AFMA

EFMA
JFMA



Americans For Medical Advancement
Europeans For Medical Advancement
Japanese For Medical Advancement


 





Your support gives
AFMA
a stronger voice.

NIH Research Protocol for Silver Spring Monkeys: A Case of Scientific Misconduct (Part II)

Neal D. Barnard, M.D., Roy Selby, M.D., Daniel N. Robinson, Ph.D., Gervasia M. Schreckenberg, Ph.D., Carol Van Petten, M.D.


Scientific Misconduct in NIH Protocol

The following sections review evidence of misconduct in the NIH protocol.

Definition of Scientific Misconduct

Scientific misconduct is investigated by the NIH's Office of Scientific Integrity (OSI) as well as the Office of Scientific Integrity Review (OSIR) in the Office of the Assistant Secretary of Health, Department of Health and Human Services. According to OSIR Director Lyle W. Bivens, "misconduct in science" is defined in Title 42, part 50 of the Code of Federal Regulations as "fabrication, falsification, plagiarism, or other practices that seriously deviate from those that are commonly accepted within the scientific community for proposing, conducting or reporting research."

Evidence suggests that serious and willful misconduct was involved in the writing and advancement of the NIH protocol. It appears that among the motives behind this fraudulent proposal was the intent to deceive members of Congress; the proposal also had the effect of deceiving judges in federal court.

Representation of the NIH Protocol as Duly Authorized Research

In his July 1, 1988 letter, Raub represented the NIH protocol to Congressman Dornan as a "carefully" drawn plan that would be carried out in conjunction with euthanasia. Raub's letter stated:

Staff of the Department of Health and Human Services (HHS) and collaborators therefore have prepared a plan for the euthanasia (Enclosure B). This plan is drawn carefully to ensure due recognition of both the special protected status of the animals and the unique neurological information they can provide with respect to improved treatment for human victims of stroke or spinal cord injury.

Documents submitted to the U.S. Court of Appeals for the Fifth Circuit by attorneys for the NIH show that the NIH encouraged the court to believe that a properly designed and approved research protocol existed. The apparent motivation was a desire to prevail in a legal proceeding over the fate of the monkeys in question. The NIH had previously indicated its plan to kill three of the animals, but the plan was blocked by a temporary restraining order. When the NIH's appeal to the U.S. District Court for the Eastern District of Louisiana was heard, Thomas Watson, representing NIH, stated:

First of all, the NIH does have an interest in this case, obviously, Your Honor, and that's to fulfill its mission which is to conduct research on these monkeys. In that proceeding, it was clear that the judge had been made to believe that experiments were to be conducted on the animals, as indicated by the statement of the Court: From what I understand, they will be put to death or experiments will be continued on them at the time they are being put to death.

In documents submitted to the U.S. Court of Appeals for the Fifth Circuit in July, 1989, the Brief for the Appellant (NIH) calls the protocol "important research procedures" in its Request for Oral Argument:

The District Court issued a preliminary injunction that bars NIH from euthanizing three research monkeys and from carrying out certain important research procedures on the monkeys in a nonpainful manner during the euthanasia process.

On page 13 of the Brief, NIH attorneys stated:

Through humane, nonpainful procedures that would be conducted as part of and following the euthanasia procedure involving deep surgical anaesthesia, NIH hopes to gain important knowledge that may lead to substantial improvements in rehabilitation therapy for individuals who have suffered brain or spinal cord damage.

On page 15 and again on page 37, the protocol is represented to the Court as "important research."

On September 29, 1989, NIH attorneys requested expedited oral argument in the court case, stating, "NIH has been prevented from carrying out its plans to conduct important medical research on certain experimental monkeys because of a temporary restraining order issued on January 5, 1989."

In arguments in federal court on January 10, NIH attorneys used arguments of the experiment's scientific value in order to have the temporary restraining order voided for one monkey, Billy. They stated that the experiment related to rehabilitation and pharmaceutical treatments, when in fact it does not:

The medical information NIH will obtain from those procedures is of the utmost significance, relating as it does to the rehabilitation therapy and pharmaceutical treatment of humans who have sustained brain or spinal cord damage. . . . If the preliminary injunction is not lifted in time for NIH and Tulane to conduct the painless research described in the attached declarations, the answers to the human tragedy of quadriplegia and paraplegia, among other such conditions, may remain locked in Billy's brain forever.

Evidently, no other, more detailed protocol for the experiment was submitted for approval. On October 19, 1989, Congressman Smith wrote to Raub asking "(1) whether NIH plans to adhere to the [July 1988] provisions planned for these animals; and (2) do any other documents exist regarding the future use of these animals?" Raub responded on November 17: "the July 1988 protocol remains our intent. Absent judicial constraints, we would effect that procedure immediately for at least three of the monkeys." On January 10, 1990, NIH attorneys submitted the protocol to the U.S. District Court for the Eastern District of Louisiana, indicating their wish to carry out its provisions. NIH attorneys reportedly stated in federal court in Louisiana that no other protocol exists. In U.S. District Court for the District of Columbia on January 12, 1990, NIH attorneys stated that no more finished document had ever been written.

NIH Representatives clearly wished federal officials to believe that the July 1988 document represented properly designed research that should be carried out without further amendments or approvals.

Critique of Protocol

The protocol's content is grossly deviant from accepted scientific norms in several respects. The report consists of five main sections: "Introduction," "Background of the Scientific Problem," "Goals," "Protocol for Euthanasia and Post-Mortem Analysis of Tissues," and "Summary."

