Alemtuzumab (Campath-1H), a monoclonal antibody that targets the CD52 antigen, has been approved for the treatment of fludarabine-refractory chronic lymphocytic leukemia. However, the profound immunosuppression caused by alemtuzumab has been associated with infectious complications.
The authors report on the incidence and risk factors for development of symptomatic cytomegalovirus reactivation in 113 patients with chronic lymphoproliferative disorders who received alemtuzumab-based therapy. Kaplan-Meier methods were applied to generate survival curves, and the log-rank test was used to assess the difference between groups; in addition, univariate and multivariate Cox proportional hazards models were used to estimate the hazard ratio of death including 95% confidence intervals.
Cytomegalovirus reactivation was diagnosed in 25 patients (22%), and most of those patients responded to antiviral therapy. Nine additional patients had asymptomatic cytomegalovirus viremia.
Alemtuzumab (Campath-1H) is a monoclonal antibody that targets the CD52 antigen, a protein found on both B and T lymphocytes. CD52 is highly expressed on cells from patients with chronic lymphocytic leukemia (CLL). This antibody was approved for the treatment of fludarabine refractory CLL by the U.S. Food and Drug Administration in 2001. In the pivotal study of alemtuzumab in patients with refractory CLL,1 the overall response (OR) rate was 33%, and grade 3 and 4 infectious complications were observed in 27% of patients. Alemtuzumab also has been used as front-line therapy in patients with CLL.2 It is particularly effective in clearing CLL cells from bone marrow and is less effective in reducing bulky adenopathy.2
Because of the marked lymphocyte depletion caused by alemtuzumab, opportunistic infections are a concern in patients who are treated with this antibody. Cytomegalovirus (CMV) viremia has emerged as an important infectious complication of treatment with alemtuzumab. Human CMV remains latent after a primary infection in an immune-competent host. Down-modulation of Class II major histocompatibility antigens by the virus creates a state permissive of a latent infection,3 and reactivation is suppressed by the host immune system.4 Compromised immunity favors CMV reactivation in the host with viral replication and development of clinical symptoms. Alemtuzumab therapy for lymphoproliferative disorders can lead to the release of inflammatory mediators like tumor necrosis factor α from dying malignant cells and to profound immune suppression from the killing of CD52-expressing normal T and B lymphocytes, natural killer cells, and monocytes; both situations are conducive toward CMV reactivation.
In the pivotal study among patients with refractory CLL, CMV reactivation was reported in 7.5%1 of patients; however, the incidence probably was under reported, because this infection was just being recognized as a complication of treatment with alemtuzumab. In the salvage setting, the reported incidence of CMV reactivation with alemtuzumab therapy has ranged from 20% to 43% of patients5–7; whereas, in the front-line setting, CMV reactivation has been reported in 10% of patients.2 In one prospective study8 that included a small cohort of patients who were monitored for CMV reactivation by using assessment of CMV antigen levels in peripheral blood leukocytes or by polymerase chain reaction (PCR), the reported incidence of CMV reactivation was 66%. However, no patient in that study had symptomatic CMV reactivation, because pre-emptive therapy with oral ganciclovir was started promptly when patients had laboratory evidence of CMV reactivation. In this article, we report a retrospective analysis of symptomatic CMV reactivation in patients with chronic lymphoproliferative diseases who received alemtuzumab-based therapy, and we describe our analysis of the potential factors that may predict for this event.
MATERIALS AND METHODS
Between 1999 and 2002, patients (Eastern Cooperative Oncology Group performance status ≤2) with CD52-positive, recurrent or refractory chronic lymphoid malignancies who had a < 20% predicted probability of achieving a response with conventional therapy were enrolled into 2 trials of alemtuzumab (Campath-1H)-based treatment. CD52 positivity was defined as CD52 expression in > 20% of malignant cells by flow cytometry or immunohistochemistry. Untreated patients with similar diagnoses who had a < 20% predicted probability of achieving a response with conventional therapy also were eligible. Apart from B-cell CLL (B-CLL), both studies included patients who had a diagnosis of T-cell prolymphocytic leukemia (T-PLL) or B-cell PLL (B-PLL), marginal zone leukemia/lymphoma, CLL with evidence of transformation (Richter transformation), and large granular lymphocytic leukemia (LGL). Patients who had received prior treatment with alemtuzumab but who developed recurrent disease >6 months after their last dose of alemtuzumab also were eligible.
