Last updated: April 01, 2006
Promoting Safe and Effective Genetic Testing in the United States - Appendix 3
Promoting Safe and Effective Genetic Testing in the United States
Appendix 3
State of the Art of Genetic Testing in the United States: Survey of Biotechnology Companies and Nonprofit Clinical Laboratories and Interviews of Selected Organizations*
Neil A. Holtzman** and Stephen Hilgartner***
The survey and follow-up interviews described here were conducted to provide the Task Force on Genetic Testing with three types of information: the extent of genetic testing in the United States; the policies and practices of organizations engaged in such testing, and the opinions of officials of the organizations contacted concerning matters related to genetic testing.
For our survey, we defined genetic tests as "the analysis of human DNA, RNA, chromosomes, proteins, or other gene products to detect disease-related genotypes, mutations, or phenotypes, or karyotypes for clinical purposes. Such purposes include prediction of disease risks, identification of carriers, monitoring, diagnosis or prognosis, and establishing genetic identity, but do not include tests conducted purely for research."
Methods
Survey Populations
A list of American biotechnology companies and molecular genetics and cytogenetics laboratories operated by nonprofit organizations was compiled from the Institute for Biotechnology Information (IBI) database, the Helix National Directory of DNA Diagnostic Laboratories, and the Association of Cytogenetics Technologists (ACT) International Cytogenetic Laboratory Directory. A few organizations not in these databases but known to us were added. Descriptors in the IBI and ACT databases enabled us to limit our mailing to organizations that were likely to be developing or providing new genetic test technologies.****
Survey Instrument
The questionnaire consisted of 26 questions, some with several parts, and most in multiple choice format. The major categories covered are shown in Table 1. Respondents' personal agreement or disagreement with six statements was sought, using a four point Likert scale.
The questionnaire was pilot tested with five officials from university-based clinical laboratories and biotechnology companies engaged in genetic testing activities and amended accordingly.
Survey
The questionnaire was initially mailed in December 1994 and January 1995. If we received no reply in a month despite reminders, we sent a short questionnaire, which consisted of four questions from the long questionnaire on genetic testing activities. If the organization did not return the short questionnaire, we telephoned the organization to collect the information. For cytogenetic laboratories we did not use a phone call followup.
Interview Sample
Organizations were selected for an in-depth interview from among those who completed the long version of the survey. We selected ten of them because we were aware of their activities and judged them to be significant players in genetic testing. Selection of the other companies was based on the following criteria: The company had to indicate on the survey that it:
- was either developing or providing tests for at least one of several complex disorders (Alzheimer?s, breast cancer, colon cancer, diabetes, melanoma) or single-gene disorders (cystic fibrosis, fragile X, Huntington's disease, Marfan syndrome, muscular dystrophies, or neurofibromatosis I or II), and
- devoted more than 10 percent of its financial resources to either research and development (R& D) and/or marketing of genetic tests and expected more than 30 percent of its revenue over the next five years to be derived from genetic testing, or
- expected revenue of $1 million and a biotechnology R&D budget of more than $2 million (current fiscal year), or
- expected revenue of more than $2 million with more than 5 percent of that revenue deriving from biotechnology activities.
(Data for 3 and 4 were obtained from the IBI data base.)
Twelve companies met criteria 1 and 2 and an additional six met criteria 1 plus 3 and/or 4. Five of the ten firms that we selected on the basis of our personal knowledge of their activities met criterion 1, and of these, two met both 2 and 3 and/or 4, two met 2 only and one met 3 or 4 only.*
A letter requesting an interview was sent to the company official who completed the questionnaire.** We also conducted interviews at five laboratories engaged in genetic testing at not-for-profit academic centers or managed care organizations. These labs were selected if they were in a geographic area that we planned to visit to interview companies, and if the interview could be scheduled at the time of the visit.
Beyond the companies studied, we make no claims about the quantitative frequency of the activities documented below. Some of the results presented here are intended to provide a qualitative sense of current activities in genetic testing that we could not probe in the survey. Others corroborate and extend findings of the survey.
The Interview
The interview covered four major topics: the history of the organization's efforts in genetic testing; its specific R&D programs and projects; its plans and policies regarding R&D, marketing, quality control, and regulatory issues; and the organization's policies and the respondent's views toward regulatory, policy, and ethical issues in genetic testing. All respondents were promised anonymity. Explicit refusals to answer particular questions occurred infrequently and were generally confined to certain lines of inquiry, such as queries about the volume of testing a company performs, which some respondents considered particularly sensitive.
Response Rates
Of the 594 biotechnology companies (BTCs) who were mailed questionnaires, 194 (32.7 percent) returned the long questionnaire and 267 (44.9 percent) returned the short questionnaire for a total response rate of 77.6 percent. Of 425 nonprofit organizations (NPOs) surveyed, 273 (64.2 percent) returned the long questionnaire and 80 (18.8 percent) returned the short questionnaire for a total response rate of 83 percent. Although the proportions of long and short questionnaires returned by molecular and cytogenetic NPOs did not differ, the response rate of molecular NPOs was higher than of cytogenetic NPOs (92 percent versus 75 percent). Except when these two classes of NPOs differ, the results are pooled.
