Incidence and Risk Factors for Vzv Infection after Allogeneic Hematopoietic Cell Transplantation in 1,045 Patients: Younger Age Less Than or Equal to 45, Occurrence of Cgvhd By NIH Consensus Criteria, and No T-Cell Depletion

Abstract

Introduction: Varicella zoster (VZV) infection is a common complication post allogeneic hematopoietic stem cell transplantation (HCT) associated with significant morbidity, such as post herpetic neuralgia and secondary bacterial infection. Strategies to diminish the incidence of VZV infection include the use of prophylactic antivirals with some controversy. Our previous studies have explored risk factors associated with the incidence of VZV infection in 192 patients (Kim, Transplant Infectious Diseases, 2007). We attempted to determine the incidence of VZV infection and to explore risk factors leading to the development of VZV infection in an extended cohort of 1,045 patients receiving alloHCT. Methods: A retrospective single center study was conducted at Princess Margaret Cancer Centre, Toronto, Canada. Medical record review was performed for 1,045 consecutive patients who had undergone an alloHCT from 2001 to 2013. VZV infection was determined by clinical features and/or microbiologic determination. The incidence of VZV infection was calculated using cumulative incidence method considering death and relapse as competing risks. Univariate and multivariate analyses were conducted using EZR to identify the risk factors for VZV infection. Results: Out of 1,045 patients, 142 cases were identified with VZV infection (13.6%) with 14.4% of VZV incidence at 5 years (95% CI, 12.3-16.8%). The median days to the diagnosis of VZV post allogeneic transplantation was 231 days (range, 27-1488 days). 86.6% of patients were diagnosed as having VZV within 2 years post transplantation. 52.1% of those who developed VZV developed post-herpetic neuralgia, while 14.8% developed disseminated VZV. A univariate analysis was conducted including the following risk factors: aGVHD grade 2-4, aGVHD grade 3-4, occurrence of cGVHD, cGVHD severity by NIH consensus criteria, diagnosis (lymphoid v. others), T-cell depletion for GVHD prophylaxis, donor (related v. unrelated), HLA (matched v. mismatched) and donor type (matched related v. matched unrelated v. mismatched). Risk factors that were significant on univariate analysis were cGVHD occurrence by NIH criteria (p<0.001, HR 3.441), progressive type onset of GVHD (p=0.001, HR 1.89,), and age (p=0.006, HR 0.98,). T cell depletion (p=0.002, HR 0.67) and matched related donor (p=0.005, HR 1.86) were also significant. Multivariate analysis confirmed that age (p=0.007, HR 0.98), occurrence of cGVHD by NIH consensus criteria (p<0.001, HR 3.07), and T-cell depletion (p=0.032, HR 0.55) were significant risk factors. ROC analysis was performed which revealed an age less than or equal to 45 to be a categorical risk factor for VZV infection. A risk score model was generated assigning a score to each risk factor. A score of 1 was assigned to younger patient with age less than or equal to 45, occurrence of cGVHD by NIH consensus criteria, and no T-cell depletion. Total score was calculated with risk score 0 (n=82, 10%), risk score 1 (n=264, 31%), risk score 2 (n=334, 40%), risk score 3 (n=163, 19%)(843 patients were used for risk score analysis from 845 patients due to missing information). Three risk groups were created: low (score 0-1, n=346, 41%), intermediate (score 2, n=334, 40%) and high (score 3, n=163, 19%). This risk score group could stratify the patients according to VZV infection (p<0.0001): 5.9% in low vs. 14.3% in intermediate vs. 25.8% in high-risk group for VZV infection at 5 yrs. Conclusions: The incidence of VZV infection was substantial at 14.4% at 5 yrs. The occurrence of cGVHD by NIH consensus criteria increases the risk of VZV infection. Of interest, younger age was also associated with increasing risk of VZV infection, while T-cell depletion was protective from VZV infection. Strategies to prevent VZV infection should be considered in the high risk group of patients for VZV infection. Further study is strongly warranted to confirm these risk factors in other cohorts.