Real-World Management and Outcomes of Patients With Relapsed/Refractory Diffuse Large B-Cell Lymphoma Treated in the United States

Abstract: Given new therapeutic advances that are potentially curative in diffuse large B-cell lymphoma (DLBCL), it is important to understand what has been observed to date in this population to assess improvement in outcomes; data from US community practices have been limited. A retrospective, observational cohort study of patients with relapsed DLBCL treated between January 1, 2010, and December 31, 2016, was performed. Patients treated with first-line rituximab-CHOP were classified as early relapse and late relapse based on time of second-line (2L) initiation. Descriptive statistics summarized patient characteristics and outcomes. Among 392 evaluable patients who received 2L treatment, the median age was 67 years (range: 21-84 years), and 59% were male. In 2L therapy, 31% patients received a salvage therapy intended for autologous stem cell transplantation (ASCT); 29% received a regimen not intended for ASCT; and 40% received a range of other regimens (DA-EPOCH-R, 2%; and other regimens among fewer than 5 patients each, 28%). The wide treatment variation for relapsed DLBCL in the real-world setting demonstrates a need for clinical pathways in this patient population.

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of aggressive non-Hodgkin lymphoma (NHL) in the United States, accounting for approximately 25% of NHL cases^1-3 with an annual incidence estimated at 9 cases per 100,000 persons.³ DLBCL is typically diagnosed in patients in their sixties, although it has heterogeneous characteristics and presentation with a broad age range at diagnosis.¹ First-line (1L) therapy for DLBCL has been established as treatment with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone),⁴ with dose-adjusted (DA) EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) as a less common, additional option,⁵ or participation in a clinical trial. 

Despite the treatment advances with the use of the anti-CD20 monoclonal antibody rituximab to treat DLBCL, approximately 40% of patients relapse or are refractory to chemotherapy, with low subsequent response rates and an associated poor prognosis and a loss of life expectancy of about 5 years compared with the general population.^6-8 Except for high-dose therapy followed by autologous stem cell transplantation (ASCT) in eligible patients, no clear standard of care exists after 1L therapy. Multiple treatment options are available for second-line (2L) treatment in patients who fail or are ineligible for ASCT,^9-18 but none are considered curative. While clinical and performance status may contribute to nonreferral to ASCT for hematologic malignancies, other socioeconomic factors shown to curb referral included lack of insurance, lack of social support, and substance abuse.¹⁹ Additional barriers to ASCT therapy include increasing age, travel distance to transplant facility, and racial disparities.^20,21

Given the substantial unmet medical need in the DLBCL patient population, this study sought to assess retrospective data to provide greater insight into this population regarding patient characteristics, clinical outcomes, and treatment patterns among patients with relapsed DLBCL beyond 1L R-CHOP. 

Methods

Study Design

This was a retrospective, observational cohort study that was conducted using an electronic medical record (EMR) data set, derived from multiple EMR platforms in the community oncology setting. Patients were followed over time until their last EMR encounter. 

The objectives were to describe treatment patterns and outcomes in a cohort of patients with relapsed DLBCL in the community oncology setting. Exploratory objectives were to estimate real-world overall survival (OS) for these patients and estimate the effect of ASCT and timing of relapse (ie, early vs late relapse) on survival.

Data Source

Clinical outcome data were derived from the PRA Oncology EMR database, which contains clinical data from over 126 community oncology and hematology practices across more than 707 locations. Records of patients receiving care in these facilities date back to 2004; the database contains longitudinal data on over 700,000 individual cancer patients. Data are aggregated from a variety of oncology-specific EMR systems and include the outpatient-practice encounter history of the patients under care including diagnosis type (ICD-9/10 CM), disease profile (eg, American Joint Committee on Cancer stage, tumor-node-metastasis stage, anatomic sites of involvement), biomarkers, standard laboratory records (eg, blood chemistry, lipid, renal function tests/panels), and vital status (through external death file matching). Both intravenous and oral antineoplastic agent treatment orders are captured (ie, drug name, treatment date/cycle number, and prescription date) from both the EMR data as well as from payer claims records for outpatient and pharmacy claims.

