Biology and Management of Diffuse Large B-cell Lymphoma

Contemporary Oncology®, Spring 2010, Volume 2, Issue 1

Recent statistics show lymphoid malignancies to be the fifth most common cause of cancer-related death in the United States.

Recent statistics show lymphoid malignancies to be the fifth most common cause of cancer-related death in the United States. In 2009, nearly 65,980 new cases of non-Hodgkin lymphoma (NHL) were diagnosed and 19,500 patients died despite currently available treatments.1 Diffuse large B-cell lymphoma (DLBCL), the most common type of NHL in the Western Hemisphere, accounts for 30% to 40% of all NHL cases diagnosed annually in the United States.2

DLBCL typically afflicts patients in their sixth decade of life; primary mediastinal DLBCL variant, however, primarily afflicts women in their late 20s or early 30s. The incidence of DLBCL has increased in the past few decades, and this trend is independent of the human immunodeficiency virus (HIV) epidemic.3

CLASSIFICATION OF DLBCL

Revisions in the pathological classification of lymphomas in the past 30 years or so have resulted in several variations in DLBCL nomenclature. It had been called diffuse histiocytic lymphoma(Rappaport’s classification), centroblastic lymphoma(Kiel’s classification), and large cleaved follicular center cellor large cell immunoblastic lymphoma(National Cancer Institute).4,5

Recently, the Revised European and American Lymphoma classification defined DLBCL as a specific disease. Within this classification, however, it is still heterogeneous, manifesting a variety of morphological features (eg, anaplastic, Burkitt’s-like), protein and gene expression patterns, and clinical presentations (eg, primary mediastinal or central nervous system [CNS] lymphomas). Studies of gene expression profiling (GEP) contributed greatly to our understanding of DLBCL biology. GEP studies identified and validated 3 different subtypes of DLBCL—germinal center B-cell (GCB), activated B-cell (ABC), and primary mediastinal lymphoma (PML)—all with significant differences in terms of prognosis, progression-free survival (PFS), and overall survival (OS).6-8

GCB-DLBCL, apparently derived from germinal center B cells, is driven primarily by deregulation of apoptosis by BCL6 and responds well to rituximab-based immunochemotherapy. ABC-DLBCL is believed to be driven by high levels of nuclear factor kappa-B (NFκB) activity and is associated with worse outcomes despite immunochemotherapy. PML shares GEP signatures similar to those of patients with classical Hodgkin’s lymphomas and is associated with a better prognosis than other DLBCL subtypes. Treatment-related toxicities remain a concern, even without the use of involved-field radiation therapy (IF-XRT), and are being addressed in clinical trials.

PATHOLOGICAL AND CLINICAL CHARACTERISTICS OF DLBCL

Morphologically, DLBCL is characterized by large B cells with a high proliferation index resembling that of germinal centroblasts. DLBCL usually develops de novo, but can emerge as a clonal transformation in patients with low-grade lymphomas or chronic lymphocytic leukemia (CLL). De novo DLBCL tends to respond better to standard therapy and has a better prognosis than transformed DLBCL. Several morphologic variants of DLBCL have been described: centroblastic, immunoblastic, plasmablastic, T-cell/histiocyte-rich, and anaplastic B-cell lymphoma (usually anaplastic lymphoma kinase—positive).9,10 While pathological evaluation can distinguish between these variants, their prognostic significance remains controversial.

Immunophenotype studies demonstrate that DLBCL co-expresses pan B-cell markers, including CD19, CD20, CD79a, CD45RA, and the nuclear transcription factor PAX5. The expression of additional markers may have prognostic implications. The proliferation factor Ki67 is usually high, with a mean percentage of 65. High Ki67 levels (>80%) have been associated with shorter OS.11 The germinal center—associated marker CD10 is expressed in 30% to 40% of cases, and BCL6 is expressed in 60% of cases. Retrospective studies have associated BCL6 expression with prolonged PFS and OS following rituximab-based immunochemotherapy.12,13 Only 10% of DLBCL cases express CD5, which has been associated with shorter survival. The expression of CD5 should raise suspicion of transformation from a more indolent form of NHL, such as small lymphocytic lymphoma (SLL) or CLL.14

