Enhanced clearance of leukemic lymphocytes in B-cell chronic lymphocytic leukemia with etodolac

Chronic lymphocytic leukemia (CLL) is a neoplastic disease characterized by the accumulation of small mature-appearing lymphocytes in the blood, bone marrow, and lymphoid tissues.1 Current therapy has not been shown to prolong survival.2 

Etodolac is a racemic mixture of R-(-) and S-(+)-1,8-diethyl-1,3,4,9,-tetrahydropyrano-(3,4-b)indole-1-acetic acid.3 Etodolac selectively inhibits cyclooxygenase-2,4 and it is approved for use in various parts of the world for treatment of degenerative joint disease and rheumatoid arthritis and for use as an analgesic.5 

We report the chance observation that racemic etodolac lowers the lymphocyte count in a patient with B-CLL and show that challenges with 13 other nonsteroidal anti-inflammatory drugs (NSAIDs) produced no significant effect. We also present studies to show that at standard anti-inflammatory doses, racemic etodolac achieves this action by enhancing the selective clearance rather than by the direct killing of leukemic lymphocytes.

The patient's condition was diagnosed as B-CLL in 1994 at age 57. Her leukemic lymphocyte phenotype was CD5⁺, CD19⁺, CD20⁺, CD23⁺, CD25⁺, FMC7⁺, and λ dim⁺. She was clinically staged at Rai/Binet 0/A. On October 23, 1997, her white blood cell count (WBC) was 34 300 and her lymphocyte count was 27 440. From October 25, 1997, to October 27, 1997, she took etodolac 300 mg PO twice a day for neck pain. On October 28, 1997, because of the development of petechiae etodolac was discontinued and a complete blood count was performed. Her platelet count was normal but her WBC had dropped to 13 400 with a lymphocyte count of 6 700. To determine whether this was a reproducible effect, etodolac at the same dose of 300 mg was administered twice a day from November 5, 1997, to November 11, 1997; January 21, 1998, to January 25, 1998; February 26, 1998, to March 2, 1998; and April 8, 1998, to April 13, 1998, all with similar effects (Figure 1A). No clinically significant changes in the hemoglobin level, neutrophil count, or platelet count were seen with etodolac (platelet data not shown). Also seen in Figure 1A and B are the results of the administration of 13 other NSAIDs. None produced any significant decrease in the lymphocyte count. The effect of etodolac appears not to be due to its anti-inflammatory property since none of the other NSAIDs elicited this response. Furthermore, the effect does not appear to be related to cyclooxygenase specificity since celecoxib, rofecoxib, and meloxicam, selective cylooxygenase-2 inhibitors, did not significantly affect the lymphocyte count. A 4-month course of etodolac was then administered between July 20, 1998, and November 23, 1998 (Figure 1C). Taken together with the previous data, the reduction of lymphocyte count in response to etodolac administration was prompt, with a nadir occurring after 2 days of administration. The duration of the effect after drug withdrawal was short, with a return of lymphocyte count to baseline within 2 weeks of drug cessation independent of the duration of therapy. The effect was also reproducible and did not appear to be lessened with repeated or prolonged drug challenges.

To determine whether cell killing could be responsible for the drop in circulating lymphocytes, mononuclear cells from the patient were cultured for 72 hours in presence of increasing concentrations of naproxen or etodolac up to 240 μM and analyzed at 24, 48, and 72 hours for the binding of monoclonal antibody to internucleosomal DNA by enzyme-linked immunosorbent assay (ELISA) as a measure of apoptosis using a standard kit assay (Cell Death Detection ELISA Plus, Boerhinger-Mannheim, Indianapolis, IN). No significant enhancement of apoptosis was achieved by etodolac compared with naproxen (data not shown). The same results were obtained in a separate experiment when assayed by analysis of the binding of FITC annexin V to and the uptake of propidium iodide by the patient's mononuclear cells by flow cytometry (data not shown). The fact that an increase in apoptosis or necrosis could not be detected for etodolac compared with naproxen in vitro did not exclude the possibility that a metabolite of etodolac or a serum factor might be necessary to achieve the effect in vivo. To examine this possibility, the percentages of viable, apoptotic, and late apoptotic/necrotic cells were measured by analysis of the binding of FITC annexin V to and the uptake of propidium iodide by isolated mononuclear cells after administration of etodolac to the patient (data not shown). As the lymphocyte count dropped after administration of etodolac, the percentage of apoptotic cells remained the same. These data suggested that etodolac at standard anti-inflammatory and analgesic concentrations did not achieve reduction of leukemic lymphocyte count by direct killing but by increasing the clearance of leukemic cells into other compartments.

To directly determine whether etodolac was enhancing the clearance of the patient's leukemic lymphocytes, a WBC scan was performed in the standard fashion with and without the administration of etodolac. As shown in Figure 2, an increase in the radiolabel appeared in the spleen with the administration of etodolac. No difference was detected in the amount of radiolabel that appeared in the liver.

These data show that at standard anti-inflammatory doses, etodolac reproducibly lowers the lymphocyte count in a patient with B-CLL. Furthermore, data is presented to show that etodolac achieves this action by affecting changes in leukemic cell compartmentalization. Further work should clarify the prevalence of this effect in B-CLL, the molecular mechanisms involved, and whether or not racemic etodolac, one of its chiral isomers, or similar compounds could be useful in B-CLL or other B-cell neoplasms.