Background
Philadelphia-like (Ph-like) B Lymphoblastic Leukemia (B-ALL) is a high-risk subtype of B-ALL that lacks the BCR-ABL1 fusion but has a gene expression profile similar to Philadelphia positive (Ph+) B-ALL. Gene expression profiling has previously identified Immunoglobulin Joining chain (IgJ) overexpression at the mRNA level in Ph-like B-ALL. This is surprising since IgJ is normally expressed by mature or maturing B cells. The normal function of IgJ protein is to concatenate monomers of immunoglobulin IgM or IgA into the mature pentameric and dimeric forms of these molecules respectively. IgJ also plays a crucial role in transport of IgA protein across mucosal epithelium to facilitate mucosal humoral immunity. In hematopathology, J chain immunohistochemistry (IHC) has been used to identify neoplastic cells in nodular lymphocyte predominate Hodgkin lymphoma (NLPHL) and can be used in distinguishing this disease from morphologic mimickers. It does not have known diagnostic utility outside of this context. Lymphoblasts do not typically express immunoglobulins at the protein level. Therefore, we sought to determine the protein expression of IgJ in B-ALL and to determine whether IgJ immunohistochemistry may be employed in identifying particular subtypes of B-ALL.
Methods
We selected a total of 46 B-ALL cases diagnosed from a bone marrow sample at our institution from 2016-2019 with adequate diagnostic material for IHC. This included 5 cases of Ph-like B-ALL, all with a CRLF2 rearrangement and overexpression, 7 de novo Ph+ B-ALL and 34 cases representing the other most commonly recognized WHO subtypes of B-ALL, determined based on cytogenetic studies performed at the time of diagnosis. No cases of B-ALL with t(5;14) and B-ALL with iAMP21 was represented. Our cohort included 23 pediatric cases and 24 adult cases and the patients ranged from 1 to 82 years old at the time of initial diagnosis. A total of 8 normal bone marrow cases (negative staging bone marrow biopsies for diffuse large B cell lymphoma, neuroblastoma or classic Hodgkin lymphoma) were used as controls. IgJ IHC was performed on B-plus fixed paraffin embedded bone marrow biopsy specimens using a commercially available and validated anti-IgJ monoclonal antibody (clone OTI3B3). Staining of bone marrow samples was performed at 2 different dilutions; tonsil secondary follicles and neoplastic cells from NLPHL were used as technical controls. Cellular staining in the lymphoblasts was scored in a blinded manner by a board certified hematopathologist and a pathologist in training as diffusely positive, partially positive, or negative.
Results
Cellular staining was distinguishable from background staining due to circulating immunoglobulins and there was almost perfect inter-observer concordance in identifying positive and negative cases (agreement of 98%, kappa test 0.94). All normal bone marrow controls cases were negative for IgJ cellular staining. A total of 11/46 (23%) B-ALL cases demonstrated partial or diffuse cellular staining for IgJ in the lymphoblasts. This included 4/5 (80%) Ph-like cases, 5/7 (71%) Ph+ cases, 1/3 MLL rearranged cases and 1/6 ETV6-RUNX1. All TCF3-PBX1 (0/4), hyperdiploid B-ALL (0/10), hypodiploid B-ALL (0/2), and B-ALL, NOS cases (0/9) were negative for IgJ. Diffuse IgJ staining was restricted to Ph-like (2/4) or Ph+ (2/5) B-ALL subtypes; the positive MLL rearranged and ETV6-RUNX1 B-ALL cases only showed weak partial staining. IgJ protein was significantly expressed in Ph+/Ph-like B-ALL (p<0.0001) and in our cohort, detected these cases with a 75% sensitivity, 95% specificity, 82% positive predictive value and 92% negative predictive value.
Conclusion
We conclude that IgJ protein expression occurs in a subset of B-ALL, predominantly restricted to Ph+ and Ph-like cases. Although, these findings will need to be validated in larger studies, our results suggest that IgJ IHC, in concert with routine standard cytogenetics studies may be a rapid and cost effective method in identifying Ph-like B-ALL.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.