Improprieties are noted below in two categories: content and administrative procedures. This separation is somewhat artificial. For example, the absence of an identified principal investigator on the original protocol is a deficiency of both content and procedure. To avoid redundancy, however, each deficiency was assigned to only one category.

Critique of Protocol: Content

The four-and-one-half-page document is markedly shorter than that of typical research protocols, which are comprehensive documents that may reach hundreds of pages in length. There are glaring omissions:

1. No principal investigator is listed. Research proposals are normally submitted by a principal investigator and one or more co-investigators whose qualifications are subjected to scrutiny by the inclusion of their curricula vitae. The NIH protocol is unsigned. In Appendix A of the report, four individuals are named who "will collect the physiological data during the euthanasia procedure": Preston E. Garraghty, Ph.D. of Vanderbilt University, Jon H. Kaas, Ph.D. of Vanderbilt University, Timothy P. Pons, Ph.D. of NIH, and Peter L. Strick, Ph.D. of the Veterans Administration Hospital.

These individuals were listed in alphabetical order. There is no indication which (if any) of these individuals is the principal investigator. In federal court proceedings on January 12, 1990, NIH attorneys produced a one-page document that they stated contained the names of nine individuals who had written the NIH protocol and sent it to Dr. Raub. They are Robert E. Burke, M.D. (NINCDS, NIH), Michael E. Goldberger, Ph.D. (Medical College of Pennsylvania), Edward G. Jones, M.D., Ph.D. (University of California, Irvine), Jon H. Kaas, Ph.D. (Vanderbilt University), Frederick A. King, Ph.D. (Yerkes Regional Primate Research Center), Mortimer Mishkin, Ph.D. (NIMH), Timothy P. Pons, Ph.D. (NIMH), Peter L. Strick, Ph.D. (Syracuse VA Medical Center/SUNY Syracuse), and William D. Willis, Jr., M.D., Ph.D. (University of Texas Medical Branch, Galveston). Given the protocol's extraordinary brevity and poor quality, it is difficult to believe that these nine signatories actually lent the full benefit of their expertise to writing the protocol.

Among the signatories is Frederick King, who was named in the confidential American Physiological Society memo as being involved in raising funds to execute an agreement among Tulane, the NIH, and representatives of several scientific societies in order to solve the NIH's public relations problem by transferring ownership of the monkeys to Tulane in exchange for private payments. Kifllng the deafferented monkeys was discussed in the memo as a means of reducing the cost of this arrangement.

In an application to the Institutional Animal Care and Use Committee submitted by Delta Director Peter Gerone on October 20, 1988, Gerone listed himself as the principal investigator. It appears that he assumed this responsibility after the protocol was written. There is no indication that he played any part in drafting the protocol, and he has no relevant expertise. The main responsibility for the application process and for assuring the proper conduct of research rests with the principal investigator. How Gerone assumed this role is unclear.

2. Portions of the text are virtually identical to a report of the panel making recommendations for the monkeys on March 16-18, 1988, entitled "Assessment of and Recommendations for the 'Silver Spring Monkeys." In the January 14, 1990 court hearing, NIH attorneys described the "Assessment" report as having been generated by an "independent" panel of experts, who had decided on the need for euthanasia of some of the monkeys independent of any research consideration. The NIH attorneys used this claim in support of their argument for carrying out the experiment. Nonetheless, it appears that the two documents were not generated independently. The first paragraph of the protocol section entitled "Background of the Scientific Problem" of the NIH proposal reads:

Damage to the human brain or spinal cord is regarded as irreversible. Nevertheless, it is well documented by clinicians and basic scientists alike that, in the first months after injury, some recovery of function occurs spontaneously. Although this recovery can sometimes be enhanced by efforts at rehabilitation, the results are generally disappointing. One reason that more substantial recovery cannot be induced is that we do not yet understand the mechanisms responsible for the limited improvement that does occur.

In the Delta panel report, the beginning of the first paragraph on page six reads:

Damage to the human brain and spinal cord has generally been regarded as irreversible. Nevertheless, it is well documented that some recovery of function frequently occurs. While this recovery can sometimes be enhanced by current modes of rehabilitation therapy, the results are generally disappointing. At present, we cannot induce more substantial recovery because of our limited understanding of the neural mechanisms responsible for the slight improvements that do occur.

A comparison of the corresponding second paragraphs of each report reveals more similarities in both structure and language -- similarities beyond those that could be expected in the work of independent writers. The only signatory to both is Raub, who was an "information resource" to the panel and the signatory on letters accompanying the NIH proposal sent to members of Congress. It is not, however, entirely clear which document was generated first.

3. The protocol is undated. This is significant in and of itself, but is also important from the standpoint of the independence of the decision to euthanize the monkeys. Raub's cover letter to members of Congress implies that the protocol was written after the panel decided on euthanasia for two of the monkeys in March, 1988:

On several occasions we have discussed the fact that all eight of the surgically disabled Silver Spring monkeys are likely to require euthanasia eventually. . . . The report on the March site visit to the Delta Regional Primate Research Center. . . . strongly reinforces these expectations. Staff of the Department of Health and Human Services and collaborators therefore have prepared a plan for the euthanasia.

However, the document presented by NIH attorneys in the January 12, 1990 hearing listing nine individuals was dated September 30, 1987, six months before the panel met. The attorneys represented it as a cover letter for the NIH protocol, although it was a loose document in a different typeface and it apparently was not included with the protocol when it was sent to members of Congress or to Delta's Institutional Animal Care and Use Committee.

