A Detailed Evaluation of Transplant-Related Toxicities and Outcome for Patients with CNS Lymphoma (CNSL) Consolidated with High-Dose Therapy and Autologous Stem Cell Transplantation (HDT-ASCT) Using Thiotepa, Busulfan (Bu), Cyclophosphamide (TBC) Conditioning

Background:  HDT-ASCT with TBC conditioning has emerged as a common consolidation strategy for patients (pts) with relapsed/refractory (rel/ref) primary (PCNSL) or secondary (SCNSL) (Welch et al, Leuk & Lymph 2014). In a prospective study, chemosensitive PCNSL pts in first remission after induction with R-MPV (rituximab, MTX, procarbazine and vincristine) proceeding to HDT-ASCT with TBC conditioning, experienced an encouraging 2-year PFS and OS of 75% and 81%, respectively (Omuro et al, Blood, 2015). Three of these patients experienced transplant-related mortality (TRM, 11.5%), which appears greater than HDT-ASCT for other lymphomas.  The purpose of this report is to correlate characteristic toxicities of TBC conditioning for CNSL to pre-HDT-ASCT clinical variables.

Methods: The MSKCC IRB approved this retrospective chart review. Eligible pts (n=34) were ≥ 18 years of age with PCNSL or SCNSL that was chemosensitive to induction therapy after which they proceeded to HDT-ASCT conditioned with TBC between December 2006 and April 2015. All pts included were treated outside of prospective clinical trials. Clinically significant grade 3-5 non-hematologic toxicities per CTCAE 4.0 occurring in >20% of pts were recorded from the initiation of conditioning until 6 months post ASCT (Figure 1). Pre-HDT-ASCT variables for analysis include: age, gender, disease (PCNSL or SCNSL), Karnofsky performance status (KPS), hematopoietic cell transplant comorbidity index (HCT-CI), number of prior regimens, prior use of whole-brain radiotherapy (WBRT), and disease status prior to HDT-ASCT (CR/CRu or PR). We evaluated the association of these pre-HDT-ASCT characteristics with the number of clinically significant grade 3-5 non-hematologic toxicities (≥4 vs. <4) using FisherÕs exact test. We further estimated progression-free survival (PFS) and overall survival (OS) using Kaplan-Meier methods.  

Results:  Thirty-three patients (97%) experienced ≥ 1 grade 3-5 non-hematologic toxicity. Febrile neutropenia (grade 3) occurred in 32 pts (94%). Of all pre-HDT-ASCT variables, only the number of prior regimens (>2) was significantly associated with incurring more grade 3-5 non-hematologic toxicities, p=0.04 (Table 1). With a median follow-up for survivors of 12 months (range, 1.5-86.2 months), PFS was 79% (95% CI, 65-96) and OS was 82% (95% CI, 68-98) at 1 year (Figures 2 and 3). During the follow-up period, there were 7 pt deaths: 4 died of disease, 2 died secondary to TRM (5.9%), and one died of a secondary malignancy (squamous cell carcinoma) 86.2 months after HDT-ASCT.  There were no progression events beyond 12 months.  In a limited subset analysis wherein n=22 had first dose bu pharmacokinetics evaluated, pre-HDT-ASCT variables were not associated with higher bu AUC levels, though 64% of these pts required a dose reduction.

Conclusions:  We reaffirmed that HDT-ASCT with TBC conditioning is effective consolidation for CNSL, but it is associated with more grade 3-5 non-hematologic toxicity in pts having had >2 prior regimens. Risk-adapted dose attenuation of TBC conditioning for this group of pts may mitigate observed toxicity.

Figure 1. Analysis of Grade 3-5 Non-Hematologic Toxicities

Table 1. Association of Pre-ASCT Variables & Grade 3-5 Non-hematologic Toxicities

			Number of Clinically Significant Grade 3-5 Toxicities
Pre-ASCT Variables		All (N=34)	Fewer than 4 (N=21)	4 or more (N=13)	p-value
Age					0.71
	<60	23 (68%)	15 (71%)	8 (62%)
	≥60	11 (32%)	6 (29%)	5 (38%)
Gender					0.72
	Female	13 (38%)	9 (43%)	4 (31%)
	Male	21 (62%)	12 (57%)	9 (69%)
Disease					0.30
	PCNSL	19 (56%)	10 (48%)	9 (69%)
	SCNSL	15 (44%)	11 (52%)	4 (31%)
KPS					0.99
	≥80	32 (94%)	20 (95%)	12 (92%)
	<80	2 (6%)	1 (5%)	1 (8%)
BMT HCT CI					0.99
	≤2	17 (50%)	11 (52%)	6 (46%)
	>2	17 (50%)	10 (48%)	7 (54%)
Number of Prior Regimens					0.04
	≤2	21 (62%)	16 (76%)	5 (38%)
	>2	13 (38%)	5 (24%)	8 (62%)
WBRT					0.17
	No	28 (82%)	19 (90%)	9 (69%)
	Yes	6 (18%)	2 (10%)	4 (31%)
Disease state prior					0.99
	CR/CRu	29 (85%)	18 (86%)	11 (85%)
	PR	5 (15%)	3 (14%)	2 (15%)

Figure 2. Kaplan-Meier Curve for PFS

Figure 3. Kaplan-Meier Curve for OS