Within the next two decades, numerous products with promise for revolutionizing health care are likely to come into the medical marketplace. Synthetically manufactured growth factors like transforming growth factor-beta and bone morphogenic protein-2 will become available to speedup skin and soft tissue healing, enhance fracture consolidation, and perhaps even repair damaged internal organs. Tissue engineered products, consisting of a biodegradable scaffold populated with the progeny of mesenchymal stem cells, may allow whole organ replacement. Gene therapy procedures will take direct aim at the very essence of our beings, our chromosomes, to correct genetic defects that have been passed from parent to child since the beginning of our species. As for medical devices, implants that permit attachment of artificial limbs directly to bone via transcutaneous connectors currently are being developed. Similarly, researchers are making implantable computer chips that will bypass nonfunctioning sensory organs to input external stimuli directly into the brain.
All these advancements—growth factors, tissue engineering, gene therapy, attachable prosthetic limbs, and implantable computers—are so new that physicians do not yet have any clinical experience with them. Each therapeutic strategy, therefore, will evolve into an empty landscape environment containing few guideposts to help judge their efficacy and safety. In each case, developer enthusiasm and manufacturer pressures to market must be counterbalanced by a concern regarding unforeseen short-term and long-term consequences of any truly innovative medical product.
Developers and manufacturers of innovative medical devices and pharmaceuticals often complain about the high cost of the testing and validation required by the Food and Drug Administration before a new medical product can be marketed in the United States. Food and Drug Administration officials, in response, are quick to mention their legislative mandate to assure that new drugs and devices are safe and effective before they are made available to the public.
Although the Food and Drug Administration has numerous shortcut mechanisms that can greatly reduce the cost of bringing to market a product that is similar to those already in use, it requires comprehensive evaluation for any truly innovative products. The expectation is that the testing will include, to the extent possible, a controlled clinical trial conducted by a limited number of investigators studying a predetermined number of subjects, generally in a prospective and randomized manner. The trial must have an adequate followup period and reflect meticulous data collection, usually by a dedicated surveillance staff.
The Food and Drug Administration, after receiving data from the trial, convenes an expert panel to assess the results. (Withholding adverse data from the Food and Drug Administration can result in severe civil penalties.) Once the product is marketed, the manufacturer must inform the Food and Drug Administration of any problems with the product that comes to light. For a few potentially dangerous products, the Food and Drug Administration may require postmarketing surveillance.
Given the amount and extent of Food and Drug Administration oversight, one wonders why the premarket clinical trials frequently fail to identify potentially serious problems with innovative medical products. How is it possible that so many Food and Drug Administration approved products ultimately need to be recalled? Is there any way that problems that become apparent with widespread use of a product could have been identified before that device or drug was released for general marketing?
Consider, for example, a hypothetical new drug that promises to enhance memory, but, unbeknownst to its developers, will cause a fatal cardiac arrest in one of every 5000 people who take the medication. If the Food and Drug Administration mandated a premarketing clinical trial involving 2500 subjects, half of whom are given the active drug and the other half (randomly assigned) are taking a placebo, there is approximately a three of four chance that nobody will experience a fatal reaction to the medication. As soon as the medication is released to general marketing, however, 200 fatalities can be expected among the first 10,000,000 users, making the pill far more lethal to Americans than bungee jumping or scuba diving.
By extending the period of observation and enlarging the sample size in premarket controlled clinical trials, the Food and Drug Administration could make better determinations about drug safety. However, the cost of such expanded testing would have to be added to the product’s final market price—much to the chagrin of consumers. Also, placing more stumbling blocks in the pathway of innovative developers and marketers might weaken our nation’s ability to compete with other less stringent countries in the international marketplace for new medical products.
The proposal being made here is that, because the development of such products is a manifestation of the evolution of technology in general, perhaps recent developments in evolution theory can yield insights into ways to improve testing and marketing innovative medical devices and pharmaceuticals.
The Burgess Shale
According to paleontologist Stephen Jay Gould, the consequences of a mud slide in Western Canada, near the border between Alberta and British Columbia, compel us to reexamine historic and evolutionary processes in a new light. 11 As mud slides go, the one that created Canada’s Burgess Shale quarry was not particularly large—approximately the size of a city block—but when it occurred (approximately 530 million years ago), the slide fortuitously trapped numerous soft-bodied Cambrian organisms in the ooze, thereby preserving them to a degree rarely observed in ancient rock formations.
