Methods

Top Methods Results Discussion Author & Article Info References

Eligibility Criteria

Patients resistant to or intolerant of DCF therapy with active hairy cell leukemia, morphologically confirmed on a bone marrow examination, were eligible for 2-CdA treatment. Resistance to DCF was defined as primary refractoriness to DCF administration or relapse after a partial response lasting less than 12 months. Patients intolerant to DCF had experienced life-threatening side effects with DCF therapy. Patients had to have an absolute neutrophil count less than 1.0 x 10⁹/L, a hemoglobin concentration less than 100 g/L, or a platelet count less than 100 x 10⁹/L to be entered into the study. All patients had to have discontinued -interferon or DCF therapy at least 3 months before the study and have no evidence of active infection. In addition, all patients had to be older than 18 years, have adequate renal (serum creatinine concentration less than 180 µmol/L) and hepatic function (bilirubin, alkaline phosphatase, alanine aminotransferase, and aspartate aminotransferase) less than two times normal, and be available for follow-up evaluation. Written informed consent was obtained and the study was approved by our institutional review board.

Drug Formulation

2-Chlorodeoxyadenosine was synthesized and purified as previously described [12] or supplied by Ortho Pharmaceutical (Raritan, New Jersey) and recrystallized before use. 2-Chlorodeoxyadenosine was supplied to the pharmacy as a 0.1% solution (1 mg/mL) of pyrogen-free 2-CdA in sterile 0.9% NaCl. For outpatients using an infusional pump device (Pharmacia Deltec CADD-PCA Infusion Pump; St. Paul, Minnesota), the desired dosage of 0.10 mg/kg per day was added to the 0.9% NaCl solution to make a total volume (drug plus diluent) of 100 mL; for inpatients, the same dosage was added daily to 500 mL of 0.9% NaCl.

Study Design

All patients received a single course of 0.10 mg/kg body weight of 2-CdA by continuous intravenous infusion for 7 days. For outpatients, this was administered through a peripherally inserted central venous catheter using an infusion pump. Inpatients were admitted to our General Clinical Research Center and the drug was infused using peripheral venous access. Hematologic and nonhematologic toxic responses were monitored according to standard criteria [13]. For clinically significant anemia or thrombocytopenia, packed red blood cells or platelet transfusional support or both were administered. For a temperature greater than 38.5 °C, patients were admitted and empiric broad-spectrum antibiotics were initiated.

Before therapy was begun, all patients had complete histories taken and physical examinations, reviews of their bone marrow histologic findings, reviews of splenic and lymph node tissue when available, bone marrow and peripheral blood tartrate-resistant acid phosphatase stains, complete blood counts with differential counts, 24-chemical blood chemistry panels, and imaging of the chest, abdomen, and pelvis by computed axial tomography. Serial peripheral blood lymphocyte, flow cytometric examination was done when possible.

During therapy and on subsequent follow-up, all patients had a physical examination and a complete blood count with differential counts daily until the absolute neutrophil count was greater than 1 x 10⁹/L and the platelet count was greater than 20 x 10⁹/L, then weekly until the blood counts returned to normal. Patients then were followed at 3-month intervals. Bilateral bone marrow aspiration and biopsies, together with tartrate-resistant acid phosphatase stains, were performed at 3, 6, and 12 months, and yearly thereafter, and were reviewed at Scripps Clinic and Research Foundation. Computed tomographic scans were repeated during periods of maximum response.

Response Criteria

Complete remission was defined as the absence of discernible hairy cells in the bone marrow aspirate and biopsy; the presence of tartrate-resistant acid phosphatase-positive cells lacking typical hairy cell morphologic features did not preclude a complete remission designation. The absolute neutrophil count was 1.5 x 10⁹/L or greater, the hemoglobin concentration was 120 g/L or greater, and the platelet count was 100 x 10⁹/L or greater. Neither splenomegaly nor lymphadenopathy were evident on physical examination or on computed tomographic scan.

Partial remission required a reduction of 50% or more in the absolute count of hairy cells in the peripheral blood and bone marrow and a similar decrement in the size of splenomegaly or lymphadenopathy or both by examination and imaging studies, when present. In addition, there had to be an increase in granulocytes, hemoglobin, or platelet count of at least 50% of the deviation from normal of the pretreatment value. All other responses were judged to be in the "no response" category.