Information obtained from genetic studies of DLBCL tumor specimens stresses the complexity of the disease’s biology. DLBCL expresses clonally rearranged IgH genes with somatic mutations in the variable region. Thus, it is thought that DLBCL cells are derived from antigen-exposed B cells. No single gene abnormality is pathognomonic of DLBCL. Recurrent translocations involving BCL6, BCL2, and MYC genes had been described in about 50% of cases. Chromosomal translocation leading to upregulation of BCL2 [t(14;18)] is present in 20% to 30% of DLBCL cases and specially observed GCB variants. Gene abnormalities in ABC-DLBCL are more complex and include trisomies, deletions, and chromosomal inactivation.15,16

GEP studies have established distinct molecular subtypes of DLBCL; ABC-DLBCL, which resembles activated peripheral blood cells; and GCB-DLBCL, which has a similar profile to germinal center cells. BCL6 translocations occur 3 times as often in the GCB subgroup than in ABC-DLBC patients. Activation of NFκB has been recognized as a key feature of ABC-DLBCL pathophysiology, and its downregulation effectively induces cell death in ABC-like tumor cells.6,8 Although GEP may be the best way to differentiate subtypes of DLBCL with prognostic variation, GEP studies need to be validated prospectively, and widespread use of GEP as a routine diagnostic tool is not practical nor recommended outside of a clinical trial. Recently, GEP results have been adapted to a more “user-friendly” approach that depends on immunohistochemistry (IHC). Note that differences in clinical behavior and therapeutic response in newly diagnosed patients with GCB versus those with non-GCB subtypes may not translate to the relapsed/refractory setting.17,18

CLINICAL STAGING AND RISK STRATIFICATION

Clinical staging of patients with DLBCL is fundamental to defining a treatment plan and stratifying risk according to the international prognostic index (IPI) system. Pretreatment evaluation should include obtaining a detailed history, performing a physical examination, complete evaluation of end-organ (cardiac, renal, liver, and hematologic) function, determining lactate dehydrogenase (LDH) levels, and HIV serology when indicated. As rituximab has become integral to the management of DLBCL, rare but serious adverse events such as hepatitis B reactivation have been documented. Fatal cases of fulminate hepatic failure due to reactivation or progression of hepatitis B or C have occurred in DLBCL patients receiving immunochemotherapy; therefore, pretreatment with hepatitis B and C serologies in patients with hepatitis B or C is recommended.

Staging of DLBCL patients requires functional and nonfunctional imaging studies, pathological examination of the bone marrow, and, when indicated, CNS evaluation. Most practicing oncologists routinely conduct computed tomography scanning of the neck, chest, abdomen, and pelvis. Functional imaging with positron emission tomography (PET) has become an important prognostic tool in managing DLBCL, and whole-body PET scanning is used to complement pretreatment staging and post-treatment evaluation of patients. Retrospective studies have demonstrated that PET scanning during and after treatment provides strong prognostic information in terms of PFS and OS. The optimal timing of PET scanning to assess prognosis and response is controversial and currently under study, with recommendations varying among academic institutions.19

Risk stratification is important when managing patients with DLBCL and should take place prior to initiating therapy. The IPI score system20 calculates a sum based on 5 variables: age ≥65; performance status ≥2; elevated LDH level; Ann Arbor stage III or IV; and ≥2 extra-nodal sites of disease. The IPI score determines whether the patient is classified as low, low-intermediate, high-intermediate, or high risk. OS differs for each category, ranging from 23% for the low-risk group to 75% for the high-risk group. The IPI score was validated in multiple clinical trials before and after rituximab was incorporated into frontline therapy for DLBCL.21-23 It has also been validated in relapsed, aggressive NHL.23 Other IPI score variations have similar prognostic capabilities, such as the age-adjusted IPI score for patients aged <65 years and the rituximab-IPI score.20,23-24 While the clinical value of the IPI score is extremely high, it does not provide much insight into disease biology or mechanisms of resistance. This emphasizes the need to identify more biomarkers representative of disease response.

FRONTLINE THERAPY

Unless contraindicated by significant and preexisting comorbid conditions, treatment of DLBCL should include rituximab- and anthracycline-based multiagent chemotherapy and aim to produce durable complete remission (CR). At this point, treatment is tailored to stage or bulk of disease and therapy response. In general, the approach to frontline management of DLBCL depends on whether it is localized or advanced-stage disease.14

Multidisciplinary Management of Early-Stage DLBCL

About 5% of patients with DLBCL present with early-stage disease. Localized DLBCL is defined as Ann Arbor stage I or II (non-bulky) disease and is typically managed with an abbreviated course of combined systemic immunochemotherapy followed by IF-XRT (Table 1).25,26

The Southwest Oncology Group (SWOG) study 8736 randomly assigned 401 patients with non-bulky stage I or II aggressive B-cell lymphoma to receive 3 cycles of cyclophosphamide, vincristine, doxorubicin, or prednisone (CHOP) followed by IF-XRT at 40 to 50 cGy or to 8 cycles of CHOP chemotherapy alone.27 Patients taking CHOP followed by IF-XRT versus CHOP alone had a better OS (77% vs 64%, respectively; P = .03) and PFS (80% vs 72%, respectively; P = .02). An updated analysis after 10 years of follow-up showed an increase in late recurrences, which explained the absence of plateau effect in the survival curve.28 A phase II study by the British Columbia Cancer Agency confirmed the SWOG results. In this study, 308 patients were treated with 3 cycles of chemotherapy followed by IF-XRT. At 5 years, the overall PFS and OS were 80% and 81%, respectively; and at 10 years, PFS and OS were 63% and 74%, respectively.29

Two studies by the Groupe d’Etude des Lymphomes de l’Adulte (GELA) have challenged the role of radiation therapy in the management of early-stage DLBCL. The first study randomized patients aged <60 years to 3 cycles of CHOP chemotherapy followed by IF-XRT or an intense regimen of doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone (ACVBP) for 6 cycles.30 After a median follow-up period of 7.7 years, OS and PFS were significantly better in the ACVBP arm. In contrast to previous clinical trials, this study included patients with bulky disease. The ACVBP regimen was associated with increased chemotherapy-related toxicity and a more hospitalizations. Moreover, vindesine is not commercially available in the United States. In the second study, DLBCL patients older than 60 with localized disease and no other adverse IPI factors were randomized to 3 cycles of CHOP plus IF-XRT or 8 cycles of CHOP alone. After a median follow-up of 7 years, the median OS and PFS were not significantly different between the 2 treatment arms.31

In a randomized clinical trial, the Eastern Cooperative Oncology Group (ECOG) assessed the value of consolidation radiation after 6 cycles of standard chemotherapy. The study enrolled 352 patients aged >60 years with stage I or II bulky-disease and randomly assigned patients that achieved CR to receive CHOP chemotherapy and either consolidation IF-XRT or no further treatment. After 6 years’ follow-up, improvement was observed in estimated disease-free survival, but there was no significant difference in OS.32

Monoclonal Antibodies in Early-Stage DLBCL

The use of monoclonal antibodies, particularly rituximab—a chimeric monoclonal antibody that targets the CD20 antigen present in normal and most malignant B cells—has changed the treatment paradigm for patients with B-cell NHL, including DLBCL. Rituximab elicits antitumor activity and apoptosis through a combination of antibody-dependent cellular cytotoxicity, complement-mediated cytotoxicity, and activation of intracellular pathways. Preclinical models demonstrated that rituximab potentiates the effect of several chemotherapeutic agents.33 In contrast to the activity of rituximab monotherapy observed in low-grade lymphomas, it demonstrates limited activity in DLBCL.34 The addition of rituximab to standard doses of chemotherapy in DLBCL, however, results in improved clinical outcomes without adding significant toxicity.

