Medical Device Daily Washington Editor

FDA's reviews of clinical trials sometimes bring to light huge missteps, and sometimes just plain old paperwork errors. The warning letters posted to the agency's web site last week seemed to capture some of both.

The trick is to figure out which is which.

An Aug. 29 warning letter to Keracure (Chicago), maker of devices for the treatment of diabetic foot ulcers, was dotted with citations for failure to notify FDA and institutional review boards (IRBs) of various adverse developments at clinical study sites, but the letter included one oddity. FDA acknowledges that it was in receipt of company correspondence dated 23 days earlier than the date of the warning letter, yet notes none of Keracure's responses to the inspectional findings.

The warning letter leads with a citation for failure to report unanticipated adverse events to the agency and to the relevant IRB within 10 days. This includes one instance in which a report involving a patient's hospitalization on June 12, 2007, apparently was never forwarded to either entity.

In another instance, the study monitor faxed an adverse event report to Keracure on Oct. 18, 2006, but the IRB received no word of the event until Nov. 28 that year, and FDA was in the dark until that Dec. 13.

The agency cited Keracure for failure to terminate the participation of clinical investigators (CIs) from the trial for deviations from the study protocol. The warning letter, which is extensively redacted, states that a site monitor noted such deviations from the protocol in four site visits between June 2006 and October 2007.

Among the deviations from the protocol were inclusion of subjects who should have been disqualified, and missing case report and consent forms. FDA cited the sponsor for failure to close the site until February 2008, "a year and a half after the initial report of protocol deviations."

Keracure is also said to lack data on whether a CI was "involved in an investigation or other research that was terminated." The company also apparently lacked financial disclosure forms for a CI and allowed another to enroll patients prior to signing an investigator agreement, with the lapsed time amounting to four months.

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

The Sept. 30 warning letter to Wellmont Holston Valley Medical Center (Kingsport, Tennessee), started out innocuously enough, but the tone turned sharp by page 2, at which point FDA states that the first citation "is a recurrence of a violation cited at the last IRB inspection and in the last warning letter issued to you in 2006."

This passage is in reference to a citation for holding IRB meetings without a quorum present. According to FDA, the hospital's IRB held meetings on Oct. 9 and Nov. 13, 2007, with only six of the 13 board members present.

The agency also alleged that a quorum was not present for "at least 13 of the IRB meetings held since March 2006" due at least in part to abstentions necessitated by conflicts of interest.

FDA found the IRB's Aug. 11 response inadequate because the hiring of a "full-time IRB support person" would not ensure a quorum, and FDA asked for a copy of procedures designed to deal with the situation.

The letter stated that a second citation, for incomplete meeting minutes, was also noted in the previous inspection and warning letter.

FDA says that the IRB "granted approval by expedited review of research for significant risk studies that did not meet the criteria of minimal risk or minor changes in research" protocols. According to the warning letter, this took place on four occasions after February 2007. The warning letter also states that the minutes from the IRB's July 10, 2007 meeting indicate that a Dr. Robert Bice approved a study "due to the cancellation of the June meeting," a move purportedly "ratified" by seven other board members.

In an e-mail response, the IRB told Medical Device Daily, "The issues cited by the FDA were of a clerical nature and not a clinical nature." The statement also notes that the Wellmont IRB has "been making improvements to our IRB policies and procedures the past several weeks, and we are confident these process improvements and additional changes we have recently enacted will make our board better and stronger than ever." The statement also notes that "our current clinical trials continue uninterrupted, and we are confident we will resume new clinical trials in the near future."

FDA's Nov. 13 warning letter to Christopher Saudek, MD, an endocrinologist at Johns Hopkins Outpatient Center (Baltimore), addressed his activities as a CI and as a sponsor. However, the heavily redacted warning letter omits much of the information, including the information on the device in question.

The only listing at clinicaltrials.gov that mentions Saudek is a study of the MiniMed insulin pump, made by Medtronic (Minneapolis).

The finding that cites Saudek as a sponsor states that he failed to notify FDA of changes to an investigational plan made in September 2004. The warning letter also states that Saudek informed the IRB of the change, but the agency asserts that the approval letter for the investigational device exemption (IDE) dated July 30, 2003, "included an attachment" that "describes your regulatory requirements" as a sponsor, which would require notification of the agency as well.

Saudek apparently informed the agency that he would henceforth update FDA on any future changes, but the agency requested "copies of policies/procedures that you have developed and implemented" to prevent future recurrences.

FDA also cited Saudek for failure to obtain signatures from three patients on updated consent forms. The letter states, "The study protocol was amended on Sept. 29, 2004, to add procedures for under-response to insulin." Most of the details are redacted, but the IRB apparently knew of the change.

The agency states further that the consent forms "lack a complete description of reasonably foreseeable risks and discomforts." As was the case with the other responses, FDA requested a copy of procedures designed to prevent recurrences.

According to FDA, Saudek failed to report to the agency and the IRB "at least 10 unanticipated adverse device effects" within the 10 working days required by regulations. However, the warning letter hinted at a paperwork hangup, suggesting that Saudek submit such reports "as IDE supplements rather than as MedWatch reports, which [the relevant procedure] appears to allow."

Saudek told MDD: "We are looking forward to resolving this quickly and take the necessary actions to do so." He said he is still corresponding with FDA on corrective actions.