Of the 28 companies selected, interviews were conducted at 25 between June 1995 and January 1997.* In 16 cases, one of us visited the company and conducted the interviews in person; in 8 cases, interviews were conducted by telephone. Most interviews lasted between 1 and 1.5 hours.** All five of the nonprofit organizations agreed to be interviewed. These interviews were all conducted in person.
Results of Survey
Companies Not Developing Tests
Twenty-nine BTCs who returned the long questionnaire said they had considered developing genetic test products or services, but decided not to. The following eight reasons were selected in descending order: not our area of expertise (9); regulatory hurdles (7); we lack a unique product angle (5); high costs (4); controversial area (4); patent issues (3); too competitive (2); and limited demand (2). Companies could list more than one reason. Four BTCs gave other reasons, three of which involved lack of complete technology.
Genetic Testing Activities
Almost nine-tenths of NPOs (316) and one-third of responding BTCs (147) were engaged in genetic testing activity. Table 2 provides a breakdown from the responses to both the long and short questionnaires. As virtually all of the NPOs are clinical laboratories, it is not surprising that they are predominantly engaged in service activities. Fifty-eight BTCs were developing or providing genetic tests (39.5 percent) and 89.5 (60.5 percent) were engaged in related activities.
The remaining analyses are limited to those organizations who completed the long questionnaire.
Among the 186 NPOs offering genetic testing services, 103 (55.3 percent) use tests developed in-house (home brews). Of the 23 BTCs providing such services, 11 (47.8 percent) use home brews.
Fifty-three BTCs and 212 NPOs reported developing or offering genetic tests for at least 1 of the 44 disorders listed in the questionnaire. These included three common complex disorders (Alzheimer's disease, breast cancer, hereditary nonpolyposis colon cancer (HNPCC)) and three of the most frequent single-gene disorders (cystic fibrosis, fragile X, muscular dystrophy). A significantly higher proportion of these 53 BTCs were developing or offering tests for the 3 complex disorders (64 percent) than for the 3 single gene disorders (47 percent). Only 22 percent of NPOs were developing or offering tests for the complex disorders (Table 3).
Among the 34 BTCs developing or offering tests for complex disorders, 13 (38 percent) were targeting particular populations, compared to only 3 of 47 NPOs (6 percent). Prenatal diagnosis was checked as an intended use for the three complex disorders by four BTCs (12 percent) and three NPOs (6 percent). Three companies said testing in children was an intended use for tests for breast cancer or HNPCC.
Patenting and Licensing
Not surprisingly, BTCs were almost four times more likely to hold patents, have patents pending, or say they would file patents than NPOs ((62 percent versus 16 percent); p < 0.0001). Among the 25 BTCs that were offering genetic test services, 14 (56 percent) had or expected to have licensing agreements with other organizations, as did 40 of 120 NPO (33 percent) molecular laboratories, and 13 of 111 cytogenetics laboratories (12 percent). The BTCs offering genetic test services were more likely to have licensing arrangements with academic institutions than with other companies (13 versus 9 BTCs), whereas the NPOs were more likely to have licensing arrangements with other companies (43 versus 12).
Assessment and External Review of New Tests
We were interested in how often organizations developing new tests assessed their reliability and/or validity prior to making them routinely available. Thirty-five of the 43 BTCs and 179 of the 223 NPOs who answered this question said they were involved in such assessments. Organizations conducting these assessments might have done so under an Institutional Review Board (IRB) protocol or under an investigational device exemption (IDE) from the Food and Drug Administration (FDA). However, of the 43 BTCs only 23 (53.5 percent) had ever submitted, or said they were likely to submit, protocols for any aspect of test development to an IRB and only 13 (30.2 percent) had ever contacted the FDA for reasons related to genetic test development. Among the 215 NPOs who answered this question, 117 (54.4 percent) had ever submitted or planned to submit to an IRB, but only 17 (7.9 percent) had ever contacted the FDA.
Table 4 displays the data on IRB submission and FDA contact by whether the organization was developing, currently offering, or both developing and offering tests. It was only among those BTCs and NPOs that were both developing and currently offering genetic tests that more than half submitted a protocol to an IRB. Organizations who were only developing genetic tests may have been at a very preliminary stage in their test development, one at which they would not be expected to go to either an IRB or the FDA. Organizations who were just offering but not developing may have acquired the tests from other organizations and not have needed to submit to an IRB or contact the FDA. In an effort to adjust for these possibilities, we examined IRB and FDA contact among the 14 BTCs and 95 NPOs who reported preparing their own probes, primers, enzymes or special chemicals for the genetic tests they currently offered. About three-fourths of both the BTCs and NPOs who had developed such "home brews" had either submitted to an IRB or contacted the FDA.
One reason that so few organizations had contacted the FDA may have been because most genetic tests being developed or marketed (by both BTCs and NPOs) are planned or offered as services rather than as tangible products, such as kits.* We, therefore, looked at FDA contact only for those organizations that reported developing or offering genetic test kits. Of the 23 BTCs developing genetic test kits and 1 marketing a kit, only 1 (4 percent) had obtained an investigational device exemption (IDE) from FDA and 6 (25 percent) had filed a premarket notification (510k). Only three NPOs were developing or marketing genetic test kits and only one had obtained an IDE. None of these BTCs or NPOs had filed a premarket approval application. Six BTCs and one NPO had other communication with FDA regarding genetic testing.