Few studies address the role of rituximab in managing early-stage DLBCL. The MabThera International Trial (MInT) studied the addition of rituximab to standard therapy in patients aged <60 years with stage I bulky or stage II to IV DLBCL.35 MInT enrolled 824 patients with aggressive lymphoma and a good IPI score (≤1 risk factor). Patients were randomized to receive 6 cycles of CHOP-like chemotherapy with or without rituximab. After a median follow-up of 3 years, patients who received rituximab had significantly improved rates of event-free survival (EFS) and OS. In a subset analysis of patients with early-stage or bulky disease, the addition of IF-XRT as consolidation following immunochemotherapy did not improve clinical endpoints.35

In a phase II clinical trial involving 60 patients with limited-stage DLBCL, SWOG evaluated rituximab plus an abbreviated course of CHOP chemotherapy, followed by IF-XRT.36 Perky and colleagues reported PFS of 88% and OS of 92%; these outcomes were superior to those observed in historical controls receiving chemotherapy and IF-XRT.37,38

Current literature supports using rituximab in combination with standard doses of CHOP chemotherapy in early-stage DLBCL, but the duration of therapy and use of radiation continue to be debated. Many physicians strongly recommend a short course—usually 3 cycles—of systemic immunochemotherapy followed by IF-XRT. Others consider 6 cycles of rituximab combined with CHOP to be clinically equivalent, with the added benefit of sparing the patient the delayed toxicity associated with IF-XRT. To optimize therapy toward a prolonged CR/cure, however, clinicians need to tailor the multidisciplinary approach of early-stage DLBCL according to the site of disease involvement and preexisting patient comorbidities.39

Approach to Advanced-Stage DLBCL

The use of systemic chemotherapy to eradicate disseminated DLBCL was first described in the early 1970s.40,41 Following these original reports, CHOP given every 21 days (CHOP-21) became the standard of care for aggressive lymphomas in the United States. The desire to improve survival for patients with DLBCL led investigators to evaluate more intensive regimens. Fisher and colleagues, on behalf of SWOG, reported the results of a large US clinical trial that compared CHOP-21 with 3 unique, intense chemotherapy regimens. These intense regimens failed to demonstrate superiority over CHOP-21 but were associated with greater toxicity. After 6 years of follow-up, OS and PFS did not differ significantly among the various investigational regimens.42

CHOP Versus ACVBP. In the past decade, 2 alternative chemotherapy regimens have been evaluated against standard CHOP in patients with advanced-stage DLBCL (Table 2). GELA compared the efficacy of ACVBP against CHOP-21 in 635 patients aged >60 years that had advanced DLBCL and a least 1 IPI score risk factor.37 While the CR rate was similar in both groups (58% vs 56%), EFS and OS were significantly better in patients treated with ACVBP. At 5 years’ follow-up, EFS was 39% in the ACVBP groups and 29% in the CHOP-21 group (P = .014); OS was significantly longer for patients treated with ACVBP than with CHOP-21 (46% vs 38%, respectively; P = .036). The CHOP-21 group had a higher incidence of CNS progression or relapse (P = .004).

Dose-Dense CHOP and Etoposide. The German Study Lymphoma Group (GLSG) evaluated the concept of dose-dense therapy in DLBCL. Elderly patients were randomized into 4 cohorts: CHOP-21 with etoposide (CHOEP-21), CHOP-21 without etoposide, dose-dense CHOP given every 14 days (CHOP-14) with etoposide (CHOEP-14), or dose-dense CHOP-14 without etoposide.43 The addition of etoposide to CHOP chemotherapy did not affect the clinical endpoints of the study. Dose-dense CHOP-14 had a similar toxicity profile compared with CHOP-21 and was associated with improved rates of PFS (53.3% vs 43.8%, respectively;