On January 25, 1990, Raub again stated in a letter to Senator Paul Simon that the protocol was "developed in 1988." The process of deciding whether any of the primates required euthanasia appears to have been intertwined with the desire to conduct the experiment.

4. No site is specified for the research. The experiment was conducted on Billy on January 14, 1990 at Delta. Although the proposal states that the research will occur when "the attending veterinarians at the Delta Primate Center decide that a particular deafferented animal must be euthanised," the proposal does not restrict the experiment to Delta. The four "investigators" mentioned above are described as working at various other institutions; none is listed as a Delta employee. If the research was to be conducted at Delta, it was not specified in which laboratory or under whose supervision that work would be conducted. This information is submitted in all research protocols, for two important reasons: It is critical in determining facilities' suitability for the research in question; it also indicates responsibility for adherence to federal research regulations. Failure to specify the research site is a major omission.

5. The protocol's presentation of the field is deficient. The protocol presents only selected information that does not reflect the depth of current knowledge about the sequelae of neurological damage or of the treatments that are available or under investigation. Properly written research proposals contain a review of relevant scientific literature justifying the proposed research. The NIH protocol contains no mention of any specific research in the area in question; it relies solely on vague, undocumented allusions to recent research. The introduction makes a variety of unsupported claims about the animals apparently aimed at lay readers; it bears no resemblance to the introductory comments appropriate to a scientific proposal.

The section entitled "Background of the Scientific Problem" suggests that the current experiment will investigate the rate-limiting variable in neural regeneration. It omits the fact that a number of research teams are conducting research in recovery from brain and nerve damage.3 It is precisely because regenerative processes are known to occur that numerous researchers have studied them in detail. Selzer4 has written, "In many cases considerable recovery is seen. Several factors are known to influence the probability and quality of nerve regeneration."

The protocol states that neural regeneration ("sprouting") following spinal cord damage has never before been demonstrated in primates. This is untrue. Bernstein and Bernstein5 demonstrated sprouting in the spinal cords of rhesus monkeys. The current group of monkeys have severed dorsal roots, and would, therefore, suffer degeneration of axonal processes in the spinal cord. The text of the NIH protocol indicates gross inattention or lack of current knowledge of the research area in question.

6. The protocol uses inappropriate language and unsubstantiated hyperbolic statements in an attempt to build a case for the monkeys' research value. The second paragraph on page two of the protocol presents sweeping statements of conjecture as if they were fact, in an effort to demonstrate that the monkeys in question hold the key to elucidating the mechanisms of recovery from stroke or other neurological injury:

But pursuing the original [research] plan alone would yield only a small fraction of the information regarding neural plasticity that we now know these animals can provide. That is, the partially deafferented animals have come to represent a scientific resource far beyond that anticipated when Dr. Taub prepared them.

The paragraph then lists reasons why those statements should be considered true. In summary, the reasons are:

a. It is now known that, after brain injury, there can be some functional reorganization of the cortex.

b. Methods (unspecified) for characterizing various neurochemical pathways are becoming available.

c. Techniques (unspecified) now allow a more accurate and thorough study of the changes that occur in the spinal cord after transection of the dorsal root sensory fibers.

These points provide no substantiation for the assertion that the animals in question are "a scientific resource far beyond that anticipated when Dr. Taub prepared them," or that they are even suitable subjects for any meaningful research. There is substantial reason to believe that these animals are not suitable subjects for further research and that the NIH protocol offers no suitable research design.

Moreover, although NIH officials have stated that the experiment is related to stroke, neither they, nor the protocol's author(s) acknowledged the serious problems -- well known to neurologists -- with animal models of stroke and other forms of brain injury. In Stroke editorial, Molinari6 bemoaned that results from animal experiments repeatedly fail to be replicated in human clinical trials. Neff replied:

The real problem faced by researchers in cerebrovascular disease is that many treatments for stroke have been validated in animal models but then failed in human trials. If the human trials are valid and if the laboratory studies are valid, then the problem must be that the laboratory models are not applicable to human cerebrovascular disease.

This conclusion is not surprising in light of the obvious differences in scale between the human brain and that of rabbits or small primates. . . . The repeated failure of laboratory-proven stroke therapies in humans can be due only to the inapplicability of animal models to human cerebrovascular disease.7

Neither of these authors would have attempted to totally discount animal experiments. It is clearly unfounded, however, to hold that animal models of stroke, and particularly this group of monkeys, are nearly certain to lead to rapid and major strides in neurological research or clinical treatments. Another Stroke editorial, January, 1990, decried the waste of resources caused by animal experiments in stroke treatment:

Each time one of these potential treatments is observed to be effective based upon animal research, it propagates numerous further animal and human studies consuming enormous amounts of time and effort to prove that the observation has little or no relevance to human disease or that it may have been an artifact of the animal model itself.8

Stroke can be due to many factors, and neural regeneration is one aspect of a complex process. One could make a case that animal experiments are more applicable in certain areas than in others. Nonetheless, no reservations about the applicability to humans of animal experiments in brain function were noted in the protocol.

The protocol's language is highly unusual. Rather than conforming to the norm of specific scientific terminology that unambiguously indicates the place of the proposed research in the relevant field of endeavor, the protocol uses lay terms to dramatize the project's supposed value. The language appears to aim at readers other than scientists. Curiously, at several points in the protocol, technical terms are introduced that would confuse the lay reader but are far too vague to satisfy scientists. For example, "experts in limb pathology," "state-of-the-art immunocytochemical, histochemical, and histological techniques," and "normal control material already in hand" are all used with no further description. These terms provide only a scientific aura without meaningful content. Whether the primary intended readers were the members of Congress to whom the protocol was sent or others is not discernible from the content. The protocol's major deviations from accepted norms suggest that the protocol's intended effect was other than scientific advancement.

