- © 2008 by American Society of Clinical Oncology
Analysis of Herpes Zoster Events Among Bortezomib-Treated Patients in the Phase III APEX Study
- Asher Chanan-Khan,
- Pieter Sonneveld,
- Michael W. Schuster,
- Edward A. Stadtmauer,
- Thierry Facon,
- Jean-Luc Harousseau,
- Dina Ben-Yehuda,
- Sagar Lonial,
- Hartmut Goldschmidt,
- Donna Reece,
- Rachel Neuwirth,
- Kenneth C. Anderson and
- Paul G. Richardson
- From the Roswell Park Cancer Institute, Buffalo; New York–Presbyterian Hospital, New York, NY; Alta Bates Cancer Center, Berkeley, CA; University of Pennsylvania Cancer Center, Philadelphia, PA; Emory University, Atlanta, GA; Millennium Pharmaceuticals Inc, Cambridge; Dana-Farber Cancer Institute, Boston, MA; University Hospital Rotterdam, Rotterdam, the Netherlands; Hospital Claude Huriez, Lille; Hotel Dieu Hospital, Nantes, France; Hadassah University Hospital, Jerusalem, Israel; Universitaetsklinikum Heidelberg, Heidelberg, Germany; and Princess Margaret Hospital, Toronto, Ontario, Canada
- Corresponding author: Asher Chanan-Khan, MD, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263; e-mail: asher.chanan-khan{at}roswellpark.org
Abstract
Purpose The aim of this subset analysis was to determine if bortezomib treatment is associated with increased incidence of varicella-zoster virus (VZV) reactivation in patients with relapsed multiple myeloma (MM).
Patients and Methods Incidence of herpes zoster was evaluated in 663 patients with relapsed MM from the phase III APEX trial comparing single-agent bortezomib with high-dose dexamethasone.
Results Bortezomib was associated with a significantly higher incidence of herpes zoster compared with dexamethasone treatment (13%, 42 of 331 v 5%, 15 of 332; P = .0002). Most herpes zoster infections were grade 1/2; incidences of grade 3/4 events (1.8% v 1.5%) and infections considered serious adverse events (1.5% v 0.9%) were similar between treatment arms, and no herpes zoster–related deaths occurred. Neither the time to onset of the herpes event nor the patients’ absolute lymphocyte counts at baseline differed significantly between arms. VZV reactivation was the only herpes viral event noted to be significantly elevated in the bortezomib treatment group compared with the dexamethasone treatment group (P = .0002). The incidence of non–VZV-related herpes viral infections was comparable between arms. No additional risk factors for herpes zoster reactivation were identified.
Conclusion Further studies are needed to explain these observations and their implications; however, for patients treated with bortezomib or bortezomib-containing regimens, the risk of VZV reactivation should be monitored and routine use of antiviral prophylaxis considered.
INTRODUCTION
Bortezomib (Velcade, Millennium Pharmaceuticals Inc, Cambridge, MA; and Johnson & Johnson Pharmaceutical Research & Development, LLC, Raritan, NJ) is a selective and reversible proteasome inhibitor with significant activity in patients with multiple myeloma (MM). Bortezomib has been shown to directly induce apoptosis of MM cells, as well as inhibit MM cell growth by interfering with cell adhesion to bone marrow stromal cells, altering cytokine secretion, and restricting angiogenesis.1,2 In addition, bortezomib enhances sensitivity and reverses resistance of MM cells to commonly used chemotherapeutic agents.1,3 Bortezomib has also been shown to have inhibitory effects on T-cell proliferation and dendritic cell function.4-6
In the phase III APEX study, bortezomib provided significantly longer time to progression, higher response rate, and improved survival compared with high-dose dexamethasone.7 In an updated analysis of APEX after extended follow-up (median, 22 months), median overall survival was 29.8 months in the bortezomib arm compared with 23.7 months in the dexamethasone arm, despite more than 62% of dexamethasone patients crossing over to receive bortezomib. Overall and complete response rates with bortezomib were 43% and 9%, respectively.8
Herpes zoster is the most common infection in patients receiving immunosuppressive agents, such as corticosteroids, and undergoing bone marrow or peripheral blood stem-cell transplantation for lymphoproliferative malignancies. Characterized by a localized painful vesicular rash, herpes zoster results from reactivation of latent varicella-zoster virus (VZV). Cell-mediated immunity (CMI) is believed to play a larger role than humoral immunity in prevention of reactivation.9 However, because MM is associated with defects in humoral immunity rather than CMI, patients with MM are not at increased risk for recurrent herpes and herpes zoster infections.10 The investigators in the phase II Clinical Response and Efficacy Study of Bortezomib in the Treatment of Refractory Myeloma (CREST) and Study of Uncontrolled Multiple Myeloma Managed with Proteasome Inhibition Therapy (SUMMIT) trials11,12 raised the possibility that bortezomib, which interferes with CMI,4 may increase the incidence of herpes zoster events, reported in 22 of 202 patients (11%) in SUMMIT and seven of 54 patients (13%) in CREST.12 Increased incidence of herpes zoster infection has also been reported in other clinical studies of bortezomib treatment of MM.13-15
This post hoc subset analysis was conducted to address whether bortezomib is associated with an increased incidence of VZV reactivation in patients with MM. Data from the phase III APEX study were used to determine the incidence of herpes zoster in patients with MM treated with bortezomib compared with high-dose dexamethasone.
