Herpes Zoster and Tofacitinib: Clinical Outcomes and the Risk of Concomitant Therapy

PATIENTS AND METHODS

The global tofacitinib RA development program consisted of 2 phase I, 9 phase II, 6 phase III, and 2 open‐label long‐term extension (LTE) studies and included a total of 6,192 tofacitinib‐treated patients with 16,839 patient‐years of tofacitinib exposure at the time of the datacut (April 2014). LTE study A3921024 data collection and analyses were still ongoing at the time of this analysis; because the study database had not yet been locked, some values may change in the final locked study database. Details of these trials with regard to tofacitinib dosing and study conduct have previously been published 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33. Patients with a history of recurrent HZ (>1 episode), disseminated HZ (single episode), or disseminated herpes simplex (single episode) were excluded from these trials. Patients who had previously experienced an HZ event involving a single dermatome were eligible to participate.

All studies were conducted in compliance with the Declaration of Helsinki, the International Conference on Harmonisation Guidelines for Good Clinical Practice, and the regulations of the relevant local countries. All enrolled patients provided written informed consent, and all participating institutions provided institutional review board or ethics committee approval prior to participation.

RESULTS

DISCUSSION

We evaluated the long‐term risk of HZ within the tofacitinib clinical development program for RA, with specific attention to the contribution of concomitant csDMARD therapy and GCs to the development of HZ. This analysis was performed following our preliminary evaluation of this question, in which we noted high rates of HZ among tofacitinib‐treated patients compared with the expected rates in other studies in RA patients receiving TNF inhibitor (TNFi) therapy 8. Our current analysis includes a greater number of patients and longer exposure to tofacitinib than in the preliminary study. Analysis of this RA cohort, with almost 17,000 patient‐years of tofacitinib exposure, suggests that use of concomitant GCs and age contribute to the risk of HZ, and that patients receiving tofacitinib at a dosage of 10 mg twice daily may be at higher risk than those using 5 mg twice daily. Data from phase III studies also indicated a trend toward higher rates of HZ in patients receiving background csDMARDs, particularly when combined with GCs. Our findings suggest that the use of the Food and Drug Administration–approved 5 mg twice daily dosage, and elimination of concomitant therapies, may represent potential risk‐reduction strategies for physicians and patients, provided disease activity remains controlled. Overall, the HZ events seemed to be clinically manageable and resolved in the majority of cases. Regardless of concomitant therapy or tofacitinib dose, patients from Japan and Korea have the highest risk of HZ compared with that in other regions. It has also previously been reported that the HZ incidence is consistent over time in tofacitinib‐treated patients, suggesting no increased risk for HZ with longer exposure to tofacitinib 10.

In our analysis of phase III studies, we observed that patients receiving tofacitinib at a dosage of 5 mg twice daily as monotherapy without GCs had a numerically 10‐fold lower incidence of HZ compared with the patients receiving tofacitinib at a dosage of 10 mg twice daily with concomitant csDMARDs and GCs (exposure time was lower in the former group, which may increase variance). The incidence of HZ in patients receiving tofacitinib 5 mg twice daily as monotherapy without GCs (crude IR 0.56 [95% CI 0.07–2.01]) was similar to the incidence reported in those receiving biologic therapies, including patients who were and those who were not receiving concomitant MTX and GCs 6, 12. GCs are a well‐described dose‐dependent risk factor for HZ worldwide 6, 7.

Several studies have also shown an increased risk of HZ in association with biologic therapies, with similar risk factors for HZ identified, including the use of GCs (6,11–13). There are limited data, however, showing that MTX itself increases the risk of HZ in patients with RA 4. A prior meta‐analysis of observational studies in RA failed to identify an association 14, and a subsequent population‐based study in Minnesota also demonstrated no increased risk with MTX 36. A study in patients with psoriasis showed no increased risk of HZ in patients treated with MTX but did document an increased risk when MTX was combined with TNF blockers 37, and an analysis of psoriatic arthritis had similar findings, in that MTX was a risk only when it was administered in combination with biologic agents 38. We could not evaluate whether MTX alone causes HZ, and although concomitant csDMARD treatment appeared to increase the incidence of HZ in phase III studies, csDMARD treatment was not identified as an independent risk factor for HZ based on the Cox proportional hazards regression model used to evaluate HZ risk factors across all phases of the tofacitinib program.

The mechanism by which tofacitinib increases the risk of HZ is not well understood but may be related to inhibition of interferon (IFN) signaling. Antiviral defenses rely on type I and II IFN signaling via the JAK/STAT pathway 39, which is inhibited by tofacitinib 40, 41. This hypothesis is supported by results from studies of sifalimumab, an anti‐IFNα monoclonal antibody, which showed a substantial dose‐dependent increase in the incidence of HZ following antibody treatment in patients with systemic lupus erythematosus (SLE) 42. An increase in the incidence of HZ was also observed in patients with SLE in a phase II study of the anti‐IFN inhibitor anifrolumab 43 and in a phase I study of rontalizumab 44. The association between MTX treatment, GC treatment, and an increased risk of HZ may also be explained by changes in IFN signaling, because concomitant GC and/or MTX treatment has been shown to suppress IFN signaling in patients with RA 45, 46. Other mechanisms, including T cell alterations and cytokine inhibition, may also be involved.

