Developing a Treatment Strategy Based Upon Risk Factors for Patients With High-Grade Lymphoma

Abstract & Commentary

Synopsis: The treatment outcome of patients with intermediate- or high-grade lymphoma with three poor prognostic factors remains unsatisfactory. In this study from a single institution within the United Kingdom, patients with poor-risk, high-grade lymphomas were treated initially with conventional chemotherapy followed by high-dose chemotherapy and peripheral blood stem cell rescue. An overall 64% complete remission rate was achieved. Prognostic factors that were associated with poorer outcome included bulky mediastinal mass, more than three extranodal sites, and low-serum albumin. Interestingly, remission status before HDCT PBSC was not found to influence event-free survival or overall survival, suggesting that early introduction of HDCT may be of benefit to those patients in partial remission.

Source: Lee SM, et al. Bone Marrow Transplant 1999; 24:271-277.

Treatment for diffuse aggressive lymphomas remains a challenge. Lee and associates at Christie Hospital (Manchester, UK) designed a study to identify the factors that convey the highest risk for treatment failure in patients with high-grade non-Hodgkin’s lymphoma. Sixty-six patients with intermediate- or high-grade lymphomas and two or three adverse prognostic factors as defined by the age-adjusted International Prognostic Index (IPI) received induction treatment with 7-11 weeks of doxorubicin, cyclophosphamide, vincristine, bleomycin, etoposide, prednisolone, and methotrexate (VAPEC-B) followed by three cycles of ifosfamide/cytarabine (a subgroup of these patients with Burkitt’s and lymphoblastic lymphoma had VAPEC-B followed by three cycles of high-dose methotrexate).1 Subsequently, all patients received high-dose chemotherapy (HDCT) with busulphan/cyclophosphamide followed by autologous peripheral blood stem cell (PBSC) rescue.

Of the 66 patients treated, 47% achieved complete remission (CR) before the high-dose phase of the treatment regimen. All but one of the remaining had achieved a partial remission (PR) before HDCT. Following the HDCT, 64% were in CR and 19% had a PR. One patient had progressive disease and four patients died of treatment-related toxicity. For survivors (n = 46), at a median follow-up of 27 months, the actuarial three-year estimate of overall survival was 67%, event-free survival was 65%, and freedom from progression was 70%.

Examining this series by univariate analysis of risk factors, reduced survival was associated with mediastinal bulk, more than three extranodal sites, remission status before HDCT, low-serum albumin, and elevated erythrocyte sedimentation rate. No significant difference was observed between patients with intermediate- or high- grade histology or between patients with two or three IPI-defined adverse factors. When subjected to multivariate analysis, mediastinal mass, more than three extranodal sites and low-serum albumin persisted as independent predictors of survival but remission status before HDCT was not found to be significantly associated with poor outcome.

Based upon these three adverse factors (mediastinal bulk, > 3 extranodal sites, and low-serum albumin) the data were reexamined and patients were identified in one of three groups (0, 1, or > 2 adverse prognostic factors) and these were found to have significantly different outcomes. Event-free survival was 85% for the low-risk group, 63% for the intermediate-risk group, and 29% for the high-risk group. Overall survival was 84% for the low-risk group, 64% for the intermediate-risk group, and 25% for the high-risk group.

Lee et al suggest that those patients with intermediate- or high-grade lymphomas and two or more of these particular adverse prognostic factors do not respond well to conventional dose therapy followed by HDCT and PBSC rescue, and that these individuals should be considered for alternative (experimental) approaches from the outset.


Clinicians treating lymphoma have long been aware of the myriad of adverse clinical prognostic factors that are independent of specific histology and stage. This report reemphasizes their potential importance in establishing prognosis for patients with intermediate- or high-grade lymphoma and raises the suggestion that those with two or more of the adverse factors identified here (which differ from the IPI factors) be considered for alternative therapies. This, of course, would be reasonable if potentially more effective therapies were available. However, one poor-risk patient in four is a long-term disease-free survivor with this therapy and it seems unreasonable to me to pass up even that small potential for cure to try untested approaches, unless they included therapy at least as intense as VAPEC-B plus HDCT and PBSC.

One intriguing point emerging on multivariate analysis in the current series is that partial remission to conventional-dose chemotherapy before HDCT and PBSC rescue was not found to be associated with poorer outcome than that seen in those who had achieved complete remission to conventional-dose therapy. In most other series, PR after conventional chemotherapy for patients with high-grade lymphomas has been associated with poor outcome (for example see1). This might imply that there are differences in the ability of HDCT to rescue partial responders treated with CHOP vs. partial responders treated with VAPEC-B. Alternatively, the particular preparative regimen used at Christie Hospital may be more effective than other such regimens. It is also important to distinguish results obtained at a single institution from results generated in a cooperative group. Often cooperative groups obtain poorer results, though this is not universal (results from the French GELA group are among the best in the world). While some interpret these differences as reflecting the inclusion of patients with poorer prognosis on group studies, careful analysis of results of technically demanding treatments such as HDCT and PBCS demonstrate that experience influences the outcome. Centers that use HDCT in fewer than six patients per year have a poorer outcome than those with more experience do.

As noted above, the prognostic factors defined in this series differ from those found in the IPI. The IPI did not examine serum albumin. However, in the IPI, more than one extranodal site conferred a poor prognosis while, in the Christie Hospital series, three or more extranodal sites were indicative of a poor prognosis. Of course, the IPI project involved many more patients. However, it may be of importance to note that most of the patients from whom the IPI was derived were treated with CHOP. If a therapy came along that was more effective than CHOP, one might expect that the prognostic factors would change. In this instance, more aggressive treatment appears to have shifted the threshold for extranodal sites. It takes three or more to influence prognosis when VAPEC-B and HDCT and PBSC are used as treatment, not more than one, as observed with CHOP.

Despite the notable results from this and other series, the use of HDCT and PBSC rescue for high-grade lymphomas remains controversial.2-6 There have been recent reports indicating improved remission rates and overall survival using this approach. However, convincing randomized trial results have not been obtained. Several large scale, multi-institutional trials are ongoing. It would be ideal if patients with poor prognostic factors could be entered into one of these studies.


1. Hag R, et al. J Clin Oncol 1994;12:1074-1084.

2. Colombat P, et al. J Clin Oncol 1994;11:1990;8: 630-637.

3. Freedman AS, et al. J Clin Oncol 1993;11:931-936.

4. Jackson GH, et al. Br J Cancer 1994;70:501-505.

5. Nademanee A, et al. Blood 1997;90:3844-3852.

6. Gianni AM, et al. N Engl J Med 1997;336:1290-1297.

In the series of patients with high-grade lymphoma treated at the Christie Hospital in Manchester, UK, which of the following factors was not an independent variable with regard to event-free and overall survival?

a. bulky mediastinal mass

b. number of extranodal sites

c. serum albumin

d. remission status prior to high-dose chemotherapy