Gould devotes an entire book, Wonderful Life, 11 to an analysis of the multicellular creatures that inhabited the primordial estuary buried by the mud. Gould also emphasized the importance this analysis has for understanding the evolution of life in general and the elaboration of other systems and technologies that develop under circumstances similar to those revealed in the Burgess Shale.
The organisms imbedded in the shale once were considered to be primitive arthropods (insects) or members of other phyla related to present day crustaceans, roundworms, and spiders. However, more recent studies 23,27,30 have reclassified many of the specimens into unique phyla that have no modern counterparts. Indeed, many of the Burgess Shale creatures have such bizarre body plans—with five eyes or seven pairs of appendages or a claw at the end of a proboscis—that they appear unworldly. Gould 11 and others 23,27,30 contend that the Cambrian period was characterized by an exceptionally high number of phyla, and therefore greater diversity of form, than exists today—the Cambrian Explosion.
The great biologic diversity that characterized this period occurred because animals were evolving into an empty landscape environment where they were the first complex multicellular creatures. The diversity can be seen as representing numerous innovative trial models. 23,27,30 With time, competition for resources and predatory coinhabitants combined to decimate organisms with suboptimal characteristics that might, under more favorable conditions, have evolved into substantial phyla. Gould calls the process “diversification and decimation”. 11 Ultimately, the winning species had offspring that, in time, gave rise to orders and genera within a more limited number of phyla that are constrained in body plan by their ancestral history.
Gould 11 and Kauffman 18 contend that the rapid diversification scenario associated with an empty landscape applies not only to living systems but also to the evolution of technology. Kaufmann, for example, has cited the early bicycles and computer operating systems as illustrations of the process of diversification into an empty landscape followed by the elimination of the less successful designs and survival of a limited number of variants.
Diversity in design appears to be a common way that new technologies evolve. In 1906, at America’s first automobile show, electric cars, steam driven vehicles, and automobiles with internal combustion (gasoline) motors all were displayed equally. 3 Designers of each type of engine undoubtedly expected that the future of automotive progress belonged to them. One hundred years of automotive evolution, however, have eliminated the unsuccessful contrivances and constrained the number of available options.
Total Knee Replacement
Many medical devices and drugs have undergone evolutionary development that resembles, in many ways, the Cambrian pattern. When no preexisting technology is available (an empty landscape), a wide assortment of products may appear within a short time to compete for market share. With time, the unsuccessful products disappear; products with more promise evolve and come to resemble the most successful design.
A perfect example of this phenomenon is the evolution of device designs for total knee replacement. Shortly after the total hip replacement was perfected in the 1960s, a very diverse group of total knee replacement designs were developed, marketed, and implanted in patients with arthritis during the next 10 to 15 years. 4,7,8,10,12–16,19,21,24–26,28,31,33,37–39 Some of these implants are, by modern standards, bizarre-looking.
Although initial reports on these devices enthusiastically praised the success of the implants, subsequent studies, usually written by surgeons not connected with the original design team, painted a less-than-rosy picture of clinical outcomes. 1,2,22,29,34,36 In some centers, many patients have had their earlier total knee components removed and replaced by more modern devices after the original implant loosened, prematurely wore out, or became infected. Not surprisingly, all total knee replacement implants now resemble the normal knee and consequently are difficult to distinguish from each other.
There is little doubt that each of the odd-looking total knee replacement implants of the early 1970s were designed with the best of intentions, based on the developer’s perspective of knee kinematics and related issues. What was sadly lacking at the time was a comprehensive understanding of the way the human knee actually works. (Most of the current research into knee biomechanics and pathomechanics is stimulated by a desire to reduce complications and increase the effectiveness of all types of knee surgery.)
Unfortunately, neither the patients who were recipients of the bizarre (by current standards) and ultimately ineffective total knee replacements of the early 1970s nor their surgeons had any way of knowing that the implants ultimately would fail.
The Burgess Shale experience shows us that no amount of foresight will allow us to predict which of several innovative competing designs ultimately will prove most successful. It is only through diversification and decimation that progress can be made when new products are being developed. If evolution in the medical arena must inevitably include a high number of designs that are doomed to fail, is there a way to reduce the likelihood that ineffective medical devices and drugs will be widely used before their deficiencies are fully defined?