Alemtuzumab was infused intravenously at doses of 3 mg, 10 mg, and 30 mg on Days 3, 4, and 5, respectively of Week 1 in both trials. In the first trial,5 alemtuzumab subsequently was administered at 30 mg intravenously 3 times per week for up to 12 weeks. Patients in the second trial9 received rituximab at a dose of 375 mg/m2 intravenously on Days 1, 8, 15, and 22. After the first week, alemtuzumab was administered at a dose of 30 mg intravenously on Days 3 and 5 of Weeks 2, 3, and 4 for a total of 7 doses of 30 mg each.
Patient received a trimethoprim/sulfamethoxazole double-strength tablet orally twice daily 3 times per week for pneumocystis carinii prophylaxis and received antiviral prophylaxis with valacyclovir 500 mg orally daily. Patients who were enrolled in the trial with alemtuzumab and rituximab were screened for CMV antigenemia after 4 weeks of therapy or when CMV reactivation was suspected (eg, fever of unknown origin, pneumonia); whereas enrolled patients who received alemtuzumab alone were evaluated for CMV antigenemia only if they were symptomatic.
Laboratory Testing for CMV Antigenemia
CMV antigenemia assay for pp65 antigen was performed with a CMV light kit (Chemicon International, Temecula, Calif) according to the manufacturer's instructions. This was performed by direct fluorescent antigen detection with the use of a monoclonal antibody directed against the CMV pp65 structural protein. Appropriate positive and negative controls were used.
Pretreatment characteristics were recorded prospectively for all patients. Descriptive statistics were calculated; Wilcoxon rank-sum tests were used to assess differences in continuous variables between groups; chi-square tests or Fisher exact tests were used to assess the association between categorical variables; and univariate and multivariate logistic regression models were used to estimate the odds ratios of CMV reactivation and the response to alemtuzumab based therapy along with 95% confidence intervals. In the multivariate logistic regression model, age, sex, Rai stage, number of prior therapies, fludarabine-refractory disease, response to prior therapy, number of lymph node sites involved, liver and spleen size, hemoglobin, white blood cell count, platelet count, serum albumin, lactate dehydrogenase, alkaline phosphatase, β-2 microglobulin (β2M), absolute lymphocyte count and percentage, CD4 and CD8 counts and their percentages, and percentage of bone marrow lymphocytes were included; and a stepwise model selection method was applied. Kaplan-Meier methods were used to generate the survival curves; then, the log-rank test was used to access the difference between groups. In addition, univariate and multivariate Cox proportional hazards models were used to estimate the hazard ratio of death, including 95% confidence intervals. All computations were carried out using SAS software (version 9.1; SAS Institute Inc, Cary, NC) or S-Plus software (version 7.0; Insightful Corp., Seattle, Wash).
One hundred twelve patients who were treated between May 1999 and March 2002 with alemtuzumab with or without rituximab (73 patients received alemtuzumab alone,5 and 39 patients received alemtuzumab plus rituximab9) were included in this analysis. Nine patients with asymptomatic CMV reactivation (CMV antigenemia detected on screening) were excluded from this analysis. Patient characteristics are summarized in Table 1. The median age was 62 years (range, 45–83 years). B-CLL was the most frequent diagnosis (59%) followed by T-PLL (14%). The majority of patients had Rai stage 3 or 4 disease (75%), and nearly half of patients were fludarabine refractory (49.5%). Most patients were extensively pretreated (median number of prior therapies = 3). Nine previously untreated patients (8 PLL and 1 LGL with erythroid hypoplasia) were enrolled because of their overall predicted poor response rates to conventional therapy (PLL) or lack of standard of care treatment (LGL).
Table 1. Patient Characteristics (N = 112)
With CMV reactivation (n = 24; 22%)
Without CMV reactivation (n = 88; 78%)
CMV indicates cytomegalovirus; WBC, white blood cells.
Symptomatic CMV reactivation was documented in 24 patients (21%). Reactivation occurred in 18 of 78 patients (24%) who were treated with alemtuzumab alone versus 6 of 39 patients (15.3%) who were treated with alemtuzumab and rituximab (P = .23). The median numbers of doses of alemtuzumab in these 2 treatment groups were 12 and 9, respectively. The median number of days from start of therapy to development of symptomatic CMV antigenemia was 27 days (range, 4–55 days). The numbers of cells per 1 million leukocytes in peripheral blood that were positive by immunofluorescence for pp65 antigen ranged from 1 to 2527 positive cells per 1 million leukocytes (median, 93 positive cells per 1 million leukocytes).