PATIENTS AND METHODS
APEX Study Design
The study design of the APEX trial has been previously described (Fig. 1).7 Briefly, APEX was an international, multicenter, phase III study conducted at 93 centers in North America, Europe, and Israel. A total of 669 patients with relapsed MM were randomly assigned in a 1:1 ratio to bortezomib or high-dose dexamethasone. Eligible patients had relapsed after one to three previous treatments and had measurable disease, Karnofsky performance status (KPS) ≥ 60%, platelet count ≥ 50 × 109/L, hemoglobin ≥ 7.5 g/dL, absolute neutrophil count (ANC) ≥ 0.75 × 109/L, and creatinine clearance ≥ 20 mL/min. Bortezomib was administered intravenously at a dose of 1.3 mg/m2 on days 1, 4, 8, and 11 of a 3-week cycle for a total of eight cycles, followed by three additional cycles given on days 1, 8, 15, and 22 of a 5-week cycle, for a maximum treatment duration of 273 days. Dexamethasone was administered orally at a dose of 40 mg on days 1 to 4, 9 to 12, and 17 to 20 of a 5-week cycle for a total of four cycles, followed by five additional cycles given on days 1 to 4 of a 4-week cycle, for a maximum treatment duration of 280 days.7 Lymphopenia was not a criterion for dose modification in this study. Routine use of anti-infective prophylaxis was not mandated.
CONSORT diagram.
Analysis of Herpes Zoster
This analysis was based on 663 patients who received at least one dose of study treatment during the APEX trial—331 patients in the bortezomib arm and 332 patients in the high-dose dexamethasone arm. Patients who developed herpes zoster reactivation during the study were identified from the reported adverse events, and the following parameters were determined for each patient: severity and outcome of herpes zoster, time to onset of the infection from dosing, and absolute lymphocyte count (ALC) and ANC at the time of onset or the closest time point before and after the onset of herpes zoster. The incidence of other infections and use of anti-infective prophylaxis (defined as documented administration of antifungal, antibacterial, or antiviral agents before the onset of herpes zoster) were determined for each treatment arm. Baseline patient and disease characteristics, including the history of prior therapies, were reviewed to assess potential factors that could predispose to herpes zoster infections.
Statistical Considerations
The demographic and baseline clinical characteristics of patients in the bortezomib and high-dose dexamethasone groups with and without herpes zoster events were evaluated using descriptive statistics. The proportions of patients with herpes zoster in the bortezomib and high-dose dexamethasone groups were compared by χ2 test; the time to onset of the infection and the median ALC and ANC at both the closest time point before and after the onset of herpes zoster were compared using the Wilcoxon rank sum test. All other parameters were analyzed descriptively.
RESULTS
Herpes Zoster
The incidence of herpes zoster occurring on treatment was significantly higher in the bortezomib group compared with the high-dose dexamethasone group (13% v 5%; P = .0002; Figs 1 and 2). Most cases of herpes zoster were grade 1 or 2; incidence of grade 3 or 4 herpes zoster events was similar between the bortezomib and high-dose dexamethasone treatment groups (1.8% v 1.5%; Fig 1). Six (14%) of 42 patients with herpes zoster in the bortezomib group and five (33%) of 15 patients in the high-dose dexamethasone group had grade 3 or 4 herpes zoster. Infectious events were reported as serious adverse events if they resulted in death, were life-threatening, required initial or prolonged hospitalization, resulted in persistent or significant disability, or were considered by the investigator to be important medical events. Treatment-emergent herpes zoster was reported as a serious adverse event in five patients (1.5%) in the bortezomib group and three patients (0.9%) in the dexamethasone group (Fig 2). Approximately 1% of patients in each treatment arm discontinued therapy because of herpes zoster. No deaths due to herpes zoster occurred in either treatment arm. The median time to onset of herpes zoster was slightly shorter with bortezomib (31 days) than with dexamethasone (51 days), although the difference was not statistically significant (P = .221).