We also examined the reported external review of those organizations developing or offering tests for the three complex disorders. At the time of the survey all three were still regarded as investigational by several professional societies. Eighteen of the 81 organizations (22.2 percent) developing or already offering tests for these disorders had not filed nor do they plan to file with an IRB and 62 (76.5 percent) had not contacted the FDA. However, of the 11 organizations that are already providing tests, 10 had submitted to an IRB, and 2 had contact with FDA. We cannot say, however, whether the submissions or contacts were for testing for these three disorders.
Laboratory Quality
Table 5 describes some quality assurance activities of the respondent organizations that were offering genetic test services. Most were registered under the Clinical Laboratory Improvement Amendment 1988 (CLIA). At the time of the survey and to the present, no proficiency testing program is required of CLIA-certified laboratories. The College of American Pathologists and the American College of Medical Genetics jointly administer voluntary proficiency testing programs for several types of genetic tests.
As shown in Table 5, only 3 BTCs (11.1 percent) but 16 NPO molecular laboratories (16.5 percent) fail to participate in either a formal proficiency testing or an informal sharing of unknown samples.
Targeting of Tests
Most organizations currently offering genetic tests aim their marketing at geneticists, genetic counselors and nongenetics medical specialists (Table 5). Significantly more BTCs than NPOs target managed care organizations. Many organizations also target primary care physicians but few market directly to patients or consumers.
Testing of Minors
Of laboratories offering genetic testing services who answered the question, 7 BTCs (28.0 percent) had restrictions on performing genetic tests for carrier status or adult-onset diseases in minors compared to 54 NPO molecular laboratories (57 percent) and 34 NPO cytogenetic laboratories (44 percent; p < 02).
Opinions of Individual Respondents
We asked respondents to indicate their personal agreement or disagreement with each of six statements on genetic testing (Table 6). On two statements, BTC and NPO personnel differed. A significantly higher percentage of BTC (20 percent) than NPO (6 percent) respondents personally agreed with the statement that "Most physicians can interpret genetic tests adequately...." A significantly higher percentage of BTC (53 percent) than NPO (39 percent) respondents thought that current CLIA policies "assure the quality of genetic test services." There was widespread agreement with the statement that "Some laboratories that offer genetic testing lack quality assurance programs" (84 percent). Seventy-three percent agreed with the statement that "FDA policies, or lack of policies, hinder the development of safe and effective genetic test kits or other products." There was considerable agreement that an industry-wide code would improve genetic services (85 percent of all respondents), and that additional laws or regulations are needed to ensure the privacy of genetic information (70 percent).
Results of Interviews
For analytic purposes, the organizations interviewed can be divided into three categories: providers of testing services, including five not-for-profit laboratories (n=15); developers of testing technology (n=9), who are creating or improving generic DNA technologies that are applicable to genetic testing services or kits; and gene discovery firms (n=2), who are working to identify new genes involved in human disease with the ultimate goal of using these genes as the basis of therapeutic and diagnostic products. The research and policies of these gene discovery companies, which emerged only in the 1990s, are beginning to play an important role in the arena of genetic testing. Three other companies proved, on interview, to be in adjacent markets and were not directly involved in genetic testing.
The Market for Genetic Tests
The companies engaged in testing activities operate in an extremely dynamic environment, frequently undergoing restructuring, forming new partnerships, embarking on new initiatives, or dropping projects. For example, one company, which had achieved some success in a generic DNA testing technology, partnered off the product and changed its research and development efforts.
There was considerable agreement that although genetic testing would grow progressively more important over the next decade or so, the largest market for DNA testing would always be in infectious diseases. Cancer testing, tissue typing, and personal identification (parentage, forensics) were also expected to continue to command a larger market than genetic testing. A number of respondents said that growth in genetic testing had been slower than they had anticipated in the late 1980s.
How companies answered our questions about expected growth seemed to be related to their situation in the market. Thus, to a large corporation developing testing technology the market has not "grown much at all." All that is underway is "specialty testing" and the "big players" in diagnostics "have not gotten into it."(v, 13)* In contrast, to a small company aiming to market more "esoteric" tests to a medical specialty such as oncology or neurology, the market already looks significant and poised for further expansion.
One small startup technology development company, founded in the late 1980s, originally expected its business to center around genetic testing. It ended up shifting to other applications, such as parentage in both humans and purebred animals, because raising money to develop genetic tests was too difficult. Investor confidence was reduced by regulatory uncertainties, the absence of a "well worn path to commercialization," and the realization that testing for single gene disorders, such as CF, that many had initially expected to be easy, turned out to be hard.
The commercial sector is not suffering because of competition from testing in academic centers, although genetic and cytogenetic testing were, for many years, largely in their province. Several company respondents reported that many medical school labs that perform genetic tests are losing money and require subsidies from their respective universities. This was consistent with the statements of the directors of university-based labs we visited, who all reported that their labs did not break even on genetic tests. The small scale of genetic testing sometimes made it difficult for them to get additional resources from the large parent laboratories at the medical centers.