Cohort Selection

Patients aged ≥18 years at time of DLBCL diagnosis (per diagnosis field or histology results) with relapsed disease were identified. Patients who were treated with a line of therapy following discontinuation of 1L R-CHOP were included in the study cohort. “Early” relapse was defined as progressive disease on treatment or within 365 days from completing the last cycle of 1L R-CHOP. “Late” relapse was defined as any 2L treatment 365 to 730 days from the last R-CHOP cycle. Patients participating in a clinical trial were excluded, as were patients who received rituximab monotherapy in the 2L or later setting (due to concerns that rituximab monotherapy might have represented maintenance instead of active therapy).²²

In an attempt to describe contemporary treatment patterns and clinical outcomes, only patients with a diagnosis of DLBCL who relapsed after chemotherapy since 2010 (ie, observation period of January 1, 2010, to November 30, 2016) were eligible.

Line of Therapy Assignment

For each regimen identified, the line of therapy in which it occurred was determined. The end of 1L was defined by the date of discontinuation of R-CHOP. A new line was defined as the initiation of a new drug regimen or the addition of new drug(s) to the initial regimen, outside of the first 30 days of that line of therapy. Lines were defined as distinct combinations; retreatment with the same therapy did not increment or end the line. A library of DLBCL treatments were used when identifying new lines of therapy. Regimens in 2L and later were classified as salvage therapy intended for ASCT (ie, DHAP [dexamethasone, cisplatin, cytarabine] ± rituximab; ESHAP [etoposide, methylprednisolone, cytarabine, cisplatin] ± rituximab; ICE [ifosfamide, carboplatin, etoposide] ± rituximab; GDP [gemcitabine, dexamethasone, cisplatin/carboplatin] ± rituximab or [gemcitabine, dexamethasone, carboplatin] ± rituximab), regimens not intended for ASCT (bendamustine ± rituximab; CEPP [cyclophosphamide, etoposide, prednisone, procarbazine] ± rituximab; CEOP [cyclophosphamide, etoposide, vincristine, prednisone] ± rituximab; GemOx ± rituximab; gemcitabine, vinorelbine ± rituximab), and therapies not classified by intention to proceed to ASCT (ie, all other regimens).

Treatment duration was calculated by the date of discontinuation for that line of therapy minus the date of initiation of that line of therapy with an allowable gap of up to 30 days. A gap of >30 days was not considered as part of the duration of that line of therapy. 

Statistical Methods

Patient demographic and clinical characteristics and study outcomes were summarized using descriptive statistics (eg, mean, standard deviation, count, proportion) for the study cohort and for subgroups of interest (ie, early relapse vs late relapse). Statistical comparisons for the purpose of identifying potential confounders were made using chi-square tests and t tests, and nonparametric tests were considered for non-normally distributed data in lieu of these methods. Any patient characteristic found to be statistically significant on univariate analysis at α=0.20 was considered for inclusion in multivariable models. 

Univariate analyses were performed to compare treatment patterns across subgroups. Adjusted Cox proportional hazard models were fit to estimate the hazard ratio (HR) of time to next treatment, adjusted for the following potential confounders: gender, age, early vs late relapse, ASCT, and comorbidities of interest (diabetes, hypertension, myocardial infarction, dementia, and depression) given their potential influence on treatment decisions.²³ Two-sided statistical significance was set at α=0.05.

Exploratory Analyses

Identification of patients who underwent ASCT and as well as those who died were made based on surrogate measures. As such, analyses involving ASCT and death, while validated by others,²⁴ are exploratory to depict trends in outcomes among this study cohort. The following are details on these surrogates and the associated analyses. 

ASCT. As part of the exploratory analyses, surrogates for ASCT were identified by clinical experts on the study team, including one based on patterns of care (Figure 1). Treatment failure within 1 year of completing R-CHOP is often followed by GDP (gemcitabine, dexamethasone, and cisplatin/carboplatin),^25,26 DHAP (dexamethasone, cisplatin, and cytarabine),^27-29 ESHAP (etoposide, methylprednisolone, cytarabine, and cisplatin),^30,31 or ICE (ifosfamide, carboplatin, and etoposide) ^29,32,33 alone or with rituximab regimens administered 2 to 4 times as proof of chemo-sensitivity. If any of the patterns of care described were identified, followed by a gap of >30 days and subsequent clinic office surveillance, it was assumed that this patient received an ASCT. The date of ASCT was calculated as 30 days prior to resumption of clinic office surveillance. The presence of ICD-9/10 codes (history of peripheral stem cell transplant: V42.82 or Z94.84; history of bone marrow transplant: V42.81 or Z94.81) also served as surrogates of ASCT. The date of ASCT was set as 30 days prior to the earliest code for transplant; patients identified as receiving ASCT by ICD-9/10 codes were reviewed by clinical team members to confirm the assignment of the date of ASCT. 

