Gene Therapy Regulation: The Need for Third Party Fiduciaries

© 2001 Peter Free

[This is page 2 of 6]    Page 1 ─ Page 3 ─ Page 4 ─ Page 5 ─ Footnotes

 

Research has also shown that most people have pre-existing adenovirus antibody. [26]  Consequently, the effectiveness of the initial administration of the adenoviral vector is reduced as the immune system intercepts and neutralizes the virus. Theoretically, immunoapheresis could be used to lower levels of the antibody circulating in the blood, thereby increasing the efficacy of gene delivery. [27]  Infusion of an anti-CD4 monoclonal antibody might accomplish the same goal. [28]

A corollary problem is the fact that adenovirus tropism is not entirely organ or site specific, and efforts to prevent systemic spread have failed. [29]  This means that therapeutic efficiency is reduced [30] and the potential for a widespread immune response increased.  Although immunoapheresis or monoclonal antibody might control the response, there remains the risk of introducing the engineered gene or its remaining wild-type companions where they are not wanted.

The Gelsinger Clinical Trial

            Gelsinger was one of eighteen patients with partial OTC deficiency chosen to participate in the IHGT phase I [31] clinical trial. [32]  Eleven patients were symptomatic before the trial began, meaning they had episodes of hyperammonemia, [33] and seven were not.  The group was divided into six cohorts of three patients each.  Adenovirus dosing escalated from 2 x 109 to 6 x 1011 particles/kg via the hepatic artery.

Three of the eleven symptomatic patients showed partial corrections of their OTC deficiency, but these were neither dose related nor statistically significant.  Every patient produced an antibody response to the adenovirus.  Low platelet and phosphate counts appeared to be dose related.  Two patients had liver enzyme levels (used as a surrogate marker for liver damage) between five and six times the upper limit of normal. [34]

Gelsinger received treatment about two and a half years after the phase I began.  His third generation adenoviral dose was at the highest level.  He became essentially comatose by the second evening.  Life support was withdrawn on the fifth day.  Autopsy showed hemorrhage and necrosis [35] in 40 to 50 percent of the liver, edematous [36] lungs with leakage of white blood cells into the alveolar [37] air spaces, acute tubular [38] necrosis in the kidneys, a completely infarcted [39] spleen, and diffuse anoxic insult to the brain.  The pathology was consistent with death from acute respiratory distress syndrome (ARDS) and disseminated intravascular coagulation (DIC) brought on by a systemic inflammatory response (immune revolt). [40]

In other words, the known toxic effect that adenovirus elicits in patients had somehow been magnified in Mr. Gelsinger.  Subsequent investigation indicated that Gelsinger's IL-6 cytokine [41] level rose earlier than that of the other subjects, and, unlike theirs, failed to return to baseline.  Adenovirus vector was found in his spleen, lymph nodes, and bone marrow.

In January 2001, the director of IHGT announced in closed meeting that a protein on the viral coat of the adenovirus vector used in Gelsinger had apparently triggered the massive immune response that killed him. [42]  The new information was based on monkey studies conducted after his death.  Researchers injected mice and monkeys with adenoviral vector tagged with fluorescent dye that enabled them to trace the virus' course.  Immune system macrophages absorbed most of the vector and set off a cytokine signaling response that activated the immune system.  The mice were unharmed, but within 24 hours, monkeys were demonstrating a significant immune reaction.  They began bleeding into their skin, indicating that disseminated intravascular coagulation was in progress.  This occurred even when the virus was killed prior to infusing it.  The research group concluded that surface proteins on the adenoviral vector were triggering the immune reaction.

Purely Medical Criticisms of the Gelsinger Protocol

            Even aside from regulatory and informed consent criticisms, the IHGT trial can be faulted for sloppy medical science that foreseeably endangered participants.  The potential for massive immune system responses in at least some patients should have been foreseen.  This is particularly true since the adenoviral vector was already known to elicit immune responses in animals and people.

Significant animal testing of the OTC vector should have preceded phase I of the human trial.  Within the limits of an animal model, [43] the results would have assisted in the assessment of the potential for adverse effects in humans.  The fact that this sort of testing apparently [44] occurred after Gelsinger's death, rather than before, means the IHGT protocol as practiced fell below an acceptable standard of care in medical experimentation.

            The basis for this criticism has to do with widely understood characteristics of the human immune system.  Using only one component of the immune system as example, it is generally accepted that the body is capable of producing antibodies to virtually any foreign substance. [45]  Small numbers of these antibodies pre-exist introduction of any invader into the body.

The question becomes not whether the invader will be recognized, but what will happen when it is.  In most cases, a controlled clonal expansion of pre-existing immune cells occurs and intra-system signaling recruits other components to repel or control the foreigner.  In some people, however, there is no response and in others an overly aggressive response.  Both of these are life threatening.

Given the high proportion of patients who have autoimmune responses to their own tissues or who have life-threatening hypersensitivity reactions [46] to substances that do not affect other people, it is clear that significant variation in immune system function exists.  One cannot make blanket predictions about how an individual will respond to immune system challenge.  This is true whether the vector is a new one or an old one administered in massive doses.  It is particularly true where the patient has likely been previously exposed to some of the antigens [47] on the vector to be administered.  If this subject has an undiscovered hypersensitivity to antigens on the therapeutic agent, the initial dosing may trigger a potentially overwhelming immune response.  In Gelsinger's case, since previous exposure to adenovirus is common among the population, hypersensitivity should have been considered. [48]

It is arguably true that people who participate in clinical trial leave behind the informal cost-benefit analysis that underlies most medical practice.  In clinical trials of a new drug, there is no way to reliably quantify the unknown.  That is, after all, the whole purpose to the trial.  Patients and researchers forfeit probabilistic certainty in forecasting what might happen.  Nevertheless, patient protection outweighs the desirability of unconstrained research, and predictable and unnecessary risks should be avoided.  In the IHGT OTC protocol, the significant risk posed by immune system mobilization was not properly anticipated and tested for.  Researchers also ignored warning points deliberately built into the human study.  This was bad medicine, bad science, and nothing more.

A Compendium of Errors Aggravated Poor Study Design

            The gist of the above evidence demonstrates that the IHGT protocol began testing human subjects before it should have and without realizing that such a step required rigorous caution in the human study.  Researchers then

(a) deviated from the experimental protocol originally approved for the study; [49]

(b) removed toxicity stopping points that had been included in the original protocols; [50]

(c) failed to incorporate patient exclusionary criteria that had already been agreed upon; [51]

(d) delayed reporting two toxicity events they had previously agreed to report immediately;

(e) failed to record in the 1999 annual report that Grade III toxicities had occurred as early as the fourth cohort; [52]

(f) apparently attempted to conceal these toxicities by omitting or averaging some test levels; [53]

(g) failed to notify the FDA of parallel animal tests that suggested significant vector risk to humans; [54]

(h) did not develop case report forms that would have enabled research staff to verify that each subject met required inclusion criteria; [55]

(i) omitted from subsequent consent forms the fact that four subjects in the fourth cohort had experienced Grade III liver enzyme elevations; [56]

(j) improperly witnessed consent forms. [57]

(k) admitted trial subjects who did not meet inclusion criteria; [58]

(l) did not conduct required lab tests on some subjects before, during, and after the trial; [59]

(m) kept incomplete and inaccurate records; [60]

(n) did not keep the institutional review board apprised of what was going on; [61] and

(o) failed to adequately monitor the study. [62]

 

[This is page 2 of 6]    Page 1 ─ Page 3 ─ Page 4 ─ Page 5 ─ Footnotes