Similar to the observations in our prior analysis 8, these data showed the risk of HZ to be substantially higher in certain areas of Asia. The IRs were shown to be highest in Japan and Korea, whereas rates elsewhere in Asia were comparable to those seen in Western Europe or North America. This raises the possibility that a genetic predisposition toward HZ under the condition of JAK inhibition might exist in some of these populations. However, regional differences in HZ reporting due to different access to health care professionals may also contribute. Bing et al reported 2 polymorphisms that were suggested to be associated with a higher risk of HZ in tofacitinib‐treated patients, although these polymorphisms were relatively rare and explained only a fraction of the increased risk observed in the Japanese patients 47. It is possible that other genetic factors could explain this. However, it is unlikely that this is attributable to a difference in background varicella rates, because the seroprevalence of varicella in these Asian regions is similar to that in North America 48.

Although we are uncertain how many of the patients enrolled in this program in North America or elsewhere received shingles vaccination, because we did not collect a vaccination history prior to study entry, it is likely that very few or no patients received the vaccination prior to entering these studies. Population‐based studies suggested that vaccine uptake during the period in which these studies were conducted was very poor among patients with RA in the US 49, and the vaccine was not available for use in some countries where enrollment took place (e.g., Japan). Furthermore, the vaccine is currently contraindicated for use during tofacitinib or biologic therapy, so it is unlikely that any patients were vaccinated during the LTE studies. However, studies are ongoing to evaluate the safety and immunogenicity of this vaccine when it is given before starting tofacitinib treatment 50.

Tofacitinib appears to be associated with an increased risk of HZ compared with TNFi therapy, and other JAK inhibitors, such as baricitinib, have also been associated with higher rates of HZ 9, 51. However, direct comparisons with baricitinib are limited by the considerably shorter follow‐up period and number of patient‐years of exposure to baricitinib compared with tofacitinib. A large proportion of patients continued or only temporarily discontinued tofacitinib treatment during their HZ event, although almost all patients received antiviral therapy while receiving tofacitinib, which could potentially have protected against dissemination. According to the tofacitinib product label, treatment should be temporarily withdrawn during any serious infection, including HZ, until the infection is resolving, and patients should be closely monitored during any infection 52.

A recent analysis showed that postherpetic neuralgia occurred in 9.1% of patients with RA in whom HZ developed, with the incidence of postherpetic neuralgia increasing with age 53. In the tofacitinib development program, 7.4% of patients developed postherpetic neuralgia following their first HZ event. However, because reports of postherpetic neuralgia were not systematically requested, our findings may represent an underestimation of this complication. Interestingly, our study showed that current smoking was protective against HZ. There is some evidence that smoking can influence IFN signaling in mice 54. However, to our knowledge, our study is the first to demonstrate that smoking is protective against HZ. This finding is in contrast to at least 1 study of zoster ophthalmicus, in which smokers had a much earlier onset of HZ compared with nonsmokers 55.

In this analysis, our ability to evaluate the influence of GC dosage on the risk of HZ was limited. The exact dose of GCs and the total duration of GC use was not reported for many patients. In addition, we were unable to evaluate the influence of MTX dosage on the risk of HZ. Another limitation of this analysis was the substantially smaller number of patient‐years of tofacitinib exposure in the monotherapy group compared with the combination therapy groups, leading to less robust incidence estimates for the monotherapy group.

In summary, we have described the long‐term risk of HZ among patients treated with tofacitinib. Importantly, we found that the risk of HZ is likely to be greater in patients receiving tofacitinib in combination with GCs compared with those receiving monotherapy without GCs. Given that similar efficacy has been observed with tofacitinib in phase III clinical studies regardless of whether it is administered as monotherapy or in combination with csDMARDs and/or GCs 56, the use of tofacitinib monotherapy without GCs could represent a risk‐reduction strategy for physicians and patients with regard to HZ and provide an effective treatment strategy for reduction of the signs and symptoms of RA, provided the patient's RA remains controlled. It is also notable that HZ events were mostly nonserious and resolved with standard antiviral treatment. Furthermore, physicians should continue to consider shingles vaccination prior to starting tofacitinib or biologic therapy. Further research is necessary to understand why Japanese and Korean patients are at increased risk, as well as to understand the mechanism by which JAK inhibition combined with GCs leads to higher rates of VZV reactivation.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Winthrop had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

ROLE OF THE STUDY SPONSOR

This study was sponsored by Pfizer Inc. The manuscript was drafted, under direction from the authors, by Alice MacLachlan at Complete Medical Communications, funded by Pfizer Inc., with all authors providing subsequent critical revision. All authors interpreted the results, approved the final draft, and had the final decision to submit the manuscript for publication. Pfizer Inc. did not control the analysis or interpretation of the study results. Publication of this article was not contingent upon approval by Pfizer Inc.

Supporting information