This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Morrison, V. A. Articles by Schiffer, C. A. Search for Related Content PubMed PubMed Citation Articles by Morrison, V. A. Articles by Schiffer, C. A. Social Bookmarking                            What's this?

Impact of Therapy With Chlorambucil, Fludarabine, or Fludarabine Plus Chlorambucil on Infections in Patients With Chronic Lymphocytic Leukemia: Intergroup Study Cancer and Leukemia Group B 9011

From the Veterans Affairs Medical Center, Minneapolis, MN; Long Island Jewish Medical Center, New Hyde Park, and North Shore University Hospital, Manhasset, NY; Cancer and Leukemia Group B Statistical Center and Veterans Affairs Medical Center, Durham, NC; University of New Mexico Health Sciences Center, Albuquerque, NM; Fred Hutchinson Cancer Research Center, Seattle WA; The Case Western Reserve University School of Medicine, Cleveland, OH; University of Chicago Medical Center, Chicago, IL; Wayne State University School of Medicine, Detroit, MI; and National Cancer Institute of Canada Clinical Trials Group, Queens University, Kingston, Ontario, Canada.

Address reprint requests to Vicki A. Morrison, MD, Sections of Hematology/Oncology and Infectious Disease, 111E, Veterans Affairs Medical Center, One Veterans Dr, Minneapolis, MN 55417; email: morri002{at}tc.umn.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
PURPOSE: We sought to determine whether therapy with single-agent fludarabine compared with chlorambucil alone or the combination of both agents had an impact on the incidence and spectrum of infections among a series of previously untreated patients with B-cell chronic lymphocytic leukemia (CLL).

PATIENTS AND METHODS: Five hundred fifty-four previously untreated CLL patients with intermediate/high-risk Rai-stage disease were enrolled onto an intergroup protocol. Patients were randomized to therapy with chlorambucil, fludarabine, or fludarabine plus chlorambucil. Data pertaining to infection were available on 518 patients. Differences in infections among treatment arms were tested with the Kruskal-Wallis, Wilcoxon, and ² tests.

RESULTS: A total of 1,107 infections (241 major infections) occurred in 518 patients over the infection follow-up period (interval from study entry until either reinstitution of initial therapy, therapy with a second agent, or death). Patients treated with fludarabine plus chlorambucil had more infections than those receiving either single agent (P < .0001). Comparing the two single-agent arms, there were more infections on the fludarabine arm (P = .055) per month of follow-up. Fludarabine therapy was associated with more major infections and more herpesvirus infections compared with chlorambucil (P = .008 and P = .004, respectively). Rai stage and best response to therapy were not associated with infection. A low serum immunoglobulin G was associated with number of infections (P = .02). Age was associated with incidence of major infection in the combination arm (P = .004).

CONCLUSION: Combination therapy with fludarabine plus chlorambucil resulted in significantly more infections than treatment with either single agent. Patients receiving single-agent fludarabine had more major infections and herpesvirus infections compared with chlorambucil-treated patients.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
INFECTIONS HAVE a significant impact on the clinical course and outcome of patients with chronic lymphocytic leukemia (CLL), despite advances in therapy and supportive care. Up to 80% of CLL patients will sustain infectious complications at some time during their disease course, ranging from moderate to life-threatening in severity, and infection accounts for deaths in at least 50% to 60% of these patients.^1-4 This population is predisposed to infections not only caused by the immunodeficiency that is inherent to the primary disease process (both hypogammaglobulinemia and cellular dysfunction), but also to immunosuppression from cytotoxic therapy or corticosteroids. The use of the purine analog fludarabine has introduced the potential for a new spectrum of infections caused by the prolonged T-lymphocyte depletion produced by this agent.⁵ Although there are reports of opportunistic infections in fludarabine-treated patients, no prospective comparative studies have been undertaken to determine whether the incidence and type of infections are different than infections in CLL patients treated with traditional alkylator-based regimens. We had the opportunity to collect data systematically regarding infections experienced in a large cohort of previously untreated patients with B-cell CLL and to compare the infectious complications associated with standard alkylator therapy to those occurring with fludarabine treatment. This analysis included the incidence and relative risk of infection as well as the spectrum of infectious complications. The clinical outcome results of this trial have been reported separately.⁶

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
Five hundred forty-four patients with previously untreated B-cell CLL were enrolled onto an intergroup protocol entitled, "A Phase III Comparison of Fludarabine Phosphate Versus Chlorambucil Versus Fludarabine Phosphate + Chlorambucil in Previously Untreated B-Cell Chronic Lymphocytic Leukemia" (CALGB 9011). Participants in the study included the Cancer and Leukemia Group B (CALGB), the Southwest Oncology Group, the Eastern Cooperative Oncology Group, and National Cancer Institute of Canada Clinical Trials Group. This protocol was open for accrual from October 1, 1990 to December 7, 1994. Eligible patients had received no prior cytotoxic therapy; however, prior corticosteroid therapy for autoimmune problems was allowed. In addition, patients were required to have either intermediate-stage (stage I or II) disease with clinical evidence of progressive disease, or high-risk disease (stage III or IV) by the modified Rai staging system.⁷ Patients were required to have a serum creatinine 1.5 x normal and AST/ALT 1.5 x normal.

