Shalom G, Zisman D, Bitterman H, Harman-Boehm I, Greenberg-Dotan S, Dreiher J, Feldhamer I, Moser H, Hammerman A, Cohen Y, Cohen AD. Systemic Therapy for Psoriasis and the Risk of Herpes ZosterA 500 000 Person-year Study. JAMA Dermatol. 2015;151(5):533-538. doi:10.1001/jamadermatol.2014.4956
The risk for herpes zoster (HZ) in patients with psoriasis treated with biologic medications or other systemic treatments has been given little attention to date.
To describe the risk for HZ in patients with psoriasis and its relation to treatment.
Design, Setting, and Participants
A cohort study was performed using the administrative database of Clalit Health Services, the largest public health care provider organization in Israel, in the setting of general community clinics, primary care and referral centers, and ambulatory and hospitalized care. We extracted information for all patients who received a psoriasis diagnosis from January 2002 to June 2013. Follow-up was conducted until the end of July 2013. The study included 95 941 patients with psoriasis in the analysis, with 522 616 person-years of follow-up. Incidence of HZ events was calculated for each systemic antipsoriatic medication provided, during a follow-up period of 11 years and 7 months. We used a generalized estimating equation Poisson regression model to examine the effect of each systemic treatment for psoriasis on HZ incidence, adjusting for age, sex, psoriasis severity, Charlson comorbidity index, steroid treatment, and socioeconomic status.
Main Outcomes and Measures
Incidence of HZ associated with systemic therapies.
In a multivariate analysis, it was observed that treatment with phototherapy (rate ratio [RR], 1.09 [95% CI, 0.62-1.93]; P = .99), methotrexate (RR, 0.98 [95% CI, 0.78-1.23]; P = .83), cyclosporine (RR, 1.16 [95% CI, 0.48-2.80]; P = .49), and biologic medications as a single agent (RR, 2.67 [95% CI, 0.69-10.3]; P = .14) was not associated with HZ. The use of combination treatment with biologic medications and methotrexate was significantly associated with an increased incidence of HZ (RR, 1.66 [95% CI, 1.08-2.57]; P = .02). The use of acitritin was associated with decreased incidence of HZ (RR, 0.69 [95% CI, 0.49-0.97]; P = .004).
Conclusions and Relevance
Physicians may need to consider offering an HZ preventive vaccine to patients receiving combination treatment with biologic medications and methotrexate, particularly if they have additional risk factors for HZ.
Herpes zoster (HZ), or shingles, is the consequence of reactivation of endogenous latent varicella zoster virus (VZV) infection within the sensory ganglia. Herpes zoster is a major public health concern. It is a substantial contributor of morbidity, disability, and chronic pain in the form of postherpetic neuralgia. Moreover, HZ complications such as dissemination and death in immune-compromised individuals are well established.1,2 Known risk factors for HZ are age older than 50 years and immune suppression. The incidence of HZ increases with age and ranges between 4 per 1000 patient-years in the fifth decade of life and 11 per 1000 patient-years in the eighth decade, with documented predilection for females.3,4 The lifetime risk of HZ is estimated at 10% to 20%.5
The association between use of biologic medications and HZ is not straightforward.6- 8 Observational studies based on several registries have reported that among patients with rheumatoid arthritis, inflammatory bowel disease, or a combination of these diagnoses, treatment with tumor necrosis factor (TNF) inhibitors was associated with an increased risk for HZ.9- 13 Complications, such as dissemination, higher rates of hospitalization, and frequent persistent postherpetic neuralgia were associated with treatment with TNF inhibitors.14,15 A US database of 407 319 patients with immune-mediated inflammatory disorders reported that patients who initiated TNF inhibitor treatment were not at higher risk for HZ compared with patients who received nonbiologic treatment regimens.16 In our previous report, there was a nonsignificant increased risk of HZ during treatment with TNF inhibitors in a cohort of 22 330 patients with psoriasis observed for 9 years.17 The purpose of the present study was to determine the risk for HZ in a cohort of patients who received a psoriasis diagnosis and assess its relation to use of traditional systemic treatments, phototherapy, and biological medications.
The study was conducted in accordance with the Declaration of Helsinki and all applicable amendments and was approved by the Clalit Health Services (CHS) Helsinki Committee. Due to the retrospective nature of the study, informed consent was not obtained. Clalit Health Services is the largest public health care provider organization in Israel, serving a population of more than 4 000 000 patients, covering 52% of the population. The CHS administrative database is an inclusive database with continuous real-time input from computerized pharmaceutical, medical, and executive operating systems. The database has accurate information on each medication prescribed and filled by the patient, including its date, dose, and mode of administration.
From January 2002 to June 2013, data regarding all patients with psoriasis diagnosed by a dermatologist in the community or during hospitalization were collected. The data included patient demographic characteristics, sex, age, Charlson comorbidity index,18 steroid exposure, drug prescriptions that were filled by patients from the pharmacy, and HZ episodes as defined in the Herpes Zoster Diagnosis subsection. Socioeconomic status was defined according to the mean socioeconomic status of patients at the clinic where each enrollee was registered.
Because the Psoriasis Area and Severity Index was not available in the CHS database, use of psoriasis-related health care services was used as an alternative in place of psoriasis severity assessment. Psoriasis-related health care services utilization burden was estimated on the basis of the number of dermatologist community clinic visits per year, outpatient clinic visits, ambulatory day-care dermatology unit hospitalizations, and dermatology department hospitalizations. Severity was calculated by the sum of scores of all visits for each patient.
The extracted information included all prescriptions for treatment of psoriasis including (1) systemic medications (acitretin, methotrexate, cyclosporine); (2) biologic medications: TNF inhibitors (infliximab, adalimumab, etanercept), ustekinumab, alefacept, and efalizumab; and (3) phototherapy (psoralen with UV-A, UV-B). To be considered “on treatment,” the patient had to have been treated with the same agent for at least 3 consecutive months for the systemic and biological medications or for at least 2 consecutive months with phototherapy. Exposure window was defined as the period from treatment initiation until 30 days following the discontinuation of that treatment and, in the case of ustekinumab, 90 days. In the case of phototherapy, exposure window was defined as the period from treatment initiation to the end of the first week after treatment termination. Treatment-free period was defined by at least 6 consecutive months without any of the listed medications, and the most recent treatment given just before the treatment-free period was considered the last dose.
An episode of HZ was defined as a diagnosis of HZ coupled with prescription of specific anti-VZV treatment. Recurrent episodes of HZ within the same 180-day period were counted as 1 episode. Episodes of HZ occurring during the exposure window were considered to be treatment related. Episodes of HZ occurring in a patient following at least 6 consecutive treatment-free months or in a patient who did not receive any of the listed medications at any time during his entire follow-up were counted as HZ episodes in the control group. None of the patients received a prophylactic VZV vaccine because it had not yet been approved for use in Israel.
Patient characteristics at baseline (age, sex, and socioeconomic status—low, intermediate, and high) and all the treatments for psoriasis (with control periods serving as a reference category) were described. For each treatment, the crude incidence of HZ was calculated as the number of events per 1000 person-years of follow-up. The effect of each treatment for psoriasis on HZ incidence was tested in a generalized estimating equation Poisson regression model adjusting for age, sex, psoriasis severity (as defined in the Methods section), Charlson comorbidity index, steroid treatment (by number of steroid prescriptions purchased per year per patient), and socioeconomic status. Categorical parameters were compared by means of χ2 tests whereas t tests were used for comparison of continuous variables among the different groups. Person-time denominators such as crude incidence rates were compared by Poisson model. All statistical tests performed were 2 sided; P < .05 was considered statistically significant; in the univariate analysis, P values were adjusted by Bonferroni correction for multiple comparisons. Statistical analysis was performed using SPSS for Windows software, version 20.0.
A total of 95 941 patients with psoriasis were included in the analysis, with 522 616 person-years of follow-up (Table 1). Male sex was predominant in the acitretin group and in the combination treatment with acitretin and phototherapy group compared with the other study groups. Female sex was predominant in the combination treatment with methotrexate and biologic medications group compared with the other study groups. Low socioeconomic status of the clinic population was more prevalent in the acitretin group compared with the UV-B, methotrexate, etanercept, and combination treatment with methotrexate and biologic medications groups. Patients treated with acitretin or methotrexate were significantly older compared with the other treatment groups or the control group. Patients treated with ustekinumab or with any of the combination treatments were more likely to have a higher psoriasis severity score. Greater Charlson comorbidity index and larger number of steroid prescriptions per year were observed in patients treated with methotrexate, cyclosporine, biologic medications as a single agent, or combination treatment with methotrexate and biologic medication.
