Nonpeghylated liposomal doxorubicin combination regimen (R‐COMP) for the treatment of lymphoma patients with advanced age or cardiac comorbidity

Abstract Doxorubicin is the most effective single agent in the treatment of non‐Hodgkin's lymphoma (NHL). Its use is limited because of the cardiac toxicity primarily in elderly patients (pts) and in pts with history of cardiac disease. Liposomal doxorubicin has been proven to reduce cardiotoxicity. The aim of this retrospective study was the use of nonpeghylated liposomal doxorubicin (NPLD) in term of efficacy, response rate and incidence of cardiac events. We retrospectively collected the experience of 33 Hematological Italian Centers in using NPLD. Nine hundred and forty‐six consecutive pts treated with R‐COMP (doxorubicin was substituted with NPLD, Myocet) were collected. Median age was 74 years, the reasons for use of NPLD were: age (466 pts), cardiac disease (298 pts), uncontrolled hypertension (126 pts), other reasons (56 pts). According to clinicians' evaluation, 49.9% of pts would not have used standard doxorubicin for different situations (age, cardiomyopathy, previous use of doxorubicin, and uncontrolled hypertension). Overall 687 pts (72.6%) obtained a complete remission (CR). About 5% (n = 51) of subjects developed major cardiotoxic events including heart failure (N = 31), ischemic heart disease (N = 16), acute heart attack (N = 3), and acute pulmonary oedema (N = 1). After a median follow‐up of 32 months, 651 pts were alive and the overall survival (OS) was 72%. After a median observation period of 23 months disease free survival (DFS) was 58%. Either in univariate or in multivariate analysis OS and DFS were not significantly affected by age or cardiac disease. Our findings strongly support that including R‐COMP is effective and safe when the population is at high risk of cardiac events and negatively selected. Moreover, the use of this NPLD permitted that about half of our population had the opportunity to receive the best available treatment.


| INTRODUCTION
Anthracyclines play a key role in the treatment of lymphoproliferative disorders and its introduction since 1970 1 in CHOP regimen led to very important results in the cure of lymphomas. Particularly in diffuse large B cell lymphoma (DLBCL) CHOP regimen has confirmed its superiority over many other schemes used. 2 The combination of monoclonal anti-CD20 antibody (rituximab) with CHOP has significantly improved the efficacy of the regimen and R-CHOP is nowadays the standard therapy for DLBCL. 3 Unfortunately, anthracyclines are associated with cardiotoxicity, mainly congestive heart failure (CHF), that can occur during treatment (acute toxicity) or several years after treatment. 4 Anthracycline toxicity is dose dependent and a cumulative dose of doxorubicin of 550 mg/m 2 is associated with 30% risk of cardiovascular disease but CHF or other cardiotoxicities are now known to occur at a lower cumulative dose of 200 mg/m 2 . Using treatment without anthracyclines to mitigate the cardiovascular risk is associated to a poorer disease control and worse survival 5 suggesting that it is not possible to propose a first line therapy without anthracyclines when the objective of the treatment is cure. Different approaches were used to avoid the cardiotoxic risk such as limiting anthracyclines dosage but this procedure could be reasonable in very elderly patients to reduce the whole toxicity 6,7 whereas is often associated with a worse outcome in younger population. Drugs as for example dexrazoxane, have been demonstrated their protective effect in pediatric but not in adult population. [8][9][10] The use of lyposomal non peghilated anthracyclines (Myocet) firstly used in the treatment of patients affected by breast cancer showed in clinical studies a reduction of cardiotoxicity. 11,12 This drug was subsequently used in the treatment of aggressive lymphoma substituting hydroxydaunorubicin in the CHOP schema (called COMP). Several studies in Hematology setting confirmed in high risk and highly selected population the efficacy and particularly the safety of the COMP Scheme. [13][14][15][16][17][18] Due to a large territorial experience in Italy with this drug, thanks to the possibility to use it according to Italian law 648/96, we retrospectively collected clinical information from a large number of Hematological Italian Centers with the aim to evaluate the effectiveness and safety of this regimen in collaboration with the Fondazione Italiana Linfomi.

| PATIENTS AND METHODS
According to Italian law 648/96 and starting from 2008 in Italy liposomal nonpeghilated anthracyclines could be used in elderly patients (over 65 years) or in cardiopathic patients. Then we had the opportunity to collect a large amount of data from several Hematological Italian Centers, including patients treated with this type of anthracycline within the classical regimen R-CHOP.
Data were collected in a centralized data set from medical records of the single Institutions collecting: some essential clinical characteristics, response to therapy, overall survival and event free survival, cardiologic events observed during therapy and during the first 2 years of follow-up. Moreover, we wanted to register why patients were treated with liposomal anthracycline, to know if the clinicians should have used standard anthracycline or they have declined the use of anthracycline.
The primary end point was the number of cardiologic complications. Secondary end points were response to therapy, overall survival, disease free survival, and event free survival.