After signed informed consent was obtained, a centralized randomization process at the CALGB Data Management Center was used to assign patients to one of the three following treatments: arm 1 = chlorambucil, 40 mg/m² orally on day 1, every 4 weeks; arm 2 = fludarabine phosphate, 25 mg/m²/d intravenously, days 1 to 5, every 4 weeks; or arm 3 = chlorambucil, 20 mg/m² orally on day 1 plus fludarabine phosphate, 20 mg/m²/d intravenously, days 1 to 5, every 4 weeks. Therapy was continued for 3 months after a clinical complete remission or for 2 months after a stable best response. The maximum duration of therapy was 1 year. Patients on either arm 1 or 2 who had evidence of disease progression on therapy or an intolerable or allergic reaction to the treatment drug were crossed over to the alternate treatment arm. The combined therapy arm (arm 3) was closed to further accrual on May 2, 1994, because an interim analysis by an independent data and safety monitoring board showed that the toxicities were significantly more frequent on this arm than on either of the single-agent treatment arms (arms 1 or 2). The concomitant use of corticosteroids was prohibited on this study. Prophylactic antibiotics were not recommended; antimicrobial therapy of established or suspected infections was provided as a component of full supportive care. The use of growth factors to maintain dose-intensity or dose schedule was not allowed. The clinical results of this trial have been reported in detail separately.⁶

The primary objective of our infection study was to determine whether the incidence of infection was different among the three treatment arms. Secondary objectives were designed to determine whether there were differences among the three arms in (1) the spectrum of infection (eg, causative agent or site of infection); (2) the occurrence of infection with regard to the therapeutic cycle (eg, during the first three cycles of therapy or after completion of induction chemotherapy); and (3) percentage of infections that occurred when patients were neutropenic. Another secondary objective was to determine whether, within each arm, the frequency of infection was related to patient age, Rai stage of CLL, baseline renal function, baseline immunoglobulin (Ig) G level, or best response to initial chemotherapy.

Case report forms from the 544 enrolled patients were retrospectively reviewed for the occurrence of infectious complications. No differentiation was made as to whether infections were treatment-related or disease-related. Available data from 518 patients (95%) were complete for infection evaluation and analysis. Reasons for exclusion of the remaining 26 patients from the infection analysis were as follows: infection data were incomplete on available flowsheets²² and cases did not meet eligibility criteria for the protocol.⁴

The infection follow-up period was defined as the interval from study entry until either reinstitution of initial therapy because of disease progression, therapy with a second chemotherapy agent, or death. Guidelines for cross-over to an alternative arm of therapy in this study included the following: evidence of disease progression while on initial protocol therapy, intolerable allergic reactions to the treatment drug, and relapse within 6 months after response to initial therapy. However, for this infection review, the infection follow-up period was arbitrarily terminated at the time of cross-over. This was done so that infections that occurred in these patients would be influenced by only a single type of therapy, which is pertinent with regard to the differential effects of fludarabine and chlorambucil on immune function. In the subset of cross-over patients who received sequential single agents (fludarabine followed by chlorambucil or vice versa), it would not have been possible to ascertain any predisposition to a particular type of infection based on prior therapy. Because the infection follow-up period varied among the three treatment arms, the mean number of infections per month of follow-up, rather than the total number of infections occurring in a given patient regardless of the length of follow-up, was used primarily for purposes of statistical analysis. However, it was not possible in our analysis to control for patients potentially being at increased risk for infections as a function solely of disease duration, regardless of therapeutic intervention.

Each episode of infection was numbered chronologically, and a series of data was collected for each discrete infection. The causative agent, when one was identified, was classified as bacterial (Gram-positive, Gram-negative, or other); fungal (Candida species, Aspergillus species, or other); viral (varicella zoster virus [VZV], herpes simplex virus [HSV], or other); parasitic; clinical infection with negative laboratory culture (or no culture obtained); and other. The site of infection was classified as lower respiratory tract (pneumonia or bronchitis); upper respiratory tract (pharyngitis, sinusitis, or rhinitis); urinary tract; skin and/or soft tissue; blood (septicemia); and other. Isolated fever was defined as the presence of fever without an identifiable focus of infection. If a site of infection was later defined in the clinical course of these patients, the infection was categorized as being from that site and not as an isolated fever. Laboratory data collected at the time of infection included the WBC count and the absolute neutrophil count (ANC) (defined as the number of segmented neutrophils and bands). At study enrollment, a baseline IgG level and serum creatinine were obtained. Creatinine clearance was estimated using the Cockcroft-Gault equation: (140 - age) x LBW_kg/(72 x serum creatinine_mg/dL).⁸ A 0.85 correction factor was used for female patients. Lean body weight for males was 50 kg + 2.3 kg/each inch over 5 feet in height, and for females, it was 45.5 kg + 2.3 kg/each inch over 5 feet in height. The cycle number of therapy during which the infection occurred was determined. If the infection occurred after completion of therapy, the number of posttreatment months was noted. In addition, the total number of cycles of initial therapy received was tabulated. A major infection was defined as an infection for which the patient required either hospitalization for care of the infection and/or treatment with parenteral antimicrobial agents.

Data Audits
As part of the quality assurance program of the CALGB, members of the Data Audit Committee visit all participating institutions at least once every 3 years. The auditors verify compliance with federal regulations and protocol requirements, including those pertaining to eligibility, treatment, toxic effects, tumor response, and outcome in a sample of protocols at each institution. Such on-site review of medical records was performed for a randomly selected subgroup of 130 patients (24% of the total) treated under this study in CALGB. Similar quality assurance audits were conducted by the other cooperative groups.

Statistical Methods
Treatment arm differences on the following measures of infection as a toxicity were tested: (1) number of infections per patient, (2) number of infections per patient per month of follow-up, (3) number of major infections per patient, (4) number of major infections per patient per month of follow-up, (5) percent (%) of patients with at least one infection, (6) % of patients with at least one major infection, (7) % of patients with at least one infection caused by a specific causative agent, and (8) % of patients with at least one infection at a specific site. For each of these measures, two different comparisons were tested: a 2 df test of differences among the three arms, and a 1 df test of a difference between the single-agent fludarabine and single-agent chlorambucil arms.