Incidence of herpes zoster by severity. Grade 3/4 events were defined by National Cancer Institute Common Toxicity Criteria version 2.0. A serious event was one that resulted in death, was life-threatening, required prolonged hospitalization, resulted in persistent or significant disability, or was an important medical event. In this study, no herpes zoster events resulted in death.
Baseline median ALC (1.45 × 109 v 1.37 × 109) and ANC (2.94 × 109 v 2.98 × 109) were comparable between bortezomib- and dexamethasone-treated patients, respectively. In each arm, baseline ALCs and ANCs were comparable between patients regardless of herpes zoster state (Table 1). Most patients who developed herpes zoster infection had below-normal ALC (< 1.5 × 109/L) during the days leading up to the onset of the event; however, ALC was not assessed on the day of onset for some patients, making it difficult to determine the correlation between the onset of herpes zoster and ALC. Notably, the median ALCs and ANCs measured at the closest time point before the herpes event were significantly lower in bortezomib-treated patients compared with dexamethasone-treated patients (P < .001; Table 2).
Baseline Characteristics of Patients
Median ALC and ANC of Patients With Herpes Zoster
Patient Characteristics
Demographic and baseline characteristics were similar between the bortezomib and high-dose dexamethasone groups in both subsets of patients with herpes zoster reactivation and without (Table 1). A multivariate analysis did not show an association between risk of herpes zoster and KPS; prior history of VZV; baseline levels of beta-2 microglobulin, hemoglobin, or platelets; prior lines of therapy; or baseline ALC or ANC values.
The frequencies of prior therapies containing corticosteroids, alkylating agents, anthracycline, or thalidomide were comparable among patients with herpes zoster in both treatment groups and were also comparable with those of the subgroup of patients who did not develop herpes zoster within the corresponding treatment arms (Table 1). The median times from transplant to study entry were also comparable among patients with herpes zoster (28 months v 20 months in the bortezomib and dexamethasone groups, respectively, and 29 months v 26 months in the corresponding groups without herpes zoster).
The proportion of patients with a reported prior history of herpes zoster was 13% in the bortezomib group and 14% in the dexamethasone group. In the subgroup of patients who developed herpes zoster, twice as many patients in the bortezomib arm (six of 42 patients; 14%) as in the dexamethasone arm (one of 15 patients; 7%) had a prior history of this infection. Of all patients with a history of reactivation, only seven of 90 patients experienced reactivation after treatment, suggesting that prior herpes zoster is not associated with reactivation.
Other Infections
Although the incidence of herpes zoster was higher in the bortezomib arm, the overall incidence of infections was similar between the two treatment groups (Table 3). Overall, 77 patients (23%) in the bortezomib group had an infection compared with 67 patients (20%) in the high-dose dexamethasone group. Analysis of non-VZV herpes viral infections showed a comparable rate between arms. Twenty-eight (8%) of 331 patients in the bortezomib group had non-VZV herpes viral infections, compared with 20 of 332 patients in the dexamethasone group (6%; P = .2264). Interestingly, patients treated with dexamethasone (35 of 332 patients; 11%) had a significantly higher incidence of Candida and other fungal infections than those in the bortezomib group (13 of 331 patients; 4%; P = .001). Although glucocorticoids, such as dexamethasone, have previously been associated with increased risk of fungal infections,16 these results demonstrate a distinction between the immunosuppressive effects of bortezomib and dexamethasone.
Incidence of Herpes Viral and Fungal Infections in the APEX Trial
Prophylactic treatment with systemic antifungal, antiviral, and antibacterial agents was less common among patients treated with bortezomib than among those treated with dexamethasone (25% v 46%, respectively). Specifically, antiviral therapies were documented to have been given for prophylaxis in 12 patients (4%) in the bortezomib group and 23 patients (7%) in the dexamethasone group. Within the subgroup of patients who developed herpes zoster, two (5%) of 42 and one (7%) of 15 patients in the bortezomib and dexamethasone arms, respectively, had received prophylactic treatment with systemic antiviral therapies.
