Immunoglobulin Sequencing Biologically Distinguishes B-Lymphoblastic Lymphoma from Acute Lymphoblastic Leukemia and Reveals a Spectrum of Disease Dissemination across Clinical Stages

Background: Molecular characterization of B-lymphoblastic lymphoma (B-LLy) has yet to identify prognostic biomarkers for use in treatment stratification or driver mutations to explain its extramedullary clinical phenotype distinct from B-lymphoblastic leukemia (B-ALL). Patients with B-LLy (with morphologically identical lymphoblasts but <25% bone marrow (BM) involvement) are empirically treated on B-ALL protocols based on clinical stage (Murphy I/II: localized/standard risk (SR); III/IV: disseminated/high risk (HR)), although debate remains as to whether B-LLy represents a separate entity. The unique clinical characteristics of B cell cancers are often presumed to reflect their stage of development at malignant transformation. In normal B cells, ordered rearrangement of immunoglobulin (Ig) heavy (IgH) and light chain genes is linked to stage of development. Upon malignant transformation, distinctive clonal Ig genomic rearrangements expand, and thus afford disease tracking. We applied Ig high-throughput sequencing (HTS) to compare the clonal composition of B-LLy with B-ALL to test whether Ig rearrangement state indicates distinct biology between these entities, and to test whether minimal disseminated disease (MDD) measured by HTS at diagnosis can refine currently limited B-LLy clinical staging.

Methods: We performed Ig HTS (Adaptive ClonoSEQ) to define and compare the Ig rearrangement state of B-LLy with that of B-ALL using extracted gDNA from paraffin-embedded tumor slide preparations from 34 patients with B-LLy (N=15 SR; N=17 HR; N=2 stage unknown) from Children’s Oncology Group (COG) protocols APEC14B1 and AALL0932, and diagnostic BM from 283 patients with B-ALL from COG protocols AALL0331 (N=141; SR) and AALL0232 (N=142; HR). We also measured B-LLy MDD in 31 patients by tracking tumor-associated Ig rearrangements in pre-treatment BM and peripheral blood (PB).

Results: All 34 B-LLy and 281/283 B-ALL samples had dominant IgH and/or Ig light chain clone(s). We detected IgH clones in 91.1% of B-LLy, 86.7% of which reflected complete V(D)J rather than partial diversity (D)-joining (J) rearrangements (vs. 74.8% complete V(D)J in B-ALL). In contrast, patients with B-ALL had more incomplete DJ rearrangements (P=0.04) indicative of a transformation state prior to variable (V) gene recombination. B-ALL V(D)J clones were enriched for usage of the most D-proximal V genes, IGHV06-01 and IGHV01-02, consistent with data indicating preferential D-proximal V gene usage in immature lymphoid progenitors and leukemia. B-LLy did not share this preferential V gene usage and lacked representation of IGHV06-01, distinguishing it from B-ALL. Further, 82.4% of B-LLy (vs. 59.7% of B-ALL; P=0.017) had dominant Ig light chain clone(s), indicative of a more mature rearrangement state. Ig HTS data from the BM and/or PB were available for 31 patients: 80.6% were MDD+, including 69.2% (N=9 of 13) with localized/SR B-LLy. BM/PB MDD level did not significantly vary between patients with localized/SR (median 0.012%; range 0-88.13%) vs. disseminated/HR (median 0.052%; range 0-24.31%) clinical staging. There was also no significant difference in MDD level between the BM and PB among 16 patients with evaluable samples from both sites. While a low event rate limited association with outcome, 0 of 6 patients without measurable MDD experienced relapse/progression/death (vs. 3 of 25 MDD+ patients).

Conclusions: B-LLy Ig clonal composition reflects a more mature developmental state than B-ALL and a spectrum of BM/PB MDD across clinical stages. Developmentally ordered Ig rearrangement first involves IgH D-J recombination, followed by V gene rearrangement, and finally Ig light chain V-J rearrangement at kappa and then lambda loci. Thus, Ig clonal composition provides biologically informative data relevant to B cell maturation stage, which here supports classification of B-LLy as a distinct entity from B-ALL. Further, the spectrum of MDD observed in this cohort agnostic of clinical stage suggests that Ig HTS may have novel utility in refining currently limited B-LLy risk stratification by distinguishing select patients who lack measurable disease dissemination or for tracking treatment response via PB HTS. In sum, Ig HTS may have a role in informing therapeutic strategies based on the unique features of B-LLy and in characterizing it as a biologically distinct entity from B-ALL.