.720 --> 00:25:15.570 align:middle line:84% When I was talking about when you're 00:25:15.570 --> 00:25:17.160 align:middle line:90% feeding the genome [INAUDIBLE]. 00:25:17.160 --> 00:25:19.170 align:middle line:84% The answer would not be these are the levels 00:25:19.170 --> 00:25:20.340 align:middle line:90% of your various things. 00:25:20.340 --> 00:25:23.340 align:middle line:84% It would be you have a 30% increased 00:25:23.340 --> 00:25:25.947 align:middle line:90% risk for prostate cancer. 00:25:25.947 --> 00:25:27.780 align:middle line:84% The following test is the right thing to do, 00:25:27.780 --> 00:25:29.100 align:middle line:90% that kind of thing. 00:25:29.100 --> 00:25:33.240 align:middle line:84% Because, again, no primary care physician can know it. 00:25:33.240 --> 00:25:38.880 align:middle line:84% As I said, almost none of them know fully all the literature 00:25:38.880 --> 00:25:43.440 align:middle line:84% and prostate-specific antigen screening for prostate cancer 00:25:43.440 --> 00:25:45.420 align:middle line:90% and let alone for 30,000 tests. 00:25:45.420 --> 00:25:50.280 align:middle line:84% So I think it would have to be a distilled, utility-based, 00:25:50.280 --> 00:25:52.390 align:middle line:90% sensitivity, specificity-based-- 00:25:52.390 --> 00:25:54.280 align:middle line:90% AUDIENCE: That's what I mean. 00:25:54.280 --> 00:25:55.800 align:middle line:84% They get a [INAUDIBLE] from the lab. 00:25:55.800 --> 00:25:56.727 align:middle line:90% They get it from the-- 00:25:56.727 --> 00:25:58.185 align:middle line:84% ISAAC SAMUEL KOHANE: --oh, it would 00:25:58.185 --> 00:25:59.352 align:middle line:90% be simplified in that sense. 00:25:59.352 --> 00:26:00.210 align:middle line:90% Absolutely. 00:26:00.210 --> 00:26:02.870 align:middle line:90% I think that is the only way. 00:26:02.870 --> 00:26:05.460 align:middle line:84% And endocrinologists have made a big career 00:26:05.460 --> 00:26:08.670 align:middle line:84% out of just interpreting three numbers-- 00:26:08.670 --> 00:26:13.008 align:middle line:84% TSH, TBGI, and T4 in the thyroid test. 00:26:13.008 --> 00:26:13.800 align:middle line:90% Because guess what? 00:26:13.800 --> 00:26:15.060 align:middle line:84% Most medical students can never figure out 00:26:15.060 --> 00:26:17.018 align:middle line:84% which way means hypothyrodism, hyperthyroidism, 00:26:17.018 --> 00:26:21.300 align:middle line:84% or changes in the amount of binding protein. 00:26:21.300 --> 00:26:22.800 align:middle line:84% AUDIENCE: Doesn't that then relegate 00:26:22.800 --> 00:26:26.630 align:middle line:84% the medical geneticists to just the category of specialist 00:26:26.630 --> 00:26:28.500 align:middle line:90% once again? 00:26:28.500 --> 00:26:31.860 align:middle line:90% the lack of a specific disease-- 00:26:31.860 --> 00:26:39.840 align:middle line:84% ISAAC SAMUEL KOHANE: I'm claiming that the system cannot 00:26:39.840 --> 00:26:44.130 align:middle line:84% have a expert medical genomicist. 00:26:44.130 --> 00:26:47.250 align:middle line:84% Because a lot of what we were talking about is primary care, 00:26:47.250 --> 00:26:48.360 align:middle line:90% is prevention. 00:26:48.360 --> 00:26:52.290 align:middle line:84% This has to be in every primary care situation. 00:26:52.290 --> 00:26:55.770 align:middle line:84% So it's either outsourced, essentially, to the internet, 00:26:55.770 --> 00:26:58.440 align:middle line:84% to some other group of people, have 00:26:58.440 --> 00:27:02.790 align:middle line:84% a low-cost parallel assistant, whether it's a human or, more 00:27:02.790 --> 00:27:04.290 align:middle line:90% likely, a computer program. 00:27:04.290 --> 00:27:07.050 align:middle line:84% And the third is redoing education. 00:27:07.050 --> 00:27:09.981 align:middle line:90% 00:27:09.981 --> 00:27:13.270 align:middle line:84% To your scenario, which, I think, is one of the good ones, 00:27:13.270 --> 00:27:14.850 align:middle line:84% which is to have a very simplified 00:27:14.850 --> 00:27:19.110 align:middle line:84% set of recommendations come to the primary care physician. 00:27:19.110 --> 00:27:21.060 align:middle line:84% It's the primary care physician, still, 00:27:21.060 --> 00:27:23.670 align:middle line:84% who just has to learn how to look intelligently 00:27:23.670 --> 00:27:26.310 align:middle line:84% at that report, that set of recommendations. 00:27:26.310 --> 00:27:31.140 align:middle line:84% I don't see them having to learn the 30,000 genes and all 00:27:31.140 --> 00:27:32.850 align:middle line:84% the epigenetic and genetic effects. 00:27:32.850 --> 00:27:34.650 align:middle line:90% That does not seem reasonable. 00:27:34.650 --> 00:27:36.810 align:middle line:84% But this is the fundamental problem 00:27:36.810 --> 00:27:38.850 align:middle line:84% that we have in genomic medicine. 00:27:38.850 --> 00:27:42.750 align:middle line:84% And that's going to be, by far, the rate-limiting step 00:27:42.750 --> 00:27:46.680 align:middle line:84% for any penetration to managing human disease 00:27:46.680 --> 00:27:51.540 align:middle line:84% than any other aspect of the genomic enterprise. 00:27:51.540 --> 00:27:54.450 align:middle line:84% We're going to discover lots of interesting things. 00:27:54.450 --> 00:27:56.240 align:middle line:90% We're going to find good drugs. 00:27:56.240 --> 00:27:57.990 align:middle line:84% We're going to make interesting diagnoses. 00:27:57.990 --> 00:28:03.970 align:middle line:84% But translating it into cares will be the real challenge. 00:28:03.970 --> 00:28:06.420 align:middle line:84% So how are we going to pay for those roles? 00:28:06.420 --> 00:28:07.590 align:middle line:90% Extremely unclear. 00:28:07.590 --> 00:28:10.125 align:middle line:84% Right now, you can't get a lot of genetic tests paid for. 00:28:10.125 --> 00:28:14.630 align:middle line:90% 00:28:14.630 --> 00:28:17.810 align:middle line:84% AUDIENCE: [INAUDIBLE] because it hasn't been developed 00:28:17.810 --> 00:28:20.645 align:middle line:90% enough in terms of [INAUDIBLE] 00:28:20.645 --> 00:28:21.770 align:middle line:90% ISAAC SAMUEL KOHANE: Right. 00:28:21.770 --> 00:28:25.778 align:middle line:84% AUDIENCE: [INAUDIBLE] combined [INAUDIBLE] result. 00:28:25.778 --> 00:28:27.320 align:middle line:84% ISAAC SAMUEL KOHANE: With the result. 00:28:27.320 --> 00:28:30.470 align:middle line:84% And, finally, because it's very expensive. 00:28:30.470 --> 00:28:32.820 align:middle line:84% And if it became free, everybody would use it. 00:28:32.820 --> 00:28:35.930 align:middle line:84% I mean, right now, as I told you the first lecture, 00:28:35.930 --> 00:28:41.600 align:middle line:84% in a high-throughput lab, it costs $0.10 to do a genotype. 00:28:41.600 --> 00:28:47.850 align:middle line:84% Do you know how much the system is built for, for that test? 00:28:47.850 --> 00:28:49.700 align:middle line:84% So I was talking about, at first, 00:28:49.700 --> 00:28:52.940 align:middle line:84% $0.10 in order to tell you, wow, the genomic future is now. 00:28:52.940 --> 00:28:54.620 align:middle line:90% We can do very cheap screening. 00:28:54.620 --> 00:28:56.930 align:middle line:84% But how much is our insurance going 00:28:56.930 --> 00:29:01.886 align:middle line:84% to be billed for these tests, for that single genotype? 00:29:01.886 --> 00:29:03.620 align:middle line:90% AUDIENCE: Over $1,000. 00:29:03.620 --> 00:29:05.540 align:middle line:84% ISAAC SAMUEL KOHANE: Over $1,000. 00:29:05.540 --> 00:29:11.827 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:29:11.827 --> 00:29:13.910 align:middle line:84% ISAAC SAMUEL KOHANE: And that's not the real cost. 00:29:13.910 --> 00:29:17.190 align:middle line:90% That's a huge profit margin. 00:29:17.190 --> 00:29:19.190 align:middle line:84% AUDIENCE: There's been a similar debate going on 00:29:19.190 --> 00:29:21.917 align:middle line:84% with these full-body MRI scans and things, right? 00:29:21.917 --> 00:29:23.000 align:middle line:90% ISAAC SAMUEL KOHANE: Sure. 00:29:23.000 --> 00:29:25.490 align:middle line:84% AUDIENCE: And so, I mean, where are 00:29:25.490 --> 00:29:29.990 align:middle line:84% the voices in the field for this and for those saying, 00:29:29.990 --> 00:29:36.300 align:middle line:84% preventatively, if we do a whole battery of non-invasive, 00:29:36.300 --> 00:29:39.830 align:middle line:84% one-day tests- submit a sample that you 00:29:39.830 --> 00:29:41.930 align:middle line:90% can do with genomic testing. 00:29:41.930 --> 00:29:44.360 align:middle line:84% You sit in the MRI and do all the stuff. 00:29:44.360 --> 00:29:46.560 align:middle line:84% And then, hey, we found this particular disease. 00:29:46.560 --> 00:29:49.372 align:middle line:84% And we just saved 60, 70 maybe $100,000 worth of treatment. 00:29:49.372 --> 00:29:51.330 align:middle line:84% ISAAC SAMUEL KOHANE: Well, first of all, that's 00:29:51.330 --> 00:29:52.622 align:middle line:90% not the way the system sees it. 00:29:52.622 --> 00:29:58.430 align:middle line:84% The system sees that you spent $4,000 bucks. 00:29:58.430 --> 00:30:02.528 align:middle line:84% And my insurance company is going 00:30:02.528 --> 00:30:04.820 align:middle line:84% to reward me for what I do the next two or three years. 00:30:04.820 --> 00:30:08.450 align:middle line:84% If you drop dead of cancer 10 years from now, 00:30:08.450 --> 00:30:11.330 align:middle line:84% that's going to be the balance on the insurance company 00:30:11.330 --> 00:30:12.080 align:middle line:90% 10 years from now. 00:30:12.080 --> 00:30:13.538 align:middle line:84% There's not a lot of payback to me. 00:30:13.538 --> 00:30:15.560 align:middle line:84% But more importantly, speaking as a clinician, 00:30:15.560 --> 00:30:17.600 align:middle line:84% I think these total body MRIs are terrible. 00:30:17.600 --> 00:30:20.773 align:middle line:84% Because you're finding a lot of incidental findings. 00:30:20.773 --> 00:30:22.940 align:middle line:84% In fact, the medical establishment is against those, 00:30:22.940 --> 00:30:23.990 align:middle line:90% I think, for good reason. 00:30:23.990 --> 00:30:28.670 align:middle line:84% Because, again, the probabilities 00:30:28.670 --> 00:30:33.950 align:middle line:84% that we have around the meaning of bumps on the MRI 00:30:33.950 --> 00:30:36.050 align:middle line:84% is around people presenting with certain symptoms. 00:30:36.050 --> 00:30:37.940 align:middle line:84% Because these studies are aimed around that. 00:30:37.940 --> 00:30:40.700 align:middle line:84% And if you just have Joe Blow off the street who just has 00:30:40.700 --> 00:30:43.415 align:middle line:84% a birthday present, a full-body MRI scan-- 00:30:43.415 --> 00:30:45.290 align:middle line:84% which, by the way, haven't you heard 00:30:45.290 --> 00:30:47.540 align:middle line:90% the advertisements for it? 00:30:47.540 --> 00:30:49.427 align:middle line:90% Yeah, it's your 40th birthday. 00:30:49.427 --> 00:30:50.510 align:middle line:90% Don't you want to do this? 00:30:50.510 --> 00:30:53.990 align:middle line:90% Or your dad's birthday-- 00:30:53.990 --> 00:30:57.650 align:middle line:84% and the advertisement is Joe Blow, the patient, 00:30:57.650 --> 00:30:59.120 align:middle line:90% reports, oh, I'm so reassured. 00:30:59.120 --> 00:31:00.740 align:middle line:84% On my 40th birthday, seemed nothing's wrong with me. 00:31:00.740 --> 00:31:02.750 align:middle line:84% What they don't report on is all the people 00:31:02.750 --> 00:31:04.250 align:middle line:84% who got bumps in their head and then 00:31:04.250 --> 00:31:06.200 align:middle line:84% had to go through craniotomy, God 00:31:06.200 --> 00:31:09.450 align:middle line:84% knows what to investigate these things. 00:31:09.450 --> 00:31:12.890 align:middle line:84% So I promise you that we will stick to the pragmatics. 00:31:12.890 --> 00:31:15.080 align:middle line:84% So this is why there's not penetration today. 00:31:15.080 --> 00:31:20.420 align:middle line:84% But today, how do you order a genetic test 00:31:20.420 --> 00:31:21.950 align:middle line:90% for a clinical problem? 00:31:21.950 --> 00:31:22.730 align:middle line:90% What do you do? 00:31:22.730 --> 00:31:24.920 align:middle line:90% 00:31:24.920 --> 00:31:26.170 align:middle line:90% AUDIENCE: How do you order it? 00:31:26.170 --> 00:31:26.460 align:middle line:90% ISAAC SAMUEL KOHANE: Well, yeah. 00:31:26.460 --> 00:31:28.010 align:middle line:90% What do you do? 00:31:28.010 --> 00:31:30.150 align:middle line:84% AUDIENCE: You go to geneclinics.org and you check 00:31:30.150 --> 00:31:36.745 align:middle line:84% if it has it and whether it's a research lab or clean up-- 00:31:36.745 --> 00:31:38.620 align:middle line:84% ISAAC SAMUEL KOHANE: Did everybody hear that? 00:31:38.620 --> 00:31:40.300 align:middle line:90% That's very, very important. 00:31:40.300 --> 00:31:43.870 align:middle line:90% Gene clinics and genetests.org. 00:31:43.870 --> 00:31:46.840 align:middle line:84% It's one of the things better stick to you after this class. 00:31:46.840 --> 00:31:51.130 align:middle line:84% Because that there are thousands of clinics and labs 00:31:51.130 --> 00:31:53.390 align:middle line:84% across the United States that have one or two tests. 00:31:53.390 --> 00:31:56.920 align:middle line:84% There's no single, giant aggregator of genetic tests. 00:31:56.920 --> 00:31:58.870 align:middle line:84% And there's this one website, maintained 00:31:58.870 --> 00:32:01.210 align:middle line:84% by my colleague, Peter Tarczy, at the University 00:32:01.210 --> 00:32:03.890 align:middle line:84% of Washington, that has all that. 00:32:03.890 --> 00:32:06.832 align:middle line:90% So that's exactly right. 00:32:06.832 --> 00:32:09.127 align:middle line:84% AUDIENCE: I mean, just a question of the market stuff. 00:32:09.127 --> 00:32:10.960 align:middle line:84% Do you think that's going to eventually roll 00:32:10.960 --> 00:32:12.970 align:middle line:90% into the [INAUDIBLE] 00:32:12.970 --> 00:32:14.230 align:middle line:90% ISAAC SAMUEL KOHANE: 100%. 00:32:14.230 --> 00:32:16.135 align:middle line:90% So that's an excellent question. 00:32:16.135 --> 00:32:21.610 align:middle line:90% 00:32:21.610 --> 00:32:24.400 align:middle line:84% OK, let me just get to that question in two minutes. 00:32:24.400 --> 00:32:28.338 align:middle line:90% 00:32:28.338 --> 00:32:29.380 align:middle line:90% Sorry, it's an old slide. 00:32:29.380 --> 00:32:30.670 align:middle line:90% It's a year-old slide. 