Treatment arm differences in measures no. 1 to 4 (above) were tested with the 2 df Kruskal-Wallis test and the 1 df Wilcoxon test. Although these nonparametric tests were used for testing, the effect sizes were illustrated with means instead of medians, because the medians often resulted in a distorted picture of the data (eg, in one test, although a significant difference among the arms existed, all three medians were equal to 0.) Treatment arm differences in measures no. 5 to 8 (above) were tested with the ² test for a difference in proportions.

To examine treatment arm differences in % of infections (and % of major infections) that occurred when patients were neutropenic (defined as an ANC < 1.0 x 10⁹/L), we used as our unit of analysis not the patient, but the infection. We used repeated measures logistic regression to examine the association of treatment arm with a dichotomous variable indicating whether or not the ANC was less than 1.0 x 10⁹/L during the (major) infection. Generalized estimating equations were used to fit these models.⁹

In addition to these tests of treatment arm differences, we also tested some correlates of infection within each treatment arm (ie, age, IgG, Rai stage, renal function, and clinical response). Spearman correlation, the Wilcoxon test, and the ² test of proportions were used to test these associations.

A two-sided Type I error rate (alpha) of 0.05 was used for all statistical tests, and no attempt was made to adjust the alpha level for the large number of statistical tests performed for this study. Thus, some tests that are called significant may be a result of chance alone, and this fact should be kept in mind when interpreting as significant test statistics with P values only slightly smaller than .05. Lastly, the interquartile range (the 25th and 75th percentiles) was used to describe the distribution of IgG levels, creatinine clearance, and length of infection follow-up (Table 1).

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

Total Number of Infections
A total of 1,107 infections occurred in the 518 patients over the infection follow-up period (Table 2). The total number, mean, and median numbers of infections per patient were greater in patients who received fludarabine, either alone or in combination with chlorambucil. However, when examining the mean number of infections per month of infection follow-up, a more meaningful comparison, patients on the fludarabine plus chlorambucil arm had the highest incidence of infection (P = .008). Comparing only the single-agent chlorambucil and fludarabine arms, the incidence of infection was higher in fludarabine-treated patients (P = .055).

Fungal infections were not common in the study population. No Aspergillus infections were reported in these patients. The majority of fungal infections were caused by Candida species; both localized cutaneous and invasive infections were found. A small number of dermatophyte infections occurred. Candida infections (both total and major infections) were most common in patients who received fludarabine plus chlorambucil (P < .0001 and .01, respectively). When comparing the single-agent fludarabine and chlorambucil arms, Candida infections were more common in the fludarabine-treated patients (P = .07). Pneumocystis carinii infections were remarkably uncommon. All three reported cases occurred in patients receiving a fludarabine-containing regimen (two on fludarabine plus chlorambucil and one on single-agent fludarabine).

The majority of viral infections in the study population were herpesvirus infections. Patients receiving therapy with fludarabine plus chlorambucil were most likely to sustain a herpesvirus infection during the study period (P < .0001). When comparing the single-agent treatment arms, herpesvirus infections occurred more often in the fludarabine-treated patients (P = .0006). Taken alone, varicella zoster virus infections were also more common in patients receiving fludarabine plus chlorambucil (P = .0007). Likewise, VZV infections were more common in patients receiving single-agent fludarabine than chlorambucil (P = .004). HSV infections were also more common in patients receiving single-agent fludarabine than chlorambucil (P = .08). The majority of the VZV/HSV infections were minor, requiring neither hospitalization nor parenteral antiviral therapy. Only three cases of disseminated VZV occurred, two in patients treated with fludarabine plus chlorambucil and one in a patient who received single-agent fludarabine. The timing of the herpesvirus infections with regard to therapy was also examined. Only 10% of VZV infections were diagnosed during the first three cycles of therapy; 50% of the VZV infections occurred after completion of initial therapy. Of the HSV infections, 34% were diagnosed during the first three cycles of therapy and 29% after completion of initial therapy.

Parasitic infections were uncommon. One fludarabine-treated patient (a resident of Milwaukee) developed cryptosporidiosis, perhaps as a part of that community’s epidemic. A case of scabies occurred in one patient treated with fludarabine plus chlorambucil.

Site of Infection
The sites of infection are listed in Table 5. Lower respiratory tract infections more commonly occurred in patients treated with fludarabine plus chlorambucil than with either single agent (P = .03). Skin and soft tissue infections were most common in patients receiving fludarabine plus chlorambucil (P = .001). Comparing the two single-agent treatment arms, fludarabine-treated patients sustained more skin and soft tissue infections than did those patients receiving chlorambucil (P = .002). Likewise, isolated fever was most frequent in patients receiving fludarabine plus chlorambucil (P < .0001); fever was also more common in patients treated with single-agent fludarabine than with chlorambucil (P = .01). As few patients had indwelling catheters, catheter-related infections did not significantly contribute to infectious morbidity in this study.

The 114 episodes of isolated fever without an obvious focus of infection were examined further. Of these, 47 (41%) occurred in patients receiving single-agent fludarabine, 18 (16%) in chlorambucil-treated patients, and 49 (43%) in patients receiving fludarabine plus chlorambucil. Across all three treatment arms, approximately two thirds of these episodes occurred during the first three cycles of chemotherapy. Likewise, neutropenia (ANC < 1.0 x 10⁹/L) was present in 22% of cases of isolated fever in each of the three treatment arms. In the setting of febrile neutropenia and apparent clinical infection, these patients were generally admitted to the hospital and begun on broad-spectrum antimicrobial agents.

