Dexrazoxane does not affect treatment outcome in non‐Hodgkin's lymphoma patients

Dexrazoxane is an iron chelator designed to reduce anthracycline-induced cardiac toxicity. We aimed to investigate whether the use of dexrazoxane affects treatment efficacy in non-Hodgkin's lymphoma (NHL) patients treated with anthracycline-containing chemotherapy. The database of patients with NHL treated in Soroka University Medical Center between 2000 and 2010 was reviewed. Included patients received anthracycline-based therapy with or without dexrazoxane as part of first-line treatment. Outcome was compared between the two groups. One hundred ninety-three patients were included; 48 of them received dexrazoxane with first-line chemotherapy. Rates of complete remission (CR), partial remission (PR), relapse, death, and a composite endpoint of relapse or all-cause mortality were similar between the groups. Cox regression analysis of risk factors for relapse revealed no effect of dexrazoxane on relapse risk. A subgroup analysis of diffuse large B-cell lymphoma (DLBCL) patients yielded the same results. Use of dexrazoxane during anthracycline-containing chemotherapy does not have a significant effect on treatment outcome in lymphomapatients. The anthracyclines are highly effective anticancer drugs, used for a variety of solid and hematological malignancies [1]. However, use of these agents is limited by a dose-dependent cardiotoxicity, which may be the dose-limiting factor in treatment. The incidence of chronic cardiotoxicity is most closely related to the cumulative dose of anthracycline. Additional risk factors for developing heart damage include preexisting cardiovascular disease (including coronary artery disease, hypertension, peripheral vascular disease, and diabetes), radiation therapy involving the heart region, type of tumor, concomitant administration of other cardiotoxic chemotherapeutic agents, female sex, and age (children and the elderly have a higher risk) [2-6]. There is a wide variation in the reported frequency of both clinical and subclinical cardiotoxicity after anthracycline therapy [7]. Rates of cardiac toxicity that have been reported in two studies of elderly patients receiving anthracycline-based chemotherapy were 21 and11%, respectively [8, 9]. In a study of 141 lymphoma patients, subclinical late cardiomyopathy was found in 27.6% [10]. In a meta-analysis of eight trials [11], use of anthracyclines was associated with a significantly increased risk of both clinical (odds ratio (OR) 5.43) and subclinical (OR 6.25) cardiotoxicity. An analysis of four studies showed that the rate of cardiac-related deaths was also significantly higher (OR 4.94). Dexrazoxane is an ethylenediaminetetraacetic acid-like chelator that effectively protects against anthracycline-induced cardiac toxicity [1]. Dexrazoxane significantly reduces the incidence of anthracycline-induced congestive heart failure (CHF) and adverse cardiac events [response rate (RR) 0.18 for clinical heart failure and RR 0.29 for clinical and subclinical heart failure combined] compared with no therapy with dexrazoxane, as shown in clinical trials and meta-analyses [7, 12-24]. On the basis of its efficacy, a set of clinical practice guidelines was published by the American Society of Clinical Oncology recommending considering the use of dexrazoxane in adult patients who have received more than 300 mg/m2 doxorubicin-based therapy [25]. Despite the apparent cardiac benefit, a number of issues remain uncertain regarding the role of dexrazoxane in both adults and children, including the possibility of a lower RR to chemotherapy [15, 16]. Multiple other randomized trials and a pooled analysis have not confirmed these findings [7, 12, 17, 21-24, 26]. However, these studies enrolled mainly patients with solid tumors; the majority of them are adults with advanced breast cancer. To the best of our knowledge, there are no similar studies in NHL patients. In this study, we investigated whether the use of dexrazoxane affects treatment efficacy in this population. One hundred ninety-three patients were included in the study; 48 of them received dexrazoxane with first-line chemotherapy. The mean dose of dexrazoxane per cycle was 1,050 mg, and 61.7% of the patients started dexrazoxane in the first or second chemotherapy cycle. Baseline patient characteristics are described in Supporting Information Table 1. Patients who received dexrazoxane were older (mean age 70.35 versus 54.1 years, P < 0.001), had more cardiovascular risk factors, and an increased incidence of chronic ischemic heart disease. One hundred thirty-seven of the 193 patients were diagnosed with DLBCL. Patients who received dexrazoxane tended to have a higher International Prognostic Index (IPI) score at diagnosis compared with the group that did not receive