Analysis of 483s/EIRs for GMP validation issues.

A change in ownership may also have contributed to the firm's problem in avoiding deviations, which ended up on a warning letter dated Dec. 11, 2000 (Doc 9764W). The firm's president-CEO, Michael Archibong, acquired Chemrich from Art Perez in 1995 and according to the EIR, the previous inspection in 1990, was rated "N/N." FDA did not explain that acronym.

Chemrich annotated the 483 as "promised to correct" each of the violations. The EIR described Chemrich as a contract manufacturer of OTC drugs, vitamins/dietary supplements and one prescription drug, an oral solution of potassium chloride.

The 483 broadly cited failure to conduct process, cleaning and water system validation as well as failure to qualify equipment for installation, operation and performance.

Hiramine opted to go into no detail in the

EIR on the process validation and equipment qualification findings, stating only that "there has to be validation of all critical systems." The agency released no document by Chemrich addressing the 483.

The Checklist Chemrich, L.A. District

Water systems Cleaning validation 

Process validation The warning letter cited Chemrich for the process validation finding separately from the water system validation and the equipment qualification, stating that the firm had no validation data for any of its products.

Later in the warning letter, district director Alonza Cruse commented that he found several observations, including the validation lapses, to be "highly significant" and informed the firm that the agency may not approve any applications for drugs until Chemrich got the operation into compliance.

Archibong's Jan. 3, 2001, response to the warning letter said that to assure that drug products had the required properties, "production procedures are written and employed during the manufacturing process" and that "instructions as to the examination of the drug product are described in the production batch records," with reviews for pH, taste, odor and homogeneity to be performed at specified points in the process and at the culmination of production.

FDA says Chemrich doesn't get it

Archibong added that Chemrich contracted with an outside entity, the name of which was redacted, to evaluate product potency. He promised the firm would continue to retain samples of drug batches for "continuing process validation."

The district's compliance officer, Mark Tucker, Ph.D., wrote on Jan. 19 that Chemrich apparently did not "clearly understand validation and equipment qualification as defined by FDA," and briefly outlined the objectives of a validation procedure. He gave Chemrich 15 days to reply.

Archibong's Feb. 5 letter to the Los Angeles district Compliance Director Thomas Sawyer noted his firm's understanding of validation did not "encompass the entire range of anticipated occurrences" and informed the agency that the firm was sorting through several consulting firms to help bring the operation into conformity. He added that Chemrich would have a consultant selected within 20 days and a plan of corrective action on paper within 30 days of contracting with a consultant.

The paper trail ended with a Feb. 7 letter from Tucker to the firm acknowledging the Feb. 5 letter and suggesting the company keep the district updated on the status of the corrections.

Hiramine's evaluation of the water system disclosed four dead legs in the system. The EIR stated that Chemrich's chief chemist, Irv Levy, informed Hiramine that Chemrich flushed the system for a purged number of minutes prior to production and that the firm conducted tests for microorganisms and total organic carbon (TOC) at "random."

The warning letter echoed the finding, adding no further detail, and Archibong's Jan. 3 letter included information on a revision of the water system so as to included isolation valves on the dead legs and "several additional tests" on system water. No other documents address this finding.

At press time, the company had not responded to calls for comment.

Chemrich Holdings, Los Angeles, CA, 10/24-26, 30 and 11/2/00, Doc. 109051M, $28.00 plus retrieval.

Pre-validation biobatch a recurring ache for Stason

What was supposed to be a follow-up inspection on Stason Pharmaceuticals of Irvine, CA, in July 2000 turned into a Sept. 26, 2000, warning letter (Doc. 9576W) for deviations from a validation protocol and for out-of-specification (OOS) outcomes in blend uniformity testing.

FDA investigators Katherine Jacobitz, Ryan Borges, Thanh Andrews and supervising investigator Terri Dodd, all of the Los Angeles district, handled the audit. Andrews did not return for the follow-up review, which took place July 28. The 483 from the July 28 evaluation ran only one page, revisiting blend uniformity issues relating to Stason's production of selegiline HCL.

The 483 for the earlier inspection cited the validation batches for selegiline for three problems, including that the biobatch was produced before the validation protocol had been finalized and that Stason did not include an evaluation of content uniformity during tableting, although "data did exist." Jim Fang, technical operations officer, stated that the company opted to employ the biobatch as a validation element because production batches were identical in size, equipment used and processing conditions.

Perhaps the major hangnail here was that "three out of five initial samples taken from the biobatch's blender failed specification" but that the investigation into the variance never made it to the validation report. The FDAers wrote further that Stason drew samples from the biobatch after the second blending and took samples from the other two batches after the third and last blending.

According to the EIR, Stason sampled the biobatch after processing but that those samples were for "information" only. Some of Stason's documents suggested that the biobatch samples used to analyze blend uniformity were from the final blending rather than from the second, and the confusion over this matter apparently served as the lynchpin for the July 28 follow-up inspection.

The Checklist Stason, L.A. District

OOS Blend Uniformity Monica Tinio, the firm's manager of regulatory affairs, wrote on July 28 that Stason would revalidate the procedure upon the next scheduled production for selegiline, but gave no date for this. The 483 from July 28 hit Stason for failure to assess composite uniformity "during tableting's beginning, middle and end" for the biobatch.

The biobatch also garnered the second 483 item of the two from July 28, which stated that Stason did not investigate the blend uniformity deviation, but conducted a second analysis of the blend, which evidently passed. The 483 indicated that this was the analysis included in the validation report and the EIR stated that Fang admitted that the data from the first analysis was omitted.

The agency purged few but key elements from the warning letter, which led off by citing the firm for "significant deviations" from the Dec. 6, 1996 validation protocol in the processing of a Nov. 17, 1995 batch of a product the name of which was redacted. This citation did not specify whether the batch in question was a biobatch.

The warning letter also noted that Stason did not investigate an OOS outcome for blend uniformity for the final blend of a product, the name of which was again purged. The Los Angeles district, which was the source of the warning, noted further the corrective actions undertaken prior to the date of the warning, but expressed concerns that Stason had merged quality control and quality assurance functions into the quality control unit. Acting district director insisted that the inspections indicated that the quality control unit was "inadequate to perform its required functions," a finding she deemed "highly significant."

The Sept. 26 letter from Harry Fan, the company's chief operating officer, echoed Tinio's promise to revalidate selegiline production at a later date, but offered no further comment. Fan's Oct. 10 reply to the warning letter complained that Stason had produced the validation batches before FDA handed out the draft guidelines on blend uniformity in August 1999 and that the agency's investigators did not explain "how the blend uniformity data failed to support our existing process validation studies." He added that the investigators did not communicate the severity of the 483 findings despite what he termed an offering of a "pro-active corrective and preventive action plan."

Fan admitted that the agency's regulatory actions caught him by surprise and requested a meeting "to better understand how the decision to issue a warning letter was reached."

Compliance officer Mark Tucker, Ph.D., of the Los Angeles district, wrote Nov. 14 that the district found proposed corrections adequate. They would be verified in a future inspection. At press time, the firm had not responded to calls for comment.

Stason Pharmaceuticals, Irvine, CA, 7/10-8/10/00, 9/26, 10/10 and 11/14/00, Doc. 109053M, $26.00 plus retrieval.