Subgroups of interest derived from the identification of ASCT (ie, ASCT vs no ASCT) and combined subgroups (ie, early relapse with ASCT, early relapsed without ASCT) were summarized by descriptive statistics. Statistical comparisons were made using chi-square tests and t tests and their nonparametric counterparts for non-normally distributed data. 

Survival Endpoints. Mortality information (alive or deceased) was not available for all patients in the dataset. As a result, survival was an exploratory endpoint based on validated surrogate measures. The last observed claim/EMR encounter followed by a period of 180 days of no clinical activity was used as a proxy for the date of death when mortality data were not available.²⁴ All other patients without these ending periods of clinical activity were marked as alive at end of their follow-up. Additionally, a date of disease progression was not available from the EMR or claims record. As such, time to next treatment was used as a surrogate endpoint for progression.^34-36 The date of the initiation of next treatment was used as the date of the progression event. 

Exploratory analyses of OS involved the count and proportion of patients defined as deceased, with time to event analyses using Kaplan-Meier method. Adjusted Cox proportional hazard models were fit to estimate risk (HR) of death, adjusted for the following potential confounders: gender, age, early vs late relapse, ASCT, and comorbidities of interest (ie, diabetes, hypertension, myocardial infarction, dementia, and depression) given their potential influence on treatment decisions.²³ 

Results

Patients

Of 9156 patients with a DLBCL diagnosis in their record, a final cohort of 392 patients (early relapse, n=333; late relapse, n=59) met all study selection criteria, including exclusion of patients with rituximab monotherapy in 2L or higher treatment, as well as requiring follow-up after 1L R-CHOP involving a 2L therapy (Table 1). Overall 186 patients had ≥6 months of follow-up from the start of 2L therapy, with 97 of these patients having ≥12 months of follow-up. Median follow-up time from diagnosis, start of 1L therapy, and start of 2L therapy were 12.7 months, 11.6 months, and 4.6 months, respectively.

Baseline characteristics are summarized in Table 2. Overall, 59% of the 392 patients were male, mean age at relapsed disease was 65.6 years, and the most common comorbidities were hypertension and type 2 diabetes at 19% and 3%, respectively. Baseline lactate dehydrogenase (LDH) values were available for 240 patients (61%), for whom the mean was 296.6 U/L with 66 (16.8%) patients having a baseline level above the upper limit of normal (>280 U/L).

Treatment Patterns

All 392 patients received a 2L therapy: 66% (n=257) of patients continued on to third-line (3L) therapy, 24% (n=93) continued to fourth-line (4L) therapy, and 11% (n=42) continued to fifth-line therapy. In the 2L setting, 31% received salvage therapy intended for ASCT, 29% received regimens not intended for ASCT, and 40% received therapies not classified by intention to proceed to ASCT; corresponding rates in the 3L setting were 16%, 31%, and 46%, respectively, and in the 4L setting were 21%, 26%, and 52%, respectively. Regimen categories by subgroup are summarized in Figure 2. Of the 392 patients with relapsed DLBCL, 51 (13%) patients underwent ASCT, including 42 and 9 patients with early and late relapse, respectively.

In 2L therapy, ICE ±R (ifosfamide, carboplatin, etoposide, with/without rituximab) was the most common salvage therapy intended for ASCT given in 102/122 patients with salvage therapy intended for ASCT; bendamustine ±R was the most common 2L regimen not intended for ASCT (69/113) followed by GemOx ±R (gemcitabine, oxaliplatin, with/without rituximab; 28/113) and CEOP ±R (cyclophosphamide, epirubicin, vincristine, prednisone, with/without rituximab; 16/113). Other regimens in 2L therapy were R-CHOP or DA-EPOCH-R (58/392) and a variety of other regimens among fewer than 5 patients each (127/392). 

Among all 392 patients, the duration of treatment was longest in the 1L setting and was shorter in each subsequent line received. Median duration of 1L R-CHOP was approximately 3 months, with median durations of each subsequent line of approximately 1 to 2 months (Table 3). On average, 5 cycles of R-CHOP were received in the 1L setting (Table 3).

