Myeloperoxidase Expression in B-Lymphoblastic Leukemia by Immunohistochemistry as a Diagnostic Confounder

Abstract

Background B-lymphoblastic leukemia (BLL) is a B lineage neoplasm that expresses typical immature B cell markers, but may also express myeloid-associated antigens. Myeloperoxidase (MPO) is a heme-containing peroxidase, which has been used as the single most specific myeloid marker when assigning lineage to acute leukemia. MPO positivity by immunohistochemistry (IHC) in BLL has been historically described in a subset of patients; however, further detailed comparison of IHC to flow cytometry and IHC methodology is described herein. Design The University of Washington pathology database was searched for new or relapsed adult BLL from 2011–2019. Cases with blasts >30% of marrow cellularity by morphology were selected. MPO IHC was performed using the Dako, polyclonal antibody (rabbit) on a Ventana instrument with streptavidin-biotin (SB) detection method. MPO IHC SB positive cases were also stained using the same antibody and platform, but a different detection method, multimer-optiview (MO). MPO IHC was called positive when >10% of neoplastic blasts showed expression. MPO positive blast percentage, staining intensity and pattern were assessed. Intensity: 0=negative, 1=mild, 2=moderate, 3=bright/same level as myeloids. Cytoplasmic pattern: homogenous or granular. Concurrent flow cytometry results and cytogenetic and/or molecular studies were reviewed. Results 35 cases were identified. Positive MPO IHC expression was present in 7/35 cases by SB method, with the percentage of MPO positive blasts ranging from 20–90% (majority >70%), all ranged from 1 - 2+ intensity and most (5/7) were homogenous pattern. 4/5 MPO SB positive cases were negative by MO detection method. MPO evaluation by flow cytometry was negative in 3 of 3 cases and myeloid associated antigens were negative or low on a subset. 2/8 BLL, BCR-ABL1 cases were MPO IHC positive by SB. Conclusion MPO staining by IHC in BLL is present in 17% of cases, often present in the majority of blasts, and positive using the SB detection system. This aberrant staining can be negated in most cases with the MO detection system. MPO expression by flow cytometry in MPO IHC SB positive cases were negative (3/3) and were low to absent for other myeloid antigens. We agree with prior studies that MPO IHC using the SB method can be a confounder for lineage assignment in acute leukemia.

Isolated Myeloperoxidase Immunohistochemical Expression in Bone Marrow Biopsy Depicts Clinical Outcomes in Adults with Typical B-Acute Lymphoblastic Leukemia

“…In the current study, isoMPO expression was detected in 26.7% of otherwise typical B-ALL marrow samples using IHC at a threshold of ≥3% and most of cases revealed a diffuse IHC expression in ≥10% of blasts, while 18.3% of FCM studied samples were positive, indicating a better detection-accuracy for IHC than FCM. Likewise, preceding studies reported a high frequency of MPO immunoreactivity in adult ALL of B-but not of T-cell lineage (17-23%) involving the majority of blasts in positive cases (Arber et al, 2001;Du et al, 2020), and negative or lower FCM results as well (Ahuja et al, 2018;Du et al, 2020). Reporting this finding seems of utmost importance to avoid misdiagnosis of ALL cases immunoreacting with MPO as AML or MPAL (Arber et al, 2001, where isoMPO expression in IHC can be a confounder or a diagnostic challenger for lineage assignment in AL (Li et al, 2019;Du et al, 2020).…”

“…Likewise, preceding studies reported a high frequency of MPO immunoreactivity in adult ALL of B-but not of T-cell lineage (17-23%) involving the majority of blasts in positive cases (Arber et al, 2001;Du et al, 2020), and negative or lower FCM results as well (Ahuja et al, 2018;Du et al, 2020). Reporting this finding seems of utmost importance to avoid misdiagnosis of ALL cases immunoreacting with MPO as AML or MPAL (Arber et al, 2001, where isoMPO expression in IHC can be a confounder or a diagnostic challenger for lineage assignment in AL (Li et al, 2019;Du et al, 2020). Due to the discordant MPO expression results between IHC and FCM, lowering the MPO cutoff for FCM is suggested so as to increase its sensitivity (Ahuja et al, 2018).…”

“…Outstandingly, isolated MPO expression (isoMPO) can be detected, most commonly by IHC but also sometimes by FCM in a subset of typical B-ALL specially in adults (Borowitz et al, 2017a), without fulfilling the diagnostic criteria of MPAL. Some investigators considered MPO expression in B-ALL as a diagnostic confounder in IHC (Du et al, 2020), or as a false positive FCM finding (Savasan et al, 2018). Nevertheless, the prognostic relations of this phenomenon were recently studied in pediatric B-ALL, and the data were controversial suggesting an association with higher risk and relapsing disease in some cohorts (Oberley et al, 2017, or better clinical outcomes in other cohorts (Raikar et al, 2018;McGinnis et al, 2020).…”