PD-L1 and PD-L2 Genetic Alterations Define Classical Hodgkin Lymphoma and Predict Outcome

Introduction. Classical Hodgkin Lymphomas (cHL) include small numbers of malignant Reed-Sternberg (RS) cells within an extensive but ineffective inflammatory/immune cell infiltrate. In cHL, chromosome 9p24.1 alterations increase the abundance of the PD-1 ligands, PD-L1 and PD-L2, and their further induction via JAK2-STAT signaling. PD-1 ligands engage the PD-1 receptor on T-cells and induce PD-1 signaling and T-cell exhaustion. Tumor cells expressing PD-1 ligands on their surface utilize the PD-1 pathway to evade an effective immune response. In recent pilot studies, PD-1 blockade was associated with high response rates and durable remissions in relapsed/refractory cHL.

The unique composition of cHL limits its analysis with high throughput genomic assays. Therefore, the precise incidence, nature and prognostic significance of PD-L1 and PD-L2 alterations in cHL remain undefined. Herein, we utilize a recently developed fluorescence in situ hybridization (FISH) assay to characterize 9p24.1/PD-L1/PD-L2 alterations in a cohort of 108 newly diagnosed cHL patients (pts) who were uniformly treated with Stanford V (a combined modality therapy regimen) and have longterm followup.

Methods. Pts were characterized as Ann Arbor early stage I/II favorable risk (ES-F), early stage unfavorable risk (bulk ≥ 10cm or ≥ .33 mediastinal dimension and/or B symptoms) (ES-U) or advanced stage III/IV (AS). ES-F pts received 8 weeks of Stanford V and 30 Gy involved field radiation (IFR); ES-U and AS pts received 12 weeks of Stanford V and 36 Gy IFR to initial sites > 5 cm.

FISH was performed on formalin-fixed paraffin-embedded diagnostic biopsy specimens using bacterial artificial chromosome probes which covered CD274/PD-L1 (labeled with spectrum orange) and PDCD1LG2/PD-L2(labeled with spectrum green) and a control centromeric probe (spectrum aqua-labeled CEP9, from 9p11-q11). Malignant RS cells were identified by their nuclear morphologic features and 50 RS cells/case were analyzed. Nuclei with a target:control probe ratio of at least 3:1 were defined as amplified (amp), those with a probe ratio of more than 1:1 but less than 3:1 were classified as relative copy gain, and those with a probe ratio of 1:1 but more than 2 copies of each probe were defined as polysomic for chromosome 9p. In each case, the percent and magnitude of disomy, polysomy, copy gain and amp were noted. In accordance with clinically approved diagnostic criteria, cases were classified by the highest observed level of 9p24.1 alteration. Specifically, cases with polysomy lacked copy gain or amp and cases with copy gain lacked amp.

Immunohistochemical staining for PD-L1/PAX5 was performed as previously described and PD-L1 expression in PAX5 dim⁺ malignant RS cells and PAX5^-infiltrating normal cells was assessed separately.

Results. Almost all newly diagnosed cHL pts in this series had concordant alterations of the PD-L1 and PD-L2 loci; disomy was found in only 1% (1/108), polysomy in 5% (5/108), copy gain in 56% (61/108) and amp in 36% (39/108) of study pts. There was a correlation between intensity of PD-L1 protein expression and relative genetic alterations in this series. Two additional pts had translocations of PD-L1 or PD-L2(2%, 2/108).

We next assessed the association between specific types of PD-L1/PD-L2 alterations, clinical risk factors and outcome. Overall, the progression-free survival (PFS) was significantly lower for AS pts compared to ES-F/U pts (p=0.017). A model of PFS for the cHL pts by genetic alteration indicated that PFS was also significantly lower for pts with amp (p=0.02). Consistent with these findings, the incidence of 9p24.1 amp increased by clinical risk group: ES-F, 24%; ES-U, 34%; and AS, 50% (p=0.024, Kruskal-Wallis test). Therefore, we fit a full model of clinical and genetic factors including B-symptoms, bulk, stage and amp. Despite the association of amp with increased clinical risk groups, the genetic alteration further delineated PFS in the multivariate model (p=0.075).

Conclusions. PD-L1/PD-L2 alterations are a defining feature of cHL with rare polysomy and more frequent copy gain and amp. There is an increased incidence of amp in pts with AS disease and a highly significant association of PD-L1/PD-L2 amp with PFS. These findings underscore the importance of genetically defined PD-1 mediated immune evasion in cHL and provide a rationale for the efficacy of PD-1 blockade in this disease.