The total crude incidence rate was 8.9/1000 patient-years for the entire study population, 8.8/1000 patient-years for the treatment-free periods (control), and 11.1/1000 patient-years for the treated patients (Table 2). Patients treated with methotrexate or combination treatment with methotrexate and biologic medications had a significantly higher incidence of HZ, compared with the control group (Table 2). The incidence of HZ in patients treated with all of the other medications (except acitretin) was statistically insignificantly higher than the incidence of HZ among the control group. In the generalized estimating equation Poisson regression model analysis, after adjustment for age, sex, psoriasis severity, Charlson comorbidity index, steroid administration per year, and socioeconomic status, combination treatment with methotrexate and biologic medications was significantly associated with an increased risk of HZ (Table 3). An increased risk of HZ was associated with age at baseline, female sex, use of corticosteroids, high and intermediate socioeconomic status, Charlson comorbidity index, and psoriasis severity. Insignificantly increased risk of HZ was observed in the infliximab treatment group (P = .07). Acitretin treatment was associated with a decreased risk for HZ.
This large population-based study comprising more than 500 000 person-years’ follow-up of 95 941 patients with psoriasis found that treatment with biologic medications as a single agent or monotherapy with a systemic agent was not associated with an increased risk for HZ. Combination treatment with methotrexate and biologic medications was associated with an increased risk for HZ. An increased risk for HZ was associated with age at baseline, female sex, use of corticosteroids, high and intermediate socioeconomic status, Charlson comorbidity index, and psoriasis severity. Acitretin treatment was associated with a decreased risk for HZ.
There is abundant literature concerning the risk for HZ in patients with rheumatoid arthritis,6,7,12,13,16,19 inflammatory bowel disease,20 or both,8 with or without the use of biologic medications. It is well documented that patients with rheumatoid arthritis are known to be at increased risk for HZ reactivation, with 2-fold to 3-fold elevated risk.12,21 Illustratively, the reported rates observed in some other immune-compromised patient groups are 30 to 50 cases per 1000 person-years in HIV-positive individuals and 30 cases per 1000 person-years in solid-organ transplant recipients.22 An extensive search of the literature for studies assessing the risk for HZ in patients with psoriasis and its relationship to the different treatment modalities yielded relatively limited data. Langley et al23 reported that less than 2% of 2475 patients treated with efalizumab for psoriasis developed HZ. Antoniou et al24 reported that 2 of 72 patients with psoriasis treated with efalizumab had an HZ episode, with a crude incidence of HZ in 28 per 1000 patients.
The present study is a follow-up study on our initial report of the first 9 years of follow-up,17 which found no significant association between treatment with biologic medications and HZ. Since the 2012 report, we have added 73 569 patients and 303 130 person-years of follow-up and our current approach differentiated between monotherapy with biologic medications and combination treatment with biologic medications and any other systemic treatment modality. Moreover, we did not overlook transient treatment cessations. As a result, we have intensified the study power and deepened its validation.
Previous studies have demonstrated that old age and treatment with corticosteroids are risk factors for HZ.5,7,12,22 These findings are consistent with our observations. Female sex was found to be associated with higher risk for HZ in the present study, as in previous reports.8,20 It is unclear whether this observation stems from an actual sex predilection rather than the possibility that women are more likely to seek medical advice for their HZ compared with men.25
Combination treatment with any of the biologic medications and methotrexate was associated with an increased risk for HZ. This observation can be explained either by additive immunosuppressive effect of both methotrexate and biologic medications or by the drug interaction leading, presumably, to increased biologic medication blood levels.
Acitretin treatment was associated with a decreased risk for HZ. This protective effect was not reported previously, and additional observations are needed to confirm this finding. We assume that the protective effect may be related to psoriasis amelioration with minimal immunosuppressive effect of the drug compared with other treatment modalities in our model.
Recommendations regarding prophylactic vaccination in patients treated with biologic medications are not uniform. Some authors suggest the use of prophylactic immunization according to serological VZV status, history of infections, or contact with varicella.25,26 Others support prophylactic vaccination against VZV in all rheumatology patients.10 Our study results might suggest the need for prophylactic vaccination against VZV in patients with psoriasis and multiple risk factors for HZ (eg, age, female sex) as part of the preparation for biologic medication treatments in the appropriate clinical setting.