Concomitant radiologic changes in the lung were observed in 4 patients, and 1 of those patients had cytopathic changes associated with CMV reactivation documented at bronchoscopy. Bronchoscopy was not done in the other 3 patients; 1 of them had Xanthomonas bacteremia. No patient developed CMV colitis or hepatitis.
Treatment Outcomes of CMV Reactivation
Four patients with CMV reactivation died, including 1 patient with bacteremia (Staphylococcus aureus and Clostridium difficile colitis), 1 patient with polymicrobial pneumonia (Xanthomonas and atypical mycobacteria), and 2 patients with progressive CLL. Eleven patients became negative for the CMV antigen with ganciclovir treatment alone (10 patients received intravenous ganciclovir, and 1 patient received oral ganciclovir), and 5 patients required the addition of intravenous immune globulin (IVIG). One patient responded to initial treatment with foscarnet and IVIG, and 1 patient responded to oral valganciclovir. The only patient who had both CMV reactivation and CMV-associated cytopathic changes in a bronchoscopic specimen did not respond to ganciclovir but eventually responded to foscarnet and IVIG. One patient who had fever and CMV antigen detected from a nasal wash improved on antibacterial antibiotics alone without requiring antiviral therapy.
Of the 9 patients with asymptomatic CMV antigenemia who were not included in the analysis, 5 patients became negative for the CMV antigen with antiviral therapy, 3 patients remained asymptomatic without any antiviral therapy, and 1 patient died of Richter transformation.
In univariate analysis, only a low serum albumin level (entered as a continuous variable) predicted for CMV reactivation (P = .01): The risk of CMV reactivation decreased by 70% for each 1-g/dL increase in serum albumin level (Table 2). Whereas the incidence of CMV reactivation was 18% for patients who had serum albumin levels ≥3 g/dL, it was 50% for patients who had serum albumin levels <3 g/dL (P = .027; odds ratio, 4.6; 95% confidence interval, 1.2–17.4) (Table 3). The range and median values of serum albumin were comparable between patients with and without CMV reactivation (Table 1). Serum β2M levels did not predict for CMV reactivation.
Table 2. Univariate Logistic Regression Model Estimating Association Between Pretreatment Characteristics and Cytomegalovirus Reactivation
OR (95% CI)
OR indicates odds ratio; 95% CI, 95% confidence interval; WBC, white blood cells.
Age, 1-y increase
Sex, men vs women
Percentage CD4, 1% increase
Percentage CD8, 1% increase
Hemoglobin, 1 g/dL increase
Platelets, 109/dL increase
WBC, each 109/dL increase
Albumin, 1 g/dL increase
Absolute lymphocyte count, increase by 1000/dL
No. of prior treatments, increase by 1
Rai stage, 3–4 vs 0–2
Response to prior therapy, refractory vs not refractory or no prior therapy
Response to prior fludarabine, refractory vs not refractory or no prior fludarabine
Table 3. Incidence of Cytomegalovirus Reactivation According to Pretreatment Serum Albumin Level
There were no differences in overall survival (P = .87) (Fig. 1) or response to treatment with alemtuzumab-based therapy between patients who did and did not develop CMV reactivation (P = .7). The pretreatment serum albumin level, along with the number of prior treatments, response to any prior treatment (refractory vs not refractory or no prior therapy), and fludarabine-refractory disease (fludarabine refractory vs not refractory or no prior fludarabine), was associated significantly with response to alemtuzumab-based therapy. With a 1-g/dL increase in albumin, the rate of response increased 1.9 times (P = .019); with >1 prior treatment, the rate of response decreased approximately 31% (P = .001). Patients who were refractory to prior treatment had a 71% decreased rate of response compared with patients who were not refractory to prior treatment or who did not receive prior treatment (P = .002). Patients who were refractory to fludarabine had an approximately 68% lower rate of response compared with patients who were not refractory to fludarabine or who did not receive prior treatment (P = .005).