Factors Affecting Growth of Genetic Testing
Technological immaturity.Both the providers of testing services and the companies developing testing technologies saw the development of cheap, high-throughput testing systems as critically important to DNA testing in general and to genetic testing in particular. Many people's "wish lists" seemed to feature technological platforms that would allow one to test for large numbers of mutations at a reasonable cost. One scientist, active in test development, argued that the lack of such systems was a major reason that genetic testing remained an "expensive" and "underutilized" technology.
If there are 400 different mutations in the gene, ... [that disease is] not amenable to ... very inexpensive techniques. ... Those techniques, on the other hand, that give you all of the information that a gene contains?sequencing at the ultimate extreme?are intrinsically low throughput, costly, and time-consuming. That's the dilemma. ... The techniques that give you all the information are not suited for the clinical laboratory, either in terms of format, cost, throughput, turnaround time. Those that have more the feel of a clinical laboratory test in terms of turnaround time, throughput and cost, don't give you all the information you want and sometimes don't even give you enough information. (iv, 19)
A multinational company, successfully selling DNA-based tests for several infectious diseases in the United States, had developed a prototype test kit for a single-gene disorder and had placed the kit into a clinical trial in Europe with the goal of eventually marketing it to clinical laboratories as a replacement to home brew tests in the U.S. However, the European laboratories preferred their own home brews over the company's standardized product because they could flexibly adapt their home brew tests to the mutations prevalent in their respective countries. Accordingly, the company shifted the project back into the research phase, with the goal of eventually offering a test, perhaps based on a new testing technology, that would cover a much larger number of mutations.
Because of rapid change, companies that were primarily offering services did not want to invest a lot in new technologies. Companies were more interested in acquiring new test technologies from academe and altering them to get the test to perform in a reliable and cost-effective way at higher volume. An official at one large clinical lab explained:
We made the decision that we could not invent the tests and so we went to established academic centers in which the tests were up and running, validated against clinical specimens and had a track record, and we negotiated licensing agreements to basically buy the technology. When we got it, we made some changes to make it work better in our hands, to satisfy ourselves that it behaved the way we thought it should. ...(i, 8)
Clinical considerations. Genetic tests for inherited disease and predispositions need only be conducted once per patient, placing an upper limit on the market size, whereas people are tested repeatedly throughout their lifetimes for infectious diseases. Similarly, in the oncology context, many envision the use of DNA testing for ongoing monitoring of cancer patients, which would require running repeated assays. For this reason, some firms involved in DNA testing were devoting few resources to genetic testing.
A recurrent theme was the problem of test sensitivity, the ability of the test to detect all disease-causing mutations. A company offering genetic testing services, said:
It's rarely just one mutation equals one genetic problem. So the problem [is] having really good screening tests. And what makes a good screening test? Are you happy with 70 percent [sensitivity]? Does it need to be 85 to 90 percent accurate? ... CF is pretty good. I mean, 70 percent with one [mutation] and...[with] a panel of up to 20 more you can get 85 percent. But you are never going to get 100 percent. ... Then we went into fragile X and triple repeats because that essentially is something [where] ... one test is going to give you your answer....There really weren't a lot of genetic diseases that we could screen cost effectively and produce effectively, just because there's so many mutations that are involved. (x, 7)
Clinical utility is not always evident in testing for inherited disorders for which treatments have not yet been developed.
If you are diagnosing something in human genetics that there's no treatment for, and no one's quite sure what the outcome is because you're looking at predisposition to disease, no one's going to buy your test. (ii, 10)
Financial considerations. Because most single-gene disorders are rare, the demand for some genetic tests is likely to be very low. One provider of testing services commented that some of these tests were very low volume tests. You'd get one a year or one a month to do. So from a business perspective it was not cost effective to do, and you just have to drop it. (viii, 9)
Another way of attempting to deal with low volume tests for rare diseases is to send them out to "boutique" labs. But this is not profitable; by law, the sending lab can only charge what it is charged plus a small handling fee. The alternative, establishing a new test in house is also costly, so the company needs to determine: "Do we lose more setting it up or sending it out?" (xi, 8)
High costs also reduce demand. One provider of genetic testing services noted:
Patients don't want to pay $200 or $600 for a genetic test. So we want to find technology platforms that allow us to drive the costs down so that maybe we're talking $35, $50 for a test rather than several hundred. So we've got to be making investments in technology platforms and looking for ways to streamline processes. (i, 15)
The director of one university laboratory pointed out that the costs of clinical testing go far beyond the technology itself.
Somebody [has] to review the results and interpret the results and issue the reports, and all the phone calls back and forth, and stuff like that. ... It's just my feeling from personal experience that those are a significant part of our total cost.(xiv,33)
A number of companies consider reimbursement when they decide to develop a new test. One provider of services said,
If the test is going to cost us $1,000 and we get so few referrals we have to do them one at a time, and we're only going to get reimbursed $200, then I'm not going to do it.(xv, 10)
Another Commented:
We have a lot of sales reps out in the field talking to physicians and oncologists who say, "you know it would be nice if we had such and such a test." Eventually all of that filters back and we make some effort into investigating would it be possible for us to do. Would it fit? Is there a market? Is there a reimbursement for it? Is it potentially going to wind up making money? And then [we] go ahead and try and make a deal, bring the technology, offer it. (vii, 8)
In considering whether to offer a particular test, some respondents said they would consider how it fit with their existing offerings. A low volume test might be added even if it was not cost effective by itself if it "completed an offering" in a particular clinical area.
