Age-dependent increase of treatment-related mortality in older patients with aggressive B cell lymphoma: analysis of outcome, treatment feasibility, and toxicity in 1171 elderly patients with aggressive B cell lymphoma—data from phase II and III trials of the DSHNHL (German High-Grade Non-Hodgkin’s Lymphoma Study Group)

Aggressive B cell lymphomas (aNHL) are a heterogeneous group of aggressive B cell neoplasms including diffuse large B cell lymphoma (DLBCL), the most frequent lymphoid neoplasia accounting for one-third of all malignant lymphomas. R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone) is the current treatment standard, leading to high cure rates especially in younger patients [1]. The incidence of aNHL, however, increases with age, and almost 50% of patients are older than 60 years [2]. In older patients, prognosis is almost identical to that of younger individuals if patients can be treated with full-dose immunochemotherapy, as demonstrated in several large phase III trials [2, 3]. However, administration of fully dosed chemotherapy is limited by the age-dependent increase of comorbidities and higher treatment-related mortality (TRM) [4]. Dose reductions are frequent and correlate with an inferior prognosis [5]. Supportive measures such as growth factor support and anti-infective prophylaxis are recommended by the American Society of Clinical Oncology (ASCO) [6] in order to facilitate treatment at full dose and reduce toxicity. Here we analyzed the feasibility and toxicity of the R-CHOP-14 regimen within consecutive prospective DSHNHL phase II and III trials (RICOVER-60, DENSE-R-, SMARTE-R-, and SEXIE-R-CHOP-14) for patients between 60 and 80 years of age. We investigated the impact of age on the outcome, feasibility, and toxicity of lymphoma therapy by dividing the patient population into four age groups (61–65 years, 66–70 years, 71–75 years, and 76–80 years). Furthermore, we analyzed whether the implementation of a stringent anti-infective prophylaxis as implemented in the most recent trial populations is able to reduce the treatment-related mortality (TRM).

In 1144 of 1171 patients (98%), a reference pathological diagnosis was available, and 946 of 1144 patients were diagnosed with DLBCL (83%). The clinical characteristics and IPI risk factors of the four age groups are shown in Tables 1 and 2 and the supplement Table 1. Subdividing the two study cohorts RICOVER-60 and CHOP-R-ESC into the four age groups within RICOVER-60 trial (6/8-CHOP-14 + 8xR) revealed significantly increased numbers of female patients (p < 0.001), more elevated LDH values (p = 0.003), and worse ECOG > 1 performance scores (p = 0.001) in elderly age groups (supplement Table 1) and within CHOP-R-ESC trials significantly more female patients (p = 0.004) and a trend to worse ECOG > 1 performance scores (p = 0.097) for the patients from elderly age groups (Table 2). Dose reductions were most frequent in the age group 76–80 years, however, to a smaller extent in the CHOP-R-ESC trial (Fig. 1 and Fig. 2); up to 70% of these pts from RICOVER-60 trial did not receive the complete chemotherapy as planned (Table 4). Hematotoxicity, infections, and TRM increased with age. TRM was significantly higher (p = 0.029) for the age group 76–80 years (amounting to 11% in the CHOP-R-ESC trials) as compared to 4% for the age group 71–75 years (Tables 3 and 4). In comparison, in the RICOVER-60 trial, TRM for 6-CHOP-14 + 8xR was 20% and 8%, respectively (p = 0.081). Higher TRM (but not higher mortality due to lymphoma) resulted in significant inferior EFS, PFS, and OS in the age group 76–80 years as compared to patients aged 71–75 years in both trials (RICOVER-60 3 years EFS, 44% (95% CI: 28–59) versus 68% (95% CI: 56–78) p < 0.001; PFS, 49% (95% CI: 33–64) versus 78% (95% CI: 68–88) p < 0.001; OS, 51% (95% CI: 36–67) versus 80% (95% CI: 70–89) p < 0.001 (Table 3, Fig. 3); R-CHOP-ESC 3-years EFS, 53% (95% CI: 42–64) versus 71% (95% CI: 63–78) p = 0.011; PFS, 57% (95% CI: 46–68) versus 74% (95% CI: 67–81) p = 0.011; OS, 61% (95% CI: 50–72) versus 77% (95% CI: 70–84) p = 0.045) (Table 3, Fig. 4). Multivariate analysis of the four age groups adjusted for IPI-factors and gender showed similar results with significant hazard ratios (HR) only between age groups 76–80 years and 71–75 years (RICOVER-60: EFS, HR = 1.9 (95% CI: 1.1–3.1) p = 0.020; PFS, HR = 2.3 (95% CI: 1.3–4.1) p = 0.004; OS, HR = 2.2 (95% CI: 1.3–4.0) p = 0.006; CHOP-R-ESC: EFS, HR = 1.7 (95% CI: 1.1–2.7) p = 0.013; PFS, HR = 1.8 (95% CI: 1.1–2.8) p = 0.011; OS, HR = 1.6 (95% CI: 1.0–2.5) p = 0.051).