Time to Next Treatment

After 1L therapy, time to next treatment was significantly shorter among those with early relapse compared to those with late relapse (Figure 3A). After 2L therapy and 3L therapy, risk of progression or death was similar across the 4 subsets (Figure 3B and 3C).

Exploratory Results

ASCT. The majority of patients who underwent ASCT in early relapse did so after 2L therapy (24/42, 57%), while the majority of patients who underwent ASCT in late relapse had first received 2L or 3L therapy (5/9, 55%). Thirteen (31%) of the patients with early relapse and 4 (45%) of the patients with late relapse underwent ASCT in the 4L or later setting. After 1L therapy, time to next treatment was similar among those who underwent ASCT compared to those who did not (Figure 3A). From 2L therapy, time to next treatment was significantly longer among those who underwent an ASCT (4.1 months; 95% CI, 2.7-6.5) compared to those who did not (2.3 months; 95% CI, 1.8-2.9; P< .01) (Figure 3B).

Overall Survival. In total, 183 (46.7% of all patients) deaths were estimated to have occurred based on using the proxy clinical activity measure previously described. Median OS from 2L therapy was similar between early vs late relapsed DLBCL patients (11.8 vs 14.4 months; P=.78), who also had a similar adjusted risk of death (HR, 0.94; 95% CI, 0.63-1.41). When OS from 3L therapy was analyzed, there were no significant differences between early vs late relapse (data not shown). The adjusted median survival from 2L therapy was significantly longer among those who underwent an ASCT (not reached; 95% CI, not estimable) compared to those who did not (10.1 months; 95% CI, 7.4-14.5), representing a 48% reduced adjusted risk of death among those who underwent an ASCT (HR, 0.52; 95% CI, 0.33-0.82) (Figure 4A). Similarly, median survivals from 3L and 4L therapies were longer among those who underwent an ASCT (18.9 months and 11.1 months, respectively) compared to those who did not (9.4 months and 4.7 months, respectively), albeit not statistically significantly different after adjustment for patient mix (Figure 4B and 4C).  

Discussion

Managing DLBCL at time of relapse in community oncology practice remains a challenge. Understanding the current treatment patterns and outcomes of these patients has the potential to identify suitable candidates for more recently approved therapies. We found that nearly two-thirds of all relapsed patients received only 2L therapy, mostly with regimens not intended for ASCT or unclassified with respect to the intention to proceed to ASCT. Wide variation in regimens existed in 2L therapy, suggesting a lack of standard of care or clear guidance in the treatment of R/R DLBCL patients. This presents health care decision-makers  and policymakers with an opportune setting in which to develop clinical pathways to aid clinicians in identifying the best treatment for each patient. Further research is warranted to identify unnecessary variation in care among these patients and to identify optimal treatment approaches associated with superior patient outcomes.  

Interestingly, nearly half of patients were excluded from this study when restricting the cohort to those who received 1L R-CHOP. The current study focused on the standard regimen of R-CHOP to maximize the specificity of DLBCL diagnosis in this cohort. Among adult patients with a diagnosis of DLBCL in their record not on clinical trial, 1L DA-EPOCH-R was observed among only 38 patients. For this reason, it was omitted from this analysis, which instead focused on the dominant 1L regimen of R-CHOP. The nonstandard regimens observed in those patients were 1L CHOP without rituximab (15%), rituximab monotherapy (14%), bendamustine plus rituximab (9%), cyclophosphamide, vincristine, and rituximab (4%), and the rest (2% or less) with other regimens, which may depict a different patient population—one with different comorbid burden. The additional requirement for study inclusion of treatment with a 2L therapy (ie, treatment in the relapsed/refractory setting) further selected a population who required subsequent treatment following R-CHOP; this criterion excluded patients who were successful on R-CHOP and did not receive a 2L treatment, ever or during the study period. 