If you could offer the whole gamut [of tests for a number of related diseases] there might be a value, an added value to the [physician] ... because he is going down the differential [diagnosis]. ... So it's not just numbers driven; it's also how it completes sort of a profile or presentation.(ix, 16)
Ethical considerations. The controversial nature of testing for inherited disease or predispositions, and the ethical questions raised by predictive genetic testing, also seemed to discourage a focus on this area. One company with a generic technology hesitated to enter the genetic testing arena.
I mean, the minute you go into human inherited disease and predisposition, then there's ethical issues. So what we did is look at the three market segments - inherited disease, cancer, and infectious disease - and said from a business point of view, what are the largest markets today? Basically almost all the pie is in infectious disease. ... And there's no ethical issues and you're replacing the initial products by bringing in something that is easier to use and cheaper and is just better for the health care system. So our primary market focus is on infectious disease for that very reason. (iii,5)
A biotechnology startup company founded in the late 1980s, was more specific. This company had developed technology for detecting mutations that it expected to apply to some of the more common Mendelian diseases, such as cystic fibrosis and neurofibromatosis.
CF was going to be a very large focus for the company. ... The management and the board...were for a time much more interested when they thought that population screening was going to be a reality. Now I guess our opinion is that it won't be ever, so that means the market is much smaller. (vii, 8)
Like the other startup discussed above, this company has redirected its efforts, in this case toward somatic cell testing and gene therapy.
Regulatory considerations. Regulatory issues loom large, particularly to small startups interested in developing kits or reagents that clinical labs could use to perform tests. Regulatory uncertainty in the case of genetic tests stemmed in part from the novelty of the tests. One small test technology developer told us:
In genetic diagnostics there never was a reliable test for a lot of those diseases. ... So you could never test for those things before. So no one was clear on how it would ever be regulated and what level of sensitivity you would need.(ii, 9)
The company thus shifted its strategy away from genetic testing, placing highest priority on parentage testing, followed by infectious disease, cancer, and genetic testing in that order.
Another technology development company was using its technology to test livestock for inherited susceptibility to disease?an application it chose in part because it is at present unregulated by government agencies (vi).
One technology development company experienced longer regulatory delays than it had previously when it sought FDA approval for an infectious disease test using new DNA technology. Despite evidence provided by the company, FDA was skeptical that the sensitivity of the test, which had previously been used exclusively in research, was as good a clinical test as the company maintained.
Laboratory Quality Assurance
All of the test providers seemed confident that their own efforts to maintain quality were effective, appropriate, and in compliance with the relevant laws. Several respondents said they believed that most laboratories did good work, but also suggested that laboratory quality was uneven and that some companies and medical center laboratories were not performing adequate work. One official at a technology development company who had been active in a proficiency testing program said:
We had a couple of places that really didn't do all that well [in proficiency tests]. They shouldn't have been testing for certain things that they were. Some of these people were people who now have the diploma - they're clinical molecular geneticists - but in the proficiency testing they didn't score all that well. ... A lot of places are good, but some places are not so good. All of them give diagnoses. [laughs] (f, 37)*
Another respondent stated that he had seen poor results used to make important decisions and argued that physicians often had little choice but to rely on word-of-mouth assessments of laboratory quality:
"Well, did you send it to such and such a laboratory?" "No, we sent it over there." "Well, you know so and so's laboratory is a better shop." And I think that that's the way, I think that's the status of genetic testing in the United States right now. There are some laboratories that are considered very good and there are some that are considered not so good. (g, 14ff)
As the above discussion suggests, there were differences in attitudes about the extent to which uneven laboratory quality was a serious problem.
Most providers of testing services were wary of additional FDA involvement in regulating home brew testing. According to some respondents, regulation of home brew testing would reduce the availability of testing for rare diseases, would slow innovation, and would lead tests to be underutilized. Several respondents argued that the procedures in place at most clinical laboratories for maintaining quality were by and large adequate and that regulation "would delay work and would cost a lot of money and really wouldn't do that much for quality." (b, 37)
Institutional Review Boards (IRBs)
IRB review occurred at a variety of organizations, including companies that provide testing services, technology development companies, and medical center laboratories. At the companies, officials typically said that they relied on the IRBs of collaborating universities to review their research protocols. One company, however, described how it had established its own IRB using external advisors to provide review of its protocols for conducting studies of the validity of tests.
Companies that acquire the rights to tests developed and validated elsewhere, and who then modify the protocols to make them work in their hands or on a large scale, often do not classify this as research which would be subject to IRB review. However, respondents described several other forms of internal review. Some of the largest organizations had formal procedures or committees for reviewing new test protocols to determine if they were ready to be offered as a clinical service. At some smaller organizations, the senior personnel collectively or, in some cases individually, made such assessments.
