Background/Purpose: Tofacitinib is an oral Janus kinase inhibitor for the
treatment of rheumatoid arthritis (RA). Rates of herpes zoster (HZ) were higher
than observed with other agents used to treat RA in tofacitinib RA development
program, especially in Japan and Korea.¹ Higher than expected HZ
rates were also recognized in the tofacitinib psoriasis (PsO) studies in Japan. It
is important to identify mechanisms of increased risk of HZ infection for
patients receiving tofacitinib. The objectives of this study were to determine
if the increased risk of HZ seen with tofacitinib are associated with genetic
risk factors, and whether the genetic risk factor can inform the mechanisms of
increased HZ rate under tofacitinib that varied across ethnicities.

Methods: We performed a trans-ethnic and trans-indication
meta-analysis of genome-wide association study in a total of 5,246
tofacitinib-treated subjects in RA, PsO Phase 2, 3 and long-term extension
(LTE) studies (data cut-off April 2014; LTE ongoing, database not locked), by
evaluating ~8 million genetic variants on HZ event (case vs control) and time
to HZ event via logistic and Cox regression, respectively.

Results: After adjusting for age, baseline absolute lymphocyte
count, genetically defined ethnicity and concomitant methotrexate use (RA
only), we identified five loci associated with either increased HZ event rate
and/or shorter time to HZ event at genome wide significance (P < 5 x
10^-8). Two associations were noteworthy based on biological
plausibility. A SNP near CD83 (and LINC01108) was associated with
increased risk of HZ events (meta-analysis OR = 3.7, P = 2.1 x 10^-8).
The frequency of the risk allele was low in Caucasians (<3%) and rare in
Japanese (<0.1%). CD83 is expressed on multiple immune subsets, is a marker of
dendritic cell maturation and B cell activation and is down-regulated in
dendritic cells by VZV infection suggesting a plausible mechanism. The second
association was near IL-17RB, and this SNP was associated with faster
rate of HZ onset (meta-analysis HR = 3.6, P = 7.6 x 10^-10).
This association was driven by a risk allele common in Japan (~17% in the
general Japanese population) but <0.2% in Caucasians. IL-17RB is
highly expressed on iNKT2 and iNKT17 cells, whereas iNKT1 cells are IL-17RB
negative and dependent on JAK-dependent IL-15. iNKT cell deficiency with low
production of interferon gamma by iNKT cells is linked to disseminated
varicella in response to vaccination in two case reports, despite a generally
intact immune system. Overexpression of IL-25, the ligand for IL-17RB,
also leads to recurrence of varicella, suggesting that the ratio of iNKT1 to
iNKT2 and 17 cells is important.

Conclusion: Genetic analysis of ~5,300 individuals treated with
tofacitinib has identified multiple loci that are associated with increased
risk of HZ. The risk allele of one of these polymorphisms is significantly more
prevalent in the Japanese population. A possible iNKT cell skewing mechanism is
suggested that will be evaluated in future studies.

REFERENCE: 1. Winthrop KL et al. Arthritis Rheumatol 2014; 66:
2675-84.

ACKNOWLEDGMENTS: All aspects of this study were funded by Pfizer Inc.