While ASCT is a potentially curative modality after failure of 1L rituximab-based chemotherapy for DLBCL, we found an unexpectedly low rate of real-world eligibility and use. The 13% ASCT rate, despite the limitations of the surrogate measure, was markedly lower than expected given that ASCT is recommended for all eligible patients.³⁷ The older age of the patients included in this analysis (mean 65.5 years old at relapsed/refractory disease) may begin to explain this observed rate, among whom its use is associated with slightly increased risk of toxicities and lower effectiveness.³⁸ In exploratory analyses, we found that patients who relapsed early and subsequently underwent ASCT had a longer duration of response after 2L therapy, almost to the same extent as late relapse patients with ASCT. Similarly, median OS from 2L therapy was significantly longer among those who underwent ASCT compared to those who did not (21.4 vs 10.5 months; P<.01), with the risk of death among ASCT recipients significantly reduced by 48%, after adjusting for patient characteristics (HR, 0.52; 95% CI, 0.33-0.81), and subsequently by 37% (HR, 0.63; 95% CI, 0.34-1.16) and 45% (HR, 0.55; 95% CI, 0.19-1.61) in the 3L and 4L setting, respectively. The adjusted risk of death among early relapsed patients who underwent ASCT was significantly reduced by 53% compared to patients with late relapse who did not undergo ASCT (HR, 0.47; 95% CI, 0.26-0.86). 

Initial searching within the EMR also revealed a surprisingly high proportion of identified patients (20%) who were treated with rituximab monotherapy. While some of this use may have been in the 2L or subsequent settings, as it was not possible to determine which patients were receiving it as maintenance therapy (which was not a population of interest), these patients were excluded from the subsequent analyses. Benefits of maintenance rituximab have not been elucidated and clinical trials have yielded mixed results, some showing no benefit and others prolongation of event-free survival and OS after 1L treatment, especially in men.³⁹ An overuse of maintenance rituximab and associated increased risk of toxicity in older adults with NHL were evident in a prior Surveillance, Epidemiology and End Results (SEER)-Medicare analysis.²² Another SEER-Medicare analysis focused on older adults with DLBCL showed that there is a propensity to use rituximab monotherapy (in place of rituximab plus chemotherapy) in patients who are older, female, and/or have earlier stage disease, and—albeit not a guideline-recommended option—OS was improved relative to no treatment in these patients.⁴⁰

The poor prognosis of patients with relapsed DLBCL was evident, particularly for patients who did not undergo ASCT, for whom median OS from time of relapse was 10.5 months. Other groups have recently reported median survival durations <12 months in patients with relapsed DLBCL. A recent meta-analysis of 635 patients with chemotherapy-refractory DLBCL, derived across institutions (MD Anderson Cancer Center and the Mayo Clinic/University of Iowa SPORE) and two randomized clinical trials (Canadian Cancer Trials Group LY.¹² and the LYSARC CORAL studies), reported a median OS of 6.6 months (95% CI, 6.1-7.5 months) and a 2-year survival rate of 21% from the start of salvage therapy.⁶ By refractory/relapse status, median survival was 7.1 months in patients with primary refractory disease, 6.4 months in patients refractory to 2L or higher therapy, and 6.4 months in patients relapsing within 1 year of ASCT. Median survival was 7 months for clinical trial participants and 6.2 months for patients identified via database search, with 2-year OS rates of 24% and 17%, respectively and corresponding 5-year rates of 20% and 7%, respectively. The approximately 4-month, observed, higher median survival in our cohort likely reflects our use of a proxy definition of death, as well as the heterogenous mix of relapsed and refractory patients, given the lack of differentiation between relapsed and refractory patients. In addition, our study captured more patients in first relapse and, therefore, is weighted more heavily in 2L treatment. As such, a portion of our sample likely included nonrefractory patients in the early relapse group resulting in the higher observed, although of similar magnitude, estimated median OS. 

We acknowledge several limitations to our study design, including those inherent to retrospective analyses, such as the potential for residual and unobserved confounding in multivariate analysis, despite adjustment for variables identified as potential confounders, and the potential for missing data and/or misclassified data points. Since the EMR data were derived from a variety of EMR platforms used in community oncology practices, ASCTs were not captured in the outpatient EMR dataset used given that they are hospital-based procedures. The validity of the chosen surrogate was only informally assessed, based on treatment patterns among those identified as recipients of ASCT, compared to those identified as not having received ASCT, as defined by surrogate measures, although clinical staff reviewed individual treatment records for verification of likely ASCT and confirmed with claims when available. Future work is warranted to validate the algorithm used to identify ASCT. Regarding the statistical comparisons, they may have been confounded by the small sample sizes, especially for some of the subgroups. There was substantial attrition in patient selection after requiring a subsequent line of therapy following progression after or discontinuation of 1L R-CHOP. This attrition is likely largely due to the additional follow-up time after 1L R-CHOP required in the dataset. Attrition due to long follow-up requirements is common in retrospective, secondary data analyses. All of these aspects may impact the generalizability of the study results. In particular, the study describes treatment and outcomes prior to recently approved therapies. Nevertheless, this study establishes a baseline observed regarding the treatment and management of DLBLC patients in this time period. The size of the database and the compilation of data across multiple EMR platforms increases the generalizability of the study results. As previously described, the cohort was restricted to those treated with 1L R-CHOP, so generalizability to DLBCL patients treated with other 1L therapies are not appropriate. 