It must be recognized that, for a time after an innovative product is released for marketing, the public will be exposed to unforeseen risks. 33 The Food and Drug Administration should not unduly delay marketing of new products to ascertain all possible risks. Try though they might, their efforts are almost certain to fail. In a premarket clinical trial, with a limited number of patients, all the factors mentioned earlier conspire to conceal deficiencies in the product being evaluated.
In the open market, however, with a larger number of patients and longer followup, more skeptical clinicians and their patients will, in time, find the hidden flaws in any drug or device. The Food and Drug Administration and Congress have, in a sense, recognized this fact by requiring postmarket surveillance, tracking, and reporting of products whose failure might prove catastrophic. As the federal government fine-tunes the postmarket surveillance system, adding and removing categories of products, many other devices and drugs not subjected to such scrutiny have proven harmful with time.
The Constraint of Liability
There exists a parallel system of restraint to overly enthusiastic or premature marketing of potentially dangerous products: tort liability. Every manufacturer, distributor, and provider in our nation’s healthcare enterprise is well aware of the open-ended liability exposure when a new product goes bad. With crystal-clear hindsight, trial lawyers present the injured patient to a jury as the hapless victim of a greedy medical establishment that places profits above safety. But as Dutton pointed out 5 : “Many of these injuries are neither negligent nor malicious; they are simply an unavoidable part of progress through science and technology.”
Once a product is available and widely used in the marketplace, developers and manufacturers are often reluctant to admit to any problems with the drug or device for fear of providing patient-consumers (and their attorneys) with ammunition for future litigation. As Susan Bartlett Foote stated 9 :
The social function of regulation depends on information—gathering data before marketing and reporting problems once a product has been distributed. Without adequate and accurate information, the safety goals cannot be met. On the other hand, the threat of product liability suits intimidates producers … Thus there are powerful incentives, even for responsible producers, to protect and withhold information.
We certainly would like to deter the development of unsafe medical products. With this in mind, manufacturers are compelled to make large payments to patients harmed by innovative products. The current tort liability compensation system is, however, a double-edged sword, hampering the evolution of the needed and beneficial along with the unsafe. This unfortunate state of affairs will continue until we acknowledge that technologic innovation cannot occur without a certain number of suboptimal designs.
A New Product Category
What can be done about the inevitable evolution and marketing of potentially dangerous drugs and devices? Insist that manufacturers purchase more liability insurance? Require that the Food and Drug Administration formally control or slow down the release of new products in some as yet undetermined way? Imposition of even greater formal control of marketing by the federal government will, if anything, inhibit product innovation by adding to development costs.
It should be possible, however, to slow down the introduction of new products (once approved by the Food and Drug Administration for marketing) by making patient-consumers more aware of the possibility of unforeseen risks. One feasible way this might be accomplished is by placing any innovative drug or device in a New Product category for a predetermined time after that product is cleared by the Food and Drug Administration for marketing. Ideally, the length of time a drug or device stays in New Product status should be based on experts’ projections of the time it would take for unforeseen complications to become apparent as the product is used. Because even knowledgeable consultants cannot accurately predict the unpredictable, truly innovative products should be kept in the New Product status for some time.
There also must be a mechanism for notifying potential consumers of new medical drugs or devices that the product is not tried-and-true but is, instead, in the new product category. This may be as simple as a warning label, or as comprehensive as a signed new product acknowledgment. A warning label might read:
This product has been tested according to Food and Drug Administration procedures and determined to be safe and effective. Be aware, however, that because the product is new on the market, unforeseen problems may become apparent with its widespread use with time.
A more comprehensive consent could include an acknowledgment by the patient of the drug or device’s New Product status and verification of the patient’s understanding that potentially serious low frequency or unforeseen problems may occur. Such a consent form would need to be signed by the patient-consumer and retained by the provider as part of the patient’s permanent medical records.