P= .003) and OS (40.6% vs 32.5%, respectively; P<.001). The NHL-B1 study subsequently evaluated these regimens in young patients with advanced DLBCL and confirmed the value of dose-dense CHOP. In younger patients, CHOEP-14 and CHOEP-21 were associated with better PFS and OS compared with CHOP-21. Despite an increased degree of myelosuppression, these regimens were well tolerated.44

Rituximab Plus Standard CHOP. Although CHOEP-14 and ACVBP demonstrated clinical superiority to CHOP-21, the parallel development of rituximab and its incorporation into standard therapy with CHOP-21 has challenged the clinical value of using such toxic regimens (Table 2). Following results from clinical trials evaluating rituximab as monotherapy in aggressive lymphomas and the data reported by Czuczman and colleagues in patients with follicular lymphomas, Vose and colleagues conducted a phase II study to evaluate the combination of rituximab and CHOP (R-CHOP) in aggressive B-cell lymphomas.45,46 The overall response rate (ORR) was 94%, with 20 of 33 patients achieving CR. Updated data after a median follow-up of 63 months demonstrated an 82% rate of PFS and an 88% rate of OS .46

GELA conducted a landmark phase III study validating the addition of rituximab to CHOP in patients aged >60 years with newly diagnosed aggressive B-cell lymphomas. Patients were randomized to receive 8 cycles of CHOP alone or rituximab plus CHOP-21.47 The study included patients with stage I to IV DLBCL; 59% had 3 or more IPI score risk factors, and 80% had Ann Arbor stage III or IV disease. Patients that received R-CHOP had higher response rates than those treated with CHOP alone (83% vs 69%, respectively; P = .005). An interim analysis reported that, after 18-month follow-up PFS and OS were significantly better in the R-CHOP arm (P <.001 and P = .007, respectively). Analysis of the study after 5 years of follow-up confirmed the efficacy of combining rituximab with systemic chemotherapy. Compared with CHOP alone, patients receiving R-CHOP demonstrated higher rates of PFS (30% vs 54%, respectively; P<.0001) and OS (45% vs 64.5%, respectively; P = .0004).47,48

The ECOG 4944 study, conducted primarily in the United States, aimed to validate findings from the GELA study. Investigators randomized previously untreated elderly patients with DLBCL to R-CHOP or CHOP and then randomized responders to observation or maintenance therapy with rituximab. Data showed that adding rituximab to chemotherapy during induction or maintenance improved time to progression compared with CHOP alone.49

Pfreundschuh and colleagues reported that adding rituximab to CHOP or a CHOP-like regimen resulted in statistically significant improvements in PFS and OS (MInT) compared with chemotherapy alone. The randomized study included 824 patients with DLBCL. At 3 years’ follow-up, patients that received the combination had a higher rate of EFS compared with chemotherapy alone (79% vs 59%, respectively; P <.0001) and improved OS (93% vs 84%, respectively; P = .0001).35

The addition of rituximab to CHOP chemotherapy has not been formally studied in high-intermediate and high-risk DLBCL patients aged <60 years. It is used in this age demographic based on data from the GELA and MInT studies. While the addition of rituximab to standard doses of CHOP chemotherapy has improved outcomes for patients with DLBCL, many fail to respond or relapse, which highlights the need to develop novel therapeutic strategies.

Rituximab Plus Dose-Dense CHOP. In the RICOVER-60 study, the German High-Grade NHL Study Group (DSHNHL) evaluated the effects of adding rituximab to CHOP-14 (R-CHOP-14) in patients aged 61 to 80 years with DLBCL.50 This study randomized 1222 elderly patients to receive either 6 or 8 cycles of CHOP-14 or R-CHOP-14. IF-XRT was planned for extra-nodal or bulky sites.The results demonstrated that the addition of rituximab to CHOP-14 improved EFS and OS in this patient population and that 6 cycles of chemotherapy were as effective as 8. After a 3-year follow-up period, EFS was 47.2% and OS was 67% in the group that received 6 cycles of CHOP-14 compared with EFS of 66.5% and OS of 78.1% for those treated with 6 cycles of R-CHOP 14. EFS was 53.0% and OS was 66% after 8 cycles of CHOP-14 compared with EFS of 63.1% and OS of 72.5% after 8 cycles of R-CHOP-14.50

