Davis et al. (1) reported on a patient with a 17-year history of NHL1with multiple relapses. After receiving two separate four-infusion courses of Rituximab, the patient developed a diffuse large cell CD20-negative B-cell NHL. The patient had prior histological diagnoses of follicular small cleaved, follicular mixed with diffuse large cell foci,and mixed cell lymphomas. Worldwide experience of over 20,000 patients treated with Rituximab has yielded only one additional report (2) of a patient with follicular, mixed cell lymphoma who,upon transformation, relapsed with CD20-negative B-cell lymphoma.
Although it has been documented that 10% of intermediate-grade NHL patients have B-cell lymphomas that de novo are CD20 negative (3), there is a paucity of information in the medical literature to specifically address loss of CD20 expression. We have recently completed a 60-patient clinical trial of Rituximab retreatment that showed all patients to be CD20 positive on relapse,including some patients who had received two courses of Rituximab (4). These two case reports raise questions regarding the incidence and frequency of antigen loss as a result of the natural history of low-grade lymphoma or as a result of therapeutic interventions.
The mechanism of action of Rituximab results in rapid and specific B-cell depletion (5), with B-cell recovery beginning from 3–6 months posttreatment. Thus, peripheral blood and bone marrow samples will be CD20 negative by flow cytometric analysis during this period, indicating B-cell depletion rather than any change in CD20 status. We have detected measurable levels of free, circulating Rituximab for as long as 6 months after treatment (6). The presence of this antibody in the circulation will block CD20 binding sites, again resulting in apparent CD20 negativity. Circulating B cells may be coated with Rituximab and blocked from detection by commonly used flow cytometric diagnostic reagents that only detect surface CD20. Immunoperoxidase staining with L26 antibody, which recognizes the intracellular domain, is necessary to confirm the CD20 antigen status in the presence of circulating Rituximab.
Should anyone have a patient suspected of having a B-cell lymphoma with the CD20 negative phenomenon, an evaluation including the following steps is recommended.
In patients known to have B-cell lymphoma, test pre- and posttreatment samples, optimally from the same source [tumor (node);bone marrow, if marrow is involved; or peripheral blood, if circulating lymphoma cells are present]. For both pre- and posttreatment samples,determine the CD19 and CD20 status by flow cytometric analysis. A CD19-negative and CD20-negative result indicates an absence of B cells;therefore, a CD20-negative clone cannot be present. A CD19-positive/CD20-negative result indicates that B cells are present that either lack the CD20 antigen or have bound Rituximab, blocking the external portion of the antigen from detection. Testing a paraffin-embedded tissue section sample with L26 (7) will determine whether cytoplasmic CD20 is present. [Note that a recent CD20 review article (7) indicated that rare cases of low-grade lymphoma may not stain L26.] Lack of L26 staining indicates a CD20-negative clone, whereas L26 staining indicates the presence of cytoplasmic CD20. If the result is positive for cytoplasmic CD20, the sample can be tested for surface-bound Rituximab. Rituximab is an IgG1κ antibody. If the lymphoma expresses IgM or λ, detection of IgG or κ on the surface suggests bound Rituximab. This implies that the cells remain CD20 positive and that Rituximab is blocking detection.
After Rituximab therapy, peripheral blood B cells may be depleted or the CD20 antigen may be blocked by bound Rituximab. Bone marrow samples are useful only if malignant cells are present; however, bone marrow is not involved in most follicular patients [over 50% of patients do not have marrow involvement at diagnosis (8)], and bone marrow may be cleared of tumor cells after treatment. However, it is critical that the tumor be sampled, because the issue is tumor cells becoming CD20 negative.
Although this appears to be a low incidence phenomenon, we are conducting additional studies and examining our own experiences as we seek to establish a registry. We welcome information on any other cases so that a registry can be maintained to further inquire into the incidence, etiology, and pathophysiology of this phenomenon (please contact Dr. Dana Lee at IDEC Pharmaceuticals; phone number,1–888-200–1181).2
The abbreviation used is: NHL, non-Hodgkin’s lymphoma.
To whom requests for reprints should be addressed.
Reply
Drs. Grillo-López and Kunkel have clearly stated the current knowledge regarding loss of CD20 expression in patients after Rituximab therapy. It is interesting to note that both identified cases of CD20-negative relapses represent transformed histologies, and we have yet to see a case of relapsed indolent disease that no longer expresses the CD20 antigen. We are impressed that so few CD20-negative cases have been identified, considering the fact that Rituximab remains present in a patient’s serum for up to 6 months—truly a chronic exposure to the selecting agent.
We strongly support their proposal to collect data on the incidence of CD20-negative relapses after Rituximab therapy. It is important to recognize, however, that an opportunity has already been lost. We will not be able to truly identify the denominator of patients treated with Rituximab, nor can we expect to identify with any accuracy the true numerator, the true number of relapses from clones that have lost their requirement for expression of this antigen, because only rare patients will undergo biopsy of their relapsed disease. In the current economically fiscal environment, it is unreasonable to expect that relapsed patients undergo tumor sampling and analysis for CD20 expression. Where feasible, such sampling should certainly be done, and the analysis for CD20 expression defined by Drs. Grillo-López and Kunkel seems entirely appropriate. With the registry that they propose,an estimate of frequency for CD20-negative relapse will be possible.
Therapeutic agents in clinical investigation are increasingly targeted to specific proteins, enzymes, or molecules. The measurement of biological effect represented by changes in target expression or activity is an integral part of the evaluation of those agents. An evaluation of similar importance in all clinical trials should be the biopsy of a representative sample of resistant or relapsed tumors to determine whether loss of target is a mechanism of escape. This is particularly relevant to monoclonal antibodies, but the principle can be applied to all targeted therapies.
RSS Feeds