00:32:30.670 --> 00:32:34.360 align:middle line:84% And they had-- on gene clinics, gene tests-- 00:32:34.360 --> 00:32:37.030 align:middle line:84% 189 gene reviews, which are very nice 00:32:37.030 --> 00:32:41.110 align:middle line:84% monographs around a specific disease involving 00:32:41.110 --> 00:32:44.860 align:middle line:84% 1,000 different clinics and 500 laboratories covering 00:32:44.860 --> 00:32:49.000 align:middle line:90% 948 diseases. 00:32:49.000 --> 00:32:53.920 align:middle line:90% And the short answer is, yes. 00:32:53.920 --> 00:32:58.973 align:middle line:84% These are going to be rolled up into requests 00:32:58.973 --> 00:33:00.640 align:middle line:84% and the various laboratory corporations. 00:33:00.640 --> 00:33:03.430 align:middle line:90% The real question is-- 00:33:03.430 --> 00:33:05.680 align:middle line:84% ultimately, we know that what it's going to look like. 00:33:05.680 --> 00:33:10.450 align:middle line:84% It's going to be a highly-roboticized sequencing 00:33:10.450 --> 00:33:12.490 align:middle line:84% genotyping operation in these labs. 00:33:12.490 --> 00:33:16.810 align:middle line:84% But in the interim, it may still be that it's rolled up. 00:33:16.810 --> 00:33:18.700 align:middle line:84% But the actual sequencing and procedures 00:33:18.700 --> 00:33:21.320 align:middle line:84% might still be at these various places. 00:33:21.320 --> 00:33:23.748 align:middle line:84% And these companies might just contract 00:33:23.748 --> 00:33:24.790 align:middle line:90% to these various sources. 00:33:24.790 --> 00:33:27.290 align:middle line:84% Right now, for instance, if you want to do Duchenne testing, 00:33:27.290 --> 00:33:28.970 align:middle line:84% your sample goes to a place in Utah. 00:33:28.970 --> 00:33:30.100 align:middle line:90% Now why is that? 00:33:30.100 --> 00:33:31.660 align:middle line:84% Why not just roll into one place? 00:33:31.660 --> 00:33:36.160 align:middle line:84% Because the guys who are resequencing the Duchenne 00:33:36.160 --> 00:33:38.080 align:middle line:84% gene for Duchenne muscular dystrophy 00:33:38.080 --> 00:33:43.030 align:middle line:84% know which primers aka which sequence works the best 00:33:43.030 --> 00:33:45.850 align:middle line:84% to amplify different segments of the gene 00:33:45.850 --> 00:33:47.410 align:middle line:90% to get reliable results. 00:33:47.410 --> 00:33:51.040 align:middle line:84% And that know-how ultimately can be definitely rolled up. 00:33:51.040 --> 00:33:55.510 align:middle line:84% But in the interim, on the five-year timescale, 00:33:55.510 --> 00:33:58.640 align:middle line:84% may or may not be just left locally. 00:33:58.640 --> 00:34:02.230 align:middle line:84% And so what these requests will be 00:34:02.230 --> 00:34:04.900 align:middle line:84% are more aggregators on front ends, 00:34:04.900 --> 00:34:08.050 align:middle line:84% conduits or channels-- as the marketers 00:34:08.050 --> 00:34:10.330 align:middle line:84% like to say-- to these myriad labs. 00:34:10.330 --> 00:34:12.159 align:middle line:84% Ultimately, I think everybody believes 00:34:12.159 --> 00:34:14.110 align:middle line:84% they'll be one highly-roboticized facility, 00:34:14.110 --> 00:34:16.960 align:middle line:90% each one of these facilities. 00:34:16.960 --> 00:34:19.090 align:middle line:84% Because after all, a gene is a gene is a gene. 00:34:19.090 --> 00:34:23.800 align:middle line:84% But because of the things like knowledge 00:34:23.800 --> 00:34:29.020 align:middle line:84% of what primers work, knowledge of which mutations are common 00:34:29.020 --> 00:34:31.300 align:middle line:90% and what they mean-- 00:34:31.300 --> 00:34:33.790 align:middle line:84% all that knowledge management is still distributed. 00:34:33.790 --> 00:34:39.820 align:middle line:84% And rolling it up is going to be, I think, the brake 00:34:39.820 --> 00:34:43.040 align:middle line:90% on that full roll-up. 00:34:43.040 --> 00:34:46.670 align:middle line:84% So for those of you who don't know the site. 00:34:46.670 --> 00:34:48.790 align:middle line:84% So if you want to test for a particular problem, 00:34:48.790 --> 00:34:53.469 align:middle line:84% you can look at this thing and say, I want to look at obesity. 00:34:53.469 --> 00:34:55.750 align:middle line:84% And it'll tell you which laboratory 00:34:55.750 --> 00:34:59.350 align:middle line:90% will do which testing it for. 00:34:59.350 --> 00:35:03.370 align:middle line:84% In fact, probably the only obesity-related gene worth 00:35:03.370 --> 00:35:05.650 align:middle line:90% testing for right now is MC4R. 00:35:05.650 --> 00:35:12.850 align:middle line:84% It's the most common genetically associated cause of obesity. 00:35:12.850 --> 00:35:19.630 align:middle line:84% MC4R is melanocorticotropin receptor, fourth type. 00:35:19.630 --> 00:35:24.460 align:middle line:84% And some high percentage, on the order of 2% of individuals 00:35:24.460 --> 00:35:26.380 align:middle line:90% with-- 00:35:26.380 --> 00:35:29.950 align:middle line:84% according to some studies-- with morbid obesity 00:35:29.950 --> 00:35:31.660 align:middle line:90% have mutations here. 00:35:31.660 --> 00:35:34.150 align:middle line:84% Unfortunately, for the rest of us, 00:35:34.150 --> 00:35:39.640 align:middle line:84% like me, we're just fat because we eat wrong. 00:35:39.640 --> 00:35:41.080 align:middle line:90% So we're talking about-- 00:35:41.080 --> 00:35:41.590 align:middle line:90% go ahead. 00:35:41.590 --> 00:35:43.060 align:middle line:90% AUDIENCE: Geneclinics.org? 00:35:43.060 --> 00:35:45.610 align:middle line:90% ISAAC SAMUEL KOHANE: Yeah. 00:35:45.610 --> 00:35:48.520 align:middle line:84% But is this indexing, that gene clinics and gene tests 00:35:48.520 --> 00:35:50.780 align:middle line:90% are generally sufficient. 00:35:50.780 --> 00:35:55.970 align:middle line:84% So if you look at adrenal hyperplasia, 00:35:55.970 --> 00:36:00.520 align:middle line:84% which is a disease of children and adulthood. 00:36:00.520 --> 00:36:03.580 align:middle line:84% It's responsible for ambiguous genitalia 00:36:03.580 --> 00:36:08.020 align:middle line:84% in infants, hirsutism, and infertility in females, 00:36:08.020 --> 00:36:12.920 align:middle line:84% and precocious puberty for males and females. 00:36:12.920 --> 00:36:16.073 align:middle line:84% So if you look for adrenal hyperplasia, 00:36:16.073 --> 00:36:16.990 align:middle line:90% you see the following. 00:36:16.990 --> 00:36:20.890 align:middle line:84% You get 21 hydroxylase deficiency. 00:36:20.890 --> 00:36:23.770 align:middle line:84% But what's the right thing to actually order? 00:36:23.770 --> 00:36:25.840 align:middle line:84% Well, if you actually knew something 00:36:25.840 --> 00:36:29.928 align:middle line:84% about the pathway, the steroidogenesis pathway, 00:36:29.928 --> 00:36:31.720 align:middle line:84% you'd know that it starts from cholesterol. 00:36:31.720 --> 00:36:35.140 align:middle line:84% The cholesterol that we all claim to hate actually 00:36:35.140 --> 00:36:37.660 align:middle line:84% is the backbone for all the steroid hormone molecules. 00:36:37.660 --> 00:36:42.070 align:middle line:84% Cholesterol will make all the salt-retaining hormones 00:36:42.070 --> 00:36:46.000 align:middle line:84% like aldosterone, the glucose-stimulating hormones 00:36:46.000 --> 00:36:49.510 align:middle line:84% like cortisol, and the sex hormones like testosterone 00:36:49.510 --> 00:36:51.580 align:middle line:90% and estradiol. 00:36:51.580 --> 00:36:59.830 align:middle line:84% And an appreciation of what the pathways are 00:36:59.830 --> 00:37:03.130 align:middle line:84% is going to allow you to actually do focus testing so 00:37:03.130 --> 00:37:05.780 align:middle line:84% that, for instance, 21 hydroxylase deficiency is, 00:37:05.780 --> 00:37:06.950 align:middle line:90% in fact, the most common. 00:37:06.950 --> 00:37:08.860 align:middle line:84% But there are other deficiencies that will 00:37:08.860 --> 00:37:10.990 align:middle line:90% happened with some regularity. 00:37:10.990 --> 00:37:13.390 align:middle line:84% 5% of congenital adrenal hyperplasia 00:37:13.390 --> 00:37:17.110 align:middle line:84% is 11 hydroxylase deficiency and not 21 hydroxylase deficiency. 00:37:17.110 --> 00:37:21.873 align:middle line:84% And the bars here just show the block 00:37:21.873 --> 00:37:23.290 align:middle line:84% in the pathway that will be caused 00:37:23.290 --> 00:37:24.970 align:middle line:90% by the deficiency of that gene. 00:37:24.970 --> 00:37:26.830 align:middle line:84% And so the point is, you still need 00:37:26.830 --> 00:37:30.110 align:middle line:84% a lot of knowledge around that specific disease 00:37:30.110 --> 00:37:32.500 align:middle line:84% and the pathways rather than just going up 00:37:32.500 --> 00:37:36.550 align:middle line:84% to a database that says obesity-- 00:37:36.550 --> 00:37:37.558 align:middle line:90% this gene. 00:37:37.558 --> 00:37:39.100 align:middle line:84% You still have to have some knowledge 00:37:39.100 --> 00:37:41.720 align:middle line:90% about what you're looking for. 00:37:41.720 --> 00:37:43.480 align:middle line:84% So what material do we want to test? 00:37:43.480 --> 00:37:47.260 align:middle line:84% Let's take four diseases-- cystic fibrosis, 00:37:47.260 --> 00:37:51.730 align:middle line:84% a disease where you get plugging of mucosal ducts. 00:37:51.730 --> 00:37:54.520 align:middle line:84% And it's a chloride transporter problem. 00:37:54.520 --> 00:37:57.610 align:middle line:90% And people die young. 00:37:57.610 --> 00:38:03.940 align:middle line:84% McCune Albright disease-- a disease of the G protein 00:38:03.940 --> 00:38:08.230 align:middle line:84% complex where, essentially, you have an activating mutation 00:38:08.230 --> 00:38:13.000 align:middle line:84% in one of the subunits so that all the G-coupled processes are 00:38:13.000 --> 00:38:16.960 align:middle line:84% hyperactive, so that you have kids with precocious puberty, 00:38:16.960 --> 00:38:19.930 align:middle line:84% cortical adenomas of the adrenal, 00:38:19.930 --> 00:38:22.090 align:middle line:90% and precocious puberty. 00:38:22.090 --> 00:38:28.000 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:38:28.000 --> 00:38:29.750 align:middle line:84% ISAAC SAMUEL KOHANE: Well, McCune Albright 00:38:29.750 --> 00:38:32.270 align:middle line:84% syndrome, the classical description 00:38:32.270 --> 00:38:39.138 align:middle line:84% is cafe au lait spots, which are these-- 00:38:39.138 --> 00:38:41.555 align:middle line:84% basically, looks like someone spilled coffee on your skin. 00:38:41.555 --> 00:38:45.380 align:middle line:90% 00:38:45.380 --> 00:38:47.390 align:middle line:84% Fibrous dysplasia of the bone-- so your bones 00:38:47.390 --> 00:38:48.890 align:middle line:84% are screwed around because they have 00:38:48.890 --> 00:38:52.340 align:middle line:84% these sort of ropey things going through them, 00:38:52.340 --> 00:38:53.540 align:middle line:90% and precocious puberty. 00:38:53.540 --> 00:38:56.450 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:38:56.450 --> 00:39:00.270 align:middle line:84% ISAAC SAMUEL KOHANE: It's distantly. 00:39:00.270 --> 00:39:02.490 align:middle line:84% As we learn more about it, multiple systems, 00:39:02.490 --> 00:39:06.150 align:middle line:84% like thyroid and growth hormone, can also be hyperstimulated. 00:39:06.150 --> 00:39:12.130 align:middle line:84% So, basically, it's a mosaic state 00:39:12.130 --> 00:39:15.240 align:middle line:84% where these mutations can actually be spread incompletely 00:39:15.240 --> 00:39:16.270 align:middle line:90% throughout the body. 00:39:16.270 --> 00:39:18.720 align:middle line:84% And so, for instance, the melanocytes 00:39:18.720 --> 00:39:20.460 align:middle line:90% are stimulated in just patches. 00:39:20.460 --> 00:39:23.460 align:middle line:84% And if you're unlucky enough to have adrenal involved-- 00:39:23.460 --> 00:39:26.610 align:middle line:90% adrenal adenoma. 00:39:26.610 --> 00:39:28.770 align:middle line:90% Please. 00:39:28.770 --> 00:39:30.753 align:middle line:84% AUDIENCE: Can I get a scenario for-- 00:39:30.753 --> 00:39:32.670 align:middle line:84% ISAAC SAMUEL KOHANE: So a patient comes to you 00:39:32.670 --> 00:39:34.380 align:middle line:90% with McCune Albright syndrome. 00:39:34.380 --> 00:39:36.233 align:middle line:84% And you suspect it because Zach told me 00:39:36.233 --> 00:39:37.900 align:middle line:84% that they have these cafe au lait spots. 00:39:37.900 --> 00:39:41.010 align:middle line:84% Boy, this person has a two big ones. 00:39:41.010 --> 00:39:43.680 align:middle line:84% And this person is having a precocious puberty 00:39:43.680 --> 00:39:45.930 align:middle line:90% and is complaining of bone pain. 00:39:45.930 --> 00:39:47.310 align:middle line:90% What do I send to test? 00:39:47.310 --> 00:39:50.370 align:middle line:90% 00:39:50.370 --> 00:39:52.920 align:middle line:90% That's the question. 00:39:52.920 --> 00:39:56.070 align:middle line:84% AUDIENCE: But whereas with the question of cystic fibrosis, 00:39:56.070 --> 00:39:59.650 align:middle line:90% they would know that [INAUDIBLE] 00:39:59.650 --> 00:40:01.560 align:middle line:84% ISAAC SAMUEL KOHANE: Well, you suspect it. 00:40:01.560 --> 00:40:09.270 align:middle line:84% You did a sweat test, which has a certain specificity 00:40:09.270 --> 00:40:12.120 align:middle line:90% and sensitivity. 00:40:12.120 --> 00:40:14.580 align:middle line:84% or maybe it's a parent who already had a kid who 00:40:14.580 --> 00:40:16.410 align:middle line:90% died from cystic fibrosis. 00:40:16.410 --> 00:40:19.230 align:middle line:90% And they just have a newborn. 00:40:19.230 --> 00:40:20.548 align:middle line:90% What are you going to test? 00:40:20.548 --> 00:40:21.840 align:middle line:90% You don't know the N thing yet. 00:40:21.840 --> 00:40:24.852 align:middle line:90% 00:40:24.852 --> 00:40:26.060 align:middle line:90% So let's start one at a time. 00:40:26.060 --> 00:40:27.110 align:middle line:90% Cystic fibrosis. 00:40:27.110 --> 00:40:28.040 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:40:28.040 --> 00:40:29.540 align:middle line:84% ISAAC SAMUEL KOHANE: A testing gene. 00:40:29.540 --> 00:40:31.850 align:middle line:90% But I'm a I'm a stupid doctor-- 00:40:31.850 --> 00:40:32.887 align:middle line:90% indeed I am. 00:40:32.887 --> 00:40:34.220 align:middle line:90% What does it mean to test genes? 