Results

Top Methods Results Discussion Author & Article Info References

Patient Characteristics

Five patients, four men and one woman, with a median age of 55 years (range, 44 to 57 years), entered the study (Table 1). All patient biopsy specimens stained positively with tartrate-resistant acid phosphatase. All patients had received previous interferon and DCF therapy. Two patients had also undergone splenectomy, and one of these two patients (patient 5) also had received abdominal irradiation to overcome biliary obstruction from massively enlarged periportal lymph nodes. Of the five patients, four had obtained a response during interferon therapy, one had a complete response, and three had partial responses. The median response duration following interferon therapy was 14 months, ranging from 4 to 18 months.

All five patients had received DCF at 4 mg/m² body surface area every other week for a median of 5 doses (range, 2 to 23 doses) (see Table 1). Patient 1 achieved a partial response of 9 months duration following 23 doses of DCF and was therefore designated as DCF resistant. Therapy with DCF was complicated by dermatomal herpes zoster. In patient 2, DCF therapy was discontinued after three doses because his therapy was complicated by pancytopenia, fever, and a subendocardial infarction. Because of these life-threatening toxicities on DCF, he was classified as being DCF intolerant. In patient 3, DCF therapy was withdrawn after two doses because of profound pancytopenia, an associated fever, and a protracted hospital stay; she was designated DCF intolerant. Patient 4 had no response to five doses of DCF and was considered to be primary DCF resistant. Treatment was complicated by nausea, vomiting, and culture-negative pneumonitis requiring the empiric use of amphotericin B. Patient 5, who had previously demonstrated responsiveness to splenectomy and interferon, was refractory to eight doses of DCF and was also considered to be primary DCF resistant. Dermatomal herpes zoster developed in patient 5 during DCF therapy.

The median duration from diagnosis of hairy cell leukemia to the initiation of 2-CdA was 3.5 years (range, 2.3 to 14.5 years). The median duration from the completion of DCF therapy to the initiation of 2-CdA therapy was 1 year (range, 0.5 to 2 years).

Peripheral Blood Hematologic Changes after Treatment

The median leukocyte count at the initiation of 2-CdA therapy was 2.0 x 10⁹/L (range, 0.8 to 21.8 x 10⁹/L), at the nadir it was 0.4 x 10 (⁹/L) (range, 0.2 to 0.8 x 10⁹/L), and after 2-CdA treatment it increased to 3.8 x 10⁹/L (range, 2.5 to 8.1 x 10⁹/L) (Table 2). Associated with these changes in the total leukocyte count was a median pretreatment absolute neutrophil count of 0.56 x 10⁹/L (range, 0.20 to 1.01 x 10⁹/L), a nadir of 0.35 x 10⁹/L (range, 0.06 to 0.58 x 10⁹/L), and a value after treatment of 2.73 x 10⁹/L (range, 0.68 to 2.95 x 10⁹/L).

The median hemoglobin concentration before treatment was 112 g/L (range, 60 to 134 g/L), the median hemoglobin nadir was 94 g/dL (range, 84 to 108 g/L), and the median hemoglobin after treatment increased to 140 g/L (range, 105 to 155 g/L). The median platelet count before treatment was 55 x 10⁹/L (range, 12 to 95 x 10⁹/L), the median platelet nadir was 54 x 10⁹/L (range, 16 to 80 x 10⁹/L), and the median platelet count after treatment increased to 123 x 10⁹/L (range, 100 to 207 x 10⁹/L) (see Table 2).

Responses

Of five patients, four (80%) obtained complete responses and one patient (20%) achieved a partial response. Patients 2 and 3, intolerant of DCF, both achieved complete responses. Patient 2 had fewer than 1% tartrate-resistant acid phosphatase-positive cells on follow-up bone marrow examination without morphologic evidence of hairy cell leukemia infiltration; the peripheral blood lymphocyte immunophenotype showed complete disappearance of CD19-positive B lymphocytes with coexpressing CD11c, CD25, and B-Ly7. Patients 1, 4, and 5, unquestionably resistant to DCF, obtained responses. Patients 1 and 4 achieved complete responses of more than 12 months and 10 months, respectively. The complete response in patient 4 lasted 10 months, but slowly progressive splenomegaly with stable peripheral hematologic parameters has since developed. Peripheral blood lymphocyte immunophenotyping performed at 15 months showed no circulating hairy cells. Patient 5 had a partial response of 2 months duration.