The Gatekeeping Function of Laboratories
Interviewees agreed that genetic counseling and informed consent were essential aspects of medical genetics, but they disagreed about the role of the testing laboratory in performing and verifying that counseling and consent had occurred. One source of the differences in opinion seemed to be the different professional practices that have evolved in the fields of laboratory medicine and pathology, on the one hand, and human genetics, on the other. The director of a medical center clinical laboratory who had a laboratory medicine orientation explained that his laboratory's "relationship is with the physician, or whoever is ordering the test" and that "the question of obtaining informed consent for testing is the responsibility of the person who orders the test and sees the patient." (b,31)
In contrast, the director of another medical center laboratory who had a human genetics orientation, suggested that laboratory medicine and pathology should (and perhaps are beginning to) shift more toward the approach of geneticists. The main differences in orientation, she argued, concerned counseling and informed consent, for which the geneticists believed that the lab should take a more active role.
I think that, you know, testing labs have somewhat of a responsibility to track down and try to get the referrals done appropriately for the benefit of the family. For example, did [the family] know the implications of [the test], that if the child turns out to be positive, they won't get disability insurance or life insurance and will be marked for many years to come and all those implications.... I think it's really important that the family be steered into genetic counseling and that the referring physician is educated in some way that there are other critical factors than just sending the blood to the lab and getting an answer back. (h, 9)
Differences of opinion on this matter were also found among commercial clinical laboratories. One official explained:
For all of our tests we have policies about who will consent to tests.... And on top of that we are laying on a required informed consent that the patient has to fill out with their physician so that they know what we are about. (e, 17ff)
Another company had developed ethical guidelines:
For accepting samples ... we won?t do testing if we know ... that it?s just for sex selection. ...The same can be true for testing minors, [and] testing individuals who are at risk for different diseases that may not have had informed consent. (i, 30)
In contrast, several other biotechnology companies argued that the role of the laboratory should be limited to assuring the reliability of the results it reports.
I've thought a lot about this, that the way to think about this is about physician autonomy. A physician wants a piece of information to judge his patient on. What the labs ought to be required to do is to make sure that they provide the physician with the information that the physician...asked for. ...The lab ought not to be in the business of practicing medicine. (j, 23)
An official at another company contended that there was insufficient justification for treating DNA tests differently, and that to require laboratories to act as gatekeepers represents "an inappropriate placement of the gate" (k, 26ff).
Some providers of testing services employed genetic counselors and some did not. Officials at a testing company that did not, and which serves a national market, argued genetic counseling is part of "hands-on patient care," not laboratory work, and that it therefore should be part of the local care that is offered to the patient by the physician. This company always provided physicians who ordered tests with a thorough interpretation of the result (c, 43). Another company, one of the largest players in the national clinical laboratory market, operated a counseling unit that included two genetic counselors and a number of less highly trained "case managers." This firm required use of its own consent form for some but not all genetic tests (a, 7ff).
Closely related to the question of the extent to which genetic testing laboratories should serve as gatekeepers was the question of the extent to which regulatory authorities or professional groups should be able to restrict the tests available to individual patients. An official at a large clinical laboratory argued:
I don't think a regulator should determine that a patient, if they're educated, should or shouldn't get tested. I think that should be the patient and the patient's family decision of whether to get tested, but not without education. I'm not saying anybody should be able to walk in and just say "I want it ?cause I want it." I think we do have an obligation to make sure they understand the test, what its limitations are and understand what a positive and a negative will each mean, but if they, after that kind of understanding, request that the test still be done, I think that should be their decision. (a, 12)
Interpretation of Test Results
Many respondents who provide testing services said that an important aspect of their work was giving physicians the information they need to understand the meaning of tests and interpret results properly. These educational efforts were needed, according to most respondents, because genetic testing represents a new area to most practicing physicians. At the large testing companies, the task of educating physicians entailed not only providing detailed reports on the meaning of each patient's results, but also producing and distributing a variety of educational materials for physicians. (Editor's note: See Appendix 4, Cho et al: Analysis of Informational Materials About Genetic Tests.) As one official put it: "[For] a lot of physicians ... this is really a very, very new area and they really need our help and will probably always need our help." (a, 8)
Another test provider argued that changes in the health care system were making it increasingly important for test providers to help physicians understand tests:
More and more front line medicine is done by family and general providers rather than subspecialists. So we have ... to make sure that the clients can get the information they need quickly when they call about what the test means, what tests need to be done, what other samples do they need. (c, 29)
Testing of Minors
The testing of minors for adult-onset disorders or carrier status was a matter about which there was some divergence of views. Some companies fully supported prohibitions on testing of minors for adult-onset disease or carrier status. In contrast, one company official argued that he did not believe that companies necessarily should be involved in deciding such matters as testing of minors. This company had nonetheless made a commitment to follow the consensus of the genetics community. A third perspective held that the question of whether testing of minors was appropriate required a case by case, disease by disease assessment by an experienced physician, suggesting that any guidelines he would support have to allow for considerable professional discretion. For example, one argued that in the case of familial polyposis knowing the test result might greatly benefit a child by eliminating the need for repeated colonoscopies. In other cases, where the test outcome would not have immediate clinical implications, testing might not be warranted until the age of majority.
