Outcomes of Pharmacokinetically (PK) Directed Busulfan in Combination with Thiotepa & Cyclophosphamide (TBC) Conditioning with HDT-ASCT in Patients with Primary & Secondary CNS Lymphoma

Background: High-dose therapy and autologous stem cell transplantation (HDT-ASCT) with thiotepa/busulfan/cyclophosphamide (TBC) conditioning is effective consolidation for patients with newly diagnosed & relapsed/refractory primary (PCNSL) or secondary CNS (SCNSL) lymphoma. A prospective study by Omuro et al (Blood 2015) showed that chemosensitive patients proceeding to HDT-ASCT with TBC conditioning in first remission resulted in encouraging 2-year PFS and OS, but with significant toxicity & transplant related mortality compared to HDT-ASCT for other lymphomas. To our knowledge, there is limited evidence defining the optimal PK directed busulfan (bu) dosing strategy in patients with CNSL undergoing TBC conditioned HDT-ASCT. We report PK targeted bu in TBC and HDT-ASCT in 22 patients with CNSL between 2011 and 2014.

Methods: Twenty two patients with CNSL who underwent TBC conditioned ASCT from 2011-2014 with PK targeted busulfan were included. TBC is thiotepa IV 250mg/m2 on days -9,-8, -7; bu 3.2mg/kg IV on days -6, -5,-4; and cyclophosphamide 60mg/kg IV on days -3 & -2 with stem cell infusion on day 0. Adjusted ideal body weight was used in all patients >125% of ideal body weight. PK analysis of bu levels obtained after first dose was by high performance liquid chromatography with mass spectrometry. Predicted area under the curve (AUC) was reported based on 6 point kinetics. Target AUC was 4100-5200 umol*min/L (goal 4700 umol*min/L). Dosage adjustments per PK were made on the 3^rd dose of bu. All PK modeling was performed using WinNonLin® 6.0 (Centara, Princeton, NJ).

Results: Twenty-two patients with primary CNSL (PCNSL, n = 12, 55%) or secondary CNSL (SCNL, n = 10, 45%) from 2011-2014 received TBC conditioning with PK targeted bu. Thirteen (59%) were men and median age was 56 years (range 25-72). Twelve (55%) & 10(45%) patients received 1-2 & 3-6 lines of prior therapy, respectively for remission induction. All patients were chemosensitive prior to HDT-ASCT with 18(82%) patients in complete remission (CR) & 4(18%) in partial remission (PR). Median pre-transplant HCT-CI was 2.5 (range 0-4) and KPS of ≥80 in 20(91%). Median first dose bu AUC was 5550 umol*min/L (range 3268-7464 umol*min/L with median total bu exposure of 14939 umol*min/L (range 11236-19240 umol*min/L). Five (23%) patients were within therapeutic range, 3(14 %) required a median dose increase of 78%, and 14 (64%) required a median dose reduction of 55% to achieve goal bu exposure. Median time to neutrophil and platelet engraftment was 11 (11-14) & 17.5 (13-52) days respectively. All patients experienced grade 3/4 non-hematologic toxicities [11 (50 %) 0-3 & 11 (50 %) ≥ 4 non-hematologic toxicities, respectively]. The incidence of potential treatment related AST/ALT elevations >3 x ULN & t-bili >1.5 mg/dl was n=5(23 %) & n=7(32 %), respectively. Age and pretransplant HCT-CT >2 were not associated with higher bu AUC or exposure (Table 1). Patients who received ≥ 3 prior regimens had a lower initial bu AUC (p = 0.04), but no difference in total bu exposure (p=0.25). There was no difference in requirement for dose reduction by pre-transplant characteristics (e.g. age, HCT-CI, or prior regimens). Median progression free survival (PFS) by 1 year was 68% (95% CI 39-86) & overall survival (OS) 74% (95% CI 44-90).

Conclusion:

TBC conditioning is effective consolidation for CNSL. PK directed dosing of bu in our study population resulted in a higher than expected bu AUC. Receipt of > 3 prior regimens was the only pre-transplant patient factor associated with lower initial bu AUC. Although we were not able to correlate any other pre transplant patient factors and bu AUC, the overall incidence of regimen related grade 3-5 non hematologic toxicities remain high. With favorable PFS & OS, future studies targeting lower bu AUC per PK are warranted to reduce toxicity.