7. The proposal' outlined goals and methods are grossly deficient. The proposal specifies three goals, followed by a one-paragraph method section. The goals are vaguely addressed, there is no description of specific hypotheses, and the methods to be used are described in only cursory terms in language aimed at lay readers. All properly written protocols describe the scientific methods in detail. The method section is of interest with regard to the scientific design's adequacy. This design ordinarily includes any examinations or measurements to be made, statistical treatments of data, and the procedures in which animal subjects will be used. The protocol's method section contains only three sentences regarding the euthanasia; there is no mention of the anesthetic agent(s) to be used or the route of administration. This is followed by one brief sentence related to the method to be used in pursuit of the first goal. Two sentences then describe monitoring the animals to avoid pain. A final sentence is devoted to the methods to be used in pursuit of the second and third goals. Although a few more details on methods can be gleaned from the description of the goals, virtually all of the methodologic details are left to the reader's imagination. Each goal and its associated method are discussed below.

a. "To determine the extent of the physiological changes in the somatosensory cortical fields that have resulted from the prolonged deafferentation of the forelimb."

The proposal notes that if the sensory nerves from a given body part, such as a finger or an arm, are severed, the portion of the brain that had previously received input from that body part may begin to receive input from other body areas. The proposal states that it would be useful to know whether the portions of the brain that had served an arm reorganized, after deafferentation of the arm, to receive input from other body parts. The proposal asserts that this research is significant because no researcher has previously investigated reorganization of an area as large as that representing an entire arm, and because a prolonged period of time has elapsed since the experimental injuries.

The proposed method is accorded only one sentence: "During this time the somatosensoiy cortex will be exposed and a microelectrode used to detect the responses of single neurons to gentle stimulation of the body surface with a camel's-hair brush."

The proposal makes the sweeping prediction that the "euthanasia procedure will not only provide essential information regarding previously unstudied dimensions of neural plasticity following injury, but will also have broad theoretical implications." This is unsupported conjecture presented as a fact.

The protocol further states that "evidence of extensive reorganization in cases of chronic deafferentation would indicate that neural activity is important in maintaining or establishing not just topographic maps but perhaps all forms of sensory representation in the cortex."

Among the protocol's failures is the absence of any rationale that relates the brain mapping to the clinical conditions of stroke and spinal cord injury. The rearrangement of sensory fields is a normal phenomenon that has long been known to occur and would simply be passively observed by the investigators. They are not investigating methods for changing the brain representations, nor are they attempting to alter brain function in motor areas, which are of particular concern in stroke. In stroke, part of the brain is dead; the principal sequelae of concern are motor deficits. While neuropsychologists have long been curious about brain maps, there is no connection between brain mapping and any foreseeable stroke treatments.

There are several deficiencies in the protocol:

i. No hypotheses are stated.

ii. No study design is given. There is no description of what comparisons would be made or what findings would be considered significant.

iii. The methods for measuring neural function are not described. There is no mention as to which cells would be examined, what sort of microelectrode would be used, what measurements would be made, or which body parts would be stimulated in attempts to elicit responses. The type of microelectrode and specifics regarding its use are critical considerations in such an experiment: Some investigators have been concerned that the damage caused by multiple microelectrode penetrations of the cortex may actually modify the cortical representations under study.9

iv. Valid and reliable results are not likely to be obtained. In owl monkeys and squirrel monkeys, there is a great deal of variation in the cortical representations of body surfaces from one individual to the next, which may also be true for crab-eating macaques (Macaca fascicularis). Evidence suggests that this variation is as great or greater between individuals of the same species as between species.11,12

Macaca fascicularis is not the species most commonly used in cortical mapping. Most data in this area come from owl monkeys and squirrel monkeys because their neurological structures facilitate detailed mapping.11 According to Dr. Taub's 1979 grant application, crab-eating macaques were used because the rhesus monkeys Taub had been using were in short supply.

The variability of cortical representations is likely to be even greater as the brain reorganizes in deafferented subjects. Reorganization is often idiosyncratic. The subsequent experience of the animals further increases this neurological variation.

In cases of high variability, a large sample is required in order to achieve statistical significance. The group of monkeys is small and varied, and continues to dwindle through fatalities. It currently consists of a total of six deafferented Macaca fascicularis monkeys and a non-deafferented rhesus female. Of the six, five have a deafferented arm in place and one had his deafferented arm amputated (not by scientific design, but as a remedy for uncontrollable self-inflicted damage). The monkey already experimented on (Billy) was bilaterally deafferented. This small and varied group is far from that expected in standard research.

The word "sample" could be applied to the brain cells measured, rather than the group of animals, but in this case there remains little likelihood of a sufficient sample size. The protocol limits the experiment to a four-hour period prior to killing the animals. The number of measurements of brain cell function that are possible within that time period is relatively small. In court testimony on January 12, 1990, Mortimer Mishkin of the NIMH stated that, at most, 90 to 100 measurements could be made. Brain mapping usually requires extensive sampling, in part because the boundaries between some cortical areas may be poorly defined and overlapping. Selecting a small sample of neurons yields a very coarse result that is difficult to interpret.9,12

Variability among animals appears to be due to two factors, genetics and experience, neither of which has been controlled in this group of animals. Since the crab-eating macaques Taub used were captured in the Philippine forest, their genetic backgrounds are entirely unknown. Experiential factors may exert an even more profound effect on the cortical map. Researchers have suggested that the use of the arm may continually modify the cortical representations that the proposal aims to measure. Jenkins and Merzenich10 report, "In detail, these cortical representations are constantly changing. At least many of these changes must reflect the dominant use history of the animal."

For this group of animals, the use of their arms in the intervening years since their deafferentation has been unstructured, unstandardized, and undocumented, making it highly unlikely that meaningful results can be obtained. As stated in the proposal's introduction, the monkeys did not receive the full schedule of retraining in Taub's original research protocol. The surviving group represents different variants of surgical procedures, as noted above. Moreover, according to veterinary reports, the condition of the deafferented arm differs from one animal to the next. All have had a greater or lesser degree of damage to the deafferented arm, and serious illnesses and fatalities have occurred. X-rays have documented spinal and skeletal changes, which may or may not affect the experience of the animals in ways that could influence the mapping experiment's results. The protocol itself is set to begin when the subjects are so debilitated that euthanasia is required for humane reasons. At that point, serious medical conditions, including possible metabolic and nutritional abnormalities, will be present. None of these effects is accounted for in the protocol. These are highly contaminated research subjects;

In addition, animals whose clinical course has been so poor that euthanasia is required are unlikely to be models for successful recovery. As Raub wrote on October 2, 1985, "We have no research protocols, either ongoing or planned, for which these animals are appropriate."

v. There are no control animals. Taub's surviving control monkeys have been moved to the San Diego Zoo and are in group housing. These differences in setting, along with the lack of control over the experiences of the animals in both groups for the past eight years, would be expected to lead to significant differences in the brain functions to be measured. The protocol's descriptions of the use of control samples reflect little understanding of the use of controls in scientific study.

vi. Confounding effects of anesthesia, age, illness, and medical treatments are neglected in the proposal. Some researchers13,14 have found significant effects of anesthetic agents on cortical mapping experiments; others15 have not. The failure to describe the anesthetic agent intended for use and to address the issue of potential artifact introduced thereby is a serious omission. In addition, age affects the capacity for cortical reorganization in some species.16

vii. A key variable in Taub's work -- the behavioral assessment describing the capacity of the animals to use their deafferented limbs -- was abandoned several years ago. Experience is a key element in brain somatotopographic representations. Without this information, there is no basis on which to judge the significance of brain restructuring, even if it could be accurately measured, and no applicability of this experiment to the rehabilitation of nerve injuries. NIH spokespersons have stated that one of the experiment's goals is to facilitate rehabilitation of victims of spinal cord injury and stroke, yet no part of the experiment relates to treatment or rehabilitation. The proposal's claim of clinical significance ("potentially of great significance for alleviating suffering and advancing human health") would be grandiose even if the protocol were more carefully constructed and properly written. Cortical mapping is "basic research," not clinical research. The finding that intact body parts make connections with portions of the brain that previously represented deafferented body parts is not directly related to the functional recovery of the deafferented body part. There is no relationship to current or proposed clinical treatments. The proposal asserts that this research will aid in the development of "strategies aimed at promoting functional recovery from nerve injury," but provides no rationale as to why this might be an expected outcome. The experiences of the animals, who for the past eight years have been confined to small individual enclosures with no rehabilitative interventions, no exercise or even freedom of movement, and no normal social interactions, are dramatically different from the rehabilitative efforts that would normally follow neurological injury in humans.

There is no evident basis for the contention that this group of animals is appropriate to research in this area. There has been essentially no control and considerable artifact in the intervening years. The sample is so small, varied, and uncontrolled that achieving statistically significant results is extremely unlikely.

b. "To identify the immunocytochemical changes in cortex or thalamus that might underlie the physiological changes."

The protocol states that, when portions of the brain change their organization (for example, to receive input from a new body part), these changes in brain organization may relate to changes in the quantities of neurotransmitters present in those parts of the brain. It then states that the quantity of neurotransmitters present may be influenced by activity in the body parts in question. Speculations then follow, suggesting that information from this study could help in "phantom limb pain" or recovery of function following deafferentation.

The protocol then proposes, with no supporting rationale, the measurement of cytochrome oxidase, GABA, glutamate, Substance P, CAM-Il, protein kinase C, C-fos, and CAT-301. The protocol states that these measurements will be used to see "whether changes in quantities of neurotransmitters in the cortex are due to changes in metabolism or neurotransmitters in the thalamus." They will also be used to search for changes in other unspecified parts of the cortex.

As with the first goal, no model or hypothesis is specified, and there is no experimental design. The techniques by which samples will be obtained, how assays will be done, how the results will be analyzed, and what degrees of differences in measurements would achieve statistical significance are not specified. Anesthetic effects are not discussed.

The text suggests that immunocytochemical techniques would be used for the assays noted above. Although no specific fixative is noted in the protocol, the IACUC application calls for the use of glutaraldehyde. This chemical would cause great distortion of the antigenic proteins, precluding most of the intended measurements. Sternberger17 has described two major problems with the fixative:

Glutaraldehyde seems to destroy, in most instances, detectability of antigen by immunofluorescence. . . . Glutaraldehyde, by virtue of its cross-linking qualities, hardens tissues so much that it prevents its own penetration, whether in perfusion or immersion fixations.

Although no hypotheses are stated, three are implied and can be separated for discussion purposes:

i. Chemical levels in deafferented sections of the brain differ from levels on the contralateral side.

