Secondary Lymphedema: Autoimmune, Auto-inflammatory, or Both?

BACKGROUND: Activation of T-helper (Th) inflammatory responses is an established pathophysiology concept for the development of secondary lymphedema (LE). However, recent evidence of an oligoclonal T-cell response in sequenced biopsy samples from LE patients, demonstrates a repertoire of T-cell receptors (TCR’s) that recognize the self-antigen, insulin, suggesting that some patients should be classified as suffering from an autoimmune disorder rather than merely an autoinflammatory condition. The hypothesis of an autoreactive component to secondary LE is influenced by several factors: 1) single nucleotide polymorphisms (SNPs) in human leukocyte antigen (HLA) class I and II predict a higher risk of podoconiosis, a form of secondary lymphedema; 2) evidence from a large retrospective study found that LE development following axillary lymph node dissection (ALND) is significantly associated with one or more autoimmune diseases; 3) there is a proven pathogenic role of autoreactive T-lymphocytes in the most clinically relevant inflammatory skin diseases. In this study, we investigate the HLA allelic genotypes associated with LE, classify them as either risk-increasing or protective, and identify putative autoantigens with a strong affinity to these LE-associated alleles. METHODS: We conducted a genome wide association (GWAS) study in patients who did or did not develop breast cancer-related lymphedema (BCRL) following ALND. Two models were run: 1) the allelic model tested the odds of lymphedema in the presence of each HLA allele; 2) the genotypic model tested the odds of lymphedema among patients having a given HLA genotype compared to those that did not have that genotype. Using a validated HLA database, we mapped the alleles of interest to the highest affinity epitope. RESULTS: In the allelic model, the HLA class I significantly associated with LE was HLA-A-02:05 (OR 1.66, 95% CI 1.02-2.71, p=0.043). The HLA class II’s with the highest association included HLA-DQB1-3:02 (OR 1.66, 95% CI 1.02-2.71, p=0.043) and HLA-DRB1-08:04 (OR 1.65, 95% CI 1.00-2.70, p=0.046). In the genotypic model, HLA-B-07:02 (OR 1.59, 95% CI 1.05-2.40, p=0.028) and HLA-DQA1-05:01 (OR 1.63, 95% CI 1.04-2.55, p=0.034), were among the most significant HLA class I and class II’s, respectively. All three alleles were presented at similar proportions in patients that developed LE and absent in patients that did not develop LE. In the allelic model, HLA-B-44:02 (OR 0.78, CI 0.63-0.96, p=0.020) was expressed at a significantly higher proportion (n=17; n=6) in patients that did not develop disease when compared to patients who did. Allelic-epitope mapping of the risk increasing alleles, HLA-DQB1-03:02 and HLA DQA1-05:01, demonstrates a high affinity to autoantigens insulin and myelin basic protein. Mapping was inconclusive for HLA-DRB1-08:04. CONCLUSION: The role of T-cell immune responses in triggering LE development has been demonstrated in clinical specimens and mouse models. Only recently, has there been insight into the putative antigens activating these responses. Identification of an HLA, LE- risk allele is provided as evidence for an antigen-driven, possibly autoimmune pathogenesis to LE development. Additionally, it provides a unique potential to develop HLA-dependent immunotherapies to prevent progression of lymphedema.