ANC
Of the 1,107 episodes of infection, ANC data were available for 998 cases (90%) (Table 6). Neutropenia, defined as an ANC less than 1.0 x 10⁹/L, accompanied 19% of infections on the fludarabine plus chlorambucil arm but only 10% and 12% of infections on the single-agent fludarabine and chlorambucil arms, respectively (P = .003). Neutropenia was more common with major infections but did not differ among the treatment arms.

Rai Stage of CLL
Across treatment arms and within each arm, there were no associations between the incidence of infection and the Rai stage of CLL, except for a marginally significant association in the fludarabine plus chlorambucil arm. Of these 141 patients, 93% with high-risk disease compared with 80% with intermediate-risk disease had at least one infection (P = .05).

Renal Function
The median calculated creatinine clearance of the 518 patients at study entry was 60 mL/min (range, 17 to 165 mL/min). The only significant association of creatinine clearance (examined as a continuous variable) with infection was in patients receiving fludarabine plus chlorambucil, who were at greater risk of sustaining at least one major infection and of having more major infections as their creatinine clearance declined (P = .03 and .04, respectively). There was no association between creatinine clearance and any infection variables in either single-agent arm.

IgG Status
The IgG levels at study entry were available on 458 (88%) of the 518 patients. The IgG levels ranged from 0 to 6,864 mg/dL, with a median of 772 mg/dL (mean, 859 mg/dL). The IgG level was low (defined as < 500 mg/dL) in 85 patients (19%). When the IgG level was examined as a dichotomous variable (< or 500 mg/dL), it was found that patients with a low IgG had a greater number of infections (P = .02). Correlations within treatment arms were nonsignificant. No significant correlation was found between the IgG level and the incidence of either major infection or VZV infection.

Best Response to Initial Therapy
The best response to CLL therapy administered on this study (complete remission, partial remission, stable disease, and progressive disease) was examined to determine whether there was an association between best response and the occurrence of infection. Notably, within each treatment arm, responding patients (complete response and partial response) were more likely than nonresponders to have at least one infection during their infection follow-up period. However, if one controlled for either the number of cycles of therapy or the length of follow-up, these associations with infection were not significant. One possible explanation is that responding patients tended to receive more cycles of therapy and to have a longer infection follow-up period than nonresponders, thus having more time in which infections could occur. No association was found between best response to therapy and the occurrence of major infections.

Infection-Related Mortality
No significant differences in infection-related mortality were present among the three treatment arms. There were no treatment-related grade 5 (lethal) toxicities attributable to infection recorded in the study population. Survival at 12 months for all 518 patients was 87%, and survival by treatment arm was 87% for fludarabine, 89% for chlorambucil, and 84% for fludarabine plus chlorambucil.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
The intergroup study CALGB 9011 provided a unique opportunity to examine the frequency and spectrum of infections during treatment and the influence of various clinical factors on the risk for infection in a large series of previously untreated patients with B-cell CLL. In addition, the impact of therapy with fludarabine was compared with that of a conventional alkylating agent. We found that infections remain a significant cause of morbidity in CLL patients. More importantly, significant differences were present in the spectrum of infection among the three treatment arms.

In CLL patients treated with conventional alkylating agents and/or corticosteroids, bacteria are the most frequent cause of infection, and the respiratory tract is the most common site of infection.^1-3,12 However, purine analog therapy has been accompanied by a different spectrum of infections attributable to quantitative and qualitative T-cell abnormalities induced by these agents, which occur early in treatment and may persist 1 to 2 years after discontinuation of therapy.^5,13-18 Infections in fludarabine-treated patients include bacterial infections common to CLL patients in addition to a variety of opportunistic pathogens,^14-16,19-34 such as Listeria,^{2,15,24,25,33} Nocardia,^32,33 and Mycobacterium species.^21,23,33 Fungal infections, often disseminated, are most often caused by Candida or Aspergillus species.^14,28,32,33 Reports of Pneumocystis carinii infections are common.^{2,14,15,23,26-28,33,34} Concomitant corticosteroid therapy increases the risk for both Listeria and Pneumocystis infections.^15,35 Lastly, herpesvirus infections, especially VZV, are frequent in fludarabine-treated patients.^{2,14,19,21,22,30,33,36}

Our findings of a 61% infection rate (17% major infection rate) in the chlorambucil-treated patients are consistent with results of prior CLL trials using conventional alkylators.^37,38 In Sawitsky’s et al³⁷ series of advanced stage CLL patients treated with either chlorambucil plus prednisone or single-agent prednisone, 26% of patients had severe infections and 30% less severe infections. Findings are similar in series where an anthracycline is added to alkylator-based regimen.^39,40

Our patients receiving fludarabine plus chlorambucil had significantly more major and total infections than did the patients treated either with chlorambucil or fludarabine alone (P < .0001). When comparing patients treated with either single-agent fludarabine or chlorambucil, there was both a greater mean number of total infections and of major infections per month of follow-up in the fludarabine-treated patients (P = .055 and .008, respectively). Although mean and median infection numbers are numerically small, one must bear in mind that CLL is a disease with a long natural history, such that the statistically significant differences among treatment arms in these parameters are clinically meaningful. Our findings of a 29% major infection rate in the fludarabine-treated patients is similar to the 34% infection rate found by Anaissie et al³⁵ in 154 previously untreated CLL patients who received fludarabine, either alone or with prednisone. However, our findings differ from those reported by the French Cooperative Group on CLL, whose patients received either fludarabine or cyclophosphamide, doxorubicin, and prednisone.⁴¹ The overall incidence of infection was comparable in the two groups (23% and 19%, respectively), as was the incidence of severe infections. One explanation for their findings may be that adding an anthracycline to an alkylator results in a greater risk of neutropenia.

