Phase II study of dose‐adjusted gemcitabine, dexamethasone, cisplatin, and rituximab in elderly relapsed diffuse large B‐cell lymphoma patients

Abstract High‐dose chemotherapy and autologous stem cell transplantation (ASCT) are too toxic for elderly patients with relapsed or refractory diffuse large B‐cell lymphoma (DLBCL). Therefore, effective and tolerable regimens for elderly patients are urgently needed. The present phase II study assessed the efficacy and safety of dose‐adjusted therapy with gemcitabine, dexamethasone, cisplatin, and rituximab (GDP‐R) in this population. ASCT‐ineligible elderly patients with relapsed or refractory DLBCL received dose‐adjusted GDP‐R in each 28‐day cycle for up to six cycles. The primary endpoint was overall response rate (ORR), and secondary endpoints were complete response (CR) rate, progression‐free survival (PFS), and safety. Thirty‐three patients were enrolled and received dose‐adjusted GDP‐R. The median age was 75 years (range: 68‐87 years). The ORR was 82.8% (90% confidence interval [CI], 67.1‐93.0%), with a CR rate of 58.6% (90% CI, 41.7‐74.1%). At a median follow‐up of 20.9 months, the 2‐year PFS rate was 46.8% (90% CI, 30.7‐61.5%) and the 2‐year overall survival rate was 63.2% (90% CI, 45.8‐76.3%). The most frequently observed grade 4 adverse events were neutropenia (63.6%), thrombocytopenia (57.6%), and lymphocytopenia (39.4%). Dose‐adjusted GDP‐R is a promising salvage regimen for ASCT‐ineligible elderly patients with relapsed DLBCL after rituximab‐containing chemotherapy and warrants further investigation.


K E Y W O R D S
diffuse large B-cell lymphoma, elderly patients, gemcitabine, quality of life, relapsed, rituximab

INTRODUCTION
Diffuse large B-cell lymphoma (DLBCL) is the most common lymphoma subtype (45.3%) in Japan [1], and around 40% of cases occur in patients older than 70 years [2]. High-dose chemotherapy and autologous stem cell transplantation (ASCT) are standard treatments for younger patients with relapsed or refractory (R/R) DLBCL, but these regimens may be too toxic for elderly patients [3]. Although a few studies have focused on outcomes of relapsed elderly patients, attempts at conventional salvage regimens in older patients often do not result in disease control and have substantial morbidity [4]. Therefore, efficient and tolerable salvage treatment for elderly patients is urgently needed.
Gemcitabine is an analog of cytarabine with more efficient cellular kinetics, including intracellular incorporation, phosphorylation, and retention [5]. Hayashi et al reported that gemcitabine treatment synergistically increased rituximab-mediated complement-induced cell activity in vitro and suggested that a combination of gemcitabine and rituximab might enhance the antitumor effects of rituximab against DLBCL because of CD20 upregulation on lymphoma cells [6]. Phase II trials conducted in ASCT-eligible or ASCT-ineligible patients suggested that treatment with gemcitabine, dexamethasone, and cisplatin (GDP) could be administered efficaciously and safely, and would be well tolerated [7]. Although GDP has been proposed for ASCT-ineligible elderly patients with R/R DLBCL, results have not been satisfactory [8,9].
Therefore, we evaluated the efficacy and safety of the combination of GDP and rituximab (GDP-R) in ASCT-ineligible elderly DLBCL patients treated previously with rituximab in a single-arm phase II trial.

Study oversight
This study was a multicenter, open-label, single-arm phase II trial (UMIN000015492). It was conducted in 20 Japanese centers belong-ing to the Japanese National Hospital Organization (J-NHO). Protocol details have been published previously [10] (Supporting Information).
We designed this trial to evaluate dose-adjusted GDP-R as salvage chemotherapy for ASCT-ineligible elderly DLBCL patients previously treated with rituximab. The dose-adjustment paradigm was designed to reduce age-related or nonhematological toxicities, which were published previously (Supporting Information). The trial was approved by the ethics boards of all participating centers, and written informed consent was provided by all participants. An independent data and safety monitoring committee monitored the trial every 6 months. The J-NHO conducting this trial had its own financial support.