00:40:34.220 --> 00:40:36.860 align:middle line:84% What thing do I stick into a tube? 00:40:36.860 --> 00:40:39.212 align:middle line:84% AUDIENCE: Oh, just do a mucosal swab. 00:40:39.212 --> 00:40:40.670 align:middle line:84% ISAAC SAMUEL KOHANE: So [INAUDIBLE] 00:40:40.670 --> 00:40:42.413 align:middle line:90% I can do a mucosal swab? 00:40:42.413 --> 00:40:46.040 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:40:46.040 --> 00:40:48.690 align:middle line:84% ISAAC SAMUEL KOHANE: The short answer is yes. 00:40:48.690 --> 00:40:50.210 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:40:50.210 --> 00:40:50.495 align:middle line:90% ISAAC SAMUEL KOHANE: What? 00:40:50.495 --> 00:40:51.370 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:40:51.370 --> 00:40:52.993 align:middle line:90% ISAAC SAMUEL KOHANE: No. 00:40:52.993 --> 00:40:53.660 align:middle line:90% McCune Albright. 00:40:53.660 --> 00:40:59.560 align:middle line:90% 00:40:59.560 --> 00:41:01.240 align:middle line:84% Maybe the cells here are not involved. 00:41:01.240 --> 00:41:02.050 align:middle line:90% Maybe it's not-- 00:41:02.050 --> 00:41:04.930 align:middle line:84% AUDIENCE: You look for the [INAUDIBLE].. 00:41:04.930 --> 00:41:06.970 align:middle line:84% ISAAC SAMUEL KOHANE: Maybe you get a skin biopsy 00:41:06.970 --> 00:41:08.860 align:middle line:90% of the cafe-au-lait spot. 00:41:08.860 --> 00:41:10.990 align:middle line:84% 21 hydroxylase is efficient germline. 00:41:10.990 --> 00:41:12.730 align:middle line:90% You can do the blood. 00:41:12.730 --> 00:41:14.680 align:middle line:84% Clots like cancer, blood is not biopsy. 00:41:14.680 --> 00:41:16.510 align:middle line:90% You need a hunk of that cancer. 00:41:16.510 --> 00:41:20.320 align:middle line:84% Or depends, is a highly inherited cancer 00:41:20.320 --> 00:41:22.930 align:middle line:84% or do you believe it's a somatic mutation in the cancer? 00:41:22.930 --> 00:41:26.950 align:middle line:84% All the point I'm making here is it's not obvious 00:41:26.950 --> 00:41:29.620 align:middle line:90% what you put into the tube. 00:41:29.620 --> 00:41:31.660 align:middle line:84% And as dumb as it sounds, when you're 00:41:31.660 --> 00:41:34.092 align:middle line:84% a tired resident on the ward and someone says, 00:41:34.092 --> 00:41:35.050 align:middle line:90% do the geniculate test. 00:41:35.050 --> 00:41:37.720 align:middle line:84% You don't know what the hell to do. 00:41:37.720 --> 00:41:39.790 align:middle line:90% It's that simple. 00:41:39.790 --> 00:41:42.040 align:middle line:84% But this is what it comes down to in genomic medicine. 00:41:42.040 --> 00:41:43.960 align:middle line:84% These are actually answerable questions, 00:41:43.960 --> 00:41:45.130 align:middle line:90% but you need some reading. 00:41:45.130 --> 00:41:48.070 align:middle line:90% Where are the mutations? 00:41:48.070 --> 00:41:50.660 align:middle line:84% And what do they mean when they are in certain tissues? 00:41:50.660 --> 00:41:59.940 align:middle line:84% AUDIENCE: So diseases [INAUDIBLE] 00:41:59.940 --> 00:42:02.700 align:middle line:84% ISAAC SAMUEL KOHANE: For a germline disease-- 00:42:02.700 --> 00:42:03.390 align:middle line:90% any tissue. 00:42:03.390 --> 00:42:04.140 align:middle line:90% It doesn't matter. 00:42:04.140 --> 00:42:07.140 align:middle line:84% So you want one that is the least invasive-- blood, swab, 00:42:07.140 --> 00:42:08.633 align:middle line:90% even spit sometimes. 00:42:08.633 --> 00:42:09.300 align:middle line:90% AUDIENCE: Right. 00:42:09.300 --> 00:42:11.280 align:middle line:84% For some other complexes, [INAUDIBLE] people 00:42:11.280 --> 00:42:14.610 align:middle line:84% suspect you have susceptibility, which 00:42:14.610 --> 00:42:17.152 align:middle line:90% then triggers the disease. 00:42:17.152 --> 00:42:19.110 align:middle line:84% ISAAC SAMUEL KOHANE: It's going to be certainly 00:42:19.110 --> 00:42:20.800 align:middle line:90% not true for many diseases. 00:42:20.800 --> 00:42:23.550 align:middle line:84% But, for instance, when Scott Weiss talks about asthma, 00:42:23.550 --> 00:42:25.950 align:middle line:84% he's thinking of a risk factor, even if it's a small risk 00:42:25.950 --> 00:42:29.070 align:middle line:84% factor and it'll increase your risk of asthma by 1.2, 00:42:29.070 --> 00:42:33.780 align:middle line:84% he's still thinking of this as a germline polymorphism. 00:42:33.780 --> 00:42:35.742 align:middle line:84% Now it may be that people that asthma-- 00:42:35.742 --> 00:42:37.200 align:middle line:84% I'm just making up something that's 00:42:37.200 --> 00:42:39.003 align:middle line:84% completely off the beaten path-- is, 00:42:39.003 --> 00:42:40.170 align:middle line:90% in fact, a somatic mutation. 00:42:40.170 --> 00:42:41.310 align:middle line:90% I doubt it. 00:42:41.310 --> 00:42:44.970 align:middle line:84% But all these genotyping studies, haplotype studies 00:42:44.970 --> 00:42:45.900 align:middle line:90% are done off of blood. 00:42:45.900 --> 00:42:49.000 align:middle line:84% So they're assuming these are germline diseases. 00:42:49.000 --> 00:42:51.180 align:middle line:84% AUDIENCE: So for something that's not germline, 00:42:51.180 --> 00:42:53.910 align:middle line:84% something like maybe you've got a disease because you're 00:42:53.910 --> 00:42:56.500 align:middle line:84% exposed to some sort of environmental toxin 00:42:56.500 --> 00:42:57.240 align:middle line:90% or something-- 00:42:57.240 --> 00:42:58.300 align:middle line:84% ISAAC SAMUEL KOHANE: Or you have a somatic mutation. 00:42:58.300 --> 00:42:59.130 align:middle line:90% Yeah. 00:42:59.130 --> 00:42:59.838 align:middle line:90% AUDIENCE: --yeah. 00:42:59.838 --> 00:43:02.190 align:middle line:90% 00:43:02.190 --> 00:43:05.850 align:middle line:84% Would it be a good standard protocol then to take-- 00:43:05.850 --> 00:43:10.770 align:middle line:84% let's say for cancer, part of the tumor and part of just 00:43:10.770 --> 00:43:12.300 align:middle line:90% any other tissue to compare-- 00:43:12.300 --> 00:43:12.960 align:middle line:90% ISAAC SAMUEL KOHANE: Or blood. 00:43:12.960 --> 00:43:13.540 align:middle line:90% Sure. 00:43:13.540 --> 00:43:16.290 align:middle line:84% AUDIENCE: --or something like that just be able to-- 00:43:16.290 --> 00:43:18.207 align:middle line:84% ISAAC SAMUEL KOHANE: The short answer is cost. 00:43:18.207 --> 00:43:20.773 align:middle line:84% But, yes, I mean, the answer is-- 00:43:20.773 --> 00:43:23.190 align:middle line:84% AUDIENCE: --wasn't it Weiss who said that in the next five 00:43:23.190 --> 00:43:25.960 align:middle line:84% years this cost thing is not going to be an issue-- 00:43:25.960 --> 00:43:26.460 align:middle line:84% ISAAC SAMUEL KOHANE: Well, it's not 00:43:26.460 --> 00:43:28.293 align:middle line:84% going to be an issue for him doing research. 00:43:28.293 --> 00:43:31.830 align:middle line:90% 00:43:31.830 --> 00:43:38.790 align:middle line:84% What I just told you is $0.10 a genotype today for Scott Weiss. 00:43:38.790 --> 00:43:42.000 align:middle line:84% For you, my friend, $1,000 in the clinic. 00:43:42.000 --> 00:43:44.520 align:middle line:84% And by the way, that's a huge commercial opportunity 00:43:44.520 --> 00:43:45.733 align:middle line:90% obviously. 00:43:45.733 --> 00:43:46.650 align:middle line:90% You're absolute right. 00:43:46.650 --> 00:43:47.190 align:middle line:90% That's going to drop. 00:43:47.190 --> 00:43:48.480 align:middle line:90% It's like a drop to $0.10. 00:43:48.480 --> 00:43:50.880 align:middle line:84% It's going to drop to $10 or $100. 00:43:50.880 --> 00:43:52.710 align:middle line:84% And the market leaders in that, who 00:43:52.710 --> 00:43:56.100 align:middle line:84% can figure out how to make the bucks on this, at that level, 00:43:56.100 --> 00:43:57.915 align:middle line:90% are going to be extremely rich. 00:43:57.915 --> 00:44:00.600 align:middle line:90% 00:44:00.600 --> 00:44:02.010 align:middle line:84% OK, so if you ordered a test, you 00:44:02.010 --> 00:44:04.320 align:middle line:84% going to need to get advice on that genetic test. 00:44:04.320 --> 00:44:06.780 align:middle line:90% Where are you going to get that? 00:44:06.780 --> 00:44:08.400 align:middle line:90% Found either in your office-- 00:44:08.400 --> 00:44:10.655 align:middle line:84% let me call you off the bat-- extremely unlikely. 00:44:10.655 --> 00:44:12.030 align:middle line:84% There's no one in my office who's 00:44:12.030 --> 00:44:14.040 align:middle line:84% going to give you genetic counseling 00:44:14.040 --> 00:44:16.230 align:middle line:84% or other licensed counselors-- maybe 00:44:16.230 --> 00:44:17.730 align:middle line:90% it's part of your institution. 00:44:17.730 --> 00:44:18.730 align:middle line:90% I'm talking about today. 00:44:18.730 --> 00:44:20.940 align:middle line:84% We're not talking about the medical genomics. 00:44:20.940 --> 00:44:22.530 align:middle line:90% Today, what are you going to do? 00:44:22.530 --> 00:44:26.310 align:middle line:84% Or licensed counselors at Brigham, or you outsource it-- 00:44:26.310 --> 00:44:29.970 align:middle line:84% and here, it says, there's a pre-test phone consultation 00:44:29.970 --> 00:44:32.355 align:middle line:90% provided patients and so on. 00:44:32.355 --> 00:44:34.230 align:middle line:84% And that might be part of the business model, 00:44:34.230 --> 00:44:36.300 align:middle line:84% that they understand that a lot of institutions 00:44:36.300 --> 00:44:38.680 align:middle line:90% don't have that facility. 00:44:38.680 --> 00:44:40.860 align:middle line:90% So they can outsource it. 00:44:40.860 --> 00:44:42.990 align:middle line:84% So how do you send a store for something. 00:44:42.990 --> 00:44:46.260 align:middle line:84% Again, these are stupid things that most residents don't know. 00:44:46.260 --> 00:44:48.120 align:middle line:84% 310 milliliters of whole blood, if it's 00:44:48.120 --> 00:44:51.780 align:middle line:84% a germline in a purple top tube, also acceptable green, 00:44:51.780 --> 00:44:53.760 align:middle line:84% which is sodium heparin, or light blue-- 00:44:53.760 --> 00:44:56.460 align:middle line:90% sodium citrate tubes. 00:44:56.460 --> 00:44:58.680 align:middle line:90% It comes down to that. 00:44:58.680 --> 00:45:00.090 align:middle line:90% What tube you put it in? 00:45:00.090 --> 00:45:04.170 align:middle line:84% If you put it in the wrong tube, you may not 00:45:04.170 --> 00:45:06.922 align:middle line:90% have an adequate extraction. 00:45:06.922 --> 00:45:08.880 align:middle line:84% And it's totally different, of course, for RNA. 00:45:08.880 --> 00:45:12.390 align:middle line:84% If you're interested in RNA, you better 00:45:12.390 --> 00:45:15.120 align:middle line:84% flash freeze it, is a short answer, as soon as you can. 00:45:15.120 --> 00:45:16.155 align:middle line:90% So what kind of DNA-- 00:45:16.155 --> 00:45:18.013 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:45:18.013 --> 00:45:19.680 align:middle line:84% ISAAC SAMUEL KOHANE: Because, basically, 00:45:19.680 --> 00:45:23.290 align:middle line:84% the amount the stability and extractability 00:45:23.290 --> 00:45:27.210 align:middle line:84% is going to depend on what is inside the tube. 00:45:27.210 --> 00:45:31.200 align:middle line:84% And the color of the tube is a code of what's inside the tube. 00:45:31.200 --> 00:45:33.930 align:middle line:84% So these tubes are not all featureless glass. 00:45:33.930 --> 00:45:36.880 align:middle line:84% They have a little bit of chemical in them. 00:45:36.880 --> 00:45:40.030 align:middle line:84% So what kind of DNA testing should be done? 00:45:40.030 --> 00:45:41.970 align:middle line:90% So let me ask you guys. 00:45:41.970 --> 00:45:44.940 align:middle line:84% We've just got together in a happy company 00:45:44.940 --> 00:45:49.050 align:middle line:90% called 512 Genomic Testing. 00:45:49.050 --> 00:45:51.180 align:middle line:90% We've got our stock options. 00:45:51.180 --> 00:45:55.008 align:middle line:90% And what are we going to do? 00:45:55.008 --> 00:45:56.550 align:middle line:84% Are we going to sequence these genes? 00:45:56.550 --> 00:45:59.910 align:middle line:84% When someone sends us a gene to 512 Genomics, 00:45:59.910 --> 00:46:02.940 align:middle line:84% are we going to sequence it or are we 00:46:02.940 --> 00:46:05.460 align:middle line:84% going to genotype it for all the known mutations? 00:46:05.460 --> 00:46:07.545 align:middle line:90% What are we going to do? 00:46:07.545 --> 00:46:13.140 align:middle line:84% AUDIENCE: [INAUDIBLE] mutation different levels [INAUDIBLE] 00:46:13.140 --> 00:46:14.113 align:middle line:90% and more complicated. 00:46:14.113 --> 00:46:15.780 align:middle line:84% ISAAC SAMUEL KOHANE: So let me push you. 00:46:15.780 --> 00:46:21.150 align:middle line:84% So just arrived in our inbox is a blood sample 00:46:21.150 --> 00:46:26.680 align:middle line:84% in the right tube from Joe Schmo, 00:46:26.680 --> 00:46:28.128 align:middle line:84% but it's really her baby's blood. 00:46:28.128 --> 00:46:29.670 align:middle line:84% And she wants to know, does this baby 00:46:29.670 --> 00:46:32.640 align:middle line:90% have 21 hydroxylase deficiency? 00:46:32.640 --> 00:46:36.018 align:middle line:84% And they heard that 512 Genomics is a very smart company 00:46:36.018 --> 00:46:38.310 align:middle line:84% because they've been trained at Harvard Medical School. 00:46:38.310 --> 00:46:40.900 align:middle line:84% And, therefore, they trust us to do it. 00:46:40.900 --> 00:46:44.860 align:middle line:84% So you're going to do a panel of what? 00:46:44.860 --> 00:46:45.360 align:middle line:90% Genes? 00:46:45.360 --> 00:46:49.324 align:middle line:90% 00:46:49.324 --> 00:46:50.940 align:middle line:90% Let's simplify your life. 00:46:50.940 --> 00:46:55.470 align:middle line:84% Well, first of all, are we going to test all genes? 00:46:55.470 --> 00:46:59.775 align:middle line:84% AUDIENCE: No, we'll start testing the genes that are 00:46:59.775 --> 00:47:00.990 align:middle line:90% known and that are frequent. 00:47:00.990 --> 00:47:01.710 align:middle line:84% ISAAC SAMUEL KOHANE: OK, I want us 00:47:01.710 --> 00:47:04.110 align:middle line:84% to have a very nice corporate vacation in Bermuda. 00:47:04.110 --> 00:47:06.480 align:middle line:84% So all the money that we spend is 00:47:06.480 --> 00:47:11.580 align:middle line:84% going to take away from the bennies I can give you 00:47:11.580 --> 00:47:14.042 align:middle line:90% on our vacation. 