The spectrum of infection also differed among the three treatment arms. Potential explanations for Candida infections being most common in patients treated with fludarabine plus chlorambucil (P < .0001) include treatment-related myelosuppression and fungal colonization from antibiotics used to treat the increased number of bacterial infections on this arm. No Aspergillus infections were reported in our series, in contrast to prior anecdotal reports in fludarabine-treated patients.^14,28,33 However, these patients may have been more heavily pretreated and had more advanced disease than our previously untreated patients. Despite many anecdotal reports of Pneumocystis infection^{2,14,15,23,26-28,33,34} and the absence of specific prophylactic measures in our trial, we observed only three cases of Pneumocystis infection, all in patients receiving fludarabine-containing regimens. These infections, in addition to mycobacterial and herpesvirus infections, are unusual occurrences in CLL patients who receive conventional alkylator therapy. VZV infections were significantly more common in the fludarabine-treated patients, whether they received it as a single agent (P = .004) or in combination with chlorambucil (P = .0007). The three cases of disseminated VZV infection occurred in patients receiving fludarabine-based regimens. The relationship between the CD4 count and herpesvirus infections was examined in a prior series of fludarabine-treated patients, most of whom also received corticosteroids.³⁵ The risk for VZV infection was significantly greater in patients with a CD4 count less than 50 cells/µL after the third cycle of therapy. We were not able to make similar comparisons in our study because CD4 counts were not routinely obtained during therapy. As T-cell dysfunction persists after discontinuation of fludarabine, it is not surprising that many of our VZV/HSV infections occurred after completion of therapy. The predominance of herpesvirus infections after fludarabine therapy was also found in a series of 174 CLL patients who received initial therapy with fludarabine, either alone or with prednisone, at the M.D. Anderson Cancer Center.⁴² Dermatomal zoster was the most common infection during the remission follow-up period in that series, occurring in 19 patients. There was also one case each of Listeria and cytomegalovirus infection; no episodes of Pneumocystis pneumonia occurred. Lastly, the respiratory tract was the most common site of infection in our patients, as in past series.^1-3 There were no significant differences in the occurrence of septicemia among the three treatment groups, despite neutropenia being more common in patients receiving fludarabine plus chlorambucil.

In the cases of isolated fever, the question arises as to whether the fever is caused by an infectious process or related to fludarabine administration. Hospital admission generally occurred in the setting of neutropenia and when patients seemed to be infected on medical assessment. It is not uncommon to be unable to define a precise focus of infection in febrile neutropenic patients. To evaluate the impact of these cases on the study findings, the statistical analyses conducted for this study were repeated with all cases of isolated fever excluded. It was found that our study findings were comparable, whether the cases of isolated fever were included with or excluded from the analyses. Thus, the finding of differences in infectious complications between the single-agent fludarabine and the chlorambucil arms was not influenced by these cases of isolated fever.

Several clinical variables had an impact on infection risk in our series. Although neutropenia may increase the risk of infection for patients on the combination arm, it does not seem to explain differences in infection among patients treated with chlorambucil or single-agent fludarabine. In a series of 402 CLL patients treated with fludarabine alone or with prednisone, Anaissie et al³⁵ found that an ANC less than 1.0 x 10⁹/L was a significant risk factor for major infection. In that series, infections were more common among 248 previously treated patients (58%), than among 154 patients with no prior therapy (34%). The large number of previously treated patients in that series, in contrast to our chemotherapy-naive patients, may complicate comparison of these series.

Although hypogammaglobulinemia is the predominant risk factor for infection in CLL patients,^1-5 the relationship between a specific Ig class and risk for infection is not well-established.^1,10-12 In our series, a low serum IgG level was associated with an increased number of infections. A relationship between a low IgG level and infection was also found by Molica et al.⁴ The risk of developing a severe infection at 5 years was 26% overall in that series but was 57% in patients with a low IgG and 68% in patients with both a low IgG and Binet stage C disease.

We found that advancing age correlated with the incidence of major infection only in the fludarabine plus chlorambucil arm. In the series by Twomey et al,³ infection risk was highest among CLL patients older than 70 years of age. However, that analysis is several decades old, and advances in supportive care, such as the expanded spectrum of antimicrobial agents, have since been made.

Baseline renal function had an impact on infections only in patients treated with fludarabine plus chlorambucil in our series. In contrast, Anaissie et al³⁵ found that a serum creatinine 1.4 mg/dL was a significant risk factor for major infection in his series of fludarabine-treated patients. As patients with poor renal function (serum creatinine > 1.5 times normal) were excluded from our study, we possibly did not detect a substantial effect of creatinine clearance on infection in our single-agent fludarabine arm. We have evaluated renal function in relation to fludarabine toxicity during the first cycle of treatment in a separate report (Martell et al, manuscript submitted for publication).

Although disease stage had no significant impact on infection in our series, such correlations have been previously reported. In Travade’s et al⁴³ series, more than 50% of infections occurred in patients with Binet stage C disease. Likewise, Itala et al¹¹ found an 82% occurrence of severe infection in Binet stage C patients compared with 33% in stage A patients. In Molica’s et al⁴ series, at least one severe infection occurred in 52% of Binet stage C patients compared with 33% of stage B patients. Anaissie et al³⁵ also found that advanced-stage disease was associated with an increased risk of major infection in fludarabine-treated patients. Potential reasons for these differences may relate to supportive care practices or to prior therapy resulting in immunosuppression. Our patients were previously untreated, whereas some series included previously treated patients (eg, 62% in Anaissie’s et al³⁵ series).

