Oral fludarabine in untreated patients with B-cell chronic lymphocytic leukemia

Fludarabine phosphate is among the most active drugs in the treatment of B-Cell Chronic Lymphocytic Leukemia (B-CLL), and as monotherapy, or in combination with cyclophosphamide and rituximab, has been widely adopted as frontline therapy for this disease. Pharmacokinetic studies show that oral fludarabine at a dose of 90 mg/m²/day produces an AUC_{(0–24 h)} (AUC standing for area under the curve) comparable to that achieved with an intravenous dose of 50 mg/m²/day but with a lower C_max ¹ with a reported efficacy comparable to the intravenous drug.² In view of the convenience of administration and potential impact on resource utilization, the National Cancer Institute of Canada Clinical Trials Group initiated a phase II study of oral fludarabine in previously untreated patients with B-CLL who had an indication for therapy. The trial accrued 126 patients between 2002 and 2004. Confirmation of the diagnosis by flow cytometry was required. Analyses of known prognostic and predictive biomarkers including cytogenetic abnormalities, ZAP-70 expression and IgV_H mutational status were performed. Cytogenetic abnormalities in peripheral blood mononuclear cells were identified by interphase FISH. The loci tested were ATM at 11q22.3, D135319 at 13q14.3, D12Z1 at 12p11.1-q11.1, and p53 at 17p13.1. A 13q34-specific sequence probe distinguished between monoallelic or biallelic loss and monosomy for chromosome 13.

All patients received oral fludarabine phosphate at a dose of 40 mg/m²/day for 5 days every 28 days for a planned six cycles, to a maximum of 8 cycles for patients not in CR after 6 cycles. Allopurinol was recommended for lymphocytosis ⩾100×10⁹/l at the institution of therapy and hematopoietic growth factor support to permit scheduled dose delivery was not allowed. Dose reductions for hematologic toxicity graded according to the National Cancer Institute Working Group (NCI-WG) criteria were made starting in the second cycle.³

All patients receiving ⩾1 cycle of therapy were evaluable for response and treatment responses were determined according to NCI-WG criteria. The primary efficacy end point was the overall response rate (CR+PR) at the time of treatment completion analyzed on an intent-to-treat basis.

Cytogenetic abnormalities were detectable in 80% of cases, consistent with the rates of 77–82% reported in previously untreated and treated patients^4,5 (Table 1). Del 13q was the most common genetic abnormality (58%), followed by del 11q (23%), trisomy 12 (16%), and del 17p (5%) with 31% of patients having >1 abnormality.

Table 1.

Cytogenetics at baseline

Parameter	Patients, n (%) (N=124)
Normal	25 (20)
Abnormal	99 (80)
Single abnormality	61 (49)
>1 abnormality	38 (31)
Single abnormality
del 13 monoallelic	34 (27)
Trisomy 12	15 (12)
del 11q	7 (6)
del 17p	3 (2)
del 13 biallelic	2 (2)
Multiple abnormalities
del 13 and del 11	19 (15)
del 13 biallelic and monoallelic	11 (9)
del 13 and trisomy 12	3 (2)
del 13 and del 17	3 (2)
del 11 and trisomy 12	2 (2)

IgV_H sequencing on 101 samples showed an unmutated status (>98% homology to known IgV_H genes) in 62 patients and a mutated status in 39. ZAP-70 expression was evaluated in 125 cases with 55 positive (44%) and 70 negative (56%). ZAP-70 was expressed in 40 of 62 cases (65%) with unmutated IgV_H, compared with 5 of 39 cases (12.8%) with mutated IgV_H. ZAP-70 expression closely correlated with the presence of unmutated IgV_H genes, a profile associated with a more aggressive clinical course of B-CLL.⁶

Ninety-one patients completed protocol-defined therapy. Thirteen patients discontinued therapy due to treatment-related toxicity, 8 from progressive disease, and 14 for other reasons (including 1 patient withdrawn due to hemolytic anemia). Myelosuppression was the most frequently observed toxicity, with grade 3 or 4 neutropenia (50%), thrombocytopenia (11%), and febrile neutropenia (3%, grade 3 only). Twenty-two patients (17%) developed grade 3 infections.

Responses were observed in 81 of the 126 patients (64%) with 22 (18%) achieving CR and 59 (47%) PR. The overall response rate declined with higher Rai stage at baseline (P=0.025). The overall response rate was 77% in patients with ZAP-70 negative/mutated IgV_H CLL cells compared with 63% in patients with ZAP-70 positive/unmutated IgV_H cells but the difference between groups was not statistically significant (P=0.22).ZAP-70 expression was unchanged after fludarabine therapy in 47 of 52 cases (90%) in which samples were available for analysis.

After a median follow-up of 31.6 months, median progression-free survival (PFS) was 15.2 months (95% CI, 13.6–16.7 months), and median treatment-free survival was 18.6 months (95% CI, 13.0–28.2 months) (Figure 1). PFS and overall survival (OS) were unaffected by del 13q, del 11q, or trisomy 12 status. However, patients with del 17p had significantly shorter PFS and OS than those without the del 17 genetic abnormality (P⩽0.001) (Figure 2). ZAP-70 expression did not influence either PFS (P=0.63) or OS (P=0.52), nor did IgV_H mutational status. However, patients with mutated IgV_H had a longer treatment-free survival compared with those with unmutated IgV_H (Figure 3).

Figure 1.

Progression-free survival (PFS; a) and treatment-free survival (TFS; b) of all patients treated with oral fludarabine.

Figure 2.

Progression-free survival (PFS; a) and overall survival (OS; b) of patients based on del 17p status.

Figure 3.

Treatment-free survival (TFS) of patients based on IgV_H mutation status.

The overall response rate of 64% in this cohort of 126 patients with previously untreated B-CLL is comparable to results reported with intravenous fludarabine. Using the same response criteria, Rossi et al.² reported an overall response rate of 80% with the same schedule of oral fludarabine in a smaller group of previously untreated B-CLL patients but the proportion of patients in Rai stage III or IV in our study was higher (31% versus 22%) than in the Rossi trial.

This study again confirms that the efficacy and toxicity of oral fludarabine are comparable to that of IV fludarabine given on a similar schedule. Although several randomized trials have confirmed a better overall and complete response rate for the combination of fludarabine and cyclophosphamide compared with fludarabine alone, no differences in the OS were observed.^7,8 In the British study, the oral formulation of fludarabine was introduced in both fludarabine arms when the drug became available, and differences in response rates between the two routes of administration did not appear significant. Oral fludarabine is an effective treatment for patients with previously untreated B-CLL, with less health resource utilization compared with IV administration. In an older population, the hematologic toxicity of the combination of fludarabine with cyclophosphamide and the absence of a demonstrable survival advantage, may favor initial therapy with single agent fludarabine and deferring combination therapy. In fludarabine-based combination therapies, the oral formulation should be considered an acceptable alternative.