FDA wants revalidated Paddock drug put on stability.

Paddock Labs of New Hope, MN, managed to revalidate a drugmaking procedure after a change in their talc supplier, but ended up with a 483 and a warning letter (Doc. 10070W) hitting the firm for failure to run the new validation lots through stability.

Investigators Marie Fadden and Fannie Harrell handled the January-February 2001 review, coming up with 25 deviations on the three-page 483 (the EIR was not released). The firm's president, Bruce Paddock, put together a March 2 response to the 483, which evidently put the agency's fears to rest regarding a sizeable number of the violations. Still, the Minneapolis district volleyed back speedily, issuing the warning letter on March 6 with no mention of Paddock's letter.

The fast-flying warning letter may have been prompted by what acting Minneapolis district director Cheryl Bigham described as the firm's "long history of GMP violations." She suggested the firm's management meet with FDA on March 15.

According to Fadden and Harrell, Paddock did not put any of the revalidation lots of Nystop (nystatin) on stability and committed only one lot to stability to assess bulk holding times. The 483 indicated the revalidation was prompted by a switch in the firm's supplier of talc, but no dates were given in regard to either of these events.

Bruce Paddock insisted his company had data for stability for three lots, but admitted the revalidation lots were not part of that effort. He added that one of the stability lots was the second produced after the revalidation and that five other lots had contributed samples toward stability studies. The warning letter offered no further detail, commenting only that the revalidation lots were not part of the stability program and that Paddock dropped only one lot into stability to check bulk holding times.

FDA cited the validation for Compro (prochloperazine) because Paddock failed to take samples at "start-up of the fill run" before each day's production or "after prolonged shutdowns to demonstrate" product uniformity. Paddock insisted that his firm had followed the protocol, which he stated required interval sampling determined by batch size, and that these samples returned numbers within specifications. He stated further that batches were mixed The letter added that tanks of a batch were mixed continuously overnight and was circulated through a filling head before filling operations. This citation did not appear on the warning letter.

Another item that died before the warning letter went out was that Paddock personnel could not locate any information describing a correlation between "tunnel temperature and its relationship to tunnel speed" in the Compro validation.

Fadden and Harrell wrote the firm also failed to include shut-down and start-up procedures for multiple-day packaging runs in the validation work. The firm insisted, however, that shut-down and start-up procedures were covered in the validation and that resulting data appeared in master control batch records.

On the matter of tunnel speed and temperature behavior, Paddock pushed the view that validation data gave the firm enough data to assess appropriate temperature and fill speed ranges in the cooling tunnel and that the firm had posted the data in master control batch records along with information on the relationship between tunnel temperature and speed. The company's March 2 letter stated that Paddock possessed worksheets listing the data, and indicated an enclosed document which contained the pertinent data.

Shut-down and start-up data was also described as absent for a 1999 validation for Colocort (hydrocortisone) along with documentation that samples were pulled from two lots. Bruce Paddock promised future validations would describe shut-down and startup work and commented the firm's records indicated it had retained "extra samples" from one of the two lots in question. He also committed to implementing use of "chain of custody" forms in order to track samples.

Paddock obtained a corrected-not verified annotation for a finding that data for gas chromatographs was insecure because of passwords which were posted on a "computer controller."

In an interview, Bruce Paddock said he had met with FDA and that as of June 26, "all issues have been resolved to the mutual satisfaction" of the agency and his company.

Paddock Labs, New Hope, MN, 12, 13, 14, 18-21, 1/2-5, 10, 17/01, Doc. 109103M, $7.00 plus retrieval.

RELATED ARTICLE:

[check] The Checklist - Paddock Labs

[check] validation

[check] documentation