Is Radiotherapy Necessary for Treating DLBCL With R-CHOP?

The combination of involved-field radiotherapy (RT) and chemotherapy has been considered a standard treatment for patients with early-stage diffuse large B-cell lymphoma (DLCBL), but results from a phase III trial published in Blood challenge the value of RT for this patient population, suggesting that R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone) alone is non-inferior to R-CHOP plus RT.

“This first randomized study in the rituximab era reports a similar outcome with these two options in [patients with non-bulky, early-stage DLBCL],” Thierry Lamy, MD, PhD, from the Department of Hematology at the Hôpital Pontchaillou in Rennes, France, and co-authors wrote. “With more than five years’ follow-up, event-free survival and overall survival are indeed identical between the two arms.”

The researchers enrolled 334 patients (range = 18-75 years) from 42 French institutions between May 2005 and June 2014 to determine differences in event-free survival (EFS) from randomization. All patients had previously untreated, non-bulky (defined as a tumor <7 cm in diameter), Ann Arbor stage II or II limited DLBCL. Patients were excluded if they had transformation of a previously indolent lymphoma, primary cerebral lymphoma, kidney or liver failure, or contraindication to doxorubicin.

Patients were randomized according to Miller modified International Prognostic Index (mIPI), lactate dehydrogenase (LDH) levels (normal vs. elevated), Eastern Cooperative Oncology Group performance status score (0-1 vs. 2-3), age (<60 vs. >60 years), and disease stage (I vs. II) to receive either:

  • R-CHOP alone (doxorubicin 50 mg/m2, cyclophosphamide 750 mg/m2, vincristine 1.4 mg/m2 on day 1, and prednisone 40 mg/m2 on days 1-5, plus rituximab 375 mg/m2 on day 1) at 14-day intervals, for 4 consecutive cycles (for those without adverse prognostic factors) or 6 consecutive cycles (for those with ≥1 prognostic factors) (n=165)
  • R-CHOP plus RT (40 Gy, in 20 fractions of 2 Gy 5 days per week) four weeks after the last R-CHOP cycle (n=169)

At baseline, patients underwent positron emission tomography (PET) and computed tomography (CT) scans; response was evaluated via PET/CT scans at two weeks after the fourth course of R-CHOP. A total of 319 patients were evaluable for response (159 in the R-CHOP group, and 160 in the R-CHOP plus RT group).

The median relative dose-intensity of cytotoxic drugs was 97 percent (95% CI 97-98) of the planned schedule. There was a median of four days of delay between cycles four and five, “mainly due to the time needed for response assessment by PET,” the authors wrote. The median time between R-CHOP and first day of RT was 36 days (range = 10-132 days).

During R-CHOP, 66 episodes of febrile neutropenia were reported in 50 patients. Four severe infections and one death due to septic shock occurred. Eighteen patients (6%) required red blood cell transfusions. After RT, two patients reported grade 3 mucositis and one reported jaw radionecrosis. A few severe (n=5) and moderate (n=8) cardiac toxicities were reported.

Most patients (n=281; 88%) achieved a complete response (CR) after four treatment cycles, including 137 in the R-CHOP group and 144 in the R-CHOP plus RT group. Another 38 patients (12%) achieved a partial response (PR; TABLE 1).

After a median follow-up of 64 months (range = 24-132 months), the five-year event-free survival (EFS; primary endpoint) was not statistically significant different between the two treatment arms: 89±2.9% for R-CHOP alone and 92±2.4% for R-CHOP plus RT (hazard ratio [HR] = 0.61; 95% CI 0.3-1.2; p=0.18).

Five-year overall survival (OS; secondary endpoint) also was similar between the groups: 92±2.5% versus 96±1.7% (HR=0.62; 95% CI 0.3-1.5; p=0.28). (The study was powered to detect differences in OS to an upper limit of 8%.)

R-CHOP alone remained non-inferior to R-CHOP plus RT whether patients achieved a CR or PR, the authors added, and EFS was significantly impacted by patient mIPI score (HR=2.8; 95% CI 0.6-14.1; p=0.04).

“In this selected group of limited-stage DLBCL, the relapse rate was very low (23 patients; 13 in the R-CHOP alone arm and 8 in the R-CHOP plus RT cohort), with a median time of 21 months (range = 4-93 months),” the authors reported. Twenty-one deaths occurred (12 in the R-CHOP arm and 9 in the R-CHOP plus RT arm), and were related to disease progression (n=11), secondary malignancies (n=3), accident (n=2), stroke (n=1), and late-onset infection (n=1); the other three deaths had unknown causes.

“This trial prompts [us] to recommend that patients with non-bulky limited stage DLBCL who reach CR based on PET evaluation after four or six R-CHOP cycles should be spared additional RT, thus avoiding radiation-related toxicity,” the authors concluded. “PET could help to deliver RT to the minority of patients with residual PET-positive tumors.” They added that the ongoing 09-1B LYSA (Lymphoma Study Association) trial is investigating whether patients with adverse prognostic factors require six cycles of R-CHOP, or if they can safely and effectively reduce R-CHOP to four cycles.

The study is limited in that it only included patients with non-bulky tumors, so it may not be generalizable to the entire DLBCL patient population. Also, the trial was designed at a time when the interval between R-CHOP cycles was uncertain. “Although we did not observe increased toxicity with a two-week interval schedule (R-CHOP 14), a three-week delivery (R-CHOP 21) appears now to be the best option,” the authors wrote.

Roche provided financial support for the study.

The authors report receiving funding from Roche Genentech.


Lamy T, Damaj G, Soubeyran P, et al. R-CHOP 14 with or without radiotherapy in non-bulky limited-stage diffuse large B-cell lymphoma (DLBCL): results of the prospective randomized phase III 02-03 trial from the LYSA/GOELAMS. Blood. 2017 October 11. [Epub ahead of print]