Tumour infiltrating lymphocytes (TILs) and immune composition in breast cancer patients from Kenya: Spatial distributions and associations with risk factors and tumour …

Document Type

Conference Paper


Pathology (East Africa)


Background: The immune landscape of breast cancer in patients from Sub Saharan Africa is understudied and findings are mainly extrapolated from studies in Caucasians and African Americans. Given the prognostic value of Tumour Infiltrating Lymphocytes (TILs) in breast cancer subtypes, the paucity of specific TIL marker data among BC patients from Africa, and the limited data on TIL spatial heterogeneity, our aim was to describe the distribution of TILs within the intratumoral stroma and the leading/invasive edge stroma, and evaluate their association across BC subtypes with established risk factors in Kenyan women with BC.

Methods: Visual quantification of overall TILs within the intratumoral stroma (sTILs) and the stroma at the leading edge/invasive edge of the tumour (LE-TILs) were performed on Haematoxylin and Eosin stained whole slide sections of pathologically confirmed breast cancer based on the guidelines of the International TIL working group. Tissue Microarrays (TMAs) of tumour were also were constructed and stained with immunohistochemistry for CD3, CD4, CD8, CD68, CD20 and Fox P3. The TMA slides were scanned on the Aperio Digital pathology slide scanner, imported into the Aperio, assigned a template, segmented and annotated before analysis using a nuclear algorithm. Linear and logistic regression models were used to assess the associations between risk factors and tumor features with IHC markers and total TILs, after adjusting for other covariates.

Results: A total of 226 invasive breast cancer cases were included in the final analysis for TILs and immune cell markers. The mean age at diagnosis of these breast cancer patients was 48.3 years. The distribution of the BC molecular subtypes defined by IHC markers was 56.2%, 16.8%, 9.3%, and 17.7 % (luminal A, luminal B, HER2-enriched, and TN breast cancers, respectively). On the whole we found that LE TIL proportions were higher than sTIL (41.3% versus 15% for TILs ≥30 and 19.3% and 4.4% for TILs ≥50 respectively). After adjusting for other tumor characteristics, we found higher TILs associated with high KI67 and high grade (for both sTIL and LE-TIL), TNBC (for sTIL only), and HER2 status, luminal B subtype, and smaller tumor size (<4 cm) (for LE-TIL only). Both sTILs and LE-TILs were predominantly composed of CD3, CD8 and CD68 with a much smaller contribution of CD4 and CD20. The percent positive cells for Fox3 were markedly lower compared to other IHC markers for both sTILs and LE-TILs.

Conclusions: Our findings with regards to enrichment of TILs in more aggressive breast cancers are similar to what has been previously published. The distinct associations between intratumoral stroma and leading edge TIL measures for some factors highlight the importance of spatial TIL evaluations in future studies. Our findings add to the body of knowledge regarding TILs and TIL subtypes in breast cancer, which underscore the need to conduct studies specific to African populations.

Publication (Name of Journal)

Cancer Research