dexrazoxane. Other disease characteristics at diagnosis were similar between the two groups (Supporting Information Table 2). First-line chemotherapy was CHOP (cyclophosphamide, adriamycin, vincristine, and prednisone) or Rituximab–CHOP in 181 patients. The rest of the patients were treated with other anthracycline-based protocols. Mean anthracycline dose was 265.36 (±68.96) mg per body surface area (BSA) in patients that did not receive dexrazoxane and 256.92 (±79.38) mg per BSA in patients that did (P = 0.482). Comparison of treatment complications between patients treated with and without dexrazoxane showed that only diarrhea was significantly more frequent in the group that received dexrazoxane (27.1 versus 8.3%, P = 0.001). All other complications were similar in incidence between the groups (Supporting Information Table 3). Comparison of treatment outcome between the groups showed that there was no statistically significant difference in rates of CR, PR, relapse, death, and the composite endpoint of death and relapse between patients treated with or without dexrazoxane. Among patients who did not relapse, the death rate was also similar (Table I). Cox regression models were used to investigate the association between dexrazoxane treatment and relapse risk, and between dexrazoxane treatment and a composite endpoint of relapse or all-cause mortality. The following variables were included in the model: gender, age, IPI score, histology, staging, extranodal involvement at diagnosis, and bone marrow involvement at diagnosis. Dexrazoxane treatment was forced into the model. Survival models revealed that treatment with dexrazoxane, adjusted for the aforementioned baseline characteristics, was not significantly associated with either relapse risk [hazard ratio 1.31 with 95% confidence interval (CI) of 0.64–2.70, P = 0.46] or a composite endpoint of relapse or all-cause mortality (hazard ratio 1.25 with 95% CI of 0.65–2.41, P = 0.50). Comparison of Kaplan–Meier estimates for relapse showed that at every time point checked, there was no statistically significant difference in relapse rate between the group that received dexrazoxane and the group that did not (Table II). Figure 1a shows Kaplan–Meier plots for relapse as a continuous variable. There was no statistically significant difference in relapse rate over time between the group of patients treated with dexrazoxane and the group not treated with dexrazoxane (log-rank test, P = 0.275). (a) Kaplan–Meier plots of relapse over time for patients treated with dexrazoxane (continuous line) and without dexrazoxane (interrupted line). Log-rank test shows no statistically significant difference in relapse rate between the groups (P = 0.275). (b) Kaplan–Meier plots of relapse over time for DLBCL patients treated with dexrazoxane (continuous line) and without dexrazoxane (interrupted line). Log-rank test shows no statistically significant difference in relapse rate between the groups (P = 0.125). One hundred thirty-seven patients in our cohort were diagnosed with DLBCL; 36 of them received dexrazoxane. We performed a separate analysis on this group. The percentage of patients who received Rituximab with chemotherapy was similar between the group of DLBCL patients who received dexrazoxane and the group that did not (75 versus 79.2%, P = 0.6). A separate comparison of outcome for DLBCL patients showed that relapse rate, death rate, and the composite endpoint of either relapse or all-cause mortality were not statistically different between the groups, as shown in Table I. The Cox regression model was applied to DLBCL group and showed that treatment with dexrazoxane was not significantly associated with either relapse risk (hazard ratio 1.889 with 95% CI of 0.827–4.316, P = 0.131) or a composite endpoint of relapse or all-cause mortality (hazard ratio 1.783 with 95% CI of 0.832–3.824, P = 0.137). Kaplan–Meier estimates for relapse showed no statistically significant difference between DLBCL patients treated with and without dexrazoxane (Table II). Figure 1b shows Kaplan–Meier plots for relapse in DLBCL patients. There was no statistically significant difference in relapse rate over time between the groups treated with or without dexrazoxane (log-rank test, P = 0.125). Clearly, there is a benefit in favor of the use of dexrazoxane for the prevention of both clinical and subclinical heart failure. However, an important question regarding any cardioprotective intervention during anthracycline therapy is whether the cardioprotective drug could decrease the heart damage by anthracyclines without reducing the antitumor efficacy and without negative effects on toxicities other than cardiac damage. At the moment, dexrazoxane is not routinely used because of a possible interference with antitumor efficacy. Swain et al. [15] detailed