The protocol calls for separate measurements of eight chemicals in five monkeys, assuming that only unilaterally deafferented monkeys with arms in place are used. Each subject provides his own control. This requires a minimum of 40 comparisons. If the significance level is set at .05, then by chance alone results would appear significant in two of the 40 comparisons. The investigators are particularly vulnerable to this error given that there is no prediction as to the expected results; they are, in essence, "fishing." To minimize this source of error, a stricter criterion for statistical significance (e.g., .01) is required. With the small sample, as discussed above, and a strict criterion for significance, achieving significant results will be difficult.

A further flaw is the absence of any discussion of the differences that may normally be expected between the two cerebral hemispheres of Macaca fascicularis.

ii. Brain chemical levels in experimental subjects are outside the range for normal subjects.

If a difference is found between the experimental monkeys and "normals," there is no way to attribute the cause of the difference. The protocol indicates that measurements will be compared between the deafferented monkeys and unspecified "normal control material already in hand." The word "control" does not refer simply to normal tissue or normal individuals. "Control" refers to a comparison sample that has undergone precisely the same treatments as the experimental group, with the exception of the experimental variable in question. For the "material already in hand" to act as a control, it would have to have come from monkeys who had experienced the same treatment, other than the experimental manipulations, and the same subsequent experience as the animals who had undergone the experimental procedures. Age, medical conditions and all other factors that could influence results should also be essentially the same. There are no monkeys who meet these criteria, particularly since very few of Taub's original control monkeys are still alive, and all of these are in an entirtly different setting from Delta and have long deviated from the procedures Taub had planned.

The protocol calls for taking samples in highly unusual circumstances. The samples would be taken from monkeys who have experienced the prolonged stress of years of isolation caging, illness so severe that euthanasia is necessary, and a final four hours of anesthesia. All of these factors can be expected to affect brain chemical levels. There is no mention as to whether control animals would have undergone similar treatment, nor is it likely that any such animals are available.

iii. There is a relationship between brain chemical levels and the brain representation of body parts.

The protocol seeks "correlations between functional and structural maps for a wide variety of neurochemical factors." The word correlation appears to be used improperly, since it is not possible to calculate a correlation between a linear variable, such as a chemical level, and a structural map.

In summary, gross methodologic problems are apparent for all three implied hypotheses. Without a carefully planned method for comparing results of sufficient numbers of subjects in experimental and control groups who are comparable on relevant variables, measurements are not meaningful.

The text makes two references to anatomical measurements in the cortex and thalamus, once in the title "Anatomy and Cytochemistry of the Somatosensory Cortex and Thalamus" and once in a text sentence: "It should be pointed out that the major value of these anatomical and chemical studies will derive from a comparison with functional data." There is, however, no mention whatsoever of any anatomical examination of the cortex or thalamus in the description of the goals or methods, or anywhere else in the protocol; it appears to have simply been overlooked. In spite of Raub's contention that "This plan is drawn carefully," it appears that the proposal was hastily conceived and written.

There is no reason to believe that the protocol can in any way assist in the two clinical problems noted, nor is there any basis for this speculation in the protocol. The anonymous author did not distinguish between phantom limb pain and phantom limb syndrome, and has not attempted to review or even comment on the literature on either condition that might relate to the protocol.

c. "To look for evidence of 'sprouting' in the spinal cord as originally intended by Dr. Taub, and to determine whether anatomical and cytochemical changes at this level of the system are correlated with reorganizational changes at higher levels."

The protocol states that when afferent nerves of cats or rats are damaged, the remaining intact nerves may "sprout" new connections that will provide a degree of compensation. It states that electron microscopic examination of the spinal cord will assess the presence of "sprouting" in the monkeys. In addition, the biochemical assays which were alluded to above for the cortex will also be applied to samples of the brain stem and spinal cord, and the monkeys' forelimbs will be examined "by experts in limb pathology." The method of tissue fixation is described in the following terms: "the monkey will be perfused with normal saline followed by a dilute aldehyde fixative." The examinations will use "state-of-the-art immunocytochemical, histochemical, and histological techniques."

As noted above, the protocol states that neural regeneration following spinal cord damage has never before been demonstrated in primates. Bernstein and Bernstein,5 however, published a detailed experiment demonstrating sprouting in the spinal cords of rhesus monkeys.

As in the discussion of the first two goals, there is a marked lack of specificity of the research design and methods. There is no indication as to which parts of the brain stem and spinal cord will be sampled. The "experts in limb pathology," the methods they would use, and what they might be looking for all remain unidentified. Previously discussed concerns regarding the absence of controls apply here as well.

The terms "dilute aldehyde fixative" are not sufficiently specific. Formaldehyde is useful for preparing tissues for examination by light microscopy, but it is inappropriate for electronic microscopy. Glutaraldehyde or a glutaraldehyde/para-formaldehyde solution would be more appropriate for that application but would be inappropriate for the immunocytochemical measures, as noted above. According to the IACUC application submitted by Gerone, glutaraldehyde was to be used.

It appears that the fixative is intended to act as the killing agent. Failure to specify the agent is gross omission. This issue would be of interest to an Institutional Animal Care and Use Committee. It is possible that the saline infusion would wash out certain anesthetics and that, just before death, the aldehyde fixative would be infused into animals no longer properly anesthetized. The lack of specificity in this area prevents the reader from rendering an informed judgment. As is true of the remainder of the protocol, the description's vagueness represents a serious deviation from scientific norms.