00:47:14.042 --> 00:47:15.000 align:middle line:90% OK, we're going to do-- 00:47:15.000 --> 00:47:17.250 align:middle line:84% I'm going to tell you this-- five genes that are 00:47:17.250 --> 00:47:17.750 align:middle line:90% [INAUDIBLE]. 00:47:17.750 --> 00:47:19.480 align:middle line:90% We're doing all five genes? 00:47:19.480 --> 00:47:24.880 align:middle line:84% AUDIENCE: No, then you look at the frequency 00:47:24.880 --> 00:47:26.338 align:middle line:84% of a particular mutation occurring. 00:47:26.338 --> 00:47:28.380 align:middle line:84% ISAAC SAMUEL KOHANE: OK, so I'm going to tell you 00:47:28.380 --> 00:47:30.610 align:middle line:84% there's two genes that account for 99%. 00:47:30.610 --> 00:47:33.380 align:middle line:90% Good enough for us? 00:47:33.380 --> 00:47:34.780 align:middle line:90% OK, let's say yes. 00:47:34.780 --> 00:47:35.890 align:middle line:90% Let's say yes. 00:47:35.890 --> 00:47:38.680 align:middle line:84% AUDIENCE: Why can't we just do one gene, if it comes 00:47:38.680 --> 00:47:42.690 align:middle line:84% back negative-- if we know two compromise 99%, [INAUDIBLE] 00:47:42.690 --> 00:47:44.440 align:middle line:84% do the one, comes back negative, we do it. 00:47:44.440 --> 00:47:47.185 align:middle line:84% But if it comes back positive, it's like, hey, we got this. 00:47:47.185 --> 00:47:48.102 align:middle line:90% [INAUDIBLE] 00:47:48.102 --> 00:47:50.560 align:middle line:84% ISAAC SAMUEL KOHANE: We're going to go back to the patient? 00:47:50.560 --> 00:47:51.460 align:middle line:90% That doesn't happen. 00:47:51.460 --> 00:47:53.740 align:middle line:90% We're a lab. 00:47:53.740 --> 00:47:55.000 align:middle line:90% We're not the doctor. 00:47:55.000 --> 00:47:56.710 align:middle line:90% That does not happen. 00:47:56.710 --> 00:47:59.350 align:middle line:84% We cannot ask, unless you really want to change-- 00:47:59.350 --> 00:48:01.145 align:middle line:90% you live in the dream world. 00:48:01.145 --> 00:48:02.770 align:middle line:84% AUDIENCE: No, I'm just saying-- no, you 00:48:02.770 --> 00:48:04.812 align:middle line:84% can't go back to the patients and get more blood. 00:48:04.812 --> 00:48:06.277 align:middle line:90% But-- 00:48:06.277 --> 00:48:08.860 align:middle line:84% ISAAC SAMUEL KOHANE: Just first sequence the more common gene. 00:48:08.860 --> 00:48:09.735 align:middle line:90% Is that we're saying? 00:48:09.735 --> 00:48:11.840 align:middle line:84% AUDIENCE: Yeah, so if you got two genes, I guess-- 00:48:11.840 --> 00:48:12.910 align:middle line:84% ISAAC SAMUEL KOHANE: OK, so let me tell you. 00:48:12.910 --> 00:48:13.743 align:middle line:90% Listen to the facts. 00:48:13.743 --> 00:48:19.010 align:middle line:84% 21 hydroxylase accounts for 95%, 11 hydroxylase 4%. 00:48:19.010 --> 00:48:20.480 align:middle line:84% So we got 99% with those two genes. 00:48:20.480 --> 00:48:21.490 align:middle line:90% So what am I actually doing? 00:48:21.490 --> 00:48:22.907 align:middle line:84% I'm not the laboratory technician? 00:48:22.907 --> 00:48:24.550 align:middle line:84% AUDIENCE: --yeah, do the 95% one. 00:48:24.550 --> 00:48:26.660 align:middle line:84% ISAAC SAMUEL KOHANE: OK, and it's negative. 00:48:26.660 --> 00:48:27.946 align:middle line:90% What do I do now? 00:48:27.946 --> 00:48:29.404 align:middle line:84% AUDIENCE: Then you do the next one. 00:48:29.404 --> 00:48:32.256 align:middle line:90% It comes back positive-- 00:48:32.256 --> 00:48:34.050 align:middle line:84% I don't know enough about the disease. 00:48:34.050 --> 00:48:36.062 align:middle line:90% So assuming that the positive-- 00:48:36.062 --> 00:48:38.020 align:middle line:84% ISAAC SAMUEL KOHANE: The positive is important, 00:48:38.020 --> 00:48:39.520 align:middle line:90% positive is very important. 00:48:39.520 --> 00:48:43.702 align:middle line:84% AUDIENCE: --then you just running that actual test. 00:48:43.702 --> 00:48:45.910 align:middle line:84% ISAAC SAMUEL KOHANE: OK, so you're doing a phase one. 00:48:45.910 --> 00:48:47.077 align:middle line:90% That's perfectly acceptable. 00:48:47.077 --> 00:48:49.240 align:middle line:84% It's probably means that, most of the time, 00:48:49.240 --> 00:48:50.770 align:middle line:90% we'll run both tests. 00:48:50.770 --> 00:48:52.900 align:middle line:84% Because most of the time, we'll probably 00:48:52.900 --> 00:48:55.670 align:middle line:84% get referrals of patients who don't have that disease. 00:48:55.670 --> 00:48:58.150 align:middle line:84% So it's a small but important incremental savings 00:48:58.150 --> 00:48:59.050 align:middle line:90% you just gave us. 00:48:59.050 --> 00:49:03.278 align:middle line:84% You took Jose's advice and said let's only do one gene. 00:49:03.278 --> 00:49:05.320 align:middle line:84% And if that's negative, we'll do the second gene. 00:49:05.320 --> 00:49:09.430 align:middle line:84% But as I explained, most kids this was sent for 00:49:09.430 --> 00:49:12.340 align:middle line:84% will, in fact, not have the disease 00:49:12.340 --> 00:49:14.060 align:middle line:90% because it'll be something else. 00:49:14.060 --> 00:49:19.600 align:middle line:84% And we'll still have to do the second one. 00:49:19.600 --> 00:49:20.440 align:middle line:90% You had a question? 00:49:20.440 --> 00:49:22.607 align:middle line:84% AUDIENCE: The question is this is a PCR test, right? 00:49:22.607 --> 00:49:23.540 align:middle line:90% Just to clarify. 00:49:23.540 --> 00:49:24.670 align:middle line:84% ISAAC SAMUEL KOHANE: Well, we haven't gotten there yet. 00:49:24.670 --> 00:49:25.212 align:middle line:90% AUDIENCE: OK. 00:49:25.212 --> 00:49:26.060 align:middle line:90% [INAUDIBLE] 00:49:26.060 --> 00:49:27.610 align:middle line:84% ISAAC SAMUEL KOHANE: OK, so let's 00:49:27.610 --> 00:49:29.860 align:middle line:84% say that we take the modification of the Jose 00:49:29.860 --> 00:49:31.330 align:middle line:90% protocol. 00:49:31.330 --> 00:49:32.960 align:middle line:90% So what are we going to do? 00:49:32.960 --> 00:49:35.502 align:middle line:84% So we look at one gene and then two genes. 00:49:35.502 --> 00:49:37.210 align:middle line:84% So what actually are we going to measure? 00:49:37.210 --> 00:49:41.770 align:middle line:84% Are we going to measure the full sequence of these genes? 00:49:41.770 --> 00:49:47.830 align:middle line:84% Are we going to just look at the known published mutations 00:49:47.830 --> 00:49:48.400 align:middle line:90% or SNPs? 00:49:48.400 --> 00:49:49.400 align:middle line:90% What are we going to do? 00:49:49.400 --> 00:49:53.290 align:middle line:90% 00:49:53.290 --> 00:49:58.760 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:49:58.760 --> 00:50:01.190 align:middle line:84% ISAAC SAMUEL KOHANE: Now what if I tell you 00:50:01.190 --> 00:50:06.590 align:middle line:84% that 2/3 of the mutations are extremely common 00:50:06.590 --> 00:50:10.953 align:middle line:84% and 1/3 of them are one of a kind. 00:50:10.953 --> 00:50:12.620 align:middle line:84% AUDIENCE: I'm assuming we know something 00:50:12.620 --> 00:50:18.500 align:middle line:84% about the frequency of having the disease associated 00:50:18.500 --> 00:50:22.340 align:middle line:84% with the frequency of having that particular mutation. 00:50:22.340 --> 00:50:30.731 align:middle line:84% So if we know that a particular mutation affects the one person 00:50:30.731 --> 00:50:34.820 align:middle line:84% and has a very high incidence of progressive disease, 00:50:34.820 --> 00:50:38.080 align:middle line:90% then that's the [INAUDIBLE] 00:50:38.080 --> 00:50:39.080 align:middle line:90% ISAAC SAMUEL KOHANE: OK. 00:50:39.080 --> 00:50:41.390 align:middle line:90% I'm Joe Schmo. 00:50:41.390 --> 00:50:46.490 align:middle line:84% My kid actually had one of those rare ones. 00:50:46.490 --> 00:50:49.880 align:middle line:84% I'm going to assume your pants off. 00:50:49.880 --> 00:50:53.960 align:middle line:84% I'm taking the vacation to Bermuda because you guys took-- 00:50:53.960 --> 00:50:57.020 align:middle line:90% there was a known rare mutation. 00:50:57.020 --> 00:50:58.550 align:middle line:90% Or even-- 00:50:58.550 --> 00:51:02.090 align:middle line:84% AUDIENCE: Right, again, this is back 00:51:02.090 --> 00:51:04.460 align:middle line:84% to what I said in the Jose modification. 00:51:04.460 --> 00:51:08.270 align:middle line:90% If you come back negative-- 00:51:08.270 --> 00:51:09.410 align:middle line:90% ISAAC SAMUEL KOHANE: I see. 00:51:09.410 --> 00:51:10.862 align:middle line:90% AUDIENCE: --then you go back-- 00:51:10.862 --> 00:51:12.320 align:middle line:84% ISAAC SAMUEL KOHANE: All I'm saying 00:51:12.320 --> 00:51:13.790 align:middle line:84% is that refinement is still going 00:51:13.790 --> 00:51:15.890 align:middle line:84% to cause-- most of the time, you're 00:51:15.890 --> 00:51:19.310 align:middle line:84% going to end up doing all the mutations. 00:51:19.310 --> 00:51:22.370 align:middle line:84% AUDIENCE: But the alternative is to just do all the mutations 00:51:22.370 --> 00:51:22.955 align:middle line:90% all the time. 00:51:22.955 --> 00:51:24.688 align:middle line:90% 00:51:24.688 --> 00:51:26.480 align:middle line:84% ISAAC SAMUEL KOHANE: What I'm telling you-- 00:51:26.480 --> 00:51:28.097 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:51:28.097 --> 00:51:29.180 align:middle line:90% ISAAC SAMUEL KOHANE: What? 00:51:29.180 --> 00:51:31.055 align:middle line:84% AUDIENCE: A third of them are kind of unique. 00:51:31.055 --> 00:51:34.285 align:middle line:84% ISAAC SAMUEL KOHANE: Unique, are private mutations. 00:51:34.285 --> 00:51:35.660 align:middle line:84% So the short answer is, I believe 00:51:35.660 --> 00:51:39.170 align:middle line:84% the right answer is, you've got to sequence the whole thing. 00:51:39.170 --> 00:51:39.830 align:middle line:90% I mean-- 00:51:39.830 --> 00:51:42.230 align:middle line:84% AUDIENCE: There's no value in preemptive screening, 00:51:42.230 --> 00:51:44.647 align:middle line:90% say, for instance-- 00:51:44.647 --> 00:51:45.980 align:middle line:90% ISAAC SAMUEL KOHANE: It depends. 00:51:45.980 --> 00:51:51.505 align:middle line:84% If it's for a given patient around the specific problem, 00:51:51.505 --> 00:51:52.880 align:middle line:84% that's a very different situation 00:51:52.880 --> 00:51:54.410 align:middle line:84% from I'm a public health authority 00:51:54.410 --> 00:51:56.683 align:middle line:84% and I want to screen the population 00:51:56.683 --> 00:51:58.100 align:middle line:84% or I want to screen the population 00:51:58.100 --> 00:52:00.480 align:middle line:84% and just detect people at risk for this. 00:52:00.480 --> 00:52:04.680 align:middle line:84% But if you're a doctor with a specific question, 00:52:04.680 --> 00:52:09.830 align:middle line:84% I'm duty-bound to actually do maximum for you. 00:52:09.830 --> 00:52:15.600 align:middle line:90% And consequently-- period. 00:52:15.600 --> 00:52:16.980 align:middle line:90% AUDIENCE: And it just seems-- 00:52:16.980 --> 00:52:20.570 align:middle line:84% maybe I misunderstood this-- but it seems kind of counter 00:52:20.570 --> 00:52:24.200 align:middle line:84% to what the current practices were, again, standard testing 00:52:24.200 --> 00:52:25.710 align:middle line:90% for things. 00:52:25.710 --> 00:52:29.417 align:middle line:84% when you have thrombosis, the first thing they do 00:52:29.417 --> 00:52:31.250 align:middle line:84% is they test me for the most common things-- 00:52:31.250 --> 00:52:34.430 align:middle line:84% [INAUDIBLE] deficiency, and all these different things. 00:52:34.430 --> 00:52:39.050 align:middle line:84% Because, well, if you've got that, then [INAUDIBLE] 00:52:39.050 --> 00:52:41.390 align:middle line:84% then we don't really need to continue. 00:52:41.390 --> 00:52:43.350 align:middle line:84% Because we already know you had a big problem. 00:52:43.350 --> 00:52:47.420 align:middle line:84% So that's, I guess, the big basis for my approach. 00:52:47.420 --> 00:52:51.560 align:middle line:90% 00:52:51.560 --> 00:52:54.230 align:middle line:84% Joe Schmo came and said, one of my kids has this thing. 00:52:54.230 --> 00:52:56.270 align:middle line:84% The fastest, cheapest thing to do 00:52:56.270 --> 00:52:58.062 align:middle line:84% would be to test them for the most common-- 00:52:58.062 --> 00:52:59.520 align:middle line:84% ISAAC SAMUEL KOHANE: So you see you 00:52:59.520 --> 00:53:02.408 align:middle line:84% can do that because the doctor can actually cut a protocol 00:53:02.408 --> 00:53:04.700 align:middle line:84% and actually talk this through with the patient-- we'll 00:53:04.700 --> 00:53:05.908 align:middle line:90% go over the most common ones. 00:53:05.908 --> 00:53:09.305 align:middle line:90% We're 512 Genomics. 00:53:09.305 --> 00:53:11.930 align:middle line:84% And 512 Genomics does not have a relationship with the patient. 00:53:11.930 --> 00:53:15.188 align:middle line:90% 00:53:15.188 --> 00:53:16.980 align:middle line:84% And the doctor doesn't understand genetics, 00:53:16.980 --> 00:53:19.200 align:middle line:90% so we already established. 00:53:19.200 --> 00:53:21.020 align:middle line:90% So again, I'm trying-- 00:53:21.020 --> 00:53:25.008 align:middle line:84% 2004, if there is a private mutation 00:53:25.008 --> 00:53:27.050 align:middle line:84% that the patient subsequently finds, we're toast. 00:53:27.050 --> 00:53:29.950 align:middle line:90% 00:53:29.950 --> 00:53:31.783 align:middle line:84% AUDIENCE: Even if we don't [INAUDIBLE] back. 00:53:31.783 --> 00:53:33.867 align:middle line:84% ISAAC SAMUEL KOHANE: When are we going to go back? 00:53:33.867 --> 00:53:35.398 align:middle line:84% No one's going to come back to us. 00:53:35.398 --> 00:53:36.940 align:middle line:84% AUDIENCE: Take the genes that account 00:53:36.940 --> 00:53:38.778 align:middle line:84% for those qualifications and people. 00:53:38.778 --> 00:53:40.570 align:middle line:84% ISAAC SAMUEL KOHANE: Well, basically, we're 00:53:40.570 --> 00:53:43.210 align:middle line:84% going to follow doctor's orders if they 00:53:43.210 --> 00:53:46.480 align:middle line:84% say we want if they say if they say 00:53:46.480 --> 00:53:48.880 align:middle line:84% I want to look at congenital adrenal hyperplasia. 00:53:48.880 --> 00:53:51.820 align:middle line:84% I think we have to look at every gene known-- 00:53:51.820 --> 00:53:53.720 align:middle line:84% we'll build up accordingly, of course. 