Conclusion

Our retrospective observations demonstrate wide variation in care among patients with relapsed DLBCL outside the clinical trial setting, warranting the need for clinical pathways in this patient populations. Given new therapeutic advances that are potentially curative in DLBCL, guidance on the treatment and management of patients with relapsed DLBCL is increasingly needed.  

References

1. Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood. 2016;127(20):2375-2390. doi:10.1182/blood-2016-01-643569

2. Friedberg JW, Fischer RI. Diffuse large B-cell lymphoma. Hematol Oncol Clin North Am.  2008;22(5):941-952, ix. doi:10.1016/j.hoc.2008.07.002

3. Shenoy PJ, Malik N, Nooka A, et al. Racial differences in the presentation and outcomes of diffuse large B-cell lymphoma in the United States. Cancer. 2011;117(11):2530-2540. doi:10.1002/cncr.25765

4. Coiffier B, Thieblemont C, Van Den Neste E, et al. Long-term outcome of patients in the LNH-98.5 trial, the first randomized study comparing rituximab-CHOP to standard CHOP chemotherapy in DLBCL patients: a study by the Groupe d’Etudes des Lymphomes de l’Adulte. Blood. 2010;116(12):2040-2045. doi:10.1182/blood-2010-03-276246

5. Wilson WH, Dunleavy K, Pittaluga S, et al. Phase II study of dose-adjusted EPOCH and rituximab in untreated diffuse large B-cell lymphoma with analysis of germinal center and post-germinal center biomarkers. J Clin Oncol. 2008;26(16):2717-2724. doi:10.1200/JCO.2007.13.1391

6. Crump M, Neelapu SS, Farooq U, et al. Outcomes in refractory aggressive diffuse large B-cell lymphoma (DLBCL): Results from the international SCHOLAR-1 study. Blood. 2017;130(16):1800-1808. doi:10.1182/blood-2017-03-769620

7. Ekberg S, Jerkeman M, Andersson PO, et al. Long-term survival and loss in expectancy of life in a population-based cohort of 7114 patients with diffuse large B-cell lymphoma. Am J Hematol. 2017. doi:10.1002/ajh.25147 

8. Add Koh JJ, Lim ST, Sultana R, et al. Predictors of early vs late diffuse large B cell lymphoma (DLBCL) relapses in the rituximab era. J Clin Oncol. 2018;36(suppl 15):e19553. doi:0.1200/JCO.2018.36.15_suppl.e19553

9. Weidmann E, Kim SZ, Rost A, et al. Bendamustine is effective in relapsed or refractory aggressive non-Hodgkin’s lymphoma. Ann Oncol. 2002;13(8):1285-1289. doi:10.1093/annonc/mdf189

10. Vacirca JL, Acs PI, Tabbara IA, Rosen PJ, Lee P, Lynam E. Bendamustine combined with rituximab for patients with relapsed or refractory diffuse large B cell lymphoma. Ann Hematol. 2014;93(3):403-409. doi:10.1007/s00277-013-1879-x

11. Ohmachi K, Niitsu N, Uchida T, et al. Multicenter phase II study of bendamustine plus rituximab in patients with relapsed or refractory diffuse large B-cell lymphoma. J Clin Oncol. 2013;31(17):2103-2109. doi:10.1200/JCO.2012.46.5203

12. Chao NJ, Rosenberg SA, Horning SJ. CEPP(B): An effective and well-tolerated regimen in poor-risk, aggressive non-Hodgkin’s lymphoma. Blood. 1990;76(7):1293-1298.

