On February 20, FDA released a Table of Pharmacogenetic Associations (updated February 25) and opened a docket for gathering feedback and comments.

As a leader in the pharmacogenetics (PGx) space, OneOme believes this is a positive step toward achieving consensus evidentiary standards for PGx testing. Our intent in providing the following comments was to help move toward this consensus by offering constructive feedback using our team’s extensive experience in evaluation of PGx evidence and our practical knowledge of how clinicians use PGx testing in patient care.*

Introduction

OneOme appreciates FDA's move to further precision medicine as well as the opportunity to review and provide feedback on the Table of Pharmacogenetic Associations. The following comments are intended to promote development and consensus around evidentiary standards and best practices in pharmacogenomics ("PGx").

General Feedback

  1. OneOme requests FDA publish the level of evidence criteria that was required for inclusion in each of Tables 1, 2 and 3, including requirements around demographics and in vitro and in vivo data, as well as how FDA assessed and reconciled limitations of the studies reviewed. Clarifying how FDA defines "sufficient scientific evidence" will help healthcare practitioners evaluate lab-developed PGx tests and laboratories assure their gene-drug associations meet FDA's standards.
  2. OneOme recognizes the challenges large organizations and regulatory agencies have in maintaining updates to their publications every time new, impactful research is published. However, citations are important to help providers understand what information was considered by FDA and therefore reconcile any differences between new or other existing data that may be applicable to patient care and the practice of medicine.
  3. OneOme has concerns that the lack of transparency around criteria for inclusion and absence of specific guidance from FDA related to PGx abrogates the clinical utility of this table. Moreover, it is likely that this will be misinterpreted by patients and providers as FDA's official guidance on PGx testing.

Specific feedback

Critical omissions

Phenytoin and HLA-B*15:02 and CYP2C9
Phenytoin (PHT) is metabolized by CYP2C9. Studies have associated the *3 allele with increased risk of severe cutaneous adverse reactions, particularly in patients of Chinese and Thai descent. Poor CYP2C9 metabolism of phenytoin is associated with increased plasma concentrations and probability of toxicity. The Clinical Pharmacogenetics Implementation Consortium (CPIC) and Dutch Pharmacogenetics Working Group (DPWG) have published specific guidelines regarding this association.1,2 In addition, CPIC, FDA, and the Canadian Pharmacogenomics Network for Drug Safety (CPND) noted that presence of the HLA-B*15:02 allele is associated with the risk of developing Stevens-Johnson Syndrome (SJS)/Toxic epidermal necrolysis (TEN).1,3,4 The high frequency of HLA-B*15:02 in Chinese, Thai, Taiwanese, and Indonesian populations (4%-36%) means it is important to note the risk of PHT-associated SJS/TEN in patients carrying this allele.

Confusing category assignments

FDA presented three tables, describing only the first as being supportive of therapeutic management recommendations. The rationale and practical implications of FDA's categorical divisions are unclear at the outset and become more so when, for example, citalopram and escitalopram are categorized separately, in Tables 1 and 3, respectively.

Citalopram is a racemic mixture in which the S-enantiomer, escitalopram, is the active component (167x more active than the R-enantiomer4) and favored target of CYP-mediated metabolism (CYP2C19, -3A4, and -2D6).5,6 Given the adverse drug event (ADE) and efficacy data available for citalopram, the pharmacokinetic data alone, which FDA acknowledges by inclusion of escitalopram in Table 3, should be sufficient to support the use of CYP2C19 PGx in therapeutic management with escitalopram and thus its inclusion in Table 1. The FDA tables also note only the likelihood of increased systemic drug concentrations in poor metabolizers but ignore the substantial data available indicating the likelihood of reduced systemic drug concentrations and lack of efficacy in rapid and ultrarapid metabolizers (carriers of CYP2C19*17).2,7-9

The confusion lies with the implication that the contents of Table 3 have less supporting evidence than those in Table 1. The case of citalopram/escitalopram belies this demarcation, because the racemic mixture of R and S citalopram is comprised of 50% escitalopram. Given the level of data supporting increased ADE risk with the racemic mixture demonstrated by FDA's inclusion of citalopram in Table 1, additional in vivo studies of escitalopram would unnecessarily expose human subjects to increased risk without demonstrable benefit.

Recommendations

As FDA works on future drafts of the Table of Pharmacogenetic Associations, OneOme offers the following recommendations:

  1. Collaborate with PGx experts in academia, industry, pharmacy, and clinical practice to achieve consensus on criteria and level of evidence for documented PGx associations.
  2. Develop and share information on the timelines for reviewing new evidence and schedule for updating the Table.
  3. Publish draft guidance for PGx testing to clarify how this table fits into FDA’s future plans regarding PGx.

References

  1. CPIC®. Guideline for Phenytoin and CYP2C9 and HLA-B. https://cpicpgx.org/guidelines/guideline-for-phenytoin-and-cyp2c9-and-hla-b/ Accessed 2/27/2020.
  2. Dutch Pharmacogenomic Working Group. Dutch guidelines, November 2018 update. https://api.pharmgkb.org/v1/download/file/attachment/DPWG_November_2018.pdf Accessed 2/27/2020.
  3. FDA. Dilantin (phenytoin sodium) package insert: https://www.accessdata.fda.gov/drugsatfda_docs/label/2009/084349s060lbl.pdf Accessed 2/27/2020.
  4. FDA. Celexa (citalopram) package insert: https://www.accessdata.fda.gov/drugsatfda_docs/label/2011/020822s038s040,021046s016s017lbl.pdf Accessed 2/27/2020.
  5. von Moltke, Lisa L., et al. "Citalopram and desmethylcitalopram in vitro: human cytochromes mediating transformation, and cytochrome inhibitory effects." Biological psychiatry 46.6 (1999): 839-849.
  6. Olesen, Ole V., and Kristian Linnet. "Studies on the stereoselective metabolism of citalopram by human liver microsomes and cDNA-expressed cytochrome P450 enzymes." Pharmacology 59.6 (1999): 298-309.
  7. CPIC®. Guideline for Selective Serotonin Reuptake Inhibitors and CYP2D6 and CYP2C19. https://cpicpgx.org/guidelines/guideline-for-selective-serotonin-reuptake-inhibitors-and-cyp2d6-and-cyp2c19/ Accessed 2/27/2020.
  8. Rudberg, I., et al. "Impact of the ultrarapid CYP2C19* 17 allele on serum concentration of escitalopram in psychiatric patients." Clinical Pharmacology & Therapeutics 83.2 (2008): 322-327.
  9. Mrazek, David A., et al. "CYP2C19 variation and citalopram response." Pharmacogenetics and genomics 21.1 (2011): 1.

*Note: Due to commenting platform constraints, the comments presented here include minor changes to formatting. References have also been included as part of the comments rather than listed in a separate reference sheet as they are on the docket.