Although such a proposal appears to create new and possibly burdensome requirements for practitioners, the concept is not entirely novel. During the mid1970s, complications from many intrauterine devices led to their withdrawal from the market. The Dalkon shield, the Lippe Loop, the Cu7, and others have disappeared from sight, whereas their competitor, the Progestasert intrauterine device (which releases progesterone), continues to be available. The manufacturer of that implant has developed a comprehensive patient information and consent booklet designed to involve the prescribing physician in the education process. The patient, by signing incorporated forms, acknowledges the risks associated with intrauterine device usage. These measures have proven acceptable to patient-consumers and their physicians. The manufacturer of Paragard, a relatively new copper releasing intrauterine device, is following a similar approach.
How would the creation of a New Product category slow down the marketing of innovative drugs and devices? In any group of patients who might benefit from a new drug or device, there exists a spectrum of need. 6,20,35 Some patients have an urgent need because existing products are unsatisfactory for them; other patients can be treated almost as well with products already on the market.
Similarly, in any population of individuals, there will be those who are more risk-adverse than others. 17 A person who is risk-adverse with a less-than-urgent need might think twice before agreeing to be an early user of a product whose potential problems have not yet been fully ascertained. A discussion with the physician about the meaning of the New Product status of the drug or device might dissuade such a patient from using the product, especially if the prescribing physician cannot present a compelling reason to do so.
If the New Product status has the effect of slowing down the introduction of innovative drugs and devices into the marketplace, how will the manufacturers of these products recover their development costs before the product’s patent runs out? The answer is simple: extend the patent protection of the product to compensate for the lower initial profits. Longer patent protection also will help reduce the cost of newly developed drugs and devices. In fact, it might be possible to adjust the length of patent protection to match the nature of the product, thereby promoting a more constrained approach to marketing new products. Also, a certain measure of product liability protection might be an added benefit of a formal new product acknowledgment system.
Evolution theory shows us that technology cannot pass through truly innovative cycles of progress, when diversification of design leads to perfection of form, without concomitant decimation of the less successful models. Exposure to potentially harmful unforeseen problems associated with new drugs and devices, to some extent, is unavoidable. Growth-factor stimulation, tissue-engineered products, gene therapy, and seemingly miraculous implants all pose unknown risks. Product developers and government agencies can do their best to discover most of the complications associated with new medical products, but these safeguards will never be infallible.
To ensure vigorous innovation in medicine, product developers must acknowledge to patient-consumers the potential for unforeseen risks associated with a particular product and provide reasonable mechanisms whereby the patient-consumer can intelligently weigh the product’s risk potential against its possible benefit, given his or her individual situation. The proposal made here, for a temporary New Product status for innovative drugs and devices, provides such a mechanism. Alexander Pope’s recommendation 32 to “be not the first by whom the new are try’d, nor yet the last to lay the old aside” is sound advice for any reasonable person. That wisdom should be embodied in the way new medical products are released by the Food and Drug Administration for marketing.
During the discussion that followed the oral presentation of this paper, several members of the Association of Bone and Joint Surgeons suggested that the most expeditious way of keeping the Food and Drug Administration informed of problems that occur with drugs and devices in the New Product category would be to have patient-consumers report complications and other suboptimal outcomes directly to the Food and Drug Administration on a form supplied by the provider-physician when the product is first administered or inserted. (Such a patient-oriented reporting form would not, however, eliminate any statutory requirements for the physician to report problems). In this way, any physician bias toward under-reporting of potentially embarrassing complications would be bypassed. Additionally, the amount of paperwork associated with the use of new products would be kept to a minimum. The author agrees that such a proposal is probably the best way for the Food and Drug Administration to keep track of problems associated with the introduction of new medical drugs and devices to the marketplace.
1. Barnes CL, Scott RD: Patellofemoral complications of total joint replacement. Instr Course Lect 42:303–307, 1993.
2. Bohm P, Holy T: Is there a future for hinged prosthesis in primary total knee arthroplasty? J Bone Joint Surg 80B:302–309, 1998.
3. Clymer F: Treasury of Early American Automobiles. New York, McGraw-Hill 1950.
4. Coventry MB, Upshaw JE, Riley LH, et al: Geometric total knee arthroplasty. Clin Orthop 94:171–176, 1973.
5. Dutton DB: Worse Than the Disease: Pitfalls of Medical Progress. New York, Cambridge University Press 1988.
6. Eddy DM: Investigational treatments. How strict should we be? JAMA 278:179–185, 1997.
7. Engelbrecht E, Siegel A, Rottger J, et al: Statistics of total knee replacement: Partial and total knee replacement, design St. Georg: A review of a 4-year observation. Clin Orthop 120:54–64, 1976.