Patents and Royalties
One policy area that respondents themselves repeatedly brought up was the question of the proper uses of intellectual property in genetic testing. In particular, a number of organizations, both companies and medical center laboratories, expressed concern that intellectual property policies would drive up the price of tests or otherwise adversely affect test availability. Several respondents worried that when patent holders grant companies exclusive licenses to genetic tests, this could limit access to necessary testing by limiting the number of laboratories and increasing the price. Respondents argued that the granting of numerous patents, each on a specific mutation in the same gene, could make testing unaffordable if each patent holder demanded a royalty. For example, an official at one testing company suggested that university technology transfer offices were filing large numbers of patents, with the result that providers of any effective genetic test might need to pay multiple, or "stacked" royalties, producing "a diagnostic test that no one can afford." (d, 40)
Discussion
In a survey conducted by one of us in 1986 only 118 biotechnology companies could be identified who were likely to be engaged in genetic testing, and of the 85 companies that replied, only 22 said they were developing or offering tests for genetic or chromosome disorders.In this survey, conducted ten years later, we identified five times as many companies likely to be involved in genetic testing. Among respondents, 6.7 times as many companies as in 1986?147 firms?were engaged in genetic testing activity. Clearly commercial interest has grown.
In 1986, it was virtually impossible to perform genetic tests for common, complex disorders. Although such testing is now possible only for a handful of disorders, it is genetic testing for the common disorders that has sparked the most commercial interest. Respondents to the earlier survey indicated that the disease characteristic of greatest importance in genetic test development was prevalence of the disease. The data ten years later (Table 3) bears this out. Although a few companies have developed tests for rare diseases, this has been mostly to gain a foothold in the field or to provide a complete battery of testing for the specialties on which they concentrate.
Despite this dramatic increase in activity, the field is far from maturity. Many more companies are engaged in research and development of genetic tests than are delivering genetic test kits or services (Table 2). Several companies perceive the market growing at a slower rate than they anticipated in the 1980s.
Scientific, clinical, technological, and ethical factors are all at work. Many diseases, including common complex ones such as breast and colon cancer, occur in the presence of any one of hundreds of different inherited mutations, or in the absence of any of them. Moreover, some people who have inherited mutations that contribute to disease occurrence in others will never get the disease over a normal lifetime. Current technology for providing tests for disorders in which multiple mutations contribute to disease causation is expensive and ponderous. Given these limitations, tests for common disorders will attract greater interest than tests for rare disorders. Several interviewees pointed to the precarious financial rewards of testing for rare diseases; they need tests for which the demand is likely to be high. This is much more probable for common disorders, even if the tests are imperfect. Further, unless they hold the patents for the key materials or processes, companies may be limited in testing or may also have to pay licensing fees on materials or processes to the patent holders. Several interviewees complained about the high costs of licensing, and a recent effort to collect royalties on the use of a patented reagent in prenatal screening bears this out.
Recent advances, particularly employing chip technology, bode well for much simpler and less expensive technology that will be capable of detecting hundreds if not thousands of different mutations simultaneously. The ability to detect protein gene products of disease-related genes may also provide a more sensitive and more functional test than direct searching for mutations at the DNA level.
From our interviews, we suspect that potential test developers are awaiting the arrival of faster throughput, less expensive, and more sensitive technologies that have a greater chance of being reimbursed by health insurers. Even then, the less-than-perfect predictability of genetic tests for several disorders, as well as the absence of interventions of proven effectiveness for many inherited diseases, may deter their entrance into the field.
Those engaged in genetic testing told us that ethical issues were a consideration. These included not only controversy over genetic testing but the role of the laboratory in assuring that patients were adequately informed of and consented to the test, and the testing of minors. Interviewees differed on how much responsibility the clinical laboratory had and how much belonged to physicians ordering tests. Concerns over ethical issues has probably contributed to relatively few organizations marketing directly to patients and other consumers, although a few do.
Many companies see other uses of recombinant DNA technology, particularly for diagnosis of infectious diseases, as a more certain market. Only 16 companies were devoting more than 10 percent of their resources to activities related to genetic testing.
The state of the technology has, in part, dictated the activities of genetic testing companies. Except when it is sufficient to test for a small number of mutations, current technology does not permit the tests to be assembled in kits and marketed to clinical laboratories. Consequently, biotechnology companies have gone into the clinical laboratory business themselves, marketing genetic testing services rather than kits. This has also proved to be less of a regulatory hurdle because FDA, although having the authority to regulate new testing services, has elected not to. One biotechnology company that is offering genetic testing states in its prospectus for common stock that "the company's genetic testing laboratory is regulated under CLIA, which imposes less complex regulatory guidelines than those required by the FDA."*,
The question arises of whether tests currently being developed and marketed have been subject to adequate external review. Under CLIA, clinical laboratories must show the analytic validity and reliability of the tests they perform, but not their clinical validity or utility. We were interested, therefore, in other means by which developers of test services subject new tests to review. This is all the more important for tests developed in-house (home brews) which are subject to no external review other than under CLIA. Approximately half of the BTCs and NPOs who offered genetic testing services reported using home brews. Although over 80 percent of commercial and university-based laboratories that were offering genetic services said their tests were assessed before being made available routinely, two mechanisms that might suggest formal review?submission to an IRB and communication with FDA?were not used by nearly that many. IRBs do not review data on test validity and utility, but submission of a protocol to an IRB indicates that the organization is systematically collecting data. Moreover, under regulations pursuant to the Medical Device Amendments to the Food, Drug and Cosmetic Act, organizations attempting to establish the clinical validity of an in vitro diagnostic device, such as a genetic test kit or specific probe, must operate under a protocol approved by an IRB and, when there is no independent means of confirming the test, must obtain an Investigational Device Exemption (IDE) from FDA. Only two respondents had obtained IDEs. The low use of IDEs probably reflects the intention of test developers to market their tests as services, which, as we have mentioned, FDA has elected not to regulate. If FDA were to regulate new devices whose developer intend to market them as services, they would probably not be deluged with applications, considering the relatively low rate of commercial development of genetic tests.