The present study used a large population-based database with data regarding all underlying diagnoses and all prescriptions, as well as primary care, inpatient, and outpatient visits with their respective diagnoses, and thus has some advantages and shortcomings. We compared patients treated with different systemic modalities for psoriasis with patients free of systemic treatment. This enables differentiation of the effect of the disease per se from the treatment modality effect. Herpes zoster diagnosis in our study was based on medical records, and therefore recall bias does not affect our observation. Although biologic agent recipients are monitored more closely, detection bias is not probable in the case of HZ because HZ is presumed to be associated with symptoms severe enough for any patient to seek medical attention at least at the primary care level.
This study has some limitations. Data regarding psoriasis severity as reflected by the common Psoriasis Area and Severity Index were not available and the severity was estimated indirectly. Despite the large size of the study population, for some medications (alefacept, efalizumab, and ustekinumab) the number of treated patients was small. Another limitation of our study is the fact that in the case of systemic medications such as methotrexate or steroids the accurate dosage was not considered and drug intake was inferred only by its pharmacy acquisition; hence, there is a possibility of overestimation of medication use or underestimation of its effect.
We observed a possible increased risk for HZ in patients with psoriasis treated by a combination of methotrexate and a biologic medication. Monotherapy with biologic medications or with systemic treatment was not associated with an increased risk for HZ. These data appear to be consistent with findings reported in studies of diseases other than psoriasis. Physicians, as well as patients treated with biologic medications together with methotrexate, at a minimum, should be aware of the risk of HZ associated with these treatments. Because early treatment of HZ may decrease the severity of symptoms and the incidence of complications (eg, postherpetic neuralgia),5 it is important to alert patients about the hazard of developing HZ. Dermatologists treating patients with combination treatment with methotrexate and biologic medications may need to consider offering HZ vaccine to patients who receive such drugs, particularly if they have other risk factors for developing HZ. The live attenuated HZ vaccine, if administered, should be given before initiation of systemic or biologic treatments that increase risk of HZ.
Accepted for Publication: November 11, 2014.
Corresponding Author: Arnon D. Cohen, MD, MPH, PhD, Chief Physician's Office, Department of Quality Measurements and Research, Clalit Health Services, 101 Arlozorov St, PO Box 16250, Tel Aviv 62098, Israel (email@example.com).
Published Online: March 22, 2015. doi:10.1001/jamadermatol.2014.4956.
Author Contributions: Drs Shalom and Zisman served as co–first authors, each with equal contribution to the manuscript. Drs Shalom and A. D. Cohen had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Shalom, Zisman, Bitterman, Dreiher, Moser, A. D. Cohen.
Acquisition, analysis, or interpretation of data: Shalom, Zisman, Bitterman, Harman-Boehm, Greenberg-Dotan, Dreiher, Feldhamer, Hammerman, Y. Cohen, A. D. Cohen.
Drafting of the manuscript: Shalom, Zisman, Bitterman, Feldhamer, Hammerman, Y. Cohen, A. D. Cohen.
Critical revision of the manuscript for important intellectual content: Shalom, Zisman, Bitterman, Harman-Boehm, Greenberg-Dotan, Dreiher, Moser, Y. Cohen, A. D. Cohen.
Statistical analysis: Greenberg-Dotan,Dreiher, Feldhamer, A. D. Cohen.
Administrative, technical, or material support: Shalom, Zisman, Moser, Hammerman, A. D. Cohen.
Study supervision: Zisman, Bitterman, Y. Cohen, A. D. Cohen.
Conflict of Interest Disclosures: Dr A. D. Cohen has served as a consultant to Abbvie, Agis, Bristol-Myers Squibb, Dexcel Pharma, Dexxon, Etwal, GlaxoSmithKline, Janssen, Leo, Lev Bar, Medison, Neopharm, Novartis, Perrigo, Pfizer, Rafa, Roche, Schering-Plough, Serono, Taro, TetraPharm, Teva, and Trima and has received research grants from Novartis. No other disclosures are reported.
Previous Presentation: This study was presented at the 73rd Annual Meeting of the American Academy of Dermatology; March 22, 2015; San Francisco, California.