FDA Cites Contaminated Loratadine Study In MDS Pharma Letter

MDS Pharma Services' "inadequate" approach to investigating contamination in a loratadine bioequivalence study will place future data submissions from the firm under increased scrutiny from FDA, according to a recent letter to the company.

Cautionary measures from the agency may include "additional inspections, third-party audits, and rejection of data where circumstances warrant," the letter states. FDA intends to determine on a "case-by-case basis" the additional steps necessary to assure the accuracy and validity of future data submitted by the Canadian research firm.

Concerns surrounding MDS Pharma surfaced in July 2003 during a routine FDA inspection of the company, which focused on a bioequivalence study comparing Schering-Plough's 10 mg Claritin loratadine tablets to a generic counterpart.

The inspection was part of FDA's Bioresearch Monitoring Program, which evaluates research conduct in order to assure the safety and welfare of human subjects. The agency writes that during the study - which measured loratadine concentration in plasma samples - 32 subject samples were contaminated due to "unexpectedly high analyte concentrations."

According to FDA, the company did not conduct a systematic evaluation to determine the source of the contamination, and instead "assumed that the contamination was limited to those samples with anomalous results."

"Based on this assumption, you selectively reassayed only the samples you assumed were affected, representing less than 1% of the total subject samples," the agency continues.

MDS notes that the letter is not a warning letter because the agency requests no response from the company.

The letter states that in a July 2003 response to FDA, MDS Pharma provided additional data and calculations intending to support the validity of the original plasma analysis.

However, the agency deemed the response "deficient" for several reasons, in part because the analyses focused on the degree of contamination (magnitude of inaccuracy in the quality control samples) as opposed to the extent of the contamination (the number of subject samples actually contaminated).

FDA notes also that while MDS Pharma's response calculated the quality control error at 6%, it "failed to characterize the overall performance of the assay in that at least 14 loratadine runs were rejected due to QC failure."

Additionally, FDA points out that the re-analysis of the data only excludes the nine subjects with recognized contamination while MDS Pharma "failed to demonstrate that the contamination was limited to those nine subjects."

Finally, FDA writes that the company did not provide evidence to suggest that the degree of contamination was equivalent across subject samples. "In fact, the data you submitted actually indicates that contamination was not equivalent because the original and repeat results for the selectively reassayed subject samples differed by a range of 2-fold to 133-fold."

In response to FDA's concerns, a second letter from the firm in October 2003 maintained that the company's data were "solid and valid" but conceded that the initial assumption about the source of contamination was "erroneous" and that "there are numerous places where contamination might have occurred."

The October letter also concurred with the agency's suggestion that all subject samples in the study should have been reassayed.

MDS Pharma re-analyzed all of the subject samples and submitted the new data to FDA in December 2003.

However, during a follow-up inspection by the agency in February 2004, the data relating to the experiments were missing and MDS Pharma claimed the data could not be found.

MDS says it "will continue to work closely with the agency" and that it believes it has "the commitment, people, technologies and processes in place to conduct studies in all phases of research in compliance with regulatory requirements around the world."

MDS notes also that it is "undertaking a thorough evaluation of our procedures for investigating and resolving unexpected bioanalytical events."