FDA Confronts Possible Generic Inequivalence

Two studies of patients who switched from Clozaril to generic clozapine are being evaluated by the U.S. Food and Drug Administration to determine whether the rating of the generic drug as bioequivalent (AB) to the brand-reference drug should be reconsidered and whether additional bioavailability assessment should be undertaken.

Ereshefsky et al. (2000) monitored clozapine serum levels in 21 patients stabilized on Clozaril (Novartis Pharmaceuticals Corp.) who were switched to two weeks of treatment with clozapine (Zenith Goldline Pharmaceuticals) in a cross-over design. In another study (Kluznik et al., in press), five of 45 Clozaril-treated patients relapsed after switching to eight weeks of treatment with Zenith clozapine.

Both of these studies were the subject of an article in The Wall Street Journal last October. They were also featured in a segment of psychLINK, a closed satellite broadcast to mental health professionals hosted by William Glazer, M.D., associate clinical professor of psychiatry at Harvard Medical School.

The news report of the studies drew prompt and harsh criticism from both the generic manufacturer Zenith and the Generic Pharmaceutical Association (GPhA). Zenith and Mylan Laboratories (the other generic clozapine manufacturer) also sent out "Dear Doctor" letters that reaffirmed the FDA bioequivalence rating of their products. Zenith stated that approximately 21,000 patients had been successfully switched to their product, and Mylan stated that after more than 22,000 clozapine prescriptions, only 20 adverse events have been received by their Drug Safety Surveillance Group.

Zenith and its parent company, Ivax Corporation, claimed in a press release that Ereshefsky et al.'s (2000) study "has serious design flaws and draws erroneous conclusions" and noted that it was funded by Novartis.

Carole Ben-Maimon, M.D., chair of the GPhA board of directors, declared in a press release, "The Wall Street Journal presented propaganda from a brand drug manufacturer, rather than the facts." Ben-Maimon, who also serves as vice president for science and public policy at generic drug manufacturer Teva Pharmaceutical, said that she was calling for an FDA statement "denouncing the brand industry's stream of one-sided news reports and self-serving research that have little or no scientific merit." In a letter to Janet Woodcock, M.D., director of the FDA Center for Drug Evaluation and Research, Ben-Maimon asked that the FDA issue its own statement on the controversy. She wrote, "An attack on the scientific merits of the FDA's generic drug approval program appearing in supposedly independent news media is far more serious and sinister than the usual and routine criticism of generic drugs by the brand name drug industry."

Ereshefsky indicated to Psychiatric Times that his study had not been funded to be as large as he would have liked; although it was sufficiently powered to statistically discern differences between product serum levels greater than 18%. He also acknowledged that the naturalistic design of the study could encounter more confounding variables than a strict bioequivalence study.

Ereshefsky added that his results indicate the need to further investigate the bioavailability of therapeutic dosages of clozapine products. He also pointed out that the FDA did not do any further investigations to determine the generic clozapine bioequivalence. Instead, the agency based its rating on a minimal amount of clozapine (one-half of a 25 mg tablet) administered to healthy volunteers.

Ereshefsky felt that his study results were validated, within the design limits, by the FDA "483" audit report he received shortly after the agency inspected his research site last October. "On the clinical side of the audit, it was 99 point something percent clean. No problems with consent, protection of rights, clinical documentation. Adherence to the protocol was appropriate," he stated. "On the laboratory side, our lab is CAP [College of American Pathologists] certified. We're not, and never have represented ourselves to be a GLP [Good Laboratory Practice]."

Ereshefsky elaborated that, rather than using CAP and pharmacokinetic monitoring parameters, the FDA audit applied GLP and bioequivalence testing standards, "some of which we don't meet, and we know we don't meet them." Despite this, he observed, "We're confident in our [clozapine] levels."

He explained that the differences in the serum levels of the drug products do not necessarily indicate differences in quality, but do reflect their different kinetics. "I distinguish between drugs being prescribable and interchangeable," he said, explaining that agents with different pharmacokinetic characteristics which can be used independently are not necessarily easily interchanged. He noted that Steven Shon, M.D., medical director of the Texas Department of Mental Health and Retardation, had considered the study findings when he issued a state-wide alert regarding the need for carefully monitoring patients whose clozapine products are changed.

In a conference call with PT, Gary Buehler, acting director of the FDA Office of Generics, commented, "Right now, we're evaluating the study [Ereshefsky et al., 2000] in detail." He elaborated, "When we get to the end of that review, and that includes a statistical review of the results, and in-depth review by one of our reviewers, and the investigation of the study site, we will come up with a course of action."

Buehler acknowledged that if the FDA determined that additional investigation of a product's bioavailability is warranted, procedure requires that the AB bioequivalent rating is withdrawn, and that a "B*" interim code is assigned to indicate that additional data are required before a bio-equivalence rating can be established.

The FDA rating of the Zenith 25 mg and Mylan 25 mg clozapine tablets as AB bioequivalent was assigned after finding comparable serum clozapine levels produced by one-half tablets (12.5 mg) in 24 healthy volunteers in a cross-over design. The requirement to test the 100 mg tablet in humans was waived, and that dosage was designated as AB bioequivalent, based on in-vitro dissolution and the satisfactory results with half of the 25 mg tablet.

Dale Conner, Pharm.D., director of the FDA Division of Bioequivalence, explained this procedure in the conference call. "We have a small list of products which are not to be given to normal volunteers, or even naive patients-patients who have never received it before-at the highest dosage strength. In general, we usually drop back, for multiple dosage strengths, to a lower strength that our medical officers in the medical community consider safe to be given to normal volunteers or naive patients."

