Commentary on Abstracts #1396, #2684, and #3114

Although responses to rituximab occur in approximately 50% of patients with follicular NHL, several studies in chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) have shown response rates in the range of only 10% to 15% (McLaughlin et al: J Clin Oncol 16:2825-2833, 1998; Foran et al: J Clin Oncol 18:317, 2000; Maloney et al: Blood 90:2188-2195, 1997; Piro et al: Proc Am Soc Clin Oncol 18:14a [abstract 49], 1999; Winkler et al: Blood 94:2217-2224, 1999; Nguyen et al: Eur J Haematol 62:76-82, 1999). At the 1999 ASH meeting, Winkler et al (abstract #1396) updated their previously published experience with an additional patient; among the 11 total patients, there was 1 very brief CR and 1 PR.

A possible explanation for the lower response rates with rituximab in CLL/SLL than in follicular NHL is that, although CLL/SLL cells are CD20-positive, they tend to express this antigen dimly. Several approaches are being investigated to improve the response rate in CLL/SLL patients. O’Brien and coworkers from the M. D. Anderson Cancer Center (abstract #2684) conducted a phase I trial of rituximab in 50 patients with CLL and 10 with prolymphocytic leukemia (PLL) or mantle cell lymphoma. All patients initially received a dose of 375 mg/m² with three subsequent weekly doses escalating by cohort up to 2,250 mg/m². The response rate was 40%, including responses in 36% of the CLL patients.

Although patient numbers were small, the authors suggested that there might be a dose-response effect; a 23% response rate was noted with doses ranging from 500 to 825 mg/m², but 8 of 10 patients responded at the highest doses. However, it is not clear how many PLL and mantle cell lymphoma patients were represented at each dose level. Prolymphocytic leukemia and mantle cells tend to express CD20 more brightly than do CLL cells and, therefore, are more likely to respond.

Byrd et al (abstract #3114) studied the pharmacokinetics of rituximab using the standard recommended dose and schedule and noted that trough levels were lower in patients with CLL/SLL than in those with follicular NHL. They hypothesized that more frequent drug administration of the antibody might have greater efficacy in these patients. Patients in this study received “stepped-up” dosing in an attempt to minimize toxicities; all patients received an initial 100-mg dose of antibody and then either 250 or 375 mg/m² three times weekly for 12 doses. Among 18 evaluable patients, there was a 50% response rate, including 1 CR. Limited data suggested a correlation between the level of expression of CD20 and response rate. Unfortunately, despite efforts to minimize toxicity, there were two serious adverse events, including a fatal pulmonary hemorrhage.

These two alternative strategies for rituximab administration should not be considered outside of a clinical trial. First, they are an extremely costly way to attain an increased number of PRs. Second, results of clinical trials are needed to demonstrate an improved outcome.

Another approach designed to increase the response rate of CLL to rituximab is augmentation of CD20 expression. In vitro studies have suggested that certain cytokines, such as granulocyte colony-stimulating factor (G-CSF), may enhance CD20 expression, and ongoing clinical trials (Venugopal et al: Blood 92:247a [abstract 1009], 1998) are evaluating rituximab administered with G-CSF (Neupogen) or granulocyte-macrophage CSF (GM-CSF [Leukine, Prokine]) in CLL patients.