Prophylactic G-CSF for Elderly Lymphoma Patients: Negative Findings

Abstract & Commentary

Synopsis: In a trial of elderly patients with non-Hodgkin’s lymphoma, the prophylactic use of granulocyte colony-stimulating factor did not improve clinical outcomes, including hospitalization rate and survival. The findings run counter to an emerging clinical trend toward the use of prophylaxis with colony-stimulating factors in susceptible populations. Certain concerns are raised about the current trial, but the findings are of great importance and need to be confirmed by additional clinical investigation before we abandon the concept of primary prophylaxis in chemotherapy-treated elderly patients.

Source: Doorduijn JK, et al. J Clin Oncol. 2003;21: 3041-3050.

The Dutch-Belgian hemato-oncology Cooperative Group undertook an investigation to determine whether the relative dose-intense cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) chemotherapy could be improved by prophylactic administration of granulocyte colony-stimulating factor (G-CSF) in elderly patients with aggressive non-Hodgkin’s lymphoma (NHL). For this, patients aged 65-90 years with aggressive histology NHL were randomly assigned to receive standard CHOP chemotherapy every 3 weeks or CHOP plus G-CSF, with the G-CSF administered on days 2-11 of each cycle.

In 389 eligible patients, the relative dose intensity of cyclophosphamide (median, 96.3% vs 93.9%; P = .01) and doxorubicin (median, 95.4% vs 93.3%; P = .04) were higher in patients with CHOP plus G-CSF. The complete response rates were 55% and 52% for CHOP and CHOP plus G-CSF, respectively (P = .63). Actuarial survival at 5 years was 22% with CHOP alone, compared with 24% with CHOP plus G-CSF (P = 0.76) at a median follow-up of 33 months. Patients treated with CHOP plus G-CSF had an identical incidence of infections and only cumulative days with antibiotics were fewer with CHOP plus G-CSF (median, 0 vs 6 days; P = .006) than with CHOP alone. The number of hospital admissions and the number of days in the hospital were not different.

Comment by William B. Ershler, MD

Occasionally in clinical medicine, a seemingly logical approach makes its way into common practice, only to have careful clinical investigation challenge its tenets. In this report, the prophylactic use of G-CSF to prevent infection, and thereby enhance survival, in elderly lymphoma patients was examined. The elderly represent a particularly susceptible at-risk population for febrile neutropenia and infection after treatment with chemotherapy with moderately intensive regimens, such as CHOP. For that reason, ASCO and NCCN guidelines have recommend that this population receive prophylactic G-CSF.1 The current trial certainly brings that recommendation into question.

Doorduijn and colleagues in the Dutch-Belgian Hemato-Oncology Cooperative Group should be congratulated for completing a difficult and challenging trial. There are concerns, however, about interpretation of their data, warranting some caution before clinicians abandon the concept of primary prophylaxis for this group of patients.

The overall survival for the population studied was less than would be expected based upon the data from other groups.2-4 Secondly, as Doorduijn et al state, patients in the G-CSF group had a higher prevalence of bulky disease, which might alone account for the lack of difference in treatment outcomes. In fact, twice as many patients in the G-CSF group had progressive disease forcing them off protocol, while 40% in the non-G-CSF group left the protocol because of chemotherapy-induced toxicity. Another concern was the higher incidence of infections, number of antibiotic days, and frequency of serious infections after the first course of chemotherapy in the non-G-CSF group. Of note, after the first course of treatment there was a puzzling decline in the number of infections in both groups, raising the question of whether the sickest patients were withdrawn from study due either to toxicity or progressive disease. Finally, the incidence of infections in the whole group was much lower than reported in other studies of older lymphoma patients. For example, in one report involving a number of US practices, the incidence of infection without growth factors was close to 40%, double the rate found in the growth factor-treated group.5 Furthermore, the common use of Rituxan® with CHOP, currently considered standard therapy for elderly patients with aggressive lymphomas, may render treated patients even more susceptible to infection.

With these concerns stated, the current results still deserve the highest consideration, as oncologists are, and will be, managing an increasing number of elderly cancer patients undergoing chemotherapy treatment. However, the aforementioned considerations, most of which were out of the control of the investigative team, should be further clarified. The findings from this group need to be confirmed before clinical practice or the current NCCN guidelines, which call for the use of growth factors with first-cycle chemotherapy in individuals aged 65 and older can be modified.

Dr. Ershler is INOVA Fairfax Hospital Cancer Center, Fairfax, VA; Director, Institute for Advanced Studies in Aging, Washington, D.C.


1. Balducci L, Yates J. Oncology. 2000;14:221-227.

2. Vose JM, et al. J Clin Oncol. 1988;6:1838-1844.

3. Zinzani P, et al. Blood. 1999;94:33-38.

4. Osby E, et al. Blood. 2003;101:3840-3848.

5. Morrison VA, et al. Clin Lymphoma. 2001;2:47-56.