13. Moccia AA, Schaff K, Hoskins P, et al. R-CHOP with etoposide substituted for doxorubicin (R-CEOP): excellent outcome in diffuse large B cell lymphoma for patients with a contraindication to anthracyclines. Blood. 2009;114(22):408. doi:10.1182/blood.V114.22.408.408

14. López A, Gutiérrez A, Palacios A, et al. GEMOX-R regimen is a highly effective salvage regimen I patients with refractory/relapsing diffuse large-cell lymphoma: a phase II study. Eur J Haematol. 2008;80(2):127-132. doi:10.1111/j.1600-0609.2007.00996.x

15. Corazzelli G, Capobianco G, Arcamone M, et al. Long-term results of gemcitabine plus oxaliplatin with and without rituximab as salvage treatment for transplant-ineligible patients with refractory/relapsing B-cell lymphoma. Cancer Chemother Pharmacol. 2009;64(5):907-916. doi:10.1007/s00280-009-0941-9

16. El Gnaoui T, Dupuis J, Belhadj K, et al. Rituximab, gemcitabine, and oxaliplatin: an effective salvage regimen for patient with relapsed or refractory B-cell lymphoma not candidates for high-dose therapy. Ann Oncol. 2007;18(8):1363-1368. doi:10.1093/annonc/mdm133

17. Papageorgiou ES, Tsirigotis P, Dimopoulos M, et al. Combination chemotherapy with gemcitabine and vinorelbine in the treatment of relapsed or refractory diffuse large B-cell lymphoma: a phase-II trial by the Hellenic Cooperative Oncology Group. Eur J Hematol. 2005;75(2):124-129. doi:10.1111/j.1600-0609.2005.00482.x

18. Xiros N, Economopoulos T, Valsami S, Rontogianni D, Fountzilas G, Raptis S. Rituximab in combination with vinorelbine/gemcitabine chemotherapy in patients with primary refractory or early relapsed T cell rich B cell lymphoma. A pilot study. Leuk Res. 2003;27(12):1097-1099. doi:10.1016/s0145-2126(03)00099-7

19. Yun HD, Dossul T, Bernal-Mizrachi L, et al. Referral patterns and clinical outcomes for transplant-eligible lymphoma and myeloma patients evaluated at an urban county hospital. J Stem Cell Res Ther. 2016;6. pii: 328. doi:10.4172/2157-7633.1000328

20. Mitchell JM, Conklin EA. Factors affecting receipt of expensive cancer treatments and mortality: evidence from stem cell transplantation for leukemia and lymphoma. Health Serv Res. 2015;50(1):197-216. doi:10.1111/1475-6773.12208

21. Majhail NS, Nayyar S, Santibañez ME, Murphy EA, Denzen EM. Racial disparities in hematopoietic cell transplantation in the United States. Bone Marrow Transplant. 2012;47(11):1385-1390. doi:10.1038/bmt.2011.214

22. Matasar MJ, Atoria CL, Elkin EB, Nabhan C. Toxicity of maintenance rituximab in older adults with non-Hodgkin lymphoma. Blood. 2015;126(23):881. doi:10.1182/blood.V126.23.881.881

23. van der Poel MW, Oerlemans S, Schouten HC, et al. Quality of life more impaired in younger than in older diffuse large B cell lymphoma survivors compared to a normative population: a study from the population-based PROFILES registry. Ann Hematol. 2014;93(5):811-819. doi:10.1007/s00277-013-1980-1

24. Wade RL, Korytowsky B, Singh P, Dev P, Bobiak S, Cariola P. The novel generation and validation of survival curves in oncology utilizing electronic medical records linked to point of service claims data. Value in Health. 2017;20(9):A421. doi:10.1016/j.jval.2017.08.138

25. Crump M, Baetz T, Couban S, et al. Gemcitabine, dexamethasone, and cisplatin in patients with recurrent or refractory aggressive histology B-cell non-Hodgkin lymphoma: a Phase II study by the National Cancer Institute of Canada Clinical Trials Group (NCIC-CTG). Cancer. 2004;101(8):1835-1842. doi:10.1002/cncr.20587

26. Gopal AK, Press OW, Shustov AR, et al. Efficacy and safety of gemcitabine, carboplatin, dexamethasone, and rituximab in patients with relapsed/refractory lymphoma: a prospective multicenter phase II study by the Puget Sound Oncology Consortium. Leuk Lymphoma. 2010;51(8):1523-1529. doi:10.3109/10428194.2010.491137

27. Velasquez WS, Cabanillas F, Salvador P, et al. Effective salvage therapy for lymphoma with cisplatin in combination with high-dose Ara-C and dexamethasone (DHAP). Blood. 1998;71(1):117-122.

28. Mey UJ, Orlopp KS, Flieger D, et al. Dexamethasone, high-dose cytarabine, and cisplatin in combination with rituximab as salvage treatment for patients with relapsed or refractory aggressive non-Hodgkin’s lymphoma. Cancer Invest. 2006;24(6):593-600. doi:10.1080/07357900600814490

29. Gisselbrecht C, Glass B, Mounier N, et al. Salvage regimens with autologous transplantation for relapsed large B-cell lymphoma in the rituximab era. J Clin Oncol. 2010;28(27):4184-4190. doi:10.1200/JCO.2010.28.1618

30. Velasquez WS, McLaughlin P, Tucker S, et al. ESHAP--an effective chemotherapy regimen in refractory and relapsing lymphoma: a 4-year follow-up study. J Clin Oncol. 1994;12(6):1169-1176. doi:10.1200/JCO.1994.12.6.1169

31. Martin A, Conde E, Arnan M, et al. R-ESHAP as salvage therapy for patients with relapsed or refractory diffuse large B-cell lymphoma: the influence of prior exposure to rituximab on outcome. A GEL/TAMO study. Haematologica. 2008;93(12):1829-1836. doi:10.3324/haematol.13440

33. Zelenetz AD, Hamlin P, Kewalramani T, Yahalom J, Nimer S, Moskowitz CH. Ifosfamide, carboplatin, etoposide (ICE)-based second-line chemotherapy for the management of relapsed and refractory aggressive non-Hodgkin’s lymphoma. Ann Oncol. 2003;14(suppl 1):i5-i10. doi:10.1093/annonc/mdg702

33. Kewalramani T, Zelenetz AD, Nimer SD, et al. Rituximab and ICE as second-line therapy before autologous stem cell transplantation for relapsed or primary refractory diffuse large B-cell lymphoma. Blood. 2004;103(10):3684-3688.

34. Roever L. Endpoints in clinical trials: advantages and limitations. Evid Based Med Pract. 2016;2:1. doi:10.4172/2471-9919.1000e111

35. Chen CC, Parikh K, Abouzaid S, et al. Real-world treatment patterns, time to next treatment, and economic outcomes in relapsed or refractory multiple myeloma patients treated with pomalidomide or carfilzomib. J Manag Care Spec Pharm. 2017;23(2):236-246. doi:10.18553/jmcp.2017.23.2.236

36. Shao C, Liu J, Zhou W, et al. Treatment patterns, health care resource utilization, and costs in patients with relapsed/refractory Hodgkin lymphoma treated with brentuximab vedotin. Leuk Lymphoma. 2019;60(4):947-954. doi:10.1080/10428194.2018.1508665.

37. Center for International Blood and Marrow Transplant. US Department of Health and Human Services, Health Resources and Services Administration. Transplant Activity Report. bloodstemcell.hrsa.gov website. https://bloodstemcell.hrsa.gov/data/donation-and-transplantation-statistics/transplant-activity-report. Updated November 18, 2019. Accessed February 3, 2020. 

38. Wildes TM, Stirewalt DL, Medeiros B, Hurria A. Hematopoietic stem cell transplantation for hematologic malignancies in older adults: geriatric principles in the transplant clinic. J Natl Compr Canc Netw. 2014;12(1):128-136. doi:10.6004/jnccn.2014.0010

39. Zhou X, Ma T, Zhang Y, Zhou N, Li J. Rituximab maintenance therapy for patients with diffuse large B-cell lymphoma: a meta-analysis. PLoS One. 2017;12(3):e0174648. doi:10.1371/journal.pone.0174648

40. Hamlin PA, Satram-Hoang S, Reyes C, Hoang KQ, Guduru SR, Skettino S. Treatment patterns and comparative effectiveness in elderly diffuse large B-cell lymphoma patients: a Surveillance, Epidemiology, and End Results-Medicare analysis. Oncologist. 2014;19(12):1249-1257. doi:10.1634/theoncologist.2014-0113