4736 - Prognostic value of tumor infiltrating lymphocytes (TILs) in patients with early-stage triple negative breast cancers (TNBC) in the absence of chemotherapy

Background

Stromal tumor-infiltrating lymphocytes (sTILs) have been shown to be a strong prognostic factor in early-stage triple negative breast cancer (TNBC). There is limited data on their prognostic value in TNBC patients in the absence of adjuvant chemotherapy.

Methods

We performed a pooled analysis of TNBC patient series not treated with chemotherapy. The percentage infiltration of sTILs was retrospectively evaluated in the most representative tumoral block of surgical specimens. Cox proportional hazards regression models stratified by series were used for invasive disease-free survival (iDFS), distant disease-free survival (D-DFS), and overall survival (OS), fitting sTILs as a continuous variable adjusted for clinicopathologic factors.

Results

We collected individual data from 518 patients from four series. Average age was 64 years (range, 24 to 96 years), and 83% of patients were node-negative. The median size was 1.6 cm (Q1-Q3: 1.6-2.5), and 82.7% of the cases were node-negative. Tumours were grade 1, 2 and 3 in 12%, 38% and 50% of the cases. The median level of sTILs was 10% (Q1-Q3, 4-30%). Higher grade was associated with larger amount of sTILs (p < 10^-3). During follow-up, a total of 173 iDFS, 118 D-DFS events and 107 deaths were observed. In the multivariable model which included the number of lymph nodes, tumour size, age, tumour grade and radiotherapy treatment, sTILs added significant independent prognostic information for all endpoints (likelihood ratio chi2= 7.14 for iDFS; p < 10^-2; chi2= 9.63 for D-DFS, p < 10^-2; chi2= 5.96 for OS, p = 0.015). Each 10% increment in sTILs corresponded to a hazard ratio of 0.90 (95% CI, 0.82 to 0.97) for iDFS, 0.86 (95% CI, 0.77 to 0.95) for D-DFS, and 0.88 (95% CI, 0.79 to 0.98) for OS. In patients with stage I tumors with sTILs≥30% (n = 74), 5-year iDFS was 91% (95% CI, 84% to 98%), D-DFS was 97% (95% CI, 93% to 100%), and OS was 98% (95% CI, 95% to 100%).

Conclusions

sTILs provide important prognostic information in systemically untreated early stage TNBC patients. sTILs may be used to identify stage I TNBC patients with excellent prognosis in whom chemotherapy may be withheld.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

ANR and CGI (RHU MyPROBE, ANR-17-RHUS-0008).

Disclosure

M.V. Dieci: Advisory / Consultancy: Lily, Celgene, Genomic Health. C. Criscitiello: Honoraria (self), for speaker engagements and consultancy roles: Pfizer; Honoraria (self), for speaker engagements and consultancy roles: Roche; Honoraria (self), for speaker engagements and consultancy roles: Lily; Honoraria (self), for speaker engagements and consultancy roles: Novartis. R.F. Salgado: Travel / Accommodation / Expenses, travel support: Roche, Merck, AstraZeneca; Travel / Accommodation / Expenses, congress support: Roche, Merck; Research grant / Funding (self), Data and safety monitoring member: Roche, Merk, Puma; Advisory / Consultancy, Advisory board: Roche, BMS. S. Loi: Research grant / Funding (institution): Novartis, Bristol Meyers Squibb, Merck, Roche-Genentech, Puma Biotechnology ,Pfizer and Eli Lilly; Advisory / Consultancy, but not compensated: Seattle Genetics, Pfizer, Novartis, BMS, Merck, AstraZeneca and Roche-Genentech. S. Kim: Research grant / Funding (institution): Novartis, Sanofi-Genzyme, Dongkook Inc. G. Curigliano: Honoraria (institution), institutional fees from them: Merck, Roche, SEAGEN, Daichi Sankyo, Novartis, Pfizer. F. André: Research grant / Funding (institution): Pfizer, AstraZeneca, Novartis, Lilly, Roche, daiichi; Advisory / Consultancy: Pfizer, AstraZeneca, Novartis, Lilