287 CNS Disease At Diagnosis May Predict Relapse of Hematologic Malignancies in Pediatric Patients After Allogeneic Hematopoietic Cell Transplantation (AlloHCT)

Track: Contributed Abstracts
Wednesday, February 13, 2013, 6:45 PM-7:45 PM
Hall 1 (Salt Palace Convention Center)
Nirali Shah, MD , Pediatric Oncology Branch, National Cancer Institute/National Institutes of Health, Bethesda, MD
Michael J Borowitz, MD, PhD , Department of Pathology, Johns Hopkins Hospital
Seth Steinberg, PhD , Biostatistics and Data Management Section, Center for Cancer Research, National Cancer Institute, National Institutes of Health
Nancy Robey, PA , Pediatric Oncology, Johns Hopkins Hospital, Baltimore, MD
Christopher Gamper, MD, PhD , Pediatric Oncology, Johns Hopkins Hospital, Baltimore, MD
Heather Symons, MD , Pediatric Oncology, Johns Hopkins Hospital, Baltimore, MD
David Loeb, MD, PhD , Pediatric Oncology, Johns Hopkins Hospital, Baltimore, MD
Alan S Wayne, MD , Pediatric Oncology Branch, National Cancer Institute/National Institutes of Health, Bethesda, MD
Allen Chen, MD, PhD, MHS , Pediatric Oncology, Johns Hopkins Hospital, Baltimore, MD

Background: Relapse is the primary cause of treatment failure post alloHCT. We sought to identify risk factors that predict relapse of hematologic malignancies after allogeneic hematopoietic cell transplantation (alloHCT) to identify those at highest risk of relapse who may benefit from novel therapies.

Design: This was a single institution, retrospective cohort study of children with acute lymphoblastic leukemia (ALL), acute myelogenous leukemia (AML), mixed phenotypic acute leukemia (MPAL) and myelodysplastic syndrome (MDS) who had undergone alloHCT between 1/1/2003 and 12/31/2010. Relapse was defined as any evidence of increasing disease post-alloHCT, including minimal residual disease (MRD). Relapse-free survival (RFS) was estimated by the Kaplan-Meier method and the log-rank test used to assess univariate associations with various characteristics. A Cox proportional hazards model was used to identify factors jointly associated with RFS.

Results: Of 70 children who underwent a myeloablative HCT for MDS or acute leukemia in complete remission at the time of HCT, 24 (34%) relapsed at a median of 214 days (range 1 month- 57 months) post-HCT. Relapse rates by disease were 14/31 (45%) for ALL; 7/26 (27%) for AML; 3/9 (33%) for MPAL; 0/4 (0%) for MDS. Univariate analysis demonstrated that black race, central nervous system (CNS) disease at diagnosis (Figure 1), greater number of regimens given to induce remission and MRD pre-HCT were associated with higher relapse probability. In a Cox model, either two or more regimens needed to achieve remission or the presence of both pre-HCT MRD and CNS disease were approximately equally predictive of increased relapse risk. In patients with ALL, CNS disease was more highly associated with relapse risk than MRD.  For those who were MRD negative, based on 19 total patients, the presence of CNS disease at diagnosis (n=2) was significantly associated with higher relapse risk (p<0.0001)  

Conclusion: We identified CNS involvement at diagnosis as a novel risk factor associated with relapse risk after alloHCT. This may be due to inherent biologic differences leading to higher risk disease, or as a sanctuary site, the CNS may be less amenable to an allogeneic effect. These patients may benefit from earlier or more intensive CNS-directed therapy to reduce relapse risk.  Validation of these risk factors in a larger population and development of a prognostic score to identify those at highest risk of relapse in addition to a biology study to evaluate for MRD in the CNS using flow cytometery is planned. The goal is for prospective use of this prognostic tool in the development of relapse prevention trials. Table 1: Relapse rates, by MRD

Disease MRD ≥0.01% Total (n) Relapse (n) Relapse Rates
Acute lymphoblastic leukemia No 20 7 35%

Yes

11

7

64%

Acute myelogenous leukemia

No

15

3

20%

Yes

11

4

36%

Mixed phenotype acute leukemia

No

8

3

38%

Yes

1

0

0%

Figure 1: RFS by CNS Disease at Diagnosis

 

Previous Abstract | Next Abstract >>