DISCUSSION
This subset analysis demonstrated that an increase in herpes zoster events was significantly associated with bortezomib therapy in the APEX trial, confirming similar observations made during phase II trials. In APEX, the incidence of herpes zoster with bortezomib was 13%, comparable with 11% reported for the heavily pretreated cohort with relapsed and refractory disease in the phase II SUMMIT trial, and 13% reported for patients with relapsed or refractory MM in the phase II CREST trial.11,12 In addition, other studies have shown an association of bortezomib with herpes zoster, with reactivation rates ranging from 12% to 57% of patients treated for relapsed MM.13-15 VZV reactivation was the only herpes viral event that was significantly increased in the bortezomib arm compared with the dexamethasone arm. This may be indicative of bortezomib sensitizing patients to viral reactivation as opposed to de novo viral infection.
Multivariate analyses using potential prognostic factors of KPS, prior history of VZV, baseline beta-2 microglobulin, baseline hemoglobin, baseline platelets, prior lines of therapy, and baseline ALC and ANC values did not show any factors associated with a risk of herpes zoster, except for the positive correlation of bortezomib treatment with an increased risk of developing HZV. Thus, no apparent predisposing factors for the development of herpes zoster could be identified.
The mechanism of VZV reactivation and herpes zoster development is not fully understood. The increased incidence of VZV reactivation seen in elderly and immunocompromised patients suggests that it may be due in part to VZV-specific host immunodeficiency.17 Patients with decreased CMI are more likely to experience disseminated VZV infection with extensive skin lesions and risks for VZV-related organ involvement.18 VZV-specific T cells appear to be necessary for suppressing VZV reactivation and preventing the development of herpes zoster. Several studies have reported that patients with MM—even with previously untreated disease—present with significantly decreased numbers of activated CD4-positive T cells and natural-killer cells compared with age-matched healthy individuals.19,20 In one study, CD4-positive T-cell subsets declined substantially with each successive line of conventional chemotherapy, suggesting that the patient's treatment status can further alter immune cell status.20 Importantly, that study demonstrated that both naive and activated CD4-positive T-cell subsets were significantly lower among patients with MM who developed opportunistic infections than among patients without such infections.
Although data are limited, it has been suggested that bortezomib treatment may alter the number and function of specific lymphocyte subsets. A recent study in 40 patients with MM demonstrated that after two cycles of treatment with bortezomib, CD56-positive natural-killer cells and CD8-positive T cells decreased from pretreatment levels by 38% and 18%, respectively.21 In that study, the CD4-positive T cells were not notably decreased with bortezomib, and no significant changes were apparent in levels of key Th1 or Th2 cytokines. It is possible that the function of VZV-specific CD4-positive T cells may be reduced or altered as a result of bortezomib therapy, without affecting the total number of CD4-positive T cells.22 Recent preclinical studies also demonstrate that bortezomib may alter the function and decrease the viability of dendritic cells, major antigen-presenting cells involved in the initiation of antiviral immune responses.5,23 It is plausible that bortezomib alters the function and/or interactions of key immune cells required for the suppression of VZV reactivation. However, it is unclear whether T cells and/or other immune cells were affected during bortezomib therapy in the APEX study because prospective measurements of different lymphocyte subsets and other immune cells were not collected.
This subset analysis showed an expected increase in total fungal infections in the dexamethasone group relative to the bortezomib group (35 of 332 patients, 11%, v 13 of 331 patients, 4%; P = .0034). Dexamethasone and other glucocorticoids increase the risk of fungal infection by depleting mononuclear leukocytes and circulating CD4-positive T cells,16 whereas thalidomide, another therapeutic option for relapsed/refractory MM, is a potent stimulator of CD8-positive T cells, synergizing with the T-cell receptor complex to increase T-cell proliferation and interferon-γ production.24 Dexamethasone and thalidomide treatments have not been reported to be associated with increased risk of VZV. These findings suggest that general immunosuppression was not responsible for increased risk of herpes zoster; rather, increased incidence appears to be a specific consequence of bortezomib therapy. It is unknown whether immune modulation by dexamethasone or thalidomide is protective against herpes zoster reactivation or if the specific immunosuppressive effect of bortezomib is sufficient for reactivation.