| Efficacy and safety response
Although a group of patients was evaluated with FDG-PET, responses to therapy (complete response: CR; partial response: PR; progressive disease: PD; non responders: NR) were defined according to 1999 Cheson et al criteria. 19 Overall response rate (ORR) was calculated as the sum of complete and partial response rate. Disease-free survival (DFS) was measured from the end of the therapy to relapse or last contact, and the curves were determined using only patients in CR after induction therapy. Overall survival (OS) was dated from the disease onset to last contact.
Event-free survival (EFS) was assessed for major cardiac events (heart failure, ischemic heart disease, acute heart attack, and acute pulmonary oedema) and was dated from the end of the therapy. If a cardiac event was observed during therapy, the treatment was stopped and the time point registered as the end of therapy date.
Safety response and all cardiac events were defined according to the National Cancer Institute's Common Toxicity Criteria. 20

| Statistical analysis
Descriptive statistics was used to summarize continuous and categorical variables.
The Kaplan-Meier method was used to estimate EFS, DFS, and OS. The log rank test was used to compare the survival curves of groups that were defined according to different prognostic factors. A competing risk analysis was carried out with death from cardiotoxicity as the primary event of interest and death from all other causes as competing event.
Cox regression models were used to investigate the predictors of mortality from all causes and the predictors of relapse. The predictive and confounders factors included both demographics (gender, age) and clinical characteristics (IPI class, symptoms (yes/no), year of diagnosis, previous cardiac disease (yes/no), LVEF (normal/abnormal), number of R-COMP cycles (≤3/>3), RT (yes/no), concomitant therapies as cardiotoxic or cardioprotective drugs (yes/no). Stratification was applied in the Cox regression model when the proportional hazard assumption was violated.
The association of CR and PR with demographic and clinical patients' characteristics was assessed by performing a multinomial logistic regression since the proportional odds assumption underlying the ordinal logistic model was violated.
A P-value <.05 was considered as statistically significant. All statistical analyses were performed using the R 3.6.0 for Windows. 21 3 | RESULTS  Among subjects that would have received standard anthracycline-based therapy also in the case of no available liposomal formulation, the distribution of the cardiotoxic events was as follow:

| Cardiac toxicity
10 heart failures, 3 ischemic heart diseases, 1 acute heart attack with no significance differences between the two groups (P = .704).
The overall probability of not experiencing a cardiotoxic event after 5 years was 92.6% [95% CI, 90.3%-94.9%]. Age greater than 75 years (n = 466) did not affect the probability of event-free condition (χ 2 = 0, P = .800; Figure 1). We did not observe any difference in OS between patients that received R-COMP therapy because of age and patients that received the same therapy because of other causes (χ 2 = 0.0, P = .900). We did not observe any difference in OS either between patients with abnormal and normal LVEF (χ 2 = 2.9, P = .09) nor between patients with and without a previous cardiopathy (χ 2 = 1.1, P = .300). Subjects that according to physicians' opinion would have received standard anthracycline-based therapy also in the case of no available liposomal formulation had both a higher overall and lymphoma-specific survival probability than subjects that would not have received it (χ 2 = 20.7, P < .0001 and χ 2 = 11.0, P < .0001, respectively; Figure 4A,B).  Table 2.

| Survival and causes of death
Among the 17 patients that died because of cardiac events, 14 were the above-mentioned subjects that experienced a cardiac event during the observation period. The chance of dying from all other causes was significantly higher in subjects older than 75 years (P < .0001) whereas no difference was seen for lethal cardiac events between the two age groups (P = .431; Figure 5).  .0064). An interpatient variability in risk of anthracyclines induced CHF confirms a genetic association as reported in several studies. [30][31][32][33][34] As reported in the literature the risk of cardiotoxicity is significant in patients without risk factors therefore patients with signs of cardiac failure are preventively excluded from treatment with R-CHOP. The T A B L E 2 Five-year estimation of survival probability stratified by IPI score and number of total R-COMP cycles Dexrazoxane is the only one drug approved in US as cardioprotective agent 39,40 but its use is limited to a small percentage of patients due to an unjustified concern that this drug may interfere with antitumor activity and increase the risk of second neoplasms. [41][42][43] The liposomal formulation showed in large randomized controlled trials and meta-analysis on metastatic breast cancer a reduction of the cardiotoxic risk, observed with conventional anthracyclines without affecting antitumor activity. [44][45][46] The safe cardiologic profile of liposomal anthracyclines was welldocumented by studies in which this drug was administered in selected patients at very high risk of cardiotoxicity (advanced age, anthracyclines pretreatment, and concomitant cordiotoxic drug). [47][48][49][50] The aim of our study was to evaluate in a multicentric and real- the bad prognosis of this group of patients was confirmed by the fact that although they were treated with an optimal regimen they achieved worse results than patients treated with standard anthracyclines. However, the use of NPLD permits us to treat patients that, with high probability, would have been treated with alternative schemes, certainly less effective than R-CHOP particularly in DLBCL.

| Response to therapy
At the same time, we have treated a group of patients considered potentially available for standard doxorubicin obtaining very high complete remission rate confirming that the use of NPLD did not reduce the efficacy of the treatment. The first objective of the study was to confirm the safety, particularly cardiac, of this NPLD even in a group of patients with high-risk situations due to age and/or preexisting cardiopathy. We observed 73% of complete remission that is completely superimposable with the results reported in other studies using R-CHOP particularly if we consider the high median age of the population treated as reported in the randomized phase III study by Fridrik et al. 17 The overall survival after a median observation of 34 months (range 1-187 month) was 70% (95% C.I.: 67.4-73.4%) and the DFS after a median observation of 24 months was 76% (95% C.I.: 74-80.7%).
In conclusion, using this liposomal formulation on a high or very high-risk population, the incidence of cardiotoxicity was absolutely comparable with the events historically reported using standard doxorubicin. The use of NPLD that significantly reduces the accumulation and consequently the toxic effect of doxorubicin on myocites permits to use the standard treatment in a large proportion of the population and the more patients we can treat adequately the more patients we can cure. Our study confirms also the effectiveness in term of complete remission rate of the COMP regimen. Finally we must consider the possible bias of selection due to the retrospective nature of the study and for this reason our conclusions should be cautious but we are confident that either the evaluation performed by a single center or the large number of patients recruited can mitigate this potential bias.