Fine Tuning of Surface CRLF2 Expression and Its Associated Signalling Profile in Childhood B Cell Precursor Acute Lymphoblastic Leukemia

Abstract 1409

Cytokine receptor-like factor 2 (CRLF2) alterations occur in 8% of unscreened children with B cell precursor acute lymphoblastic leukemia (BCP-ALL). The prognostic impact is still controversial and likely to be dependent on different clinical protocols. Biologic studies may facilitate the identification of factors contributing to these discrepant results. Moreover a better understanding of the biochemical mechanism(s) underlying such aberrancies may provide a more effective selection of patients that could benefit of signal transduction inhibitors treatments. In the present study we evaluated CRLF2 genomic rearrangements and the signalling profile involving pJAK, pSTAT5 and pS6 signalling pathways, known to be associated with CRLF2 rearrangements (Tasian SK et al, Blood 2012).

A total of 60 (54 consecutive fresh and 6 thawed) diagnostic bone marrow samples from children with BCP-ALL enrolled in AIEOP-BFM-ALL trials were analyzed for CRLF2 surface expression by flow cytometry (FCM). CRLF2 transcript levels as well as CRLF2 aberrations (i.e. P2RY8-CRLF2) were analyzed by molecular methods as recently described (Palmi C et al, Leukemia 2012). TSLP-induced (10 ng/mL for 30 minutes) pSTAT5 and pS6 response were evaluated in CD7-/CD19+/CD10+/CD45^low blasts by phosphoflow cytometry in thawed mononuclear cells obtained after ficoll gradient centrifugation.

Eight out of 60 (13.3%) samples were FCM CRLF2 'positive' (mean value of CRLF2+ cells = 87.5%) with a bright or intermediate fluorescence distribution. However, further cases showed peculiar CRLF2 expression: 2/60 (2.25%) were 'partially positive' (mean CRLF2+ cells = 2.25%), with two clearly distinguished blast populations (one negative and one positive); 9/60 (15.00%) were 'dim positive' (mean CRLF2+ cells = 3.24%) showing a clear shift-to-right distribution compared to negative control (normal B lymphocytes CD7-/CD19+/CD10-/CD45++), and 41/60 (68.30%) were fully 'negative' (CRLF2+ cells <1%). Parallel RT-PCR analysis of CRLF2 expression was performed in 51 samples: all FCM-positive samples were over expressed by PCR (i.e. fold change of >20 compared to the median value of almost 500 samples tested) whilst CRLF2 'dim positive' or 'partially positive' had a PCR fold change of < 20. Interestingly, 29/32 CRLF2 'negative' were concordantly negative by PCR, however, in 3/32 samples a fold change of >20 was obtained. Of note, in these 3 discordant cases P2RY8-CRLF2 was detected as a very weak PCR amplification (n=2) or PCR negative (n=1), suggesting the presence of minor rearranged cell populations.

Aberrant pSTAT5 response was observed in all CRLF2 ‘positive’ analyzed samples (n = 8) compared with CRLF2 'negative' cases (n = 9) showing 59.7% and 5.4% of pSTAT5+ cells (means), respectively (p =0.00001). Interestingly CRLF2 'dim positive' cases (n = 5) had an intermediate level of pSTAT5 response (mean 18.1%) significantly different compared to both CRLF2 'positive' and CRLF2 'negative' cases (p <0.05). Our results are concordant with recent findings by Taisan et al, although with some differences. We found that IL-7 (100 ng/mL) induced p-STAT5 response at variable levels in BCP-ALL blasts regardless of CRLF2 expression. Also TSLP-induced pS6 response, in addition to pSTAT5, was aberrant in CRLF2 'positive' cases compared to CRLF2 'negative' ones (mean 26.9 vs. 0.2, p = 0.0001), yet CRLF2 'dim positive' cases showed intermediate mean level of pS6 response being 8.3%. All CRLF2 'dim positive' cases were negative (or undetectable) for P2RY8-CRLF2 deletion.

In conclusion, we confirm and further extend that TSLP mediated signal transduction induces aberrant JAK-STAT and PI3K-mTOR signaling pathways in CRFL2 mutated or over expressed BCP-ALL cases. Of note, cases with dim positivity of CRLF2 (usually considered negative by standard flow cytometry criteria) and negativity for P2RY8-CRLF2 deletion showed an active pSTAT5 and pS6 signaling at intermediate level between CRLF2 'positive' and CRLF2 'negative'. Whether it refers to the pattern of minor clones or to the presence of additional mechanism(s) driving aberrant signal transduction, are still under investigation. Overall, these findings may be relevant for diagnosis and prognosis in CLRF2 positive BCP-ALLs; they further support that signal transduction inhibitors may have therapeutic relevance in this setting.