Currently, there is no consensus on the need for antiviral prophylaxis in patients undergoing therapy for MM. Although our analysis of the APEX study is limited by its post hoc nature, a higher proportion of patients in the dexamethasone group were reported to have received antiviral prophylaxis (7% v 4%). Limited data from ongoing studies suggest that prophylactic treatment with acyclovir may be effective in suppressing VZV reactivation in patients treated with bortezomib. In one recent report (n = 36), antiviral prophylaxis with acyclovir 400 mg three times daily was effective in preventing herpes zoster among patients with relapsed MM treated with bortezomib monotherapy.25 Published results from a phase I/II study of bortezomib combination therapy (with melphalan and prednisone) in the first-line treatment of elderly patients with MM suggest that antiviral prophylaxis may be beneficial in preventing the development of herpes zoster.26 The initial 38 patients enrolled in that study did not receive antiviral prophylaxis, but after the increased rate of herpes zoster (13%) was recognized, subsequently enrolled patients received prophylactic therapy with acyclovir. Among the 30 subsequently enrolled patients, the incidence of herpes zoster was reduced to 7%.26
In summary, bortezomib therapy was not associated with an overall increase in infections in our study, although increased episodes of herpes zoster infection were reported. Most of these incidents were mild or moderate compared with those occurring with high-dose dexamethasone. The reasons for the increase in herpes zoster, however, are unclear. No correlation was seen between the degree of immunosuppression in bortezomib-treated patients and the development of herpes zoster. As judged by the ALC levels, the degree of immunosuppression was not severe, with median ALC levels above 1.0 × 109/L, and no predisposing factors were identified. It has been observed that the 20S subunit of the proteasome is localized to dorsal root ganglia neuronal cell bodies. In vivo mouse studies suggest that proteasome inhibitor-associated neuropathy may result from damage to the neuronal cell body in the dorsal root ganglia. Coincidentally, latent VZV is found in the dorsal root ganglia; thus one could speculate that the VZV reactivation may be related to the bortezomib-induced damage in the dorsal root ganglia neuronal cell body. Further prospective clinical trials are needed to identify patients at risk and to determine the benefit of routine antiviral prophylaxis in those patients, as well as laboratory studies to better understand potential mechanisms involved. Nonetheless, because increased rates of herpes zoster have been reported here and in other settings, antiviral prophylaxis with acyclovir may be beneficial in reducing the incidence of herpes zoster infection and should be considered for patients with MM who receive bortezomib-based therapies.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: Rachel Neuwirth, Millennium Pharmaceuticals Inc (C) Consultant or Advisory Role: Asher Chanan-Khan, Millennium Pharmaceuticals Inc (U); Thierry Facon, Celgene Corp (U), Pharmion (U), Janssen-Cilag (U); Jean-Luc Harousseau, OrthoBiotech (U), Pharmion (U), Celgene (U); Sagar Lonial, Millennium Pharmaceuticals Inc (C), Celgene Corp (C); Donna Reece, Ortho Biotech Canada (C) Stock Ownership: None Honoraria: Pieter Sonneveld, Johnson & Johnson, Millennium Pharmaceuticals Inc; Michael W. Schuster, Millennium Pharmaceuticals Inc, Celgene Corp; Edward A. Stadtmauer, Millennium Pharmaceuticals Inc, Celgene Corp; Hartmut Goldschmidt, Celgene Corp, Millennium Pharmaceuticals Inc; Donna Reece, Ortho Biotech Canada Research Funding: Jean-Luc Harousseau, Millennium Pharmaceuticals Inc; Sagar Lonial, Millennium Pharmaceuticals Inc; Hartmut Goldschmidt, Millennium Pharmaceuticals Inc Expert Testimony: None Other Remuneration: Hartmut Goldschmidt, Celgene Corp
AUTHOR CONTRIBUTIONS
Conception and design: Asher Chanan-Khan, Paul G. Richardson
Provision of study materials or patients: Pieter Sonneveld, Michael W. Schuster, Edward A. Stadtmauer, Thierry Facon, Jean-Luc Harousseau, Dina Ben-Yehuda, Sagar Lonial, Hartmut Goldschmidt, Donna Reece, Kenneth C. Anderson, Paul G. Richardson
Collection and assembly of data: Asher Chanan-Khan, Pieter Sonneveld, Michael W. Schuster, Edward A. Stadtmauer, Thierry Facon, Jean-Luc Harousseau, Dina Ben-Yehuda, Sagar Lonial, Hartmut Goldschmidt, Donna Reece, Kenneth C. Anderson, Paul G. Richardson
Data analysis and interpretation: Asher Chanan-Khan, Rachel Neuwirth, Paul G. Richardson
Manuscript writing: Asher Chanan-Khan, Sagar Lonial, Paul G. Richardson
Final approval of manuscript: Asher Chanan-Khan, Sagar Lonial, Paul G. Richardson
Footnotes
-
published online ahead of print at www.jco.org on August 18, 2008
-
Supported by Millennium Pharmaceuticals Inc.
-
Presented at the 10th Annual Congress of the European Society of Haematology, Stockholm, Sweden, June 2-5, 2005; and at the Annual Meeting of the American Society of Hematology, Orlando, FL, Dec 9-12, 2006.
-
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received October 19, 2007.
- Accepted June 6, 2008.