00:53:53.720 --> 00:53:58.300 align:middle line:84% If they say, I want to look at 21 hydroxylase gene, 00:53:58.300 --> 00:53:59.980 align:middle line:84% we have to sequence the whole darn gene. 00:53:59.980 --> 00:54:01.480 align:middle line:84% Now if they specifically say, I want 00:54:01.480 --> 00:54:04.060 align:middle line:84% to look at just a mutation-- which I'll never tell us. 00:54:04.060 --> 00:54:07.360 align:middle line:90% I want a mutation in codon 32. 00:54:07.360 --> 00:54:09.580 align:middle line:90% Well, we'll do that genotyping. 00:54:09.580 --> 00:54:14.170 align:middle line:84% But if the doctor is a standard doctor, 2004, and says, 00:54:14.170 --> 00:54:16.420 align:middle line:90% check 21 hydroxylase gene. 00:54:16.420 --> 00:54:20.530 align:middle line:84% We've got to sequence the whole thing. 00:54:20.530 --> 00:54:22.840 align:middle line:90% There is no choice. 00:54:22.840 --> 00:54:24.355 align:middle line:90% Otherwise, we're toast. 00:54:24.355 --> 00:54:33.240 align:middle line:90% 00:54:33.240 --> 00:54:35.370 align:middle line:84% Yes, so I think the only thing I missed 00:54:35.370 --> 00:54:39.720 align:middle line:84% from the slide for epigenetic things like methylation, 00:54:39.720 --> 00:54:41.220 align:middle line:84% we need to think about doing that. 00:54:41.220 --> 00:54:44.440 align:middle line:84% But for certain diseases involve imprinting, for instance, 00:54:44.440 --> 00:54:47.790 align:middle line:90% you want to look at methylation. 00:54:47.790 --> 00:54:50.020 align:middle line:90% So pre-flight checklist. 00:54:50.020 --> 00:54:52.520 align:middle line:84% What is this going to take to get our clinic up and running? 00:54:52.520 --> 00:54:54.440 align:middle line:90% Now we're 5212 Clinic. 00:54:54.440 --> 00:54:56.580 align:middle line:90% We're not the company anymore. 00:54:56.580 --> 00:54:58.700 align:middle line:84% So would you or a staff member be able to be-- 00:54:58.700 --> 00:55:01.342 align:middle line:84% so I took this from some very good website. 00:55:01.342 --> 00:55:03.050 align:middle line:84% I wish I could tell you which one it was. 00:55:03.050 --> 00:55:07.070 align:middle line:84% But one of the things that you need to do before going, 00:55:07.070 --> 00:55:09.740 align:middle line:84% would you or a staff member be able to be an advocacy 00:55:09.740 --> 00:55:10.958 align:middle line:90% resource? 00:55:10.958 --> 00:55:12.500 align:middle line:84% Is a patient-- and when appropriate-- 00:55:12.500 --> 00:55:14.625 align:middle line:84% a family prepared for either a positive or negative 00:55:14.625 --> 00:55:15.798 align:middle line:90% test result? 00:55:15.798 --> 00:55:18.090 align:middle line:84% Does the patient understand the medical, psychological, 00:55:18.090 --> 00:55:19.760 align:middle line:84% and social ramifications of the test? 00:55:19.760 --> 00:55:21.950 align:middle line:84% Do you have a referral less than appropriate specialist and/or 00:55:21.950 --> 00:55:24.290 align:middle line:84% genetic counseling services to resolve any issues that cannot 00:55:24.290 --> 00:55:25.220 align:middle line:90% be handled in the office? 00:55:25.220 --> 00:55:27.530 align:middle line:84% And these are all things that we have to do as a clinic 00:55:27.530 --> 00:55:29.780 align:middle line:84% before we can even start ordering these tests. 00:55:29.780 --> 00:55:32.130 align:middle line:90% And the workflow is as follows. 00:55:32.130 --> 00:55:35.840 align:middle line:84% For a positive test, these are all the things we have to do. 00:55:35.840 --> 00:55:38.510 align:middle line:84% Now we're a primary care practice. 00:55:38.510 --> 00:55:41.730 align:middle line:84% And the interpretation of a positive result 00:55:41.730 --> 00:55:44.090 align:middle line:84% is different if it's a diagnostic test, 00:55:44.090 --> 00:55:46.550 align:middle line:84% a predictive test, a carrier test, prenatal testing, 00:55:46.550 --> 00:55:48.560 align:middle line:90% or newborn screening. 00:55:48.560 --> 00:55:52.930 align:middle line:84% If it's a positive test and it's clinical diagnosis confirmed, 00:55:52.930 --> 00:55:53.930 align:middle line:90% it's predictive testing. 00:55:53.930 --> 00:55:55.015 align:middle line:84% It tells you have increased risk. 00:55:55.015 --> 00:55:56.300 align:middle line:84% If it's carrier testing, it tells you 00:55:56.300 --> 00:55:57.380 align:middle line:90% that you're a carrier. 00:55:57.380 --> 00:55:59.270 align:middle line:84% It's prenatal testing, it says that the fetus 00:55:59.270 --> 00:56:00.680 align:middle line:90% has a specific condition. 00:56:00.680 --> 00:56:02.510 align:middle line:84% And if it's newborn screening, it 00:56:02.510 --> 00:56:07.340 align:middle line:84% tells you the newborn has a disease. 00:56:07.340 --> 00:56:12.350 align:middle line:84% And the follow-up includes all the things you could imagine. 00:56:12.350 --> 00:56:15.830 align:middle line:84% For negative tests, if it's a diagnostic test, 00:56:15.830 --> 00:56:18.663 align:middle line:90% the symptoms are unexplained. 00:56:18.663 --> 00:56:21.080 align:middle line:84% For predictive testing, the likelihood of showing symptoms 00:56:21.080 --> 00:56:22.460 align:middle line:90% is decreased. 00:56:22.460 --> 00:56:23.930 align:middle line:84% If it's carrier, it's highly likely 00:56:23.930 --> 00:56:25.370 align:middle line:90% that you're not a carrier. 00:56:25.370 --> 00:56:29.070 align:middle line:84% If it's prenatal testing, if the fetus was symptomatic, 00:56:29.070 --> 00:56:31.670 align:middle line:90% then it's unexplained. 00:56:31.670 --> 00:56:34.340 align:middle line:84% If it's not symptomatic, the chance for the condition 00:56:34.340 --> 00:56:35.000 align:middle line:90% is small. 00:56:35.000 --> 00:56:37.520 align:middle line:84% And newborn screening-- the newborn 00:56:37.520 --> 00:56:39.583 align:middle line:84% is not expected to have the condition. 00:56:39.583 --> 00:56:41.000 align:middle line:84% There's a whole bunch of follow up 00:56:41.000 --> 00:56:43.640 align:middle line:84% that does or does not develop from these different levels 00:56:43.640 --> 00:56:47.360 align:middle line:84% of use of the same darn test across these different clinical 00:56:47.360 --> 00:56:48.050 align:middle line:90% states. 00:56:48.050 --> 00:56:53.540 align:middle line:84% And thinking about 512 Genomics, our old company, 00:56:53.540 --> 00:56:56.348 align:middle line:84% if we didn't know which one of the situations are involved, 00:56:56.348 --> 00:56:58.640 align:middle line:84% we'd have to practice a very defensive corporate policy 00:56:58.640 --> 00:57:02.510 align:middle line:84% to make sure we'd be in the maximal informed 00:57:02.510 --> 00:57:06.450 align:middle line:90% state for all conditions. 00:57:06.450 --> 00:57:08.990 align:middle line:84% So can I send a sample for microarray testing? 00:57:08.990 --> 00:57:11.480 align:middle line:84% We've heard so much about microarrays in this class. 00:57:11.480 --> 00:57:16.160 align:middle line:84% In the USA, the answer is yes for a research protocol. 00:57:16.160 --> 00:57:18.920 align:middle line:84% This summer, you're going to over at Harvard Medical 00:57:18.920 --> 00:57:23.060 align:middle line:84% School genotyping with a resequencing 00:57:23.060 --> 00:57:29.190 align:middle line:84% array for cardic myopathies and for hearing deficiencies. 00:57:29.190 --> 00:57:33.020 align:middle line:84% These are custom chips from Affymetrix. 00:57:33.020 --> 00:57:38.270 align:middle line:84% And certainly for research protocols, 00:57:38.270 --> 00:57:40.965 align:middle line:84% we're using them for expression as well. 00:57:40.965 --> 00:57:42.590 align:middle line:84% Can you do this for commercial testing? 00:57:42.590 --> 00:57:43.910 align:middle line:90% The answer is no. 00:57:43.910 --> 00:57:47.780 align:middle line:84% Roche Diagnostics, three months ago, 00:57:47.780 --> 00:57:51.170 align:middle line:84% tried to get approved through the expedited 00:57:51.170 --> 00:57:59.180 align:middle line:84% review of the FDA, the P450 chip, which 00:57:59.180 --> 00:58:03.950 align:middle line:84% has all the genes that are involved for metabolizing 00:58:03.950 --> 00:58:05.330 align:middle line:90% toxins and drugs. 00:58:05.330 --> 00:58:08.720 align:middle line:84% And therefore, it's a very good pharmacogenomic screening. 00:58:08.720 --> 00:58:12.380 align:middle line:84% And they were blocked dead in their tracks by the FDA. 00:58:12.380 --> 00:58:14.220 align:middle line:84% Unless they can appeal it correctly, 00:58:14.220 --> 00:58:17.990 align:middle line:84% they will have to go through full bore extensive FDA 00:58:17.990 --> 00:58:19.670 align:middle line:90% approval. 00:58:19.670 --> 00:58:22.420 align:middle line:90% And So why is it the case? 00:58:22.420 --> 00:58:24.170 align:middle line:84% I can give you a lot of different reasons. 00:58:24.170 --> 00:58:25.837 align:middle line:84% But the short answer is the FDA does not 00:58:25.837 --> 00:58:28.100 align:middle line:84% understand this technology and does not 00:58:28.100 --> 00:58:29.150 align:middle line:90% know what to do with it. 00:58:29.150 --> 00:58:31.625 align:middle line:84% I was part of a panel with the FDA-- 00:58:31.625 --> 00:58:33.500 align:middle line:84% [INAUDIBLE] talking about it-- they have just 00:58:33.500 --> 00:58:35.630 align:middle line:84% agreed, right now, with big pharma, 00:58:35.630 --> 00:58:38.960 align:middle line:84% just on what is a data structure with which they can transfer 00:58:38.960 --> 00:58:43.550 align:middle line:84% results, let alone interpretation or analysis 00:58:43.550 --> 00:58:45.230 align:middle line:90% of the results. 00:58:45.230 --> 00:58:47.600 align:middle line:84% AUDIENCE: So they're restricting things 00:58:47.600 --> 00:58:51.350 align:middle line:84% just because they don't understand it, even though it's 00:58:51.350 --> 00:58:55.580 align:middle line:84% not unleashing a new drug on the population 00:58:55.580 --> 00:58:58.160 align:middle line:84% that they don't understand the side effects. 00:58:58.160 --> 00:59:00.680 align:middle line:84% This actually doesn't do anything other 00:59:00.680 --> 00:59:04.460 align:middle line:84% than maybe give patients something to worry about-- 00:59:04.460 --> 00:59:08.360 align:middle line:84% ISAAC SAMUEL KOHANE: Let me make it very concrete. 00:59:08.360 --> 00:59:10.430 align:middle line:84% It gives you a result that the patient then 00:59:10.430 --> 00:59:12.500 align:middle line:84% has ovaries and breasts removed for. 00:59:12.500 --> 00:59:15.103 align:middle line:90% 00:59:15.103 --> 00:59:16.770 align:middle line:84% AUDIENCE: I guess when you put in that-- 00:59:16.770 --> 00:59:19.530 align:middle line:84% ISAAC SAMUEL KOHANE: Yeah, I mean, that's what-- 00:59:19.530 --> 00:59:20.030 align:middle line:90% what? 00:59:20.030 --> 00:59:22.305 align:middle line:84% AUDIENCE: --it's a classic example of what I've heard from 00:59:22.305 --> 00:59:22.805 align:middle line:90% bilateral-- 00:59:22.805 --> 00:59:25.208 align:middle line:90% [INAUDIBLE] for no real reason. 00:59:25.208 --> 00:59:26.750 align:middle line:84% ISAAC SAMUEL KOHANE: --for no reason. 00:59:26.750 --> 00:59:28.130 align:middle line:90% But that's going to happen. 00:59:28.130 --> 00:59:29.300 align:middle line:84% Many things are going to happen like that. 00:59:29.300 --> 00:59:30.920 align:middle line:84% I mean, just going to the doctor-- 00:59:30.920 --> 00:59:33.190 align:middle line:84% I try to stay away from the doctor as much as I can, 00:59:33.190 --> 00:59:34.820 align:middle line:90% which is probably a bad idea. 00:59:34.820 --> 00:59:36.710 align:middle line:84% Because you probably should get routine care. 00:59:36.710 --> 00:59:41.840 align:middle line:84% But if every time you expose yourself to even 00:59:41.840 --> 00:59:44.870 align:middle line:84% an investigative procedure for the wrong reason, 00:59:44.870 --> 00:59:48.740 align:middle line:84% they're likely to find-- again, to the false positive reason-- 00:59:48.740 --> 00:59:50.660 align:middle line:84% things that are incidental but are 00:59:50.660 --> 00:59:55.580 align:middle line:84% going to create huge costs, and worry, and morbidity. 00:59:55.580 --> 00:59:59.850 align:middle line:84% Diagnostic tests themselves can kill you. 00:59:59.850 --> 01:00:03.030 align:middle line:84% So let's say you are told you have a risk for this. 01:00:03.030 --> 01:00:10.230 align:middle line:90% And then you get a colonoscopy. 01:00:10.230 --> 01:00:12.830 align:middle line:84% A tiny fraction, probably one in 200,000 people, 01:00:12.830 --> 01:00:16.460 align:middle line:84% suffer bad outcomes from that procedure. 01:00:16.460 --> 01:00:18.080 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:00:18.080 --> 01:00:23.420 align:middle line:84% ISAAC SAMUEL KOHANE: Yeah, those are pain in the you know what. 01:00:23.420 --> 01:00:27.740 align:middle line:90% 01:00:27.740 --> 01:00:29.160 align:middle line:90% So this is a problem. 01:00:29.160 --> 01:00:31.250 align:middle line:84% So the short answer is-- microarray testing, no. 01:00:31.250 --> 01:00:35.030 align:middle line:84% But there's nothing wholly about the FDA. 01:00:35.030 --> 01:00:37.490 align:middle line:90% In the Netherlands-- 01:00:37.490 --> 01:00:40.580 align:middle line:84% I don't know-- again, this is what 01:00:40.580 --> 01:00:42.583 align:middle line:84% I had heard last year around this time. 01:00:42.583 --> 01:00:44.000 align:middle line:84% And I don't know if it's happened. 01:00:44.000 --> 01:00:45.583 align:middle line:84% But they were claiming they were going 01:00:45.583 --> 01:00:49.460 align:middle line:84% into routine use of microarray for breast biopsy 01:00:49.460 --> 01:00:50.685 align:middle line:90% screening and evaluation. 01:00:50.685 --> 01:00:52.310 align:middle line:84% I have to do the research to figure out 01:00:52.310 --> 01:00:55.520 align:middle line:90% whether it's happened or not. 01:00:55.520 --> 01:00:57.560 align:middle line:84% And there's, of course, the issue of consent. 01:00:57.560 --> 01:01:01.400 align:middle line:84% And the short answer-- as we'll get to shortly-- 01:01:01.400 --> 01:01:05.100 align:middle line:84% is that you have to do a fairly extensive consent process. 01:01:05.100 --> 01:01:08.070 align:middle line:84% However, if the patient is symptomatic, 01:01:08.070 --> 01:01:10.730 align:middle line:84% if they have a cancer, or they're having a heart attack, 01:01:10.730 --> 01:01:14.270 align:middle line:84% or they have dementia-- they're actually symptomatic, 01:01:14.270 --> 01:01:16.993 align:middle line:84% it's a much more simple consent process. 