Association of Pleural Effusion and Bleeding in Patients with Chronic Myelogenous Leukemia Receiving Dasatinib.

BACKGROUND: Dasatinib is an oral potent multikinase inhibitor that has demonstrated high efficacy and safety in patients (pts) with chronic myelogenous leukemia (CML). Pleural effusion (PE) and bleeding (BL) occur in some patients receiving dasatinib, frequently resulting in therapy discontinuation. The incidence of the occurrence of both events in the same pt either concomitantly or sequentially is unknown.

METHODS: We evaluated the incidence of PE and BL either simultaneously or sequentially among 138 consecutive pts (69 female) with CML after failure of imatinib treated at our institution in phase I and II studies of dasatinib between November 2003 and January 2006.

RESULTS: The median age for the entire cohort was 57 years (range, 15–81), median time on imatinib 164 wks (range, 21–253). Among 50 pts treated in the phase I study, 23 (46%) were in chronic (CP), 7 (14%) in accelerated (AP), and 20 (40%) in blastic phase (BP) at the time of dasatinib start, whereas 88 pts received dasatinib in phase II studies, 43 (49%) in CP, 25 (28%) in AP, and 20 (23%) in BP. Fifteen (11%) pts started dasatinib at a dose <100 mg, 22 (16%) at 100 mg, 92 (67%) at 140 mg, and 9 (6%) at >140 mg daily. The median time on dasatinib was 42 wks (range, 4–120). PE occurred in 48 (35%) pts, being grade 3–4 in 23 (17%). The median time to its development was 5 wks (range, 1 to 107). PE occurred in 29% of pts in CP, 50% in AP, and 33% in BP. Thirty-seven BL episodes were recorded in 32 (23%) pts. Seven (5%) pts had grade 1, 16 (12%) grade 2, 9 (7%) grade 3, and none grade 4–5 BL. The median time to development of BL was 6 weeks (range, 0.5–38), occurring within the first 3 months of therapy in 22 (69%) pts. BL occurred in 12% of pts in CP, 31% in AP, and 35% in BP. Basic coagulation studies were normal in all pts. Fourteen (37%) BL episodes occurred with platelets >100x10⁹/L (2 with normal platelet counts). Seventeen (12%) patients had both PE and BL (2 CP, 5 AP, and 11 BP), representing 36% of those with PE and 53% of those with BL. PE preceded BL in 8 pts while BL preceded PE in 9. The median time from dasatinib start to PE was 28 days (10–230) and to BL was 34 days (3–218). The median time between PE and BL was 17 days (range, 4–216). Six pts had grade 1, 3 grade 2, and 7 grade 3 PE, whereas 4 had grade 1, 10 grade 2, and 3 grade 3 BL (r²=0.03). The median hemoglobin drop was 2.1 g/dL (range, 1.1–4.5). PE and BL were more frequent among pts treated at daily doses ≥140 mg compared to those treated at ≤100 mg (100% vs 0%) and among those receiving dasatinib twice daily compared to once daily (88% vs 12%; p=0.001). Dasatinib was discontinued for a median of 10 days (median, 0–97) and the dose reduced in 7 pts due to PE. Loop diuretics were given to 10 pts, steroids to 2, and thoracentesis was required in 4 pts with PE, all of the latter demonstrating exudative features (median 91% lymphocytes [range, 73–100]). Dasatinib dose was discontinued for a median of 10 days (range, 0–23) and the dose reduced in 6 pts due to BL. BL types included gastrointestinal in 14 (82%) pts (lower in 11, upper in 3), gingival in 2, and epistaxis in 1. Nine patients had endoscopic examination demonstrating inflammatory changes (n=4), gastric ulcer (n=1), rectal ulcer (n=1), no lesion (n=3). Twelve pts required transfusions of packed red blood cells (PRBCs) and platelets, 3 only platelets, 1 only PRBCs, and 1 none. Dasatinib was terminated due to PE in 1 and due to BL in 2.

CONCLUSION: PE and BL can occur in a subset of pts with CML during dasatinib therapy, particularly among pts with AP or BP receiving dasatinib ≥140 mg twice daily. Appropriate clinical monitoring, transient dasatinib interruption and dose reduction are required to adequately manage these complications.