Oral Presentation Advances in Neuroblastoma Research Congress 2016

Predicting “early” relapse/progression/death in children with INRGSS Stage M neuroblastoma using clinical and biologic factors: An INRG database analysis (#119)

Lucas Moreno 1 , Andrew DJ Pearson 2 , Dongjing Guo 3 , Daniel Morgenstern 4 , Claudia Pasqualini 5 , Shifra Ash 6 , Dominique Valteau-Couanet 5 , Susan Cohn 7 , Wendy B London 3
  1. Hospital Infantil Universitario Nino Jesus, Madrid, Spain
  2. The Institute of Cancer Research, London, uk
  3. Dana-Farber/Boston Children’s Cancer and Blood Disorders Center, Harvard Medical School, Boston, USA
  4. Great Ormond Street Hospital, London, UK
  5. Institut Gustave Roussy, Villejuif, France
  6. Schneider Children's Medical Center, Petach Tikva, Israel
  7. University of Chicago, Chicago, USA

Introduction

Most high-risk neuroblastoma patients who relapse or progress quickly after starting frontline therapy will ultimately die. Identification of these patients upfront would allow early treatment with novel therapies. Our objective was to identify clinical and biologic factors prognostic of early relapse/progression/death (“early event”).

Patients and Methods

From the INRG database, patients ≥18 months old diagnosed between 1998-2014 with INRGSS stage M disease were included. Patients were divided at random into a test set and a validation set to identify a cut-off to categorize patients as early event versus “later/no event”. To define “early event” multiple cut-offs were tested at 30-day intervals from 30 to 547 days . Logistic regression identified clinical and biologic factors prognostic of an early event.

Results

1820 patients met inclusion criteria: 23% ≥5 years old, 35% MYCN amplified, 48% diploid, 39% 11q LOH, 39% 1p LOH, 11% LDH≥1400 U/L, 22% ferritin≥30 ng/mL, 33% high MKI, 3% Stage 4N (metastases exclusive to distant lymph nodes) and 9% liver metastases.  Regardless of the cut-off tested, the risk of death was high (hazard ratio [HR]>5) for patients with an early event compared to later/no event. A 182-day cut-off (identified 8% of the cohort as “early events”) was implemented. 5-year OS was 10±3% for patients with early events (n=150) and 46±1% for patients with later/no events (n=1670) (p<0.0001, HR=5.1 [95%CI: 4.2-6.1]). Factors prognostic of an early event were: MYCN amplification (p=0.03; odds ratio [OR]=1.5), LDH ≥1400U/L (p=0.04; OR=1.7) and liver metastases (p=0.02; OR=3.1).

Conclusions

MYCN, LDH, and liver metastases were determined to be strongly associated with early events in this INRG cohort. However, to precisely identify patients who may benefit from alternative treatment approaches, a deeper understanding of the genomic alterations that drive rapidly progressive disease will be needed