01:01:16.993 --> 01:01:18.410 align:middle line:84% All the consent issues that you've 01:01:18.410 --> 01:01:22.085 align:middle line:84% heard about in public press are about pre-symptomatic testing. 01:01:22.085 --> 01:01:25.270 align:middle line:90% 01:01:25.270 --> 01:01:29.545 align:middle line:84% AUDIENCE: So why don't they just approve 01:01:29.545 --> 01:01:32.500 align:middle line:84% that the pre-symptomatic testing with the caveat 01:01:32.500 --> 01:01:37.225 align:middle line:84% or the [INAUDIBLE] rule that we're just going to follow 01:01:37.225 --> 01:01:39.100 align:middle line:90% and observe a patient? 01:01:39.100 --> 01:01:42.130 align:middle line:84% Let's say that we usually go in and get a full-body MRI 01:01:42.130 --> 01:01:47.680 align:middle line:84% scan [INAUDIBLE] I think you shared a story with someone 01:01:47.680 --> 01:01:48.650 align:middle line:90% about that. 01:01:48.650 --> 01:01:49.150 align:middle line:90% Found it. 01:01:49.150 --> 01:01:52.030 align:middle line:84% And try to do nothing [INAUDIBLE] complications. 01:01:52.030 --> 01:01:54.820 align:middle line:84% So rather than go in and cut someone's head 01:01:54.820 --> 01:01:59.230 align:middle line:84% open, or a colonoscopy and everything else, 01:01:59.230 --> 01:02:01.210 align:middle line:90% just observe the patient-- 01:02:01.210 --> 01:02:03.790 align:middle line:84% six months, a year, two years, three years. 01:02:03.790 --> 01:02:04.210 align:middle line:84% ISAAC SAMUEL KOHANE: So let me tell you 01:02:04.210 --> 01:02:05.252 align:middle line:90% something about medicine. 01:02:05.252 --> 01:02:06.890 align:middle line:90% That's impossible. 01:02:06.890 --> 01:02:16.210 align:middle line:84% If you find something that looks bad on an MRI-- 01:02:16.210 --> 01:02:18.940 align:middle line:84% because even if you believe that it's 01:02:18.940 --> 01:02:21.760 align:middle line:84% only one chance in 100 that is truly a bad thing 01:02:21.760 --> 01:02:23.597 align:middle line:84% and something bad happens to that patient, 01:02:23.597 --> 01:02:25.430 align:middle line:84% not only are you toast with the legal system 01:02:25.430 --> 01:02:26.920 align:middle line:90% but you'll feel terrible. 01:02:26.920 --> 01:02:28.990 align:middle line:84% I mean, it could make you very, very 01:02:28.990 --> 01:02:33.550 align:middle line:84% fidgety to have a patient with a finding 01:02:33.550 --> 01:02:38.050 align:middle line:84% that you were taught in medical school is badness. 01:02:38.050 --> 01:02:41.110 align:middle line:84% And we're told in 2004, medical education in 2004. 01:02:41.110 --> 01:02:44.120 align:middle line:84% We're taught that some bumps on MRIs mean bad things. 01:02:44.120 --> 01:02:48.490 align:middle line:84% So if you start changing your priors by screening everybody, 01:02:48.490 --> 01:02:50.140 align:middle line:84% then bumps do mean different things. 01:02:50.140 --> 01:02:51.932 align:middle line:84% But that's not the way we've been educated. 01:02:51.932 --> 01:02:53.350 align:middle line:90% So we kind of get itchy as hell. 01:02:53.350 --> 01:02:56.200 align:middle line:84% AUDIENCE: So with the advent of new detection technologies 01:02:56.200 --> 01:02:59.440 align:middle line:84% for particular things that you context intelligence 01:02:59.440 --> 01:03:05.515 align:middle line:84% specific for this sort of problem [INAUDIBLE] 01:03:05.515 --> 01:03:08.140 align:middle line:84% ISAAC SAMUEL KOHANE: Yes, except it never really does because-- 01:03:08.140 --> 01:03:11.200 align:middle line:90% 01:03:11.200 --> 01:03:12.700 align:middle line:90% well, that's not true. 01:03:12.700 --> 01:03:16.480 align:middle line:84% For certain things, we're going to get specific enough markers. 01:03:16.480 --> 01:03:19.630 align:middle line:90% I can imagine that a bump-- 01:03:19.630 --> 01:03:22.690 align:middle line:84% may be wrong-- imagine a bump plus a functional assay, 01:03:22.690 --> 01:03:25.720 align:middle line:84% like with a PET scan, will show that this is a highly rapidly 01:03:25.720 --> 01:03:29.530 align:middle line:84% dividing thingamajig and, therefore, it's a problem. 01:03:29.530 --> 01:03:32.230 align:middle line:84% Or it's totally metabolically quiescent 01:03:32.230 --> 01:03:33.980 align:middle line:90% and we can just observe it. 01:03:33.980 --> 01:03:35.080 align:middle line:90% So that's possible. 01:03:35.080 --> 01:03:36.580 align:middle line:90% Yes. 01:03:36.580 --> 01:03:39.520 align:middle line:84% AUDIENCE: I just listening to [INAUDIBLE] talking about his 01:03:39.520 --> 01:03:42.140 align:middle line:84% idea of why don't we treat cancers before they actually 01:03:42.140 --> 01:03:42.640 align:middle line:90% become-- 01:03:42.640 --> 01:03:44.432 align:middle line:84% ISAAC SAMUEL KOHANE: He's absolutely right. 01:03:44.432 --> 01:03:47.200 align:middle line:84% AUDIENCE: --and that, to me, falls right in line 01:03:47.200 --> 01:03:51.750 align:middle line:84% with this just what's the harm in doing this analysis. 01:03:51.750 --> 01:03:53.470 align:middle line:84% ISAAC SAMUEL KOHANE: Well, I have 01:03:53.470 --> 01:03:55.330 align:middle line:90% the greatest respect for Judah. 01:03:55.330 --> 01:03:57.950 align:middle line:84% And I'm sure he'd would agree with what I'm about to say. 01:03:57.950 --> 01:04:01.960 align:middle line:84% It all comes down to what is the treatment 01:04:01.960 --> 01:04:04.430 align:middle line:90% and what threshold do you pick. 01:04:04.430 --> 01:04:08.770 align:middle line:84% So if you're looking at a certain angiogenic signature, 01:04:08.770 --> 01:04:10.630 align:middle line:84% whether it's in a polymorphism or something 01:04:10.630 --> 01:04:13.240 align:middle line:84% that you're measuring, what's your false positive rate 01:04:13.240 --> 01:04:14.740 align:middle line:90% and how toxic is your treatment? 01:04:14.740 --> 01:04:17.060 align:middle line:84% If you're treatment's totally benign, of course, 01:04:17.060 --> 01:04:21.190 align:middle line:84% treat everybody and cheap and treat everybody. 01:04:21.190 --> 01:04:24.400 align:middle line:90% But that's almost never true. 01:04:24.400 --> 01:04:27.970 align:middle line:84% It's never completely benign and completely cheap. 01:04:27.970 --> 01:04:29.800 align:middle line:84% So you end up having to make decisions. 01:04:29.800 --> 01:04:33.240 align:middle line:90% 01:04:33.240 --> 01:04:36.800 align:middle line:84% So now for presymptomatic asymptomatic testing, 01:04:36.800 --> 01:04:43.410 align:middle line:84% consent is actually a very, very complex and drawn out process. 01:04:43.410 --> 01:04:46.250 align:middle line:90% So here's what is involved. 01:04:46.250 --> 01:04:47.750 align:middle line:90% Again, this is 2004. 01:04:47.750 --> 01:04:51.230 align:middle line:84% It may change to be more rigorous or less rigorous, 01:04:51.230 --> 01:04:53.480 align:middle line:84% depending on whether or not the Genetic Privacy 01:04:53.480 --> 01:04:56.120 align:middle line:84% Act, which is now somewhere in Congress, 01:04:56.120 --> 01:04:58.500 align:middle line:90% will be enacted or not. 01:04:58.500 --> 01:05:01.130 align:middle line:84% So the current state-of-the-art is, 01:05:01.130 --> 01:05:03.320 align:middle line:84% the major medical facts including the diagnosis, 01:05:03.320 --> 01:05:06.440 align:middle line:84% the prognosis, and the way the treatment of disorder tested 01:05:06.440 --> 01:05:07.970 align:middle line:84% has to be explained to the patients. 01:05:07.970 --> 01:05:09.950 align:middle line:84% The genetic facts involving-- including 01:05:09.950 --> 01:05:13.383 align:middle line:84% risks for other family members has to be explained. 01:05:13.383 --> 01:05:15.800 align:middle line:84% The chance that the test will give a correct prediction as 01:05:15.800 --> 01:05:20.270 align:middle line:84% well indeterminate or unexpected findings has explained to them. 01:05:20.270 --> 01:05:22.700 align:middle line:84% The risk of receiving an unfavorable test 01:05:22.700 --> 01:05:24.440 align:middle line:84% result and the possible consequences 01:05:24.440 --> 01:05:25.790 align:middle line:90% for yourself and your family-- 01:05:25.790 --> 01:05:26.975 align:middle line:84% in the case of prenatal diagnosis, 01:05:26.975 --> 01:05:28.433 align:middle line:84% this may include the risk of facing 01:05:28.433 --> 01:05:29.900 align:middle line:90% a decision about abortion. 01:05:29.900 --> 01:05:33.793 align:middle line:84% And you have to tell them in a way that they really recognize, 01:05:33.793 --> 01:05:34.460 align:middle line:90% they understand. 01:05:34.460 --> 01:05:39.375 align:middle line:84% Some people are videotaping these concerns now, 01:05:39.375 --> 01:05:41.000 align:middle line:84% well, defensively and also to make sure 01:05:41.000 --> 01:05:43.280 align:middle line:84% that the practitioners are doing the right thing. 01:05:43.280 --> 01:05:46.820 align:middle line:84% Also, it should be made clear that can refuse the test 01:05:46.820 --> 01:05:49.790 align:middle line:84% and informed of potential benefits and disadvantages, 01:05:49.790 --> 01:05:53.000 align:middle line:84% including unsettled questions of privacy protection dealing 01:05:53.000 --> 01:05:56.560 align:middle line:84% with insurances, banks, or employers. 01:05:56.560 --> 01:06:01.550 align:middle line:84% You might be able to say today the insurance company 01:06:01.550 --> 01:06:03.050 align:middle line:84% is not going to do anything with it. 01:06:03.050 --> 01:06:05.450 align:middle line:84% But that test is now forever in your medical record 01:06:05.450 --> 01:06:07.432 align:middle line:84% and maybe not true five years from now. 01:06:07.432 --> 01:06:09.140 align:middle line:84% It might be a different insurance company 01:06:09.140 --> 01:06:11.635 align:middle line:84% or the same insurance company will feel differently. 01:06:11.635 --> 01:06:13.760 align:middle line:84% Your care will not be jeopardized whatever decision 01:06:13.760 --> 01:06:15.020 align:middle line:90% you and your family make. 01:06:15.020 --> 01:06:17.030 align:middle line:84% Possible use of your tissue sample after testing-- they 01:06:17.030 --> 01:06:18.030 align:middle line:90% have to understand that. 01:06:18.030 --> 01:06:20.000 align:middle line:84% Whether it's destroyed or kept from an analysis 01:06:20.000 --> 01:06:23.083 align:middle line:84% and whether it can be kept for DNA banking. 01:06:23.083 --> 01:06:25.500 align:middle line:84% All of these things have to be discussed with the patient. 01:06:25.500 --> 01:06:28.910 align:middle line:84% This is not a five-minute discussion. 01:06:28.910 --> 01:06:32.930 align:middle line:84% And remember, again, back to our corporate clinic, 01:06:32.930 --> 01:06:37.580 align:middle line:84% 512 Clinics, our standard of visit for a patient 01:06:37.580 --> 01:06:40.250 align:middle line:84% is on the order of 10 to 15 minutes. 01:06:40.250 --> 01:06:42.890 align:middle line:84% And I'm telling you that this concept process requires 01:06:42.890 --> 01:06:45.620 align:middle line:90% an hour to do it adequately. 01:06:45.620 --> 01:06:48.848 align:middle line:90% 01:06:48.848 --> 01:06:50.390 align:middle line:84% Having said this, I want to tell you, 01:06:50.390 --> 01:06:54.170 align:middle line:84% we're actually performing routine comprehensive genetic 01:06:54.170 --> 01:06:58.010 align:middle line:84% testing on our entire population. 01:06:58.010 --> 01:07:00.060 align:middle line:90% Did you know that? 01:07:00.060 --> 01:07:02.300 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:07:02.300 --> 01:07:03.890 align:middle line:90% ISAAC SAMUEL KOHANE: You bet. 01:07:03.890 --> 01:07:08.720 align:middle line:84% So we're doing we're testing all kids in the United States 01:07:08.720 --> 01:07:13.010 align:middle line:84% for a genetic disease like phenylketonuria. 01:07:13.010 --> 01:07:15.590 align:middle line:84% We're now looking for all the common polymorphisms 01:07:15.590 --> 01:07:23.640 align:middle line:84% of CFTR, the cystic fibrosis responsible gene. 01:07:23.640 --> 01:07:28.940 align:middle line:84% And if you look what the Massachusetts State Laboratory 01:07:28.940 --> 01:07:32.262 align:middle line:84% is doing right now in Worcester with University 01:07:32.262 --> 01:07:34.220 align:middle line:84% of Massachusetts, I said, they're Jamaica Plain 01:07:34.220 --> 01:07:36.920 align:middle line:84% but they're affiliated with the University of Massachusetts. 01:07:36.920 --> 01:07:39.740 align:middle line:84% They're actually looking at maybe 10, 20 diseases 01:07:39.740 --> 01:07:43.640 align:middle line:84% that they're going to do risk factor screening for. 01:07:43.640 --> 01:07:46.130 align:middle line:84% Now unlike our company, 512 Genomics, 01:07:46.130 --> 01:07:49.790 align:middle line:84% they really have to come to a couple of decisions 01:07:49.790 --> 01:07:51.800 align:middle line:90% about when to do this. 01:07:51.800 --> 01:07:54.830 align:middle line:84% How is it determined what is routinely screened? 01:07:54.830 --> 01:07:57.140 align:middle line:84% Three components-- public health assessment, 01:07:57.140 --> 01:07:59.240 align:middle line:84% evaluation of tests and interventions, 01:07:59.240 --> 01:08:01.230 align:middle line:84% positive and screening implementation. 01:08:01.230 --> 01:08:03.230 align:middle line:84% The public assessment is fairly straightforward. 01:08:03.230 --> 01:08:05.647 align:middle line:84% Disease or conditions should be an important public health 01:08:05.647 --> 01:08:07.100 align:middle line:90% burden not a rare bird. 01:08:07.100 --> 01:08:08.640 align:middle line:90% And what does that mean? 01:08:08.640 --> 01:08:12.240 align:middle line:84% That means, typically, they like to see it above one in 20,000. 01:08:12.240 --> 01:08:14.240 align:middle line:84% They won't admit to that, but it's around there. 01:08:14.240 --> 01:08:16.532 align:middle line:84% Like thyroid disease, which was one of the first things 01:08:16.532 --> 01:08:19.107 align:middle line:84% to be screened, is about one in 5,000. 01:08:19.107 --> 01:08:21.649 align:middle line:84% Congenital adrenal hyperplasia, which they do screen for now, 01:08:21.649 --> 01:08:24.410 align:middle line:84% for 21 hydroxylase is about one in 20,000. 01:08:24.410 --> 01:08:26.630 align:middle line:84% The prevalence of the genetic trait has been known. 