(P 409) IMMUNOHISTOCHEMICAL STAINING FOR MYOGENIN, AP2 BETA AND HMGA2 PREDICT FOXO1 FUSION STATUS IN RHABDOMYOSARCOMA: A REPORT FROM THE SOFT TISSUE COMMITTEE OF THE CHILDREN'S ONCOLOGY GROUP

Objective: Presence of a FOXO1 fusion is a key prognostic factor in rhabdomyosarcoma (RMS). Based on prior gene expression studies, surrogate immunohistochemical (IHC) markers have been proposed that discriminate between FOXO1 fusion positive and FOXO1 fusion negative RMS with high specificity. IHC staining was incorporated into central pathology review for ARST1431 (Children’s Oncology Group intermediate risk RMS trial supported by Pfizer) to evaluate the performance of this IHC panel in a prospective cohort of patients with known fusion status.

Methods: Central pathology review for patients enrolled on ARST1431 included review of H&E morphology as well as staining characteristics for myogenin, AP2beta and HMGA2. Based on prior published algorithms, the pattern of IHC staining was used to predict the presence or absence of a FOXO1 gene fusion. Specifically, presence of a FOXO1 fusion was predicted if myogenin staining was present in greater than 50% of tumor cells and AP2beta staining was subjectively stronger than HMGA2 staining. IHC fusion prediction was then compared to the results of institutional fusion testing.

Results: Pathology review was completed for 306 patients, of whom 58%, 29% and 13% had embryonal, alveolar or other morphology, including spindle cell, mixed and NOS, respectively. 98% (299/306 cases) had sufficient tissue for central IHC testing, of which 292 had assessment of fusion status. IHC testing was not interpretable in 1 case in which the tissue was necrotic. Institutional assessment of fusion status included fluorescence in situ hybridization in 85% (248/291), RT-PCR in 7% and a variety of NGS, karyotyping or Nanostring in the remainder. Using the previously described algorithm, the sensitivity of IHC to predict fusion status was 98.7% and specificity was 97.6%. True positive cases included 71 ARMS, 5 RMS, mixed and 2 RMS, NOS. True negative cases included 170 ERMS, 7 ARMS and 29 others including 22 SSRMS and 7 RMS, NOS. False negative cases included 1 RMS, NOS. False positive cases comprised 5 morphologic ARMS; additional molecular profiling is underway in all cases with available material.

Conclusion: Surrogate marker IHC has high sensitivity and specificity to predict FOXO1 fusion status in RMS and may be useful in small biopsies with scant or poor quality material. Additional analysis of staining patterns in specific biologic subtypes (eg. MYOD1 positive RMS) and analysis of fusion variants is ongoing.