We found no association between the best response to initial therapy and infection. This is in contrast to Keating’s et al³⁶ series of 75 fludarabine-treated CLL patients, in which the incidence of infection was less in patients who achieved a complete response or a nodular partial response compared with patients with other partial response or nonresponsive disease. In a follow-up report from the M.D. Anderson Cancer Center of 174 CLL patients who received initial therapy with fludarabine (71 as a single agent and 103 in combination with prednisone), these differences in infection incidence remained significant.⁴² In this series, patients in remission were observed for infection until they either developed progressive disease or received a second therapy (usually a fludarabine-containing regimen). However, 59% of these patients received prednisone plus fludarabine, a combination with which opportunistic infections have been seen. Thus, it is difficult to compare these results with our study, in which corticosteroids were prohibited.

There are differences in the frequency of infectious complications reported in this retrospective review compared with the summary results for the intergroup clinical trial (Rai et al, manuscript submitted for publication). The protocol guidelines for CALGB study 9011 required that infectious toxicities be reported only if they were clearly related to the therapy administered. In a disease with inherent immune deficiencies such as CLL, it can be exceedingly difficult to differentiate therapy-related infections from those that are disease-related, and decisions made by treating physicians and research nurses in this regard may be arbitrary. For example, if a CLL patient develops pneumococcal pneumonia while on therapy with either chlorambucil or fludarabine, it is quite difficult to determine whether disease-related hypogammaglobulinemia or therapy-related immunosuppression was the sole predisposing risk factor for this infection. Likewise, if a fludarabine-treated CLL patient develops varicella zoster infection, although one may tend to attribute this to fludarabine-related immunosuppression, abnormalities in inherent cell-mediated immune function may also predispose CLL patients to these infections. In our review, we included all infections that were documented to occur during the defined infection follow-up period and made no attempt to categorize infection as treatment- or disease-related. In addition, some of the herpesvirus infections reported to the Data Management Center were coded as skin toxicities rather than infectious toxicities. Lastly, we included isolated fever, which sometimes, though not always, occurred in the setting of neutropenia, as an infection. For these reasons, a larger number of infections were included in this review than the summary report of the intergroup trial.

Two issues raised by our study may have important implications for infection prophylaxis. In contrast to the occurrence of Pneumocystis infection noted in prior reports, it does not seem that Pneumocystis prophylaxis is mandatory for CLL patients receiving fludarabine therapy. However, this issue may warrant examination when fludarabine is used in combination with newer agents such as alemtuzumab (Campath 1-H; Millennium and ILEX Partners, Ltd, Cambridge, MA), which also alter cell-mediated immunity. Second, the incidence of herpesvirus infections, specifically VZV, in patients receiving fludarabine is of concern. Although the majority of these VZV infections are not fatal, morbidity may be significant. As this was a retrospective review, recommendations for antiviral prophylaxis cannot be made based on our findings. The CALGB is examining the incidence of herpesvirus infections prospectively in the current fludarabine treatment protocol for CLL patients (CALGB study 9712). Consideration should be given to examining the issue of antiviral prophylaxis in a prospective randomized fashion in future treatment trials of CLL patients, especially those in which patients receive fludarabine or other agents affecting cell-mediated immunity.

	APPENDIX

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 

	ACKNOWLEDGMENTS

 
Supported in part by grant nos. CA11028, CA33601, CA35279, and CA41287 from the National Cancer Institute, Bethesda, MD; and in part, by grants from the National Cancer Institute to the Cancer and Leukemia Group B (grant no. CA31946), to the Southwest Oncology Group (grant no. CA32102), and to the Eastern Cooperative Oncology Group (grant no. CA21115), and by the National Cancer Institute of Canada. Research for CALGB 9011 was also supported by a grant from Berlex Laboratories, Inc, to the Cancer and Leukemia Group B Foundation.

	NOTES

 
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION APPENDIX REFERENCES

 
1. Chapel HM, Bunch C: Mechanisms of infection in chronic lymphocytic leukemia. Semin Hematol 24: 291-296, 1987[Medline]

2. Kontoyianis DP, Anaissie EJ, Bodey GP: Infection in chronic lymphocytic leukemia: A reappraisal, in Cheson BD (ed): Chronic Lymphocytic Leukemia: Scientific Advances and Clinical Developments. New York, NY, Marcel Dekker, Inc, 1993, pp 399-417

3. Twomey JT: Infections complicating multiple myeloma and chronic lymphocytic leukemia. Arch Intern Med 132: 562-565, 1973[Abstract/Free Full Text]

4. Molica S: Infections in chronic lymphocytic leukemia: Risk factors, and impact on survival, and treatment. Leuk Lymphoma 13: 203-214, 1994[Medline]

5. Cheson BD: Infectious and immunosuppressive complications of purine analog therapy. J Clin Oncol 13: 2431-2448, 1995[Abstract/Free Full Text]

6. Rai KR, Peterson B, Appelbaum FR, et al: Fludarabine compared with chlorambucil as primary therapy for chronic lymphocytic leukemia. N Engl J Med 343: 1750-1757, 2000[Abstract/Free Full Text]

7. International Workshop on Chronic Lymphocytic Leukemia: Chronic lymphocytic leukemia: Recommendation for diagnosis, staging, and response criteria. Ann Intern Med 110: 236-238, 1989