The overall median duration of follow-up is more than 10+ months (range, 2 to 14+ months). Three of four patients who obtained a complete response remain in unmaintained complete remission following the administration of a single course of 2-CdA.

Peripheral Blood Lymphocyte Immunophenotypic Changes

Serial immunophenotypic analyses at 3- to 6-month intervals using peripheral blood was available in three of the five patients, all of whom achieved complete responses (Table 3). A population of B lymphocytes with CD19 positivity coexpressing CD11c, CD25, and B-Ly7 was present in all three patients before initiating 2-CdA therapy. All three patients had complete disappearance of this lymphocyte subset at 3 to 6 months and serially thereafter. All three patients had a decrement in their absolute CD4 and CD8 counts following therapy. However, in two of the three patients (patients 3 and 4), follow-up CD4 and CD8 counts have exceeded the values before treatment (see Table 3).

Patient 1, who responded completely, had a peripheral blood flow cytometric examination before initiating 2-CdA treatment and again at 12 months (see Table 3). The CD4 count before treatment was 666 cells/mL and after treatment it was 654 cells/mL. The CD4/CD8 ratio, however, improved from 0.60 to 1.42.

Before treatment, patient 2 had a CD4 count of 628 cells/mL and a CD8 count of 489 cells/mL; at 6 months, a CD4 count of 159 cells/mL and a CD8 count of 414 cells/mL; and at 12 months, a CD4 count of 254 cells/mL and a CD8 count of 362 cells/mL.

Patient 3 had a pretreatment CD4 count of 242 cells/mL and a CD8 count of 204 cells/mL; at 9 months, a nadir CD4 count of 77 cells/mL and a nadir CD8 count of 74 cells/mL; and at 12 months, a CD4 count of 321 cells/mL and a CD8 count of 284 cells/mL, exceeding the pretreatment values.

Before treatment, patient 4 had a CD4 count of 167 cells/mL and a CD8 count of 306 cells/mL; at 3 months, a nadir CD4 count of 73 cells/mL and a CD8 count of 396 cells/mL; and at 15 months, a CD4 count of 283 cells/mL and a CD8 count of 336 cells/mL, exceeding the pretreatment values.

Peripheral blood lymphocyte immunophenotype profiles were not performed on patient 5.

Toxicity

Transient, reversible myelosuppression was the major toxicity encountered. Of the five patients, grade 3 neutropenia (absolute neutrophil count, 0.5 to 1.0 x 10⁹/L) developed in one, grade 4 neutropenia (absolute neutrophil count less than 0.5 x 10⁹/L) developed in four, grade 3 thrombocytopenia (platelet count, 25 to 50 x 10⁹/L) developed in two, and grade 4 thrombocytopenia (platelet count less than 25 x 10⁹/L) developed in one patient. No patient had clinical bleeding.

Neutropenic febrile episodes developed in two of the four complete responders, and they were treated with empiric broad-spectrum antibiotics, although there were no documented infections. Three months after 2-CdA treatment, bronchoscopically culture-negative pneumonitis developed in patient 3 that resolved completely following treatment with empiric antibiotics. Patient 5, the partial responder, died 6 months after receiving 2-CdA. No autopsy was performed. His course was complicated by Streptococcus pneumoniae bacteremia and dermatomal herpes zoster.

Discussion

Top Methods Results Discussion Author & Article Info References

Both DCF [14, 15] and 2-CdA [6, 7] have been shown to induce responses in patients with hairy cell leukemia who have not responded to -interferon therapy. Despite similar chemical structures and mechanisms of action, 2-CdA can induce durable complete responses in patients with hairy cell leukemia who are truly resistant to DCF or who have experienced prohibitive toxic reactions with DCF therapy. In this study, two patients received only brief courses of DCF, but intolerance manifested by life-threatening toxic reactions prevented further DCF administration. Three patients were DCF resistant; one patient, despite protracted DCF therapy, relapsed early at 9 months and was thus resistant to DCF, and two patients had primary refractoriness to DCF therapy. The responses achieved in the three patients considered to be DCF resistant suggest a lack of cross-resistance between 2-CdA and DCF in hairy cell leukemia.