01:08:26.630 --> 01:08:28.609 align:middle line:84% The natural history of the condition for susceptibility 01:08:28.609 --> 01:08:30.234 align:middle line:84% to link disease to overt disease should 01:08:30.234 --> 01:08:32.042 align:middle line:90% be adequately understood. 01:08:32.042 --> 01:08:34.500 align:middle line:84% And, basically, the safety and efficacy of the test-- blah, 01:08:34.500 --> 01:08:35.708 align:middle line:90% blah, blah-- has to be known. 01:08:35.708 --> 01:08:38.910 align:middle line:84% Now policy issues are obviously important. 01:08:38.910 --> 01:08:42.750 align:middle line:90% But the main issue is this one. 01:08:42.750 --> 01:08:45.930 align:middle line:84% The cost of the screening should be established. 01:08:45.930 --> 01:08:47.729 align:middle line:84% In other words, the screening procedure, 01:08:47.729 --> 01:08:51.779 align:middle line:84% whatever it is, whether it's genotyping or tandem mass spec, 01:08:51.779 --> 01:08:54.149 align:middle line:84% has to actually be within the budget 01:08:54.149 --> 01:08:55.920 align:middle line:90% of the State of Massachusetts. 01:08:55.920 --> 01:09:00.420 align:middle line:84% And that's why, unlike 512 Genomics, 01:09:00.420 --> 01:09:04.710 align:middle line:84% they're absolutely allowed to make decisions like we're only 01:09:04.710 --> 01:09:07.109 align:middle line:84% going to look for the genotypes that account 01:09:07.109 --> 01:09:09.660 align:middle line:84% for 99% of the disease burden, because it's a public health 01:09:09.660 --> 01:09:13.200 align:middle line:90% intervention and not our doctor. 01:09:13.200 --> 01:09:14.910 align:middle line:90% And that's very, very different. 01:09:14.910 --> 01:09:19.460 align:middle line:90% 01:09:19.460 --> 01:09:22.673 align:middle line:84% AUDIENCE: [INAUDIBLE] tests biochemical or-- 01:09:22.673 --> 01:09:25.090 align:middle line:84% ISAAC SAMUEL KOHANE: Most of those tests were biochemical. 01:09:25.090 --> 01:09:26.750 align:middle line:84% They're testing a genetic disease. 01:09:26.750 --> 01:09:30.560 align:middle line:84% But what I'm telling you is this is actually old. 01:09:30.560 --> 01:09:32.990 align:middle line:84% The cystic fibrosis test now is a DNA test. 01:09:32.990 --> 01:09:35.479 align:middle line:84% And they have it now to test that going online this 01:09:35.479 --> 01:09:41.340 align:middle line:84% is an old slide in the remaining 10 minutes. 01:09:41.340 --> 01:09:42.330 align:middle line:90% Let's roll play. 01:09:42.330 --> 01:09:46.040 align:middle line:84% Let me tell you about a disease, congenital hyperinsulinemia 01:09:46.040 --> 01:09:49.130 align:middle line:84% is the most frequent cause of severe, persistent 01:09:49.130 --> 01:09:52.370 align:middle line:84% hyperglycemia in newborn babies and children. 01:09:52.370 --> 01:09:54.710 align:middle line:84% In most countries, it occurs in approximately one 01:09:54.710 --> 01:09:59.150 align:middle line:84% in 25,000 to one in 50,000 births. 01:09:59.150 --> 01:10:01.460 align:middle line:84% About 60% of babies with hyperinsulinemia 01:10:01.460 --> 01:10:04.130 align:middle line:84% develop hypoglycemia during that first month of life. 01:10:04.130 --> 01:10:05.820 align:middle line:90% Their blood sugar crashes. 01:10:05.820 --> 01:10:07.580 align:middle line:84% An additional 30% will be diagnosed later 01:10:07.580 --> 01:10:08.330 align:middle line:90% in the first year. 01:10:08.330 --> 01:10:09.660 align:middle line:90% And the remainder half of that. 01:10:09.660 --> 01:10:14.000 align:middle line:84% So 90% of them are diagnosed in the first year, 01:10:14.000 --> 01:10:15.953 align:middle line:90% 60% in the first month. 01:10:15.953 --> 01:10:17.870 align:middle line:84% With early treatment and aggressive prevention 01:10:17.870 --> 01:10:19.980 align:middle line:84% of hypoglycemia, brain damage can be prevented. 01:10:19.980 --> 01:10:22.310 align:middle line:84% The brain damage is from having sustained low blood 01:10:22.310 --> 01:10:25.070 align:middle line:84% sugar, which is actually not harmful when you're adult 01:10:25.070 --> 01:10:27.110 align:middle line:90% and your brain is stable. 01:10:27.110 --> 01:10:28.760 align:middle line:84% But when you have a developing brain, 01:10:28.760 --> 01:10:32.330 align:middle line:84% and it's not getting a lot of glucose and, therefore, 01:10:32.330 --> 01:10:35.665 align:middle line:84% metabolism for periods of time, especially repeatedly, 01:10:35.665 --> 01:10:37.040 align:middle line:84% you can really have brain damage. 01:10:37.040 --> 01:10:39.243 align:middle line:84% However, brain damage can occur in up to 50% 01:10:39.243 --> 01:10:41.660 align:middle line:84% of children with hyperinsulinism if their condition is not 01:10:41.660 --> 01:10:44.300 align:middle line:84% recognized or if treatment is ineffective in the prevention 01:10:44.300 --> 01:10:45.350 align:middle line:90% of hypoglycemia. 01:10:45.350 --> 01:10:48.710 align:middle line:90% 01:10:48.710 --> 01:10:52.430 align:middle line:84% So remember this-- one in 25,000 and one 01:10:52.430 --> 01:10:55.310 align:middle line:84% in 50,000-- on the other hand, neonatal hypoglycemia-- 01:10:55.310 --> 01:10:57.650 align:middle line:84% so hypoglycemia in these newborn kids 01:10:57.650 --> 01:11:00.800 align:middle line:84% has an incidence of two to five per 1,000. 01:11:00.800 --> 01:11:04.310 align:middle line:84% There's a zillion reasons why kids have hypoglycemia, 01:11:04.310 --> 01:11:07.190 align:middle line:84% from being slightly immature at birth 01:11:07.190 --> 01:11:09.980 align:middle line:84% to the IV was switched off too fast and so all of a sudden 01:11:09.980 --> 01:11:12.410 align:middle line:84% the pancreas didn't have enough minutes to wake 01:11:12.410 --> 01:11:17.510 align:middle line:84% up and switch off the insulin, that they're 01:11:17.510 --> 01:11:20.070 align:middle line:90% having an infection, and so on. 01:11:20.070 --> 01:11:23.840 align:middle line:84% So the state of our knowledge in 1992, 01:11:23.840 --> 01:11:27.020 align:middle line:84% when I was finishing my residency, 01:11:27.020 --> 01:11:30.800 align:middle line:84% was there was this disease called nesidioblastosis 01:11:30.800 --> 01:11:33.890 align:middle line:84% or hyperplasia of the pancreatic islet cells, the cells that 01:11:33.890 --> 01:11:35.360 align:middle line:90% produce insulin. 01:11:35.360 --> 01:11:37.811 align:middle line:84% And for reasons that we didn't know, 01:11:37.811 --> 01:11:41.570 align:middle line:84% either the entire pancreas or spots in pancreas 01:11:41.570 --> 01:11:45.410 align:middle line:90% were hyper-producing insulin. 01:11:45.410 --> 01:11:51.147 align:middle line:90% And if we didn't treat it-- 01:11:51.147 --> 01:11:52.730 align:middle line:84% some percentage these kids would go on 01:11:52.730 --> 01:11:58.460 align:middle line:84% to be medically manageable but a whole bunch of them were not. 01:11:58.460 --> 01:12:01.970 align:middle line:84% And we'd have to take out their whole pancreas. 01:12:01.970 --> 01:12:03.428 align:middle line:84% And sometimes we saw that we didn't 01:12:03.428 --> 01:12:04.928 align:middle line:84% need to take out the whole pancreas, 01:12:04.928 --> 01:12:07.160 align:middle line:84% that there were just focal pieces of pancreas that 01:12:07.160 --> 01:12:07.520 align:middle line:90% were involved. 01:12:07.520 --> 01:12:08.728 align:middle line:90% We could have left them more. 01:12:08.728 --> 01:12:10.460 align:middle line:84% Because by taking the whole pancreas out, 01:12:10.460 --> 01:12:15.830 align:middle line:84% we made them, A, diabetic, B, with insufficiency 01:12:15.830 --> 01:12:19.045 align:middle line:84% of those enzymes that used to dissolve your foods. 01:12:19.045 --> 01:12:20.420 align:middle line:84% It was really not a pretty sight. 01:12:20.420 --> 01:12:23.090 align:middle line:84% Plus, we kept them in the hospital for weeks 01:12:23.090 --> 01:12:23.900 align:middle line:90% to diagnose them. 01:12:23.900 --> 01:12:28.760 align:middle line:84% That was the state-of-the-art when I was a Fellow. 01:12:28.760 --> 01:12:31.280 align:middle line:84% Here's the state of our knowledge in 2004. 01:12:31.280 --> 01:12:33.230 align:middle line:84% There is something called a potassium channel, 01:12:33.230 --> 01:12:36.590 align:middle line:84% of which there are many, which controls insulin release 01:12:36.590 --> 01:12:38.150 align:middle line:90% in the islet cells. 01:12:38.150 --> 01:12:40.670 align:middle line:84% Basically, there are these two proteins-- 01:12:40.670 --> 01:12:51.500 align:middle line:84% SUR and Kir6.2, which actually gate the flow of potassium. 01:12:51.500 --> 01:12:53.910 align:middle line:84% The channel determines the resting membrane potential, 01:12:53.910 --> 01:12:55.430 align:middle line:84% which is maintained at the necessary voltage 01:12:55.430 --> 01:12:57.472 align:middle line:84% to keep voltage dependent calcium channels closed 01:12:57.472 --> 01:12:58.910 align:middle line:90% in a cell that does not secrete. 01:12:58.910 --> 01:13:00.920 align:middle line:84% When glucose comes into the cell and there's 01:13:00.920 --> 01:13:04.730 align:middle line:84% a resulting change in the ratio of ATP to ADP 01:13:04.730 --> 01:13:06.740 align:middle line:84% because of metabolism, K channels 01:13:06.740 --> 01:13:09.710 align:middle line:84% close the membrane that depolarizes. 01:13:09.710 --> 01:13:12.530 align:middle line:84% Subsequently, voltage-gated calcium channels 01:13:12.530 --> 01:13:16.400 align:middle line:84% are open, initiating insulin secondary cascade. 01:13:16.400 --> 01:13:18.620 align:middle line:90% Therefore, the K channel-- 01:13:18.620 --> 01:13:20.225 align:middle line:84% the potassium channel-- functions 01:13:20.225 --> 01:13:22.547 align:middle line:84% that are linked to the metabolic state of the cell 01:13:22.547 --> 01:13:24.380 align:middle line:84% and the electrical activity of the membrane, 01:13:24.380 --> 01:13:27.390 align:middle line:84% resulting in the stimulation or inhibition of insulin release. 01:13:27.390 --> 01:13:30.410 align:middle line:84% So we have a thermostat, rheostat, 01:13:30.410 --> 01:13:34.970 align:middle line:84% which says, essentially, the more glucose I see, the more 01:13:34.970 --> 01:13:38.000 align:middle line:84% I'm going to release calcium and, therefore, 01:13:38.000 --> 01:13:40.080 align:middle line:90% insulin into the blood. 01:13:40.080 --> 01:13:42.590 align:middle line:90% So that's the basic rheostat. 01:13:42.590 --> 01:13:47.540 align:middle line:84% And we know that these two genes, which actually happen 01:13:47.540 --> 01:13:51.410 align:middle line:84% to be next to each other on chromosome 11-- 01:13:51.410 --> 01:13:53.330 align:middle line:84% if you have mutations in them, instead 01:13:53.330 --> 01:13:56.660 align:middle line:84% of coming nicely together in this octamer, 01:13:56.660 --> 01:14:00.110 align:middle line:84% they come in these various dysfunctional 01:14:00.110 --> 01:14:08.030 align:middle line:84% or poorly functional heterodimers or homooctamers. 01:14:08.030 --> 01:14:10.130 align:middle line:84% They just come together in the wrong way. 01:14:10.130 --> 01:14:12.200 align:middle line:90% And they don't work right. 01:14:12.200 --> 01:14:15.230 align:middle line:84% And it turns out, there's a lot of mutations 01:14:15.230 --> 01:14:17.550 align:middle line:90% in SUR, one of those genes. 01:14:17.550 --> 01:14:23.280 align:middle line:84% In the Kir6.2, there's only three mutations. 01:14:23.280 --> 01:14:28.350 align:middle line:84% But looks like maybe 30, 40 mutations in SUR. 01:14:28.350 --> 01:14:29.772 align:middle line:90% Some of them are common. 01:14:29.772 --> 01:14:30.480 align:middle line:90% They're hotspots. 01:14:30.480 --> 01:14:32.063 align:middle line:84% A lot of different families have them. 01:14:32.063 --> 01:14:33.510 align:middle line:90% Some of them are one-of-a-kind. 01:14:33.510 --> 01:14:35.650 align:middle line:84% Some of them are in the coding regions. 01:14:35.650 --> 01:14:40.020 align:middle line:90% These are the 39 exons of SUR. 01:14:40.020 --> 01:14:41.520 align:middle line:84% And, by the way, that's really going 01:14:41.520 --> 01:14:45.210 align:middle line:84% to cost us in our 512 Genomics company 01:14:45.210 --> 01:14:46.950 align:middle line:90% to sequence all of those. 01:14:46.950 --> 01:14:50.220 align:middle line:90% And some of them are in introns. 01:14:50.220 --> 01:14:53.190 align:middle line:90% They're at the splice junction. 01:14:53.190 --> 01:14:58.583 align:middle line:84% They're just right across the intron, exon boundary 01:14:58.583 --> 01:15:00.000 align:middle line:84% into the part of the intron, which 01:15:00.000 --> 01:15:03.810 align:middle line:84% is the donor or splice acceptor site, which determines when 01:15:03.810 --> 01:15:05.100 align:middle line:90% you're going to splice or not. 01:15:05.100 --> 01:15:09.660 align:middle line:84% And these all are known highly-penetrant causes 01:15:09.660 --> 01:15:11.460 align:middle line:90% of disease. 01:15:11.460 --> 01:15:15.750 align:middle line:90% They're autosomal recessive. 01:15:15.750 --> 01:15:19.200 align:middle line:84% They are autosomal recessive, but if you 01:15:19.200 --> 01:15:24.600 align:middle line:84% look at these two forms of hyperinsulinemia-- 01:15:24.600 --> 01:15:26.910 align:middle line:84% so if you look at the slide, there's 01:15:26.910 --> 01:15:29.700 align:middle line:84% two patients that have the two different presentations I 01:15:29.700 --> 01:15:30.360 align:middle line:90% told you about. 01:15:30.360 --> 01:15:35.430 align:middle line:84% One has these focal hyperplasia, which pushes aside 01:15:35.430 --> 01:15:37.500 align:middle line:84% the normal tissue, and the other where it 01:15:37.500 --> 01:15:40.270 align:middle line:90% diffuses throughout the tissue. 01:15:40.270 --> 01:15:42.340 align:middle line:90% We now know why that is. 01:15:42.340 --> 01:15:46.620 align:middle line:84% It turns out that there is in the diffuse form, 01:15:46.620 --> 01:15:50.670 align:middle line:84% it's germline transmission of the homozygous state. 01:15:50.670 --> 01:15:53.070 align:middle line:84% And the focal state-- what's happening is there is loss 01:15:53.070 --> 01:15:59.910 align:middle line:84% of heterozygosity so that you have a parental gene-- 01:15:59.910 --> 01:16:02.040 align:middle line:84% I think actually a paternal gene. 01:16:02.040 --> 01:16:03.625 align:middle line:84% And there's a loss of heterozygosity, 01:16:03.625 --> 01:16:05.500 align:middle line:84% so it would be a loss of the maternal allele. 