8. Cockcroft DW, Gault MH: Prediction of creatinine clearance from serum creatinine. Nephron 16: 31-41, 1976[Medline]

9. Liang KY, Zeger SL: Longitudinal data analysis using generalized linear models. Biometrika 73: 13-22, 1986[Abstract/Free Full Text]

10. Rozman C, Montserrat E, Viñolas N: Serum immunoglobulins in B-chronic lymphocytic leukemia. Cancer 61: 279-283, 1988[Medline]

11. Itala M, Helenius H, Nikoskelainen J, et al: Infections and serum IgG levels in patients with chronic lymphocytic leukemia. Eur J Haematol 48: 266-270, 1992[Medline]

12. Morrison VA: The infectious complications of chronic lymphocytic leukemia. Semin Oncol 25: 98-106, 1998[Medline]

13. Juliusson G: Immunological and genetic abnormalities in chronic lymphocytic leukaemia: Impact of the purine analogues. Drugs 47: 19-29, 1994 (suppl 6)

14. Wijermans PW, Gerrits WBJ, Haak HL: Severe immunodeficiency in patients treated with fludarabine monophosphate. Eur J Haematol 50: 292-296, 1993[Medline]

15. O’Brien S, Kantarjian H, Beran M, et al: Results of fludarabine and prednisone therapy in 264 patients with chronic lymphocytic leukemia with multivariate analysis-derived prognostic model for response to treatment. Blood 82: 1695-1700, 1993[Abstract/Free Full Text]

16. Keating MJ, O’Brien S, Robertson LE, et al: New initiatives with fludarabine monophosphate in hematologic malignancies. Semin Oncol 20: 13-20, 1993 (suppl 7)

17. Keating MJ, O’Brien S, Kantarjian H, et al: Nucleoside analogs in treatment of chronic lymphocytic leukemia. Leuk Lymphoma 10: 139-145, 1993

18. Byrd JC, Dow NS, Howard RS, et al: Marked depletion of natural killer cells in chronic lymphocytic leukemia patients receiving fludarabine: A consideration for future immune-based therapy. Proc Am Soc Clin Oncol 16: 28a, 1997 (abstr 96)

19. Keating MJ: Fludarabine phosphate in the treatment of chronic lymphocytic leukemia. Semin Oncol 17: 49-62, 1990[Medline]

20. Keating MJ, Estey E, O’Brien S, et al: Clinical experience with fludarabine in leukaemia. Drugs 47: 39-49, 1994 (suppl 6)

21. Keating MJ, O’Brien S, Kantarjian H, et al: Long-term follow-up of patients with chronic lymphocytic leukemia treated with fludarabine as a single agent. Blood 81: 2878-2884, 1993[Abstract/Free Full Text]

22. Keating MJ, Kantarjian H, Talpaz M, et al: Fludarabine: A new agent with major activity against chronic lymphocytic leukemia. Blood 74: 19-25, 1989[Abstract/Free Full Text]

23. Puccio CA, Mittelman A, Lichtman SM, et al: A loading dose/continuous infusion schedule of fludarabine phosphate in chronic lymphocytic leukemia. J Clin Oncol 9: 1562-1569, 1991[Abstract]

24. Girmenia C, Mauro FR, Rahimi S: Late listeriosis after fludarabine plus prednisone treatment. Br J Haematol 87: 407-408, 1994[Medline]

25. Anaissie E, Kontoyiannis DP, Kantarjian H, et al: Listeriosis in patients with chronic lymphocytic leukemia who were treated with fludarabine and prednisone. Ann Intern Med 117: 466-469, 1992

26. Schilling PJ, Vadhan-Raj S: Concurrent cytomegalovirus and pneumocystis pneumonia after fludarabine therapy for chronic lymphocytic leukemia. N Engl J Med 323: 833-834, 1990[Medline]

27. Bastion Y, Coiffier B, Tigaud JD, et al: Pneumocystis pneumonia in a patient treated with fludarabine for chronic lymphocytic leukemia. Eur J Cancer 27: 671, 1991

28. Bergmann L, Fenchel K, Jahn B, et al: Immunosuppressive effects and clinical response of fludarabine in refractory chronic lymphocytic leukemia. Ann Oncol 4: 371-375, 1993[Abstract/Free Full Text]

29. Kemena A, O’Brien S, Kantarjian H, et al: Phase II clinical trial of fludarabine in chronic lymphocytic leukemia on a weekly low-dose schedule. Leuk Lymphoma 10: 187-193, 1993[Medline]

30. Robertson LE, O’Brien S, Koller C, et al: A three-day schedule of fludarabine in chronic lymphocytic leukemia (CLL). Blood 80: 47a, 1992 (abstr, suppl 1)

31. Hiddemann W, Rottmann R, Wörmann B, et al: Treatment of advanced chronic lymphocytic leukemia by fludarabine: Results of a clinical phase-II study. Ann Hematol 63: 1-4, 1991[Medline]

32. Keating MJ, Kantarjian H, O’Brien S, et al: Fludarabine: A new agent with marked cytoreductive activity in untreated chronic lymphocytic leukemia. J Clin Oncol 9: 44-49, 1991[Abstract/Free Full Text]

33. Byrd JC, Hargis JB, Kester KE, et al: Opportunistic pulmonary infections with fludarabine in previously treated patients with low-grade lymphoid malignancies: A role for Pneumocystis carinii pneumonia prophylaxis. Am J Hematol 49: 135-142, 1995[Medline]

34. Sanders C, Perez EA, Lawrence HJ: Opportunistic infections in patients with chronic lymphocytic leukemia following treatment with fludarabine. Am J Hematol 39: 314-315, 1992[Medline]