2-Chlorodeoxyadenosine requires phosphorylation by deoxycytidine kinase to its putative active form, 2-chlorodeoxyATP (2-CdATP). 2'-Deoxycoformycin, a tight-binding inhibitor of adenosine deaminase, causes the intracellular accumulation of 2'-deoxyadenosine triphosphate (dATP). In dividing cells, 2-CdATP and dATP inhibit ribonucleotide reductase [16]. Ribonucleotide reductase is, however, almost undetectable in resting cells. Yet a unique feature of 2-CdA is its high level of cytotoxicity against resting cells [17]. Therefore, a different mechanism of lethal DNA damage must occur. The accumulation of deoxynucleotide 5'-triphosphate causes increased DNA strand breaks over time. The eventual outcome is the activation of two enzyme systems: a Ca+±Mg+±dependent endonuclease that produces double-stranded DNA strand breaks at internucleosomal regions and a poly-(ADP-ribose) polymerase that consumes both NAD and ATP. The lymphocyte thus exposed to 2-CdA or DCF cannot repair DNA damage. The cleavage of DNA into oligonucleosomal fragments occurs, which is characteristic of apoptosis, a form of programmed cell death [18].

In this report, we have shown that 2-CdA therapy appears to be safe to administer in patients previously treated with DCF. The two patients who had experienced life-threatening toxic reactions with DCF therapy had only transient myelosuppression after 2-CdA administration, with culture-negative fever being the only clinical consequence. However, given the immunosuppressive properties of both DCF [19] and 2-CdA [20], the risks of opportunistic infections and secondary malignancies need clarification in larger numbers of patients with longer duration of follow-up. Two of the three patients studied had overall improvement in their absolute CD4 and CD8 counts following 2-CdA therapy compared to their pretreatment values. These results compare favorably with flow cytometric studies performed in patients with hairy cell leukemia, not previously exposed to DCF, treated with single courses of 2-CdA [20]. In these 16 patients with hairy cell leukemia before 2-CdA treatment, the absolute lymphocyte counts and T/B-cell ratios were abnormal in most patients, but the ratio of helper/suppressor T cells (CD4/CD8) was normal in most. Both T and B lymphocytes were reduced after 2-CdA treatment, and among the T lymphocytes both CD4 and CD8 lymphocytes were affected. A decrease in the CD4/CD8 ratio was seen as a result of slower recovery of CD4 cells. Between 6 and 12 months after 2-CdA administration there was a tendency toward restoration of T-cell subsets in most patients with a median post-treatment CD4 lymphocyte count of 540 cells/mL.

The pattern of acute toxicity observed in these patients resistant to or intolerant of DCF was also similar to that previously reported, with transient myelosuppression and culture-negative fever being the principal adverse effects [6, 7]. Unfortunately, flow cytometric examination was not performed in patient 5, so the contribution of immunosuppression to the development of his dermatomal herpes zoster could not be determined.

The apparent lack of cross-resistance between interferon and the nucleosides DCF and 2-CdA and their individual abilities to induce durable responses should provide the therapeutic arsenal for developing a curative approach to the treatment of HCL, should relapses occur late after 2-CdA therapy. Whether the durable responses following 2-CdA will eventually be established as cures or whether combinations of these agents might be needed to develop curative strategies remains to be defined. Given that hairy cell leukemia is a disorder with a long natural history, protracted follow-up will be necessary to determine ultimate relapse rates. The advances in therapy that have already occurred will probably have a favorable effect on survival for a disorder that before the introduction of effective systemic therapy had a median survival time of only 53 months [21].

In conclusion, evidence is emerging for some lack of cross-resistance between the two nucleosides, 2'-deoxycoformycin and 2-CdA, in the treatment of hairy cell leukemia. In addition, 2-CdA is not prohibitively toxic in patients intolerant to DCF, indicating that the therapeutic-to-toxic ratio for 2-CdA may be superior to that of DCF in patients with hairy cell leukemia. This will have important implications for future protocol design. Current data must be interpreted cautiously because these observations have only been made in small numbers of cases and must be established in much larger studies. Furthermore, careful observation of the immunosuppressive effects and their clinical consequences in patients receiving treatment with multiple nucleoside agents must be studied closely.