Alemtuzumab treatment results in significant immunosuppression; normal T cells and malignant B and T cells are depleted by this antibody.10 CMV reactivation in patients who are treated with alemtuzumab likely results from profound T-cell suppression. A high incidence of CMV reactivation has been reported in patients undergoing allogeneic nonmyeloablative hematopoietic stem cell transplantation (HSCT)11 with preparative regimens that incorporate alemtuzumab. Recurrent CMV reactivations in allogeneic HSCT recipients conditioned with alemtuzumab have been linked to a failure to generate a cytotoxic T-lymphocyte response to CMV.12
The incidence of CMV reactivation in patients with chronic lymphoproliferative disorders who are treated with alemtuzumab (and who are not allogeneic HSCT recipients) varies from 10% to 66%.2, 8, 13 In a study of patients with previously treated CLL, therapy with alemtuzumab and filgrastim was associated with a 43% incidence of CMV reactivation.14 Lundin et al.6 reported a 28% incidence of symptomatic CMV reactivation in patients with advanced cutaneous T-cell disease who were treated with alemtuzumab. We previously reported a 22% incidence of CMV reactivation in patients with CLL who were treated with alemtuzumab for minimal residual disease after chemotherapy.15 The incidence is consistently higher in the HSCT setting.11, 12 To our knowledge, the current report describes the largest series of patients to date with chronic lymphoproliferative diseases who received treatment with alemtuzumab-based therapy, and we report an incidence of symptomatic CMV reactivation of 22%. Screening for CMV viremia in asymptomatic patients who are receiving alemtuzumab may result in detection of low levels of CMV in more patients; however, the question of whether antiviral therapy for such asymptomatic patients is indicated has not been answered. In our series, 9 patients were identified who had asymptomatic CMV antigenemia detected; 6 of those patients received antiviral therapy, 3 of them remained asymptomatic without any CMV directed therapy, and all 9 patients completed their planned alemtuzumab-based therapy despite low-level CMV antigenemia.
Detection of the CMV pp65 antigen in peripheral blood leukocytes and PCR-based testing for CMV DNA in blood are the 2 most commonly used methods for detecting CMV and correlate well in patients with symptomatic CMV disease.16 In a prospective comparison, CMV antigen-detection methods reportedly detected viremia earlier than routine PCR-based methods.17 Newer ultrasensitive PCR-based assays can detect CMV viremia earlier or more frequently than antigen-detection methods18 but may be handicapped by the burden of oversensitivity. For clinically significant CMV reactivations, antigen-detection methods are comparable to PCR-based methods.18
In univariate analysis, only a lower serum albumin level predicted for CMV reactivation in patients who were treated with alemtuzumab. All but 1 of our patients with CMV reactivation had a serum albumin level <4 g/dL. Nguyen et al.7 also examined age, the number of prior regimens, prior rituximab therapy, prior splenectomy, modified Rai stage at alemtuzumab therapy (low/intermediate vs high), absolute neutrophil count, and absolute lymphocyte count and observed no correlation with any of those factors and the development of CMV viremia. It is conceivable that serum albumin is a surrogate marker for latent variables, such as integrity of immune status or nutritional status. Pursuing the question of nutrition, we investigated for a possible correlation between percentage change in body weight during alemtuzumab therapy and CMV reactivation, but we did not observe any such correlation. The question of whether aggressive nutritional intervention to correct pre-existing hypoalbuminemia will reduce CMV activation in this setting needs to be examined prospectively. However, the association of hypoalbuminemia with CMV reactivation has been reported in the context of solid-organ transplantation. Serial assessment of CMV-specific antibodies and serum albumin levels after renal transplantation indicated that a postoperative drop in CMV-specific immunoglobulin G and albumin is observed in patients with CMV reactivation.19 Similarly, persistent hypoalbuminemia has been associated with CMV reactivation after simultaneous kidney-pancreas transplantation.20
With appropriate therapy, most patients achieved clearance of CMV viremia. The deaths in patients with CMV reactivation resulted from progressive disease or polymicrobial infection/septicemia. In the current study, only 1 patient had evidence of CMV pneumonia.
In conclusion, CMV reactivation may occur in patients with chronic lymphoproliferative disorders who receive treatment with alemtuzumab. Most patients respond to appropriate antiviral therapy without developing CMV disease, and a low serum albumin level is a risk factor for reactivating CMV during treatment with alemtuzumab.