Semler Research Center Private Limited (Bangalore, India) is suing the FDA for requiring that sponsors repeat BA/BE studies based on the outcome of an FDA inspection that raised concerns about the trustworthiness and correctness of data.

The legal complaint was filed January 18, 2018, in the US District Court Central District of California. The firm seeks $50,000,000 damages and a jury trial.

  • The complaint named eleven (11) individuals, including:
  • FDA Commissioner Gottlieb
  • Director of Office of Study Integrity and Surveillance
  • The Acting Director of the Office of Scientific Investigations within the Office of Compliance
  • Director of the Division of Enforcement and Postmarketing Safety
  • Acting Branch Chief of Compliance Enforcement Branch of Division of Enforcement and Postmarketing Safety

As well as other staff employed at the FDA, including all of the investigators participating in the inspection ending October 9, 2015.

It makes for interesting reading, particularly the section on the 2015-2016 inspection and actions by the FDA that begins on page 10.

It appears that the FDA had a definite idea of what they were seeking when they arrived at the site. And including two investigators (who have deep experience in GMP electronic records data integrity issues associated with laboratory instruments) seems unusual unless the FDA expected that they would need to review data from these instruments. We’ll get to this later.

For readers with subscriptions to the Pink Sheet, their January 25th 2018 article on the topic provides additional details.

US Action

The FDA issued a 3-page form 483 to the firm at the close of the inspection on October 9, 2015.  

Observation #1 begins ‘During the inspection, FDA investigators found an Excel spreadsheet on Semler’s server describing the substitution of plasma samples for studies…For example in study [redacted] the spreadsheet indicates that plasma samples from subject 10 who received reference product were substituted by the plasma samples from subject 41 who received test product.’

This form 483 is unusual because all seven observations address, for example:

  • Apparent lack of complete data
  • Lack of controls over laboratory computer systems so that analysts could delete, copy, and rename chromatographic data
  • Failure to retain raw data in the study folder
  • Failure to capture audit trails

Four investigators from the FDA, including Dipesh Shah and Daniel Roberts, conducted the inspection. These two individuals have been investigators in previous inspections at firms where they identified serious data integrity breaches.

The firm received an untitled letter in April 2016 based on the observations identified in this inspection.

The FDA inspected the same site on September 15th 2014 and December 13th 2013, and did not issue a form 483 for either inspection. One of the investigators at the 2014 and 2015 inspection was the same, Arindam Dasgupta.

EMA Action

As if US action was not enough, the EMA recommended suspension of medicines evaluated at the site based on the FDA and WHO inspections.

WHO conducted inspections at the site in January 2015 with follow up inspections in December 2015 and published a notice of concern on April 12th 2016 regarding their findings.

In conclusion, they state “The above problems observed at Semler and other problems described in the report, indicate the existence of a general or systematic deviation from commonly accepted quality standards, and cannot be ascribed to a single person or two working outside of the quality management system. On these grounds, the PQT [Pre-qualification team] recommends an immediate stop for all submissions of dossiers relying or in whole or in part on involvement from Semler until the underlying issues have been verified to have been adequately resolved.”

It seems most unusual that the inspections in 2013 and 2014 did not result in the issuance of a form 483. But, then in 2015, a reasonably short form 483 identified serious problems in the area of data management and resulted in serious consequences for the firm.

Though I have no data for this supposition, it seems to suggest the involvement of a whistleblower or the FDA gaining some information to convince them that a potentially serious problem existed at the site along with information identifying the nature of the problem and where to look for evidence of these problems. Details for this supposition are addressed in the following section.

Let’s start with the information provided on page 10 of the Semler complaint and look at some of the elements that make the conduct of this inspection unusual.