35. Anaissie EJ, Kontoyiannis DP, O’Brien S, et al: Infections in patients with chronic lymphocytic leukemia treated with fludarabine. Ann Intern Med 129: 559-566, 1998[Abstract/Free Full Text]

36. Keating MJ, Kantarjian H, Talpaz M, et al: Fludarabine therapy in chronic lymphocytic leukemia. Nouv Rev Fr Haematol 30: 461-466, 1988

37. Sawitsky A, Rai KR, Glidewell O, et al: Comparison of daily versus intermittent chlorambucil and prednisone therapy in the treatment of patients with chronic lymphocytic leukemia. Blood 50: 1049-1059, 1977[Abstract/Free Full Text]

38. Montserrat E, Alcala A, Parody R, et al: Treatment of chronic lymphocytic leukemia in advanced stages: A randomized trial comparing chlorambucil and prednisone versus cyclophosphamide, vincristine, and prednisone. Cancer 56: 2369-2375, 1985[Medline]

39. Keating MJ, Scouros M, Murphy S, et al: Multiple agent chemotherapy (POACH) in previously treated and untreated patients with chronic lymphocytic leukemia. Leukemia 2: 2369-2375, 1988

40. French Cooperative Group on Chronic Lymphocytic Leukemia: Long-term results of the CHOP regimen in stage C chronic lymphocytic leukemia. Br J Haematol 73: 334-340, 1984

41. French Cooperative Group on Chronic Lymphocytic Leukemia: Multicentre prospective randomised trial of fludarabine versus cyclophosphamide, doxorubicin, and prednisone (CAP) for treatment of advanced-stage chronic lymphocytic leukemia. Lancet 347: 1432-1438, 1996[Medline]

42. Keating MJ, O’Brien S, Lerner S, et al: Long-term follow-up of patients with chronic lymphocytic leukemia (CLL) receiving fludarabine regimens as initial therapy. Blood 92: 1165-1171, 1998[Abstract/Free Full Text]

43. Travade P, Dusart JD, Cavaroc M, et al: Les infections graves associees a la leucemie lymphoide cronique. La Presse Med 15: 1715-1718, 1986

Submitted April 3, 2000; accepted May 28, 2001.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				J. G. Gribben How I treat CLL up front Blood, January 14, 2010; 115(2): 187 - 197. [Abstract] [Full Text] [PDF]

				W. U. Knauf, T. Lissichkov, A. Aldaoud, A. Liberati, J. Loscertales, R. Herbrecht, G. Juliusson, G. Postner, L. Gercheva, S. Goranov, et al. Phase III Randomized Study of Bendamustine Compared With Chlorambucil in Previously Untreated Patients With Chronic Lymphocytic Leukemia J. Clin. Oncol., September 10, 2009; 27(26): 4378 - 4384. [Abstract] [Full Text] [PDF]

				V. A. Morrison Management of Infectious Complications in Patients with Chronic Lymphocytic Leukemia Hematology, January 1, 2007; 2007(1): 332 - 338. [Abstract] [Full Text] [PDF]

				S. W. Lane, P. Marlton, P. N. Mollee, D. W. Milligan, A. K. Burnett, and K. Wheatley Increased mortality with FLA compared with ADE chemotherapy in high-risk AML. Blood, December 15, 2006; 108(12): 3950 - 3951. [Full Text] [PDF]

				K. W. L. Yee and S. M. O'Brien Chronic Lymphocytic Leukemia: Diagnosis and Treatment Mayo Clin. Proc., August 1, 2006; 81(8): 1105 - 1129. [Abstract] [Full Text] [PDF]

				M. Sandherr, H. Einsele, H. Hebart, C. Kahl, W. Kern, M. Kiehl, G. Massenkeil, O. Penack, X. Schiel, S. Schuettrumpf, et al. Antiviral prophylaxis in patients with haematological malignancies and solid tumours: Guidelines of the Infectious Diseases Working Party (AGIHO) of the German Society for Hematology and Oncology (DGHO) Ann. Onc., July 1, 2006; 17(7): 1051 - 1059. [Abstract] [Full Text] [PDF]

				Y. L. Kasamon, I. W. Flinn, M. R. Grever, L. F. Diehl, E. Garrett-Mayer, S. N. Goodman, M. S. Lucas, and J. C. Byrd Phase I Study of Low-Dose Interleukin-2, Fludarabine, and Cyclophosphamide for Previously Untreated Indolent Lymphoma and Chronic Lymphocytic Leukemia Clin. Cancer Res., December 1, 2005; 11(23): 8413 - 8417. [Abstract] [Full Text] [PDF]

				S. N. O'Brien, N. M.A. Blijlevens, T. H. Mahfouz, and E. J. Anaissie Infections in Patients with Hematological Cancer: Recent Developments Hematology, January 1, 2003; 2003(1): 438 - 472. [Abstract] [Full Text] [PDF]

				J. C. Byrd, B. L. Peterson, V. A. Morrison, K. Park, R. Jacobson, E. Hoke, J. W. Vardiman, K. Rai, C. A. Schiffer, and R. A. Larson Randomized phase 2 study of fludarabine with concurrent versus sequential treatment with rituximab in symptomatic, untreated patients with B-cell chronic lymphocytic leukemia: results from Cancer and Leukemia Group B 9712 (CALGB 9712) Blood, January 1, 2003; 101(1): 6 - 14. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Morrison, V. A. Articles by Schiffer, C. A. Search for Related Content PubMed PubMed Citation Articles by Morrison, V. A. Articles by Schiffer, C. A. Social Bookmarking                            What's this?