Whether R-CHOP-14 is superior to R-CHOP-21 is debatable. A quick glance at data from the GELA and RICOVER-60 trials suggests that R-CHOP-14 yields better PFS and OS at 3 years’ follow-up. Patient and disease characteristics differed between these studies, however. Compared with the RICOVER-60 trial, the GELA study included more patients with DLBCL (78% vs 84%, respectively), stage I/II disease (55% vs 20%), and high-intermediate/high-risk IPI scores (39% vs 59%). In the randomized LNH03-6B study, the GELA group recently compared R-CHOP-21 to R-CHOP-14 in patients with DLBCL. Preliminary results suggest that the two regimens are equivalent, calling into question the role of dose-dense therapy in managing DLBCL.51

Rituximab with Other Chemotherapy Regimens. Additional clinical trials have explored the use of rituximab with other chemotherapy regimens. Wilson and colleagues from the National Cancer Institute studied dose-adjusted etoposide, vincristine, cyclophosphamide, doxorubicin, and prednisone (DA-EPOCH) combined with rituximab in previously untreated patients with DLBCL. In this particular regimen, the chemotherapy doses are adjusted with each subsequent cycle in an attempt to achieve an absolute neutrophil count nadir of 500 cells/mL. The study enrolled 72 consecutive patients (aged ≥18 years) with untreated stage II or higher DLBCL. Patients received 6 to 8 cycles of rituximab and DA-EPOCH; IF-XRT was not permitted. Correlative studies were performed to address biomarkers of disease response by IHC. At 5 years, PFS and OS were 79% and 80%, respectively.52,53 An ongoing randomized phase III study from the Cancer and Leukemia Group B is comparing R-CHOP-21 to rituximab in combination with DA-EPOCH in previously untreated DLBCL. Correlative studies will prospectively validate GEP studies versus outcomes in various subsets of DLBCL.

Four alternative strategies have been or are being evaluated in clinical studies, with some showing promising activity in DLBCL patients: (1) the concept of dose-dense rituximab (12 doses of rituximab delivered concomitantly with 6 cycles of CHOP-14); (2) the targeting of the ubiquitin-proteasome system with bortezomib as a means of potentiating the antitumor activity of immunochemotherapy; (3) rituximab maintenance following autologous bone marrow transplant (CORAL study); and (4) the use of high-dose chemotherapy (HDC) with autologous stem cell support (ASCS) in first remission, especially for patients with high-risk disease.54-60

CNS PROPHYLAXIS IN DLBCL

CNS relapse is a rare but significant complication in patients with DLBCL, with an estimated 2.3% to 4.5% probability of occurring after 1 year from diagnosis.61,62 Subsets of DLBCL have a 4- to 15-fold greater risk of CNS relapse, and it is imperative to identify such patients and implement prophylactic therapy (ie, intrathecal chemotherapy). Boehme and colleagues retrospectively evaluated the incidence of and risk factors associated with CNS recurrence in 1693 patients with aggressive lymphomas. A multivariate analysis identified an elevated LDH at diagnosis and lymphoma involvement of ≥2 extra-nodal sites as the only prognostic factors for CNS relapse.61

M.D. Anderson Cancer Center conducted a similar study and identified additional prognostic factors for CNS relapse: >1 extra-nodal site of disease, age >60 years, albumin levels <3.5 gr/dL, retroperitoneal involvement, and LDH >450 IU/mL.62 Similar findings have been observed in patients receiving immunochemotherapy. Boehme and colleagues conducted a retrospective study of factors associated with CNS relapse in patients treated with R-CHOP-14 and estimated the rate of CNS relapse at 2 years to be 6.9%. Having >1 extra-nodal site or B-symptoms were prognostic for CNS relapse.63 Routine intrathecal chemotherapy is strongly recommended in patients with DLBCL that have the following characteristics: ≥2 extra-nodal sites; testicular or breast involvement (any stage); lymphoblastic variants; oropharyngeal or para-spinal sites of involvement; bone marrow involvement; or concomitant HIV infection.

CHALLENGES IN R-CHOP RELAPSED/REFRACTORY DLBCL

The international phase III PARMA study confirmed the role of HDC-ASCS in treating relapsed/refractory DLBCL.23,64 Patients with relapsed/refractory DLBCL underwent salvage chemotherapy for 2 cycles; patients with chemotherapy-sensitive DLBCL were randomized to further salvage chemotherapy with cytarabine/platinum-based chemotherapy alone or with ASCS. At 5 years, EFS in the transplant arm was 46% and OS was 53% compared with 12% and 32%, respectively, in the chemotherapy-alone arm.23,64 Based on these results, salvage chemotherapy followed by HDC-ASCS has been adopted as the standard of care for transplant-eligible DLBCL patients.

Patients with relapsed/refractory DLBCL are treated with a vast number of regimens based on non—cross-resistant chemotherapy agents as opposed to chemotherapy in the frontline setting. The goal of salvage regimens is to achieve maximum tumor burden cytoreduction in preparation for HDC-ASCS. To secure adequate peripheral blood stem cell collection, we generally choose among salvage regimens with high rates of response (especially CR), low hematologic or non-hematologic toxicity, and no association with significant stem cell damage.

Rituximab has altered the treatment paradigm in DLBCL, and it has been postulated that the subset of patients with relapse/refractory disease following a rituximab-based regimen differs from the one studied in pre-rituximab clinical trials. Several investigators are evaluating response to salvage chemotherapy and HDC-ASCS in relapsed/refractory DLBCL patients previously treated with upfront R-CHOP.

The Spanish Cooperative Group for Bone Marrow Transplants in Lymphomas reported on a retrospective analysis of data for patients with DLBCL who received rituximab combined with etoposide, methylprednisolone, cytarabine, and cisplatin (R-ESHAP) as salvage therapy following frontline therapy with rituximab chemotherapy or chemotherapy alone to assess the influence of initial therapy with rituximab on outcomes.65 Martin and colleagues studied 163 consecutive patients with relapsed/refractory DLBCL who received R-ESHAP as second-line therapy; 94 had received rituximab chemotherapy in the frontline setting and 69 had been treated with chemotherapy alone. Following R-ESHAP salvage therapy, ORR was 67% for those previously treated with rituximab compared with 81% for those treated only with chemotherapy (P= .045). CR was 37% for the frontline rituximab arm compared with 56% for the chemotherapy-only group (P= .015). At 3 years, PFS was significantly higher for the patients who received only chemotherapy compared with those treated with rituximab up front (57% vs 38%, respectively; P<.0001). OS at 3 years was also significantly improved for the chemotherapy-only arm (64% vs 17%, respectively; P= .0005). The same percentage of patients in both groups underwent subsequent HDC-ASCS. A multivariate analysis found prior exposure to rituximab to predict worse PFS and OS.65 These findings suggest that DLBCL patients who relapse or fail to respond to upfront rituximab-based immunochemotherapy have more resistant disease, and this is a challenge for clinicians treating aggressive B-cell lymphoma. It also illustrates the need to define biomarkers that correspond to the mechanisms of resistance to immunochemotherapy.

CONCLUSIONS

The incorporation of rituximab into therapy with standard doses of CHOP has resulted in improved clinical outcomes when compared with standard chemotherapy alone in patients with DLBCL. It has raised expectations that developing innovative induction and salvage therapies which logically incorporate new target-specific agents will improve outcomes further. To optimize the treatment of patients with DLBCL, the time has come to reevaluate previously accepted biomarkers of responsiveness (eg, Bcl-2 expression, international prognostic index, GEP) and to test and validate novel biomarkers that correlate with response to a specific agent or regimen.