two phase III trials in advanced breast cancer patients, one of which showed no effect of dexrazoxane on tumor response. In the other trial, objective response was lower for patients who received dexrazoxane rather than placebo. There was no significant difference in time to progression and overall survival (OS) between the arms in both trials. Other randomized trials did not find a difference in RR [17, 21-24]. In the early Cochrane review by van Dalen et al. [27], there was some suggestion (from five trials) that patients treated with dexrazoxane might have a lower RR (0.88, P = 0.06). For OS and progression-free survival (PFS), no statistically significant difference was found. In the updated meta-analysis [7], data on RR (defined as the number of patients in CR or PR) could be extracted from six trials with a total of 1,021 patients. The meta-analysis showed no significant difference in RR between the dexrazoxane and control group (RR 0.89, P = 0.08 from six trials). Furthermore, no significant difference between the dexrazoxane and the control group was found for both PFS and OS. None of these studies included NHL patients. Similarly, our study confirms that dexrazoxane given with anthracyclines as part of first-line chemotherapy does not have a significant effect on treatment outcome in NHL patients. The group of patients treated with dexrazoxane was older, had more cardiovascular risk factors, and had a greater proportion of patients with a higher IPI. We performed a Cox regression analysis correcting for these negative prognostic variables in the model and found no negative effect of dexrazoxane on relapse risk or on the composite endpoint of relapse and all-cause mortality. We also performed a separate analysis specifically for DLBCL patients and found similar results, thus avoiding the confounding impact of histologic diagnosis on the results. As the percentage of patients treated with Rituximab was similar between the groups, we can assume that Rituximab treatment also had no impact on our results. In conclusion, our study suggests that use of dexrazoxane during anthracycline-containing chemotherapy does not significantly affect treatment outcome in lymphoma patients. Because of its proven efficacy as a cardioprotective agent, we should consider its use in NHL patients at high risk for developing anthracycline cardiotoxicity. Both the efficacy of dexrazoxane and its effect on treatment outcome in NHL patients need to be assessed in a randomized controlled trial. We reviewed the database of all patients with NHL treated in Soroka University Medical Center between the years 2000 and 2010. Patients diagnosed with NHL who received anthracycline-based therapy with or without dexrazoxane as part of first-line treatment were included in the study. This study was approved by the institutional review board of Soroka University Medical Center. We compared outcome of first-line treatment between the two groups. Outcome was defined according to the criteria described by Cheson et al. [28] and included the rate of CR, PR, relapse, death, and a composite endpoint of relapse or all-cause mortality. Relapse was confirmed by a second biopsy of the lymph node or organ involved, or positive findings on positron emission computerised tomography (PET-CT) or CT. Cox regression models were used to investigate the association between dexrazoxane treatment and relapse risk, and between dexrazoxane treatment and a composite endpoint of relapse or all-cause mortality. The results are presented as the mean (SD) for continuous variables and as the total patients (percentage of total patients) for categorical data. T-test was used for comparison of the continuous variables or chi-square test for categorical data, with the use of Fisher's exact test if needed. We used Mann–Whitney test for the comparison of variables without normal distribution. Cox regression model was used for relapse and for the composite endpoint of relapse or all-cause mortality. The Kaplan–Meier survival curve describes the relapse rate in the group treated with dexrazoxane compared with that of the group that was not. Variable selection in multivariable modeling was based on clinical and statistical significance. A two-tailed P value of ≤0.05 was considered significant. The statistical analysis was done using SPSS version 17. Additional Supporting Information may be found in the online version of this article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. Anat Rabinovich*, Dana Weiss , Maya Weissman , Victor Novack , Itai Levi*, * Hematology Institute, Soroka University Medical Center Beer-Sheva, Israel, Faculty of Health Sciences, Ben-Gurion University of the Negev, Soroka University Medical Center, Beer-Sheva, Israel, Clinical Research Center, Soroka University Medical Center, Beer-Sheva, Israel.