No details at all are provided on the methods of tissue examination, how they fit into a research design, how the results will be analyzed, or what will be regarded as a significant finding. Synapses resulting from sprouting can be analyzed in a variety of ways: their number, density, and type; the number of synaptic contacts made by each terminal; the transmitters they contain; etc. The mention of "state-of-the-art immunocytochemical, histochemical, and histological techniques" provides the reader with no evidence that the proposal's author understood which methods would be appropriate for such a study. Such wording is entirely inappropriate for a scientific document.

It must also be noted that the protocol aims to accomplish its task in "no more than four hours." A typical mapping procedure requires several hundred microelectrode penetrations and measurements11 and takes many hours to complete. The four-hour time limitation, which is difficult to reconcile with the proposed goals, appears to derive from public relations concerns.

The problems with this proposal go beyond its scanty presentation. There is little, if any, scientific basis for using this unusual group of monkeys in this area of research. The grandiose claims that are made are not in keeping with reasonable goals of scientific endeavor.

8. No list of references is appended to the protocol. This is highly improper, as it prevents the reader from verifying the statements in the protocol or seeking additional information.

Critique of Protocol: Administrative Procedures

On or about July 1, 1988, the NIH protocol was sent to Congressmen Robert K. Dornan and Robert Smith, along with a two-page cover letter signed by William F. Raub, then NIH Deputy Director. On August 30, 1988, it was sent to Congressman Charlie Rose. The process by which this plan was submitted is unclear. However, it appears to have deviated markedly from normal procedures required for proposals for federally funded research:

1. There is no evidence that any competitive grant applications were solicited from potential investigators. Raub had, in fact, expressed the NIH's unequivocal lack of interest in this research area in his letter to Dr. Vasapoli (cited above).

2. There is no evidence that the research area in question had been evaluated as to its priority before the proposal's appearance.

3. The Institutional Animal Care and Use Committee (IACUC) should not have been asked to approve the protocol in its present format. The protocol was described by NIH 12, 1990, as a preliminary protocol. The list of presumed authors also described it as preliminary. It is highly improper for an IACUC to approve a project without being presented detailed descriptions of the procedures that animals will undergo and a rationale for the number of animals used and the species chosen. Given that Peter Gerone was both the listed principal investigator and the Delta Center's director, it appears that the Delta IACUC's capacity to render objective judgement was compromised.

Raub clearly intended the NIH protocol to stand on its own; he described it as a finalized document. Moreover, Raub planned for its implementation with no further review. The NIH plan was not offered to the Congressmen as a draft proposal or a summary of another, more comprehensive, document. The cover letter does not state that details will follow, nor is there any suggestion in the proposal or the cover letter that a process of applying for appropriate approvals had begun. Nowhere is there any suggestion that a more detailed protocol is available for review. Raub's cover letter described the proposal as a plan which "is drawn carefully." It was represented as a plan that was to lead directly to the implementation of the research protocol described therein.

In a letter to Congressman Smith on October 13, 1988, Raub unequivocally stated his intention to immediately implement the provisions of the NIH protocol:

With respect to the most seriously disabled monkeys, we have deferred the euthanasia decision since July at your and Mr. Dornan's request. However, further delay no longer is tenable. . . . On your visit to the Delta Center in late August, you and I discussed the proposed euthanasia protocol in detail. I am grateful to you for the time you invested to understand why the experimentally disabled monkeys are of unique importance to the rehabilitation of handicapped human beings. My colleagues and I plan to proceed in accord with my July 1 letter and attachments, and my subsequent discussions with you, Mr. Dornan, and your staff members. I hope you recognize that the DHHS cannot do otherwise and still honor its mission and responsibilities.

The proposed protocol is not simply a continuation of the original experiments in progress at the time the animals were originally seized; it is a materially different project for which a new approval process is required. Raub's letters to Congressmen Doman and Smith clearly show that he intended them to believe that the plan had been carefully written and properly proposed, and that he intended to proceed. On January 14, 1990, Gerone and other investigators did so.

Conclusions

The protocol is markedly deficient in many areas. Its format and content are grossly deviant from the norms commonly accepted for scientific research. The research protocol is incomplete, poorly written, and unlikely to achieve meaningful or useful results. The frequent use of inappropriate language and extravagant claims suggests that its intent was to create an impression for lay readers rather than scientists.

This case's history shows that experiments on this group of animals have been advanced for reasons unrelated to science. James Wyngaarden described the Silver Spring Monkeys case as the most enervating part of his tenure as NIH Director. Along with his assistant William Raub, Wyngaarden played an active role in negotiations with Congress and scientific organizations on the case. Numerous scientific organizations have joined the NIH's efforts to advance various experiments on the animals, not for their scientific value, but explicitly in order to prevail in litigation or public controversies. Likewise, the NIH protocol advanced by Raub in 1988 appears to be motivated by reasons other than the advancement of scientific knowledge.

Evidence suggests that the proposal was offered with four possible aims in mind, none of which relate to scientific advancement in the area in question, but all of which assist the NIH in keeping the animals out of the activists' custody and help solve what NIH officials viewed as a major public relations problem:

1. To kill the monkeys. NIH and Delta officials have attempted to make a case for euthanasia based on poorly supported rationales. The research proposal appears to be intended to give added reason for disposing of the animals and, thus, of the possibility of setting any precedent perceived as contrary to the NIH's interests.

Killing the animals would also allow the NIH to rid itself of a longstanding public relations problem. A document from the American Physiological Society specifically calculates how killing the deafferented animals would save money for several professional societies in a plan to free the NIH from its embarrassing role in the case.

