Validation Times , 11/01/1999 1 8
Paco hit on OOS testing, process validation errors.(Paco Pharmaceutical)(Brief Article)
COPYRIGHT 1999 Washington Information Source
New Jersey District Office investigator Paul Bellamy's inspection of Paco Pharmaceutical's plant in Lakewood, NJ, unearthed five problems that filled out a one-page 483. However, the firm apparently dodged a warning letter with its response to Bellamy's observations. The inspection, conducted from July to August 1998, revealed several problems with validation of the firm's packing and processing of prescription and OTC drugs. The 483 opened with a citation for failure to validate manufacturing processes for a drug the name of which was redacted. The problem was demonstrated by out-of-specification benzoyl peroxide (BP) content.
The EIR stated that the firm had employed "downscale batch sizes" and company procedures called for one batch to be tested. However, Bellamy insisted that three batches at each strength would need to be evaluated.
In 1998, the firm changed the proportion of BP in two products, but failed to validate the change in formulation. Three of four lots of drugs based on the new formulation were rejected, but apparently went on the market.
Firm agrees to 3-batch validation
The firm's response, dated Sept. 15 and signed by Director of Quality Assurance Abraham More, indicated that the scaled-down batch size would no longer be employed and that, were the procedure to be reintroduced, the firm would draw "three consecutive validation batches for each concentration ...supported by the appropriate stability studies." More also said retained samples of the four lots in circulation were tested and met specifications.
Paco was cited with a three-part observation on the 483 concerning how it documented the handling of OOS results. According to the 483, one lot of an unnamed drug tested OOS twice, which the firm "attributed to laboratory error and the sample was retested and in-specification results were obtained." The 483 stated that the firm lacked validation data supporting the re-work procedure and that documentation was insufficient to invalidate the OOS results of the first two tests.
The EIR noted that the batch in question had been mixed for 90 minutes. Paco's SOPs called for a minimum mix time of one hour, but no maximum was designated. However, there was no data implying whether a mixing time in excess of one hour was detrimental to the mix.
Paco's response indicated a change in procedures to address the documentation deficiencies.
Another lot that tested OOS was passed along as well after a second evaluation. The firm concluded in this case that the problem was with the high-performance liquid chromatography column.
The EIR noted, however, that the difference between the OOS and the in-spec results was only .002%, which Bellamy said was indicative of a column that was "working consistently before and after the re-conditioning."
Paco's response indicated that a new column would be used in "similar future situations." The firm nonetheless insisted that a "split-peak" indicated that the column had been compromised and that the inversion and back-flushing of the column was "a common and acceptable practice."
Another instance of OOS results was attributed to "the age of the standard," which the firm replaced. The fresh standard passed the batch, but the EIR noted that there was no documentation indicating that the old standard was compromised.
FDA cited Paco for a cleaning procedure "that could not be validated." The 483 stated further that "the acceptance criteria for conductivity and residue limit could not be met." The EIR provided no additional detail.
The firm's response stated that the filler "met the acceptance criteria for active drug substance in rinse water samples; however, two of the rinse water samples did not meet the requirement for conductivity after rinsing for 15 minutes and two of the swab samples did not meet the requirement for active drug substance." The letter stated that a "re-clean" brought all results up to acceptance criteria. Paco also promised to revalidate the cleaning "during the next scheduled production."
The second cleaning validation citation was likewise lacking detail in the inspection records. The 483 and the EIR noted only that the firm lacked validation data "to support the adequacy of the cleaning procedures" for a filler, filter housing and transfer lines after a 3% Trypticase Soy Broth media fill run.
Paco said its cleaning procedures allowed filters to stay in place during cleaning. After removing the filters, cleaning efforts met acceptance criteria, but the firm noted that, "since the equipment was cleaned twice before acceptable results were obtained, the process is not considered validated."
Paco indicated that its SOP would be revised to call for "purging of the cleaning agent through air-operated valves during the cleaning cycle," which the letter stated was a validated method. Similarly, the letter promised that "a protocol will be developed to validate the cleaning of media from the filter housings and transfer lines prior to the next scheduled media fill."
Paco Pharmaceuticals, Lakewood, NJ, 7/13-17, 20-23, 27, 30, 31, & 8/3, 4, 10, 12/98, Doc. 108732M, $6.00 plus retrieval.
RELATED ARTICLE: The Checklist - Paco, New Jersey District
* OOS
* Cleaning validation
* Process validation
InfoTrac OneFile (R)
No comments:
Post a Comment