Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of NHL both in the United States and globally, representing roughly 30% of all new diagnosis. With a median age at diagnosis of 70 years, DLBCL is in many ways a disease of older individuals, a fact which will be magnified over the next decades as the population sees doubling and tripling of those older than 75 and older than 85 years, respectively. These projections indicate that over the same time, the number of NHL cases will increase by 67%. Hence, there is a pressing need to understand age-related host and disease factors, how they impact therapeutic choices and outcomes, and to develop specific strategies to address these factors.
Based on randomized controlled trials (RCTs), frontline therapy with anthracycline-based chemoimmunotherapy for DLBCL has resulted in overall survival rates of 60% to 70% and represents the best chance for cure, as the majority of relapsed or refractory patients will succumb to the disease. Without therapy, DLBCL is rapidly fatal in a matter of weeks. However, older patients frequently have comorbidities and functional impairment limiting feasibility of standard therapy. Hence, there are significant challenges facing the oncologist who must integrate disease, host, and patient factors with the goal of individualizing treatment decisions for the older DLBCL patient. In this context, oncogeriatric tools are essential to the initial assessment of these patients and augment the lymphoma-specific care plan.
From a practical standpoint, the initial assessment of the older patient with DLBCL is evolving to include parallel pretreatment disease staging, biologic and genetic risk assessment, and oncogeriatric risk assessment to inform decision making. Ideally, these assessments inform the oncologist and patient regarding the disease risk in the context of patient life expectancy, the likelihood of cure or disease control balanced against the likelihood of treatment related morbidity and mortality, and provide recommendations for treatment modification to mitigate these effects.
Biology: Effective Age, Germinal Center Versus Non-GC, and Methodology for Determining Cell of Origin
Chronologic age has factored prominently into the DLBCL landscape, as age is both associated with increased incidence of disease as well as poorer prognosis. Historically, a cutoff of 60 years was utilized in the IPI risk model, as high-dose therapy and stem cell rescue were generally limited above this age category. However, when clinical prognostic models have been revisited in the era of chemoimmunotherapy (eg, RCHOP), this dichotomous cut point has often been revised to 70 to 72 years as a prognostic component. While binary cut points are useful in clinical models, it is recognized that chronologic age is a continuous variable with vast heterogeneity regarding biologic implications for any individual. Consequently, functional definitions of age are preferable and reflect individual physiology.
Biology in older patients
It is readily acknowledged that DLBCL is a heterogeneous disease as evidenced by both varied clinical outcomes and increased understanding of the spectrum of underlying biology. Based on the WHO classification (WHO 2008), the majority of diagnoses in older patients will be classified as DLBCL, not otherwise specified (NOS); potentially adverse disease subsets such as T-cell–rich B-cell lymphoma, immunoblastic morphology, and EBV-associated DLBCL occur in greater frequency in older adults. Although associated with worse outcomes, specific treatment recommendations are lacking for these entities.
DLBCL expressing both cMyc t(8;14) and BCL2 t(14;18) by FISH are termed “double hit” lymphomas and have been associated with very poor prognosis. Double-hit lymphomas diagnosed by FISH may be present in 2% to 11% of new diagnoses. IHC can also identify DLBCL with dual expression of MYC and BCL2, coined double expressing DLBCL, which is also associated with inferior outcomes and may be present in one-third of patients. Optimal treatment is still undefined and increased treatment intensity is not a plausible strategy for most older patients.
Over the last decade, significant advances in molecular genetics have deepened our understanding of DLBCL biology. GEP of DLBCL, NOS has identified at least 2 distinct molecular subsets, namely a germinal center B-cell (GC) and an activated B-cell (ABC) phenotype, both reflecting the putative cell of origin biology. The ABC phenotype, with increased nuclear factor-κB (NF-κB) and chronic BCR receptor activation, is associated with worse outcomes under current standard of care therapy. Furthermore, both genetic complexity and the ABC phenotype increases with increasing age and account in part for worse outcomes independent of clinical predictors. The reality of DLBCL molecular changes, however, extends beyond merely assessing cell of origin, with significant genetic alterations identified in the microenvironment (stromal signals) as well as immune surveillance and antigen presentation aberrations, which are equally important.
