LAG3 is an immune checkpoint expressed on a variety of immune cells including a sub-population of 'exhausted' effector T cells and TREGs. Early-phase studies of anti-LAG3 mAb show promise in solid and haematological cancers. We have previously demonstrated LAG3 is enriched within the tumor microenvironment in Hodgkin Lymphoma (Gandhi et al. Blood 2006). Data in the aggressive B-cell lymphoma DLBCL is lacking.
We used a conventional discovery/ validation approach in two population based Australian cohorts (discovery: Brisbane/Canberra; validation: Sydney) totalling 250 patients treated with R-CHOP with >4 year follow-up. Digital gene expression (NanoString) using a consistent LAG3 cut-off showed inferior 5-year overall survival (OS) in both cohorts (discovery: 54% vs. 82%, HR 3.13, p=0.003; validation: 63% vs. 86%, HR 2.95, p=0.025 respectively). In a multivariate model, LAG3HI(p=0.001) was a predictor of OS independent of R-IPI and Cell-of-Origin (by NanoString LST assay). PD-L1 expression was also a predictor of survival though to a lesser degree than LAG3. Notably, LAG3 expression stratified PD-L1HIpatients into two sub-groups with differential survival, with dual LAG3 and PD-L1 positivity conferring particularly poor OS (PD-L1HI/LAG3HI39% vs. 81% PD-L1HI/LAG3LO, HR 3.65, p=0.023).
Next, the discovery/validation cohorts were combined with 129 additional DLBCL cases from the ALLG biobank (in whom tissue but no outcome data was available), to test for biological associations and correlations. In these 379 cases, LAG3HIwas enriched in the ABC/UC (66%) subtype vs. LAG3LO(p=0.003). LAG3 was positively correlated with numerous immune checkpoints/effectors including CD4, CD8, PD-1, PD-L1, PD-L2, TIM-3 and CD163 (all p<0.0001, r range 0.44-0.67) consistent with an adaptive immune response. High LAG3 expression was significantly more common in EBV+ DLBCL (p=0.037).
LAG3 gene expression was highly correlated with protein expression by tissue microarray based immunohistochemistry (r=0.79, p<0.001). To determine which cells expressed LAG3, de-aggregated fresh-frozen tumor-infiltrating lymphocytes were interrogated by flow cytometry. CD4+ T-cells were ~2-fold higher than CD8+ T-cells. CD4+ T-cells were further sub-divided into 'classical' CD127LOCD25HITREGs, CD127LOCD25LOinducible-TREGs, and CD4 non-TREGs. This showed that LAG3 was highly expressed within the CD8 and both TREG populations, but there was minimal CD4 non-TREG expression. LAG3HICD8+ T-cells were frequently enriched in the inhibitory checkpoints PD-1 and/or TIM3 consistent with a highly dysfunctional/exhausted phenotype.
Finally, levels of soluble LAG3 (sol-LAG3) were quantified within paired plasma of patients enrolled into the ALLG NHL21 PET-adapted prospective DLBCL study. Samples taken pre-therapy and following 4 cycles of R-CHOP (at the time of interim-PET) were compared. Sol-LAG3 levels were higher in patients with DLBCL at diagnosis compared to healthy controls (p<0.0001). Interestingly patients that became interim-PET-ve had a significant drop in sol-LAG3 levels (p=0.008) between time-points, whereas no change was observed in those that remained interim-PET+, suggesting that sol-LAG3 has utility as a disease response biomarker.
In conclusion, high expression of LAG3 in DLBCL is enriched in the non-GCB phenotype, and is associated with poor outcome independent of clinical and biological prognosticators. Dual PD-L1HI/LAG3HIexpression confers particularly poor outcome after conventional front-line immuno-chemotherapy. Intratumoral LAG3 expression is high on PD-1+ CD8+ and TREG subsets. Sol-LAG3 appears to a circulating disease response biomarker. The results combined indicate a key role for LAG3 within the immunobiology of DLBCL and provide a strong rationale for early phase clinical trials utilising anti-LAG3 and anti-PD1 mAb combinations.
Keane:BMS: Research Funding; Roche: Other: Education Support, Speakers Bureau; Celgene: Consultancy, Research Funding; Takeda: Other: Educational Meeting; Merck: Consultancy. Gould:Novo Nordisk: Other: Educational Travel. Abro:Novartis: Consultancy; Amgen: Other: education support congress attendance; Janssen: Other: education support congress attendance; Bristol-Myers Squibb: Speakers Bureau; Celgene: Other: education support congress attendance. Tobin:Celgene: Research Funding; Amgen: Other: Educational Travel. Birch:Medadvance: Equity Ownership. Talaulikar:Novartis: Honoraria, Speakers Bureau; Janssen: Honoraria, Speakers Bureau; Takeda: Research Funding; Roche: Honoraria, Speakers Bureau; Amgen: Consultancy, Honoraria. Bird:Amgen, Novartis: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Hertzberg:Amgen: Membership on an entity's Board of Directors or advisory committees; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Roche: Membership on an entity's Board of Directors or advisory committees; Takeda: Honoraria, Membership on an entity's Board of Directors or advisory committees; MSD: Membership on an entity's Board of Directors or advisory committees. Gandhi:Celgene: Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Merck: Honoraria, Membership on an entity's Board of Directors or advisory committees; Amgen: Honoraria; Takeda: Honoraria; Gilead: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; BMS: Membership on an entity's Board of Directors or advisory committees, Research Funding.
Asterisk with author names denotes non-ASH members.