01:16:05.500 --> 01:16:10.170 align:middle line:84% And so in those cells with a loss of maternal allele 01:16:10.170 --> 01:16:14.820 align:middle line:84% get focal hyperplasia due to the poor functioning 01:16:14.820 --> 01:16:17.620 align:middle line:90% of that channel. 01:16:17.620 --> 01:16:21.510 align:middle line:84% So I told you about this disease. 01:16:21.510 --> 01:16:23.580 align:middle line:84% And in the last five remaining minutes, 01:16:23.580 --> 01:16:25.750 align:middle line:90% would you order a genetic test? 01:16:25.750 --> 01:16:28.590 align:middle line:84% How about hypoglycemia in a newborn? 01:16:28.590 --> 01:16:29.700 align:middle line:90% Maybe. 01:16:29.700 --> 01:16:32.010 align:middle line:90% Jose says no. 01:16:32.010 --> 01:16:35.295 align:middle line:84% Dr. 512, are you going to order this test or not? 01:16:35.295 --> 01:16:38.650 align:middle line:90% 01:16:38.650 --> 01:16:40.900 align:middle line:84% At least explain like a doctor because you're actually 01:16:40.900 --> 01:16:41.500 align:middle line:90% making a decision. 01:16:41.500 --> 01:16:42.250 align:middle line:90% Because we have a patient. 01:16:42.250 --> 01:16:42.955 align:middle line:90% We need a decision. 01:16:42.955 --> 01:16:44.320 align:middle line:84% AUDIENCE: This is one in a 1,000? 01:16:44.320 --> 01:16:45.403 align:middle line:90% ISAAC SAMUEL KOHANE: What? 01:16:45.403 --> 01:16:46.750 align:middle line:90% AUDIENCE: You said one in 1,000? 01:16:46.750 --> 01:16:48.750 align:middle line:84% ISAAC SAMUEL KOHANE: I think I said two to 5,000 01:16:48.750 --> 01:16:50.950 align:middle line:90% of routine hypoglycemia. 01:16:50.950 --> 01:16:54.610 align:middle line:84% And I think I said one in 25,000 for 01:16:54.610 --> 01:16:56.600 align:middle line:90% persistent hyperinsulinemia. 01:16:56.600 --> 01:16:58.845 align:middle line:90% 01:16:58.845 --> 01:16:59.345 align:middle line:90% Yes? 01:16:59.345 --> 01:16:59.845 align:middle line:90% No? 01:16:59.845 --> 01:17:02.630 align:middle line:90% What are we going to do? 01:17:02.630 --> 01:17:05.210 align:middle line:90% Dr. Wolff? 01:17:05.210 --> 01:17:05.793 align:middle line:90% AUDIENCE: Yes. 01:17:05.793 --> 01:17:06.835 align:middle line:90% ISAAC SAMUEL KOHANE: Yes. 01:17:06.835 --> 01:17:08.580 align:middle line:84% So there are people like who will do it. 01:17:08.580 --> 01:17:11.041 align:middle line:84% But, thankfully, we don't give them the opportunity. 01:17:11.041 --> 01:17:14.032 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:17:14.032 --> 01:17:15.740 align:middle line:84% ISAAC SAMUEL KOHANE: Because 512 Genomics 01:17:15.740 --> 01:17:16.700 align:middle line:84% would be incredibly happy with you 01:17:16.700 --> 01:17:18.140 align:middle line:90% because we'd be making a mint. 01:17:18.140 --> 01:17:21.200 align:middle line:84% But probably you want to look for more signs and symptoms 01:17:21.200 --> 01:17:23.330 align:middle line:84% to make you think that this is happening. 01:17:23.330 --> 01:17:27.980 align:middle line:90% 01:17:27.980 --> 01:17:32.990 align:middle line:84% So my take on it would be if there's a persistent drop 01:17:32.990 --> 01:17:36.650 align:middle line:90% in glucose, if the IV runs-- 01:17:36.650 --> 01:17:39.500 align:middle line:84% I should say less than twice the basal amines. 01:17:39.500 --> 01:17:41.120 align:middle line:84% Let me give you the full scenario. 01:17:41.120 --> 01:17:43.400 align:middle line:84% You have a kid that you're maintaining a blood 01:17:43.400 --> 01:17:46.040 align:middle line:84% sugar by giving twice the normal amount of glucose. 01:17:46.040 --> 01:17:49.280 align:middle line:84% And every time you try to pull it down, 01:17:49.280 --> 01:17:50.525 align:middle line:90% the kid gets hypoglycemic. 01:17:50.525 --> 01:17:51.650 align:middle line:90% And it doesn't happen once. 01:17:51.650 --> 01:17:54.398 align:middle line:84% It happens many times over the first few days. 01:17:54.398 --> 01:17:55.940 align:middle line:84% And so, basically, convinced yourself 01:17:55.940 --> 01:17:56.750 align:middle line:90% that there's something wrong. 01:17:56.750 --> 01:17:57.740 align:middle line:84% So you give yourself two or three days 01:17:57.740 --> 01:18:00.110 align:middle line:84% to really convince yourself there's something wrong. 01:18:00.110 --> 01:18:02.810 align:middle line:90% And when you do that, well-- 01:18:02.810 --> 01:18:05.150 align:middle line:84% I'm just telling you by experience-- probably 01:18:05.150 --> 01:18:07.250 align:middle line:84% you want to do that within the first week. 01:18:07.250 --> 01:18:08.255 align:middle line:90% So what would you order? 01:18:08.255 --> 01:18:12.350 align:middle line:90% 01:18:12.350 --> 01:18:13.890 align:middle line:90% AUDIENCE: Sequences that-- 01:18:13.890 --> 01:18:15.360 align:middle line:84% ISAAC SAMUEL KOHANE: By the way, don't feel bad about it. 01:18:15.360 --> 01:18:17.943 align:middle line:84% If I asked any group of medical students or endocrine Fellows, 01:18:17.943 --> 01:18:19.282 align:middle line:90% they would not know the answer. 01:18:19.282 --> 01:18:20.490 align:middle line:90% AUDIENCE: Sequence the genes. 01:18:20.490 --> 01:18:22.990 align:middle line:90% ISAAC SAMUEL KOHANE: Which gene? 01:18:22.990 --> 01:18:23.842 align:middle line:90% AUDIENCE: Both. 01:18:23.842 --> 01:18:25.092 align:middle line:90% ISAAC SAMUEL KOHANE: Both, OK. 01:18:25.092 --> 01:18:27.750 align:middle line:90% 01:18:27.750 --> 01:18:28.800 align:middle line:90% Which part of the gene? 01:18:28.800 --> 01:18:32.030 align:middle line:90% 01:18:32.030 --> 01:18:33.550 align:middle line:90% Remember that picture of SUR. 01:18:33.550 --> 01:18:35.300 align:middle line:84% I thought we saw things everywhere, right? 01:18:35.300 --> 01:18:35.925 align:middle line:90% AUDIENCE: Yeah. 01:18:35.925 --> 01:18:37.242 align:middle line:90% [INAUDIBLE] 01:18:37.242 --> 01:18:38.825 align:middle line:84% ISAAC SAMUEL KOHANE: From what sample? 01:18:38.825 --> 01:18:42.990 align:middle line:90% 01:18:42.990 --> 01:18:44.858 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:18:44.858 --> 01:18:46.150 align:middle line:90% ISAAC SAMUEL KOHANE: You could. 01:18:46.150 --> 01:18:48.340 align:middle line:84% And eventually, three months out, if 01:18:48.340 --> 01:18:49.570 align:middle line:90% we have no better story. 01:18:49.570 --> 01:18:52.428 align:middle line:84% But parents are going to come wailing at you saying-- 01:18:52.428 --> 01:18:53.845 align:middle line:84% AUDIENCE: Presumably-- [INAUDIBLE] 01:18:53.845 --> 01:18:56.630 align:middle line:84% ISAAC SAMUEL KOHANE: No, but maybe it's not. 01:18:56.630 --> 01:18:59.770 align:middle line:84% I mean, I'm telling you that there's this-- 01:18:59.770 --> 01:19:03.050 align:middle line:84% these parents are wailing on you and saying-- 01:19:03.050 --> 01:19:04.690 align:middle line:90% and you're saying three months. 01:19:04.690 --> 01:19:06.460 align:middle line:90% I thought I came to Harvard. 01:19:06.460 --> 01:19:08.540 align:middle line:84% Don't you guys have genetic testing? 01:19:08.540 --> 01:19:09.910 align:middle line:90% So what are you going to do? 01:19:09.910 --> 01:19:13.570 align:middle line:84% AUDIENCE: Probably just stick a needle in your baby [INAUDIBLE] 01:19:13.570 --> 01:19:15.190 align:middle line:90% ISAAC SAMUEL KOHANE: You brute. 01:19:15.190 --> 01:19:16.780 align:middle line:90% So here's what I would do. 01:19:16.780 --> 01:19:19.330 align:middle line:90% 01:19:19.330 --> 01:19:22.598 align:middle line:84% If you ever come certainly to pancreatic removal, which 01:19:22.598 --> 01:19:23.890 align:middle line:90% we will have to, in some cases. 01:19:23.890 --> 01:19:26.120 align:middle line:84% Definitely won't want to look at the mutation tissue. 01:19:26.120 --> 01:19:29.050 align:middle line:84% But what I'll do is I'll actually look at the parents 01:19:29.050 --> 01:19:34.190 align:middle line:84% and see if a father or mother also have the mutation. 01:19:34.190 --> 01:19:39.560 align:middle line:84% So if the father, for instance, has the mutation in one gene-- 01:19:39.560 --> 01:19:43.280 align:middle line:84% if he's heterozygous for the mutation-- 01:19:43.280 --> 01:19:46.990 align:middle line:84% I have a very high index of suspicion. 01:19:46.990 --> 01:19:49.090 align:middle line:84% If the baby is homozygous, obviously-- 01:19:49.090 --> 01:19:54.810 align:middle line:84% so I look at baby blood and look at dad blood 01:19:54.810 --> 01:19:56.530 align:middle line:90% is what I would do. 01:19:56.530 --> 01:19:58.150 align:middle line:84% I'm pointing out it's not obvious. 01:19:58.150 --> 01:19:59.940 align:middle line:84% And every story is slightly different. 01:19:59.940 --> 01:20:02.530 align:middle line:84% And that's why this whole area of genomic medicine 01:20:02.530 --> 01:20:05.620 align:middle line:84% is incredibly fraught with a knowledge management problem. 01:20:05.620 --> 01:20:08.360 align:middle line:90% 01:20:08.360 --> 01:20:10.780 align:middle line:84% AUDIENCE: This may be a really stupid 01:20:10.780 --> 01:20:12.130 align:middle line:90% and [INAUDIBLE] question. 01:20:12.130 --> 01:20:14.170 align:middle line:84% ISAAC SAMUEL KOHANE: No, no such thing. 01:20:14.170 --> 01:20:16.050 align:middle line:84% AUDIENCE: Since, in this particular example, 01:20:16.050 --> 01:20:24.310 align:middle line:84% you [INAUDIBLE] possibly inherited gene, 01:20:24.310 --> 01:20:26.680 align:middle line:84% why wouldn't you try to build up-- 01:20:26.680 --> 01:20:29.800 align:middle line:90% 01:20:29.800 --> 01:20:39.190 align:middle line:84% get a sufficient cell sample, extract an array, [INAUDIBLE] 01:20:39.190 --> 01:20:41.650 align:middle line:84% and just basically test to see which one 01:20:41.650 --> 01:20:45.103 align:middle line:90% those guys [INAUDIBLE] 01:20:45.103 --> 01:20:47.020 align:middle line:84% ISAAC SAMUEL KOHANE: Are we looking expression 01:20:47.020 --> 01:20:49.070 align:middle line:90% or are we looking at DNA? 01:20:49.070 --> 01:20:52.605 align:middle line:84% AUDIENCE: Well, we've been working on expression. 01:20:52.605 --> 01:20:54.730 align:middle line:84% ISAAC SAMUEL KOHANE: So those were point mutations. 01:20:54.730 --> 01:20:55.855 align:middle line:90% Those were point mutations. 01:20:55.855 --> 01:20:58.112 align:middle line:90% Some of them were deletions. 01:20:58.112 --> 01:21:00.070 align:middle line:84% But most of them-- in fact, every one I showed, 01:21:00.070 --> 01:21:02.440 align:middle line:84% there was a microdeletion, three or four bases gone. 01:21:02.440 --> 01:21:05.738 align:middle line:84% Every array that I showed you would actually 01:21:05.738 --> 01:21:07.030 align:middle line:90% still say the gene was present. 01:21:07.030 --> 01:21:10.190 align:middle line:90% 01:21:10.190 --> 01:21:16.280 align:middle line:84% AUDIENCE: But if it was in one of the [INAUDIBLE] 01:21:16.280 --> 01:21:18.800 align:middle line:84% ISAAC SAMUEL KOHANE: No, that just changes the splicing. 01:21:18.800 --> 01:21:20.600 align:middle line:84% Or it's causing early termination 01:21:20.600 --> 01:21:21.530 align:middle line:90% of the gene product. 01:21:21.530 --> 01:21:24.290 align:middle line:90% 01:21:24.290 --> 01:21:26.520 align:middle line:84% Let's say for the sake of argument-- 01:21:26.520 --> 01:21:28.214 align:middle line:90% let's go back. 01:21:28.214 --> 01:21:31.110 align:middle line:84% AUDIENCE: Yeah, maybe I just [INAUDIBLE] 01:21:31.110 --> 01:21:35.510 align:middle line:84% ISAAC SAMUEL KOHANE: Let's say that it's here. 01:21:35.510 --> 01:21:39.710 align:middle line:84% And let's say that you're missing this. 01:21:39.710 --> 01:21:42.170 align:middle line:84% That you don't, in fact, have it spliced correctly. 01:21:42.170 --> 01:21:45.500 align:middle line:84% And that leads to continued translation 01:21:45.500 --> 01:21:47.480 align:middle line:90% of the intron as an exon. 01:21:47.480 --> 01:21:49.370 align:middle line:84% And, therefore, since it's mumbo jumbo, 01:21:49.370 --> 01:21:51.560 align:middle line:90% it causes premature termination. 01:21:51.560 --> 01:21:56.360 align:middle line:84% You still have 90% of the gene there that's being transcribed. 01:21:56.360 --> 01:21:59.180 align:middle line:84% And that RNA, therefore, is going 01:21:59.180 --> 01:22:03.560 align:middle line:84% to be registered by the expression array. 01:22:03.560 --> 01:22:06.185 align:middle line:84% Now you could say maybe I want to have a resequencing array. 01:22:06.185 --> 01:22:09.350 align:middle line:90% 01:22:09.350 --> 01:22:12.620 align:middle line:84% We still have to resequence the gene, no way around that. 01:22:12.620 --> 01:22:19.280 align:middle line:90% 01:22:19.280 --> 01:22:20.700 align:middle line:90% OK, where would you send it? 01:22:20.700 --> 01:22:21.690 align:middle line:90% We know the answer. 01:22:21.690 --> 01:22:24.690 align:middle line:84% We'd look up gene clinics, gene tests. 01:22:24.690 --> 01:22:26.670 align:middle line:84% What would you tell the parents when you're 01:22:26.670 --> 01:22:27.840 align:middle line:90% obtaining of the consent? 01:22:27.840 --> 01:22:30.180 align:middle line:90% Well, I'm running out of time. 01:22:30.180 --> 01:22:32.160 align:middle line:84% But I think in addition to all the things 01:22:32.160 --> 01:22:35.940 align:middle line:84% that we said before, we actually have to tell them that we 01:22:35.940 --> 01:22:37.980 align:middle line:90% might not find the cause of it. 01:22:37.980 --> 01:22:41.610 align:middle line:84% But that if the father's shows up one thing, 01:22:41.610 --> 01:22:44.320 align:middle line:84% that might be one result. And therefore, we 01:22:44.320 --> 01:22:46.320 align:middle line:84% may only have to hack out a bit of the pancreas. 01:22:46.320 --> 01:22:48.600 align:middle line:84% If the kid's homozygous for it, we 01:22:48.600 --> 01:22:50.740 align:middle line:84% may have to take out the whole pancreas and so on. 01:22:50.740 --> 01:22:53.580 align:middle line:90% So that's it for today. 01:22:53.580 --> 01:22:57.030 align:middle line:84% That's really the state-of-the-art of genomic 01:22:57.030 --> 01:22:59.390 align:middle line:90% medicine, 2004. 01:22:59.390 --> 01:23:00.000 align:middle line:90%