  • Paragraph 31 says that investigators arrived at 7:30 AM before normal operating hours.  In my experience, the FDA respects a firm’s operating hours.  
    • Also, in Paragraph 31, the complaint states that the FDA ‘insisted upon accessing the BA/BE laboratory without waiting for the responsible staff members of SRC to arrive.’
      • Again, in my experience, the FDA awaits escorts from the firm before entering non-administrative areas of the facility. This approach suggests to me that the FDA was concerned about the possible destruction of records before they could begin the inspection.
    • Also, a ‘clue’ that something unusual was afoot, Paragraph 32 in the claim states that ‘FDA Inspectors informed SRC that this was not a project-specific inspection but rather was an inspection to verify the computer validation and related aspects of BA/BE testing’ and that ‘FDA…told SRC’s staff not to retrieve any documents until they made a request for the same.’ The FDA’s previous inspections of the facility were reported to be product specific.
      • Again, the request for the firm not to retrieve any documents until the FDA requested them suggests a concern about possible record destruction.
  • Paragraph 41 gives a clue of the FDA’s focus where the complaint states ‘Multiple times FDA Inspectors went to the BA/BE instrument room and checked the Instrument Computer system connected to the instrument and compared that with the access given in the conference room.’
    • Investigators appear to be searching for a specific document.  
  • Paragraph 43 describes that an SRC employee saw one investigator looking at a picture they had taken that showed a ‘File Explorer window open to a server identified by the letter “Y” and showing a document path to a particular folder’ to which FDA investigators did not yet have access.  The complaint states that the folder in question was titled ‘DUMP.’ After more back and forth to the laboratory where a search revealed ‘the contents of one part of that file’ after which the investigator asked that the firm print out the file.
  • Paragraph 45 identifies the file as ‘…a Microsoft Excel spreadsheet file titled “March 2010 version 1xls.xls…”‘ Paragraph 46 of the complaint indicated that ‘FDA Inspectors, having seemingly found what they were looking for from the start…’
  • In Paragraph 47 of the complaint, Semler Research Corporation maintains that a rogue employee intentionally falsified data in the problematic spreadsheet referenced in the first observation of the inspection form-483. SRC believes that this act was the result of a collaboration between this ‘rogue employee’ acting with others at a neighboring firm who were perceived to be a competitor of SRC.
  • Paragraph 57 of the claim states that that ‘on June 17, 2015, someone accessing the SRC Server system…fabricated data into the Suspect Spreadsheet.’ Most of the staff were reported to have been at lunch during this time.

Regardless of whether or not this was an intentional act meant to harm a competitor, SRC appears to have deficiencies in how they manage both paper and electronic data to ensure records supporting the conduct of BA/BE studies are trustworthy and complete. For example:

  • Observation 5 identifies ‘bioanalytical data worksheets were discovered crumpled/torn in a trash pile on the floor or otherwise discarded inside the bioanalytical laboratory.’    
  • The firm failed to have the data necessary to reconstruct the study which is also a serious concern.  
  • Three LCMS instruments did not have enabled audit trails. Thus, key data on who performed specific operations and when they performed them cannot be determined. It is also not possible to determine if data was modified or deleted.  
  • Analysts had privileges in the LCMS instrument systems that allowed them to ‘delete, copy and rename chromatographic LC/MS raw data….’  
  • And, finally, it appears that the firm did not develop and maintain spreadsheets in an environment where additions, deletions, or changes could be captured in an audit trail.

Regardless of whether the addition of data to the spreadsheet was nefarious with the intent to do business harm, it does seem to indicate that SCR did not have sufficient security controls over their electronic systems. If the firm maintained the spreadsheet in a controlled environment, it should have been possible to determine who made the problematic additions to the spreadsheet in question.

We will continue to watch this case. It would be enlightening to see if the FDA EIR for the inspection reflects the same ‘story’ as the legal complaint.

SRC is not the first firm to sue the FDA. Apotex, a Canadian generic drug manufacturer, sued the FDA in 2012. They claimed to suffer harm when the FDA imposed an Import Alert based on the outcome of an inspection of their Etobicoke, Canada site ending on December 19, 2008.  In 2012, they again sued the FDA in the US District Court for the District of Columbia for what they felt were unreasonable delays in re-inspection of a manufacturing site.

It’s also not the first time that a BA/BE firm in India has been the recipient of EMA enforcement actions. In 2015, the EMA recommended that multiple medicines be suspended based on clinical studies at GVK.