In 2015, although increasingly cost-efficient, GEP remained impractical in daily practice and did not, as of yet, define therapy. As a result, surrogate IHC models of COO (eg, Hans, Tilly) have been implemented given ready availability and quick turnaround. Although prognostic value is maintained, the IHC models are fraught with misclassification and poorer reproducibility. Approaches using rapid assessment of predictive biology in paraffin-embedded samples, such as the nanostring predictors Lymph2Cx, are being validated currently and are promising technologies to anticipate in the near future. These discoveries have given us insight into the essential oncogenic pathways involved in the heterogeneity of DLBCL and offer the potential for targeted therapeutic agents.
DLBCL is an aggressive entity with the natural history characterized by rapid progression and death in the absence of treatment. Consequently, chemotherapy is frequently used to extend survival, even among patients were curative intent cannot be achieved. Anthracycline-based chemotherapy has been associated with the greatest cure rates, with CHOP chemotherapy forming the backbone for modern-day therapy. The initial decision regarding the feasibility of anthracycline-based chemotherapy is consequently paramount. This decision must integrate assessments of organ function, clinical judgment, comorbidity, and performance status as previously discussed.
Early-Stage Treatment Paradigms
Therapeutic options for early-stage DLBCL (stage I/II without bulk) are distinct from advanced-stage disease with some overlap. Patients with age as their only risk factor have very good prognosis with combined modality therapy. Treatment approaches were initially defined by the SWOG 8736 which established CHOP × 3 cycles followed by IFRT as superior to CHOP × 8 cycles for PFS and OS at 5 years (PFS 77% vs 64%, OS 82% vs 72%); this difference, however, disappears after 8 years, a fact that should be considered in the context of life expectancy. The addition of rituximab to CHOP × 3 + IFRT was explored in SWOG 0014 in patients with one or more risk factors, with PFS of 88% at 5-year median follow-up. It is worth noting data from the GELA/LYSA group question the benefit of IFRT in older patients following four cycles of CHOP (LNH 93-4), and more recently following four cycles of RCHOP when in CR. Patients with bulky (> 10 cm) disease generally are treated with full course RCHOP ± IFRT for local control.
Whether or not to include radiation consolidation may be informed by disease location, associated RT toxicity, comorbidity, and life expectancy. Radiation should be considered for definitive therapy in patients who cannot tolerate chemotherapy.
Advanced-Stage Treatment Paradigms
The standard of care for initial treatment of DLBCL in both older and younger patients is chemoimmunotherapy comprised of rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (RCHOP 21) given every 21 days for six to eight cycles. The standard of care is based on the LNH 98–5 study, which included patients age 60–80 (median age of 69); 55% of patients were 70 years or older. This study has mature data, with 5-year progression-free and overall survival significantly longer for the rituximab-containing arms (54% and 58% RCHOP compared to 30% and 45% CHOP, respectively). Dedicated RCTs designed to improve upon RCHOP21 have explored upfront consolidative transplantation, dose-dense q14 day regimens, and maintenance rituximab without improvement in survival.
Ongoing efforts are exploring alternative regimens such as the dose-adjusted R-EPOCH regimen and cell-of-origin–directed trials. The DA-R-EPOCH regimen is an infusional chemotherapy regimen that may have improved efficacy in DLBCL, particularly in GC biology, and high proliferative rate tumors (eg, Myc +, high Ki67), and is potentially less cardiotoxic owing to infusional doxorubicin. Phase II data from the CALGB was associated with 5-year TTP and OS of 81% and 84% in all patients. A CALGB RCT comparing RCHOP21 versus DA-R-EPOCH was completed but not reported as of 2015. This regimen can be safely considered in older patients and may have an advantage in higher-risk scenarios (eg, double-hit biology). Cell-of-origin directed approaches are being explored in clinical trials and have taken into consideration differential activity of certain agents in the nongerminal center (ABC) phenotype. Agents such as bortezomib, ibrutinib, and lenalidomide all have phase I and II data supporting combination with RCHOP, with phase III studies ongoing globally.
Radiation consolidation to sites of bulk or residual masses may be beneficial in selected cases. IFRT has been associated with increased local control and PFS, but not OS. However, in older patients with competing nonlymphoma comorbidities for life expectancy and a lack of effective second-line options, this initial improvement in disease control may be worthwhile. Careful discussion of these pros and cons is often needed.
Patients who relapse following RCHOP chemotherapy have a poor prognosis, even when second-line therapy and subsequent high-dose therapy and autologous stem cell rescue can be utilized. For the vast majority of older individuals, this intensive strategy is not feasible due to age-related comorbidity and increased TRM associated with transplant approaches. In this setting, goals of care shift to disease control and palliation.
Prephase concept and reduced intensity
Often at initial presentation, disease-related factors impair host performance status and are associated with treatment-associated morbidity and mortality. The German NHL high-grade study group has utilized a concept termed “prephase” to mitigate the impact of decreased functional status initially. Employing a week of corticosteroid and a single dose of vincristine before the initiation of therapy, the RICOVER60 study suggested there was a 50% reduction in cycle 1 and 2 TRM as a consequence of this maneuver. While dedicated validation of this approach is needed, other prospective studies have suggested there is clinical benefit to such a concept; clinicians frequently employ steroids to improve functional status in anticipation of starting definitive chemotherapy.
Reduced or nonanthracycline-based curative intent
In clinical scenarios where RCHOP is not deemed feasible, there exist a variety of dose-reduced regimens, with relative dose intensities of 50% to 70% standard dosing, and a number nonanthracycline regimens that maintain curative potential. In general, the trade-off includes a reduction in efficacy with the advantage of reduced toxicity and TRM. Direct comparative data are frequently lacking in this area, with data more frequently culled from retrospective cohorts or phase II trials (Table 105-5).
TABLE 105-5REDUCED DOSE OR NONANTHRACYCLINE THERAPY |Favorite Table|Download (.pdf) TABLE 105-5 REDUCED DOSE OR NONANTHRACYCLINE THERAPY
|REGIMEN ||N ||PLANNED RDI ||AGE MEDIAN (RANGE) ||ORR (CR/PR) ||EFS ||OS ||TRM |
|RCHOP21 (Phase III) ||N = 202 ||100% ||69 y (60–80) ||83% (75%/7%) ||57% @ 2 y ||70% @ 2 y ||6% |
|RCHOP21 (retrospective) ||N = 61 ||70% ||76 y ||87% (79%/8%) ||57% @ 2 y ||68% @ 3 y ||NR |
|R-mini-CHOP (Phase II) ||N = 149 ||~50% ||83 y (80–95) ||74% (63%/11%) ||47% @ 2 y (PFS) ||59% @ 2 y ||8% |
|DRCOP (Phase II) ||N = 80 ||NA ||69 y (61–92) ||86% (75%/11%) ||60% @ 3 y ||74% @ 3 y ||5% |
|Nonanthracycline regimens |
|R-miniCEOP (Phase III) ||N = 114 ||100% ||73 y (64–84) ||81% (68%/13%) ||54% @ 2 y ||~74%@ 2 y ||6% |
|R-GCVP (EF ≤ 55%) (Phase II) ||N = 61 ||NA ||76 y (52–90) ||61% (39%/23%) ||50% @ 2 y (PFS) ||56% @ 2 y ||NR |
|R-Bendamustine (Phase II) ||N = 14 ||NA ||85 y (80–95) ||69% (54%/15%) ||40% @ 2 y (PFS) ||40% @ 2 y ||0% |
|CR, complete response; EFS, event free survival; NR, not reported; ORR, overall response rate; OS, overall survival; PFS, progression free survival; PR, partial response; RDI, reduced dose intensity; TRM, treatment related mortality. |
DLBCL patients characterized as very old adults, for example, over the age of 80 years, or with associated geriatric syndrome and/or frailty are under represented in clinical trials with a lack of evidenced-based guidelines. Typically, full-dose chemotherapy is not approachable, but often curative regimens can still be considered with modifications designed to mitigate serious toxicity. A phase II GELA study specifically in patients older than 80 years (n = 149) investigated a reduced intensity R-mini-CHOP protocol with 2-year PFS and OS of 47% and 59%, respectively, and maintained at 4 years. A variety of other nonanthracycline options also exist, such as CEPP, CDOP, CNOP, CEOP, and GCVP, all given with rituximab.
Response Assessment and Follow-Up
Interim restaging is performed to identify scenarios where disease is not responding or progressing despite therapy. This confirmation is particularly important in older adults where toxicity may be significant. Interim PET scans have been demonstrated to retain excellent negative predictive value, but questionable positive predictive value that has not allowed changes in therapy; consequently, interim PET scans are not recommended outside clinical trials. PET scans at the end of therapy, however, are now standard and remain very predictive of outcome.
Surveillance imaging as a routine post therapy has been called into question, with a lack of evidence supporting improved outcomes with screening CT scans and concern over false positives and exposure to medical radiation. For advanced-stage patients, CT scans every 6 months for 2 years are recommended currently, balancing the increased risk of relapse during this time interval with a desire to limit exposure.
High-Dose Therapy and Autologous Stem Cell Rescue in Older Patients
High-dose chemotherapy followed by autologous stem cell rescue (HDT/ASCR) was established in the PARMA study as the recommended second-line approach in chemosensitive transplant-eligible patients. In the absence of HDT/ASCR, relapsed DLBCL therapy is generally not curative. Unfortunately, there are limited data reporting the safety and efficacy of HDT/ASCR in the older patient. The Center for International Blood and Marrow Transplant Research (CIMBTR) has reported a 28% increase in the number of HDT/ASCR procedures in the United States when comparing 1994–1995 to 2004–2005. Commensurate with this increase, the proportion of patients 60 years or older undergoing HDT/ASCR during the same period went from less than 7% to 35%.
Given the significant potential for morbidity and mortality with HDT/ASCR, patient selection in the older patient is paramount. Factors considered include age, comorbidities, functional status, and psychosocial support systems. The second-line aaIPI (PS, LDH, stage) has been confirmed in a number of trials as predictive of PFS and OS with an intent to transplant; 3-year estimated event-free survival (EFS) for low/intermediate-risk versus high-risk patients was 40% versus 18% (p < .001). A Hematopoietic Cell Transplantation Comorbidity Index (HCT-CI) also exists and high-risk scores are associated with decreased likelihood of receiving transplant and OS. Despite age-related physiologic changes, in general age is not a limiting factor when considering the ability to mobilize and collect stem cells, although complete count recovery may be slower post transplant in older individuals. TRM in NHL patients older than 55 years in the CIBMTR database was 15% and increased 1.86 times compared to their younger counterparts. This may be as high as 18% to 19% for individuals older than 70 years at transplant.
The largest randomized study of second-line therapy followed by HDT/ASCR (Collaborative Trial in Relapsed Aggressive Lymphoma [CORAL] study) included patients up to age 65, with a median age of 55. In this patient population, the intent to treat (ITT) EFS was 31% for all patients, with identified prognostic factors of relapse less than 12 months from prior therapy, prior rituximab, and the second-line age-adjusted IPI. For patients with early relapse following prior rituximab-based therapy, the ITT EFS was only 23%. Given these modest results in younger patients, one must be selective in recommending HDT/ASCR in the older individual. Early referral to a transplant center for evaluation, cardiac and pulmonary assessment, second-line aaIPI, and CGA should all likely precede initiation of intensive second-line regimens such as R-ICE or R-DHAP when transplant is the goal.