Once tests are marketed, they are subject to CLIA regulations. As CLIA does not have separate requirements for laboratories offering DNA-based genetic testing, a laboratory need only demonstrate general, good laboratory practices. Voluntary proficiency testing programs were available at the time of the survey, and informal sharing of unknown specimens to compare results is a longstanding clinical laboratory practice. Nevertheless, 11.1 percent of BTCs offering genetic testing services and 16.5 percent of NPO molecular laboratories did not participate in either formal or informal proficiency testing programs. In the interviews, several commercial laboratory directors told us they were aware of poor quality laboratories who were offering services. They pointed out that information on the quality of laboratories spreads by word of mouth. There is no systematic method of informing providers and patients of laboratories that have demonstrated high quality performance.
Summary and Conclusions
Compared to a decade ago, genetic testing activity has grown among both commercial biotechnology companies (BTC) and nonprofit organizations (NPOs). Growth has been in generic testing technologies, development of tests for specific diseases, and provision of testing services. Many BTCs engaged in genetic testing, and a smaller proportion of NPOs are developing or offering tests for common complex disorders.
Growth is slower than many observers anticipated for a variety of interconnected reasons. Current test technologies have limited clinical sensitivity, technology is expected to change, the market is limited because it is only necessary to test a person once for a given genetic condition, the clinical utility of some tests is questionable and that makes demand uncertain, many genetic disorders occur at a low frequency, tests sometimes pose ethical and financial considerations, and regulations are still uncertain.
Technology development companies and large clinical multipurpose clinical laboratories do not view the market for genetic testing as particularly profitable in the short term. It is not clear how rapidly they will develop and market new tests for either common or rare disorders. At the same time, one cannot conclude that these companies will not move forward on genetic testing in the long run. In particular, new technologies for rapid and low-cost genotyping, which are being demonstrated in infectious diseases and somatic cell testing, might readily spread to germline diagnostics, particularly when they provide clinical data useful enough to warrant the testing.
Tests under development often do not receive external review. Several BTCs and NPOs that are developing genetic tests, including tests for complex disorders, and some organizations using home brews in offering testing services, fail to make use of institutional review boards (IRBs), obtain investigational device exemptions from the Food and Drug Administration (FDA), or submit the tests to FDA review.
Quality assurance of laboratories offering genetic tests is uneven. Most laboratories offering molecular genetic tests are registered under the Clinical Laboratory Improvement Amendments (CLIA), but participation in proficiency testing programs for genetic tests is not required of CLIA-certified laboratories. Eleven percent of NBTCs and 16 percent of NPO molecular laboratories did not participate in either formal or informal proficiency testing programs.
Commercial laboratory directors told us that they were aware of poor quality laboratories who were offering services. They pointed out that information on the quality of laboratories spreads via word of mouth and that there is no systematic way for health care providers or patients to identify laboratories that have demonstrated high quality performance. These findings raise questions as to whether current policies assure the quality of laboratories offering genetic testing.
Both BTCs and NPOs are marketing to nongeneticist physicians. Education of physicians in the proper interpretation of genetic tests poses an ongoing challenge. The majority of respondents from both BTCs and NPOs disagree with the statement that "most physicians can interpret genetic tests adequately to the patients."
Those interviewed disagreed considerably concerning the extent to which the testing laboratories should insist on compliance with ethical and clinical guidelines, especially regarding informed consent and the indications for testing. Overall, the study suggests that patients and test providers would both benefit from clarification of policies.
About the Authors
**Neil (Tony) Holtzman is Professor of Pediatrics at the Johns Hopkins University School of Medicine. He holds joint appointments in Health Policy and Epidemiology at The Johns Hopkins School of Hygiene and Public Health. Holtzman is also director of Genetics and Public Policy Studies at The Johns Hopkins Medical Institutions. He is Acting Chair of the Maryland Advisory Council on Hereditary and Congenital Disorders and was a member of the NIH-DOE Working Group on Ethical, Legal, and Social Implications of Human Genome Research while it was extant. He was Chair of the Working Group's Task Force on Genetic Testing.
***Stephen Hilgartner is Assistant Professor in the Department of Science and Technology Studies at Cornell University. He specializes in social studies of contemporary biology, biotechnology, and medicine. Currently, his main research project is an extended ethnographic study of the social world of genome mapping and sequencing.
 | Top of page |