8. Evanski PM, Waugh TR, Orofino CF, et al: UCI knee replacement. Clin Orthop 120:33–38, 1976.
9. Foote SB: Managing the Medical Arms Race. Berkeley, University of California Press 1992.
10. Goodfellow JW, O’Connor J: Clinical results of the Oxford knee. Clin Orthop 205:21–32, 1986.
11. Gould SJ: Wonderful Life. New York, WW Norton 1989.
12. Gunston FH: Polycentric knee arthroplasty. Clin Orthop 94:128–135, 1973.
13. Hanslik L: First experience on knee replacement using the Young hinged prosthesis combined with a modification on the McKeever patella prosthesis. Clin Orthop 94:115–121, 1973.
14. Herbert JJ, Herbert A: A new total knee prosthesis. Clin Orthop 94:202–210, 1973.
15. Insall J, Scott WN: The total condylar knee prosthesis: The first five years. Clin Orthop 145:68–77, 1979.
16. Jones GB: Total knee replacement: The Walldius hinge. Clin Orthop 94:50–57, 1973.
17. Kahneman D, Tversky A: The psychology of preferences. Sci Am 246:160–173, 1982.
18. Kauffman S: At Home in the Universe. New York, Oxford University Press 1995.
19. Kenesi C: The “Caviar” Madreporic knee prosthesis. Clin Orthop 145:94–100, 1979.
20. King NMP, Henderson G: Treatments of last resort: Informed consent the diffusion of new technology. Mercer Law Review 42:1007–1050, 1991.
21. Letournel E, LaGrange J: Total knee replacement with the LL type prosthesis. Clin Orthop 94:249–256, 1973.
22. Lettin AW, Neil MJ, Citron ND, et al: Excision arthroplasty for infected constrained total knee replacements. J Bone Joint Surg 72B:220–224, 1990.
23. Levinton JS: The big bang of animal evolution. Sci Am 256:84–91, 1992.
24. Marmor L: The modular knee. Clin Orthop 94:242–248, 1973.
25. Matthews LS, Sonstegard DA, Kaufer H: The Spherocentric knee. Clin Orthop 94:234–241, 1973.
26. Mazas FB, GUEPAR: Guepar total knee prosthesis. Clin Orthop 94:211–221, 1973.
27. McMenamin MAS, McMenamin DLS: The Emergence of Animals: The Cambria Breakthrough. New York, Columbia University Press 1990.
28. Morel JR, Thomas RJ, Freeman MAR: ICLH Replacement of the knee. Clin Orthop 145:47–59, 1979.
29. Morrey BF, Adams RA, Ilstrup DM, et al: Complications mortality associated with bilateral or unilateral total knee arthroplasty. J Bone Joint Surg 69A:484–488, 1987.
30. Morris SC, Whittington HB: The animals of the Burgess Shale. Sci Am 243:122–133, 1979.
31. Phillips RS: Shiers’ alloplasty of the knee. Clin Orthop 94:122–127, 1973.
32. Pope A: Essay on Criticism. In Audry E, Williams A (eds). Poems of Alexander Pope. New Haven, Yale University Press 276, 1961.
33. Ranawat CS, Shine JJ: Duo-condylar total knee arthroplasty. Clin Orthop 94:185–195, 1973.
34. Rottger J, Heinert K: St. Georg endoprosthesis system (slide and hinge principle). Observations and results following 10 years’ experience with over 3,400 operations. Z Orthop Ihre Grenzgeb 122:818–826, 1984.
35. Sass H-M: Moral aspects of risk innovation. Artif Organs 21:1217–1221, 1997.
36. Schneider R, Abenavoli AM, Soudry M, et al: Failure of total condylar knee replacement: Correlation of radiographic clinical surgical findings. Radiology 152:309–315, 1984.
37. Sheehan JM: Arthroplasty of the knee. Clin Orthop 145:101–109, 1979.
38. Walker PS, Shoji H: Development of a stabilizing knee prosthesis employing physiological principles. Clin Orthop 94:222–233, 1973.
39. Yamamoto S: Total knee replacement with the Kodama-Yamamoto knee prosthesis. Clin Orthop 145:60–67, 1979.