Without acknowledging the specter of schizophrenia relapse as a possible consequence of inadequate bioequivalence testing of clozapine, Conner indicated that the agency is considering altering its procedures to allow for bioequivalence testing of some medications with a narrow therapeutic index in patients at therapeutic dosage.

"Actually it has little if anything to do with this situation," he said, "but more from our concern that even a 12.5 mg dose [of clozapine] may not be totally safe to give to normal volunteers."

Conner added, "When we do these studies, we try to recommend the safest study design that will get us very valid scientific results and tell us whether products are indeed bioequivalent or not."

Ereshefsky et al. (2000) identified 21 patients who had received at least three months of treatment with Clozaril and had been successfully maintained with stable dosage for at least 28 days prior to entering the study. The naturalistic study design allowed concomitant medications and dosing regimens that were unequal in morning and evening. No medication dosage changes were allowed 28 days prior to and throughout the study. After continuing Clozaril through a two-week run-in period, patients were randomized to receive either clozapine (Zenith) or Clozaril in two cross-over two-week treatment periods.

Clozapine trough serum levels were obtained on multiple days to determine steady state and multiple serum levels during the 12-hour period after the last morning dose was obtained, in order to ascertain peak drug level Cmax and the area under the curve (AUC) reflecting total drug absorbed over time.

The investigators reported significantly different peak clozapine serum levels produced by the brand and generic products. There was also a numeric difference in the AUC of the products, with a trend toward lower concentration of generic clozapine. Although this latter finding was not statistically significant, taken with the drug concentration changes measured after the morning dose, it suggested that there was a difference between the products in the amount of drug absorbed. Ereshefsky speculated to PT that if the last-day serial levels had followed the larger evening dose, rather than the morning dose that was more convenient to follow, differences in absorption might have become statistically significant.

In its press release, Zenith criticized the Ereshefsky study for the absence of blinding and for the inclusion of a patient who had used alcohol, which conflicted with the study participation criteria. Ereshefsky told PT, "Where we had exclusion criteria deviations, we had appropriate justification and approval by the sponsor."

Kluznik and colleagues (in press) compared the effectiveness and serum levels of the clozapine products in 45 patients who had taken Clozaril for a mean of 38 months. Patients, psychiatrists and psychologists who rated the patients' symptoms were blind to which medication was being administered, although nursing and pharmacy were not. After five weeks of baseline data collection with continued Clozaril maintenance, patients were randomized to two eight-week medication change phases in cross-over design. Clinical assessments were obtained every four weeks with the Clinical Global Impression (CGI) scale, the Brief Psychiatric Rating Scale (BPRS) and the Beck Depression Inventory (BDI). Serum clozapine and norclozapine levels were drawn weekly throughout the 21 weeks of study.

Kluznik and colleagues concluded that Clozaril and the generic drug were "not clinically interchangeable." The investigators found significantly lower norclozapine serum levels with the generic drug, although clozapine levels were comparable. The generic drug was associated with lower BPRS and CGI scores, but with greater improvement on the BDI.

Five patients relapsed after switching from Clozaril to generic clozapine, but none of the patients relapsed while receiving Clozaril. Kluznik et al. said that the five patients who relapsed were resistant to previous neuroleptics but had a superior response to Clozaril, "and had maintained that response without exacerbation or relapse for years in some cases."

Two of the five patients sustained complete remission of symptoms for more than a year prior to the study. In addition, 11 patients worsened during the study-nine while receiving the generic drug. These patients did not meet criteria for full relapse but exhibited increased irritability, insomnia, anger and anxiety.

Zenith stated in its press release that it had been unable to obtain a copy of Kluznik et al.'s study, but asserted, "Regardless of this study, regional treatment centers in Minnesota have successfully converted approximately 2,000 patients to Zenith Clozapine, which represents virtually the entire Minnesota population taking this medication."

A possible mechanism for symptoms to worsen after a patient is switched to a clozapine product with a slower rate of absorption is related to the relatively rapid release of clozapine from dopamine D2 receptors (Seeman and Tallerico, 1999). Ereshefsky described this as a loose D2 binding, in contrast to the relatively persistent, tight binding of haloperidol (Haldol).

Kluznik et al. (in press) concurred with a similar study (Toney et al., 2000) that changing to a clozapine formulation with a slower rate of absorption could reduce D2 receptor blockade, and thereby reduce clinical response of previously stabilized patients. Kluznik et al. also suggested, "It is possible that if the generic is absorbed too slowly, that while the AUC may not be different, the level of peak concentration never goes above a threshold needed for sufficient D2 receptor saturation."

Questions of whether brand and generic clozapine differ in pharmacokinetics, bioavailability or therapeutic effect must ultimately be resolved by the FDA. The GPhA stressed in its letter to Woodcock that the generic clozapine products were tested according to established FDA requirements prior to marketing and found to be bioequivalent to the Clozaril brand reference.

References:

References

1.

Ereshefsky L, Lam YWF, Toney GB et al. (2000), Clozapine bioequivalence in patients. New Research 710. Presented at the 153rd Annual Meeting of the American Psychiatric Association. Chicago; May 18.

2.

Kluznik JC, Walbek N, Farnsworth MG, Melstrom K (in press), Clinical effects of randomized switch of patients from Clozaril to clozapine (Zenith). J Clin Psychiatry.

3.

Seeman P, Tallerico T (1999), Rapid release of antipsychotic drugs from dopamine D2 receptors: an explanation for low receptor occupancy and early clinical relapse upon withdrawal of clozapine or quetiapine. Am J Psychiatry 156(6):876-884.

4.

Toney GB, Ereshefsky L, Lam YWF et al. (2000), Interchangeability of clozapine formulation in stabilized patients. New Research 711. Presented at the 153rd Annual Meeting of the American Psychiatric Association. Chicago; May 18.