Pathology review process
A pathology review was conducted centrally by an expert hematopathologist, who classified all patients according to the World Health Organization classification [11]. A lymphoma phenotype was determined using available tissue blocks and evaluating a standard phenotype panel and, if indicated, cytogenetic and molecular studies, including MYC rearrangements in addition to BCL2 and/or BCL6 rearrangements (detected using fluorescence in situ hybridization [FISH] or standard cytogenetics; Supporting Information), were performed to detect double-hit lymphomas [12] and confirm DLBCL subtypes using Hans's criteria (germinal center B-cell-like [GCB] and non-GCB subtypes) [13]. Local and regional phenotype data, as well as any cytogenetic or molecular results, were tabulated for review. Then, central and regional expert pathologists rendered the diagnosis and quantitated various pathologic parameters.

Treatment protocol
Eligible patients were aged 65 years or older, with R/R DLBCL, who had received at least three cycles of one standard chemotherapeutic regimen, rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone (R-CHOP

Efficacy evaluation
Disease status assessments included physical examinations, CT scans, and bone marrow analysis. Tumor response evaluation was performed according to the Revised Response Criteria for Malignant Lymphoma [14]. CT scans of the neck, thorax, abdomen, and pelvis or positron health survey [16], and patients were assessed at baseline, middle (M), end of the protocol (E), and 6 months after the end of the protocol (S). A statistically significant change in a QOL score when compared with the baseline score was considered a clinically meaningful change.

Statistical analysis
The sample size calculation was reported previously [10].

Patients
Between

Treatment outcomes
At least one cycle of protocol therapy was administered to 33 patients.
The median cycle of GDP-R was 4 (range: 1-6  Figure 1) and 63.2% (90% CI, 45.8-76.3%; Figure 2), respectively. The OS associated with the interim response after three cycles of treatment (CR + PR vs SD + PD) was significantly different (P < .001; Figure S1). In univariate analysis using prognostic variables, none of the following variables was predictive of PFS or OS to treatment: age, sex, IPI, and pathological markers, including non-GCB, CD5, MIB-1, and Bcl2 and c-Myc IHC/FISH status. Only p53 status was predictive of both PFS and OS (Table 2).

Toxicity assessments, dose adjustment, and QOL
Thirty-three patients constituted the safety analysis set. All AEs including grade 3 and 4 are summarized in Table 3

DISCUSSION
To the best of our knowledge, this is the first study to evaluate doseadjusted GDP-R for ASCT-ineligible elderly patients with R/R DLBCL.
The results confirmed our hypothesis that dose-adjusted GDP-R is an effective and feasible salvage regimen for elderly patients. The primary endpoint of the trial was met, and the ORR of dose-adjusted GDP-R was more effective than that found in previous trials.  were assessed using p53 overexpression as a surrogate, which was significantly associated with decreased survival [17]. Several secondline chemotherapy regimens designed to increase the proportion of ASCT-ineligible elderly patients with R/R DLBCL through the use of more intense chemotherapy were associated with substantial hematologic toxicity and considerable use of healthcare resources [18]. In our study, p53 expression levels correlated with PFS and OS, and highlighted the potential clinical effectiveness of using bendamustine in combination with rituximab (BR) [19], which will be evaluated in the treatment of lymphoma with higher p53 expression levels in the near future. Recently, polatuzumab vedotin combined with BR resulted in a significantly higher CR rate and reduced the risk of death by 58% compared with BR in patients with transplantation-ineligible R/R transplant-ineligible DLBCL [20], but this has not been confirmed. The 2-year PFS rate after treatment with dose-adjusted GDP-R was 46.8%, which was noninferior compared with previous trials (Table   S1). Hou   in which dose-adjusted GDP-R was administered to elderly patients with R/R DLBCL, which was well tolerated without unexpected AEs or treatment-related deaths.

Prognosis of patients with R/R DLBCL is
In summary, this novel therapy was well tolerated; incidence, severity, and type of AEs were acceptable compared with those observed previously [8,19]. Dose-adjusted GDP-R can be considered the preferred treatment option for ASCT-ineligible elderly patients with R/R DLBCL; however, this issue warrants confirmation in a larger number of patients. Organization.

DATA AVAILABILITY STATEMENT
De-identified data are available upon request. SY contributed to the study design, data analysis, and manuscript   preparation. AK, IC, HI, NS, NH, MS, TS, TK, TY, IY, SY, KS, TH, HT, KO, MT, AS, HI, and HN reviewed the manuscript. AS was responsible for data management. AK was responsible for statistical analysis.