2. To indicate to Congress and the courts that the NIH has an important research interest in the animals that justifies their staying at a federal facility, rather than a sanctuary approved by animal protection groups.

3. To justify the continued diversion of NIH resources to litigation or other expenses in the case.

4. To satisfy the curiosity of the experimental neuropsychologists regarding the monkeys' brain function (even though any findings would be compromised by confounding variables and absence of controls).

Review of the NIH protocol and other aspects of this case indicates that there is a willingness on the part of NIH and Delta officials to discard scientific rigor and integrity for the sake of financial and political expediencies. Representatives of several large scientific organizations have privately indicated their willingness to assert the value of research on these animals in order to further the aim of defeating what they perceive as a threat to their own and their organizations' interests.

The use of the plan by Raub and others shows repeated misrepresentations, deception in communications with members of Congress and the courts, and abuse of the scientific process to further other aims.

A full and objective investigation of the circumstances surrounding this case is essential to restoring confidence in the capacity of NIH administrators to make proper decisions concerning the conduct of research and to enforce animal welfare laws. Such an investigation should consider, among the many other indications of misconduct:

The process by which the proposal was developed.

The persistent advocacy activity of Raub and others in this case.

The apparent compromise of scientific judgment that led Raub to repeatedly state to members of Congress and federal judges that the protocol was "carefully" drawn.

The failure to adhere to repeated promises not to conduct invasive experiments on the animals.

The process by which a principal investigator with no expertise in the relevant research area was assigned.

The process by which the protocol (which omitted important details regarding an extended invasive procedure) was approved by the Institutional Animal Care and Use Committee at the institution of which the principal investigator was the director.

The inappropriate attempts to elicit determinations for euthanasia of the animals.

The secret agreement between Tulane University, the NIH, and scientific societies regarding the monkeys and the pressure that agreement may have exerted on the euthanasia decisions.

The grossly deviant process by which the NIH proposal was advanced warrants thorough investigation. Impropriety appears to be well-entrenched in numerous scientific organizations and at the highest levels of the NIH. Robert Windom sent the protocol to Congressman Charlie Rose on August 30, 1988, and, in an accompanying letter, attempted to convince the Congressman that his objections to it were unfounded. Secretary Louis Sullivan wrote to Louisiana Governor Buddy Roemer on May 24, 1989:

Although we have kept an open mind toward the possibility of relocating a few of the monkeys, we have no intention of entering into any agreement that. . . precludes carrying out the appropriate physiological, histochemical, and histopathological determinations at the time when humane considerations dictate euthanasia.

A full and impartial investigation is urgently needed. There is every reason to believe that research funds, personnel, and materials are easily diverted to non-research purposes, and that truth and objectivity currently have low priority in much federally funded research.

References

1. Gerone PJ: The right to animal research. Animal Care Matters, National Institutes of Health, August 1989.

2. Sweet WH: Deafferentation pain in man. Appl Neurophysiol 1988;51:117-127.

3. Waxman SG: Functional recovery in diseases of the nervous system. Adv Neurol 1988; 47:1-7.

4. SeIzer ME: Nerve regeneration. Sem Neurol 1987;7:88-96.

5. Bernstein ME, Bernstein JJ: Regeneration of axons and synaptic complex formation rostral to the site of hemisection in the spinal cord of the monkey. Internat J Neurosci 1973;5:15-26.

6. Molinari GF: Why model strokes? Stroke 1988;19:1195-1197.

7. Neff S: Clinical relevance of stroke models. Stroke 1989;20:699-700.

8. Wiebers DO, Adams HP, Whisnant JP: Animal models of stroke: Are they relevant to human disease? Stroke 1990;21:1-3.

9. Merzenich MM, Kaas JH, Wall JT, Sur M, Nelson RJ, Felleman DJ: Progression of change following median nerve section in the cortical representation of the hand in areas 3b and 1 in adult owl and squirrel monkeys. Neuroscience 1983;10:639-665.

10. Merzenich MM, Nelson RJ, Kaas JH, et al: Variability in hand surface representations in areas 3b and 1 in adult owl and squirrel monkeys. J Compar Neurol 1987;258:281-296.

11. Jenkins WM, Merzenich, MM: Reorganization of neocortical representations after brain injury: A neurophysiological model of the bases of recovery from stroke. Prog Brain Res 1987;71:249-266.

12. Robinson CJ: Somatotopographic organization in the second somatosensoly area of M.
fascicularis. J Comp Neural
1980192:43-67.

13. Duncan OH, Dreyer DA, McKenna TM, Whitsel BL: Dose- and time-dependent effects of ketamine on SI neurons with cutaneous receptive fields. J Neurophysiol 1982;47:677-699.

14. McKenna TM, Whitsel BL, Dreyer DA: Anterior parietal cortical topographic organization in macaque monkey: Re-evaluation. J Neurophysiol 1982;48:289-318.

15. Stryker MP, Jenkins WM, Merzenich MM: Anesthetic state does not affect the map of the hand representation within area 3b somatosensory cortex in owl monkey. J Compar Neurol 1987;258:297-303.

16. McKinley PA, Jenkins WM, Smith JL, Merzenich MM: Age-dependent capacity for somatosensory cortex reorganization in chronic spinal cats. Brain Res 1987;428:136-139.

17. Sternberger LA: Immunocytochemistry, 3rd Ed. New York, John Wiley & Sons, 1986. human disease? Stroke 1990:21:1-3.

Return to Part I


Contents