Because of increased infectious complications seen in the first 20 patients included in the DENSE-R-trial, the DSHNHL introduced a stringent anti-infective prophylaxis for all subsequent patients treated in all CHOP-R-ESC trials. Adherence to this schedule was almost complete (61–65 years, 93%; 66–70 years, 92%; 71–75 years. 89%; and 76–80 years, 91%) compared to the much lower percentage of patients receiving prophylaxis in the preceding RICOVER-60 trial (6xCHOP-14 + 8xR: 33%, 36%, 33%, 41%) resulting in higher numbers of patients being treated with interventional antibiotics (49%, 52%, 59%, 53%) compared to a much lower number of patients in the R-CHOP-ESC trials (26%, 31%, 36%, and 48%). The rate of grades 3 and 4 infections per patient was also different in these two patient cohorts with significantly lower numbers in the R-CHOP-ESC trials (19%) in comparison to the RICOVER-60 trial for 6xCHOP-14 + 8xR (28%; p = 0.005) (data within age groups: RICOVER-60 6xCHOP-14 + 8xR, 24%, 20%, 33%, 44%; R-CHOP-ESC, 10%, 19%, 24%, 28%). The TRM was significant lower for the R-CHOP-ESC trials in comparison to the whole RICOVER-60 cohort (4% vs. 7%, p = 0.038). The TRM was reduced by almost 50% in the more recent R-CHOP-ESC trials compared to the original RICOVER-60 trial (for details see supplement Table 2).

Choosing the best treatment for elderly patients with aggressive lymphoma is a clinical challenge. The number of patients not treated at all still increases with age [13]. More frequently, the age-dependent increase of relevant comorbidities [14] leads to a substantial increase in patients receiving less than full-dose immunochemotherapy. On the other hand, several trials incorporating R-CHOP have shown that many elderly patients can be cured, depending, among others, on the dose and dose intensity of chemotherapeutic agents actually being administered [2, 3]. If the relative dose intensity decreases below 70%, the relapse rate increases dramatically [5, 15]. As treatment-related toxicity dramatically increases with age, measures to decrease treatment-related morbidity and mortality in elderly patients are urgently needed.

Our analysis from a series of prospective multicenter trials is based on the R-CHOP-14 regimen administered at full doses. It shows that beyond the age of 75 years, there is a striking rise in TRM. One might speculate that the usage of CHOP-14 (instead of CHOP-21) regimen is responsible for this increase. However, there are currently no data available supporting this hypothesis. On the contrary, Cunningham et al. included patients in the same age group (56% of pts. > 60 years, range 19 to 88 years) comparing R-CHOP-21 with R-CHOP-14 and demonstrated g a higher infection rate in patients treated with R-CHOP-21. [16] The cut-off point of 75 years is among others supported by a retrospective analysis based on the SEER database in which the age above 75 years was identified as an independent risk factor for death within the first 30 days of treatment [17]. Therefore, measures to reduce toxicity and TRM in the elderly including a stringent antibiotic prophylaxis are of great importance for patients treated with immunochemotherapy. This might also be relevant for patients treated with R-CHOP-21 as the higher infection rate in these patients is possibly due to a lower rate of G-CSF prophylaxis compared to the CHOP-14 trials. In patients between 76 and 80 years of age, we observed a cumulative TRM in the RICOVER-60 trial of 20%. TRM dropped to 11% in the subsequent CHOP-R-ESC trials of the DSHNHL, although these trials used either higher doses and/or more frequent administration (DENSE-R, SEXIE-R) or an extended exposure time of rituximab (SMARTE-R). In the absence of randomized data, we hypothesize that a main reason for this improvement might be the introduction of a stringent infectious prophylaxis with a > 90% adherence in the CHOP-R-ESC trials. We found a TRM reduction by almost 50% (Table 3) compared to the RICOVER-60 study. The cut-off for an increased TRM in patients over the age of 75 years remained the same through all trials analyzed. We conclude that optimal supportive measures like the administration of prophylactic antibiotics, oral amphotericin B, and virostatic agents enable the application of full-dose immunochemotherapy in larger fractions of elderly patients with less toxicity, leading to improved cure rates. Well-designed strategies to select individualized prophylactic measures based on comorbidities and the history of a given patient are urgently needed. Data from prospective randomized trials for the increasing number of patients over 80 years are still missing. A phase II trial with a dose-reduced R-CHOP (R-miniCHOP) showed promising results, with a 2-year-PFS of 47% and OS of 59% [18]. Very recently, the data of a phase II trial applying ofatumumab in combination with miniCHOP and a mandatory prephase treatment with vincristine and prednisolone reported an improved 2 years OS of 64.7% [19]. The efficacy of bendamustine in elderly and frail patients is currently being analyzed. Park et al. published a smaller phase II trial with very modest results [20] and a median OS of 10.2 months and PFS of 5.4 months. The prospective B-R-ENDA phase II trial of the DSHNHL included pts over 80 years of age and pts being too frail to be included in the R-CHOP-ESC trials reported a 2-years OS of 42% [21]. Such studies and the results reported here highlight the clinical dilemma of treating elderly patients with aggressive lymphoma: efficacy is closely correlated with treatment intensity frequently resulting in high TRM. This analysis shows that a remarkable proportion of older patients, in particular those between 76–80 years, can receive treatment comparable to younger patients. This proportion might further increase by further optimizing supportive measures. Supportive measures (pre-phase treatment, close surveillance, prophylactic anti-infective therapy) are highly effective when treating elderly pts with aggressive lymphoma and result in significantly better treatment outcome without changing immunochemotherapy. The challenge remains to identify those elderly pts who will benefit from standard treatment if optimal prophylaxis is administered, but avoid such treatment and lower dose and dose intensity in those patients who will not benefit even if optimal prophylaxis is administered. Supporting clinical decision-making by wider and consequent use of geriatric assessment scores might help to improve the treatment results in older patients without changing state-of-the-art immunochemotherapy [15]. Recent studies trying to improve treatment results for elderly patients with DLBCL by adding new drugs to R-CHOP have largely failed, in part, because patients on these combinations suffered from more and previously unknown severe infections. Particularly in these situations, the addition of consequent anti-infective prophylaxis may help to make treatment more feasible and safe enabling patients to benefit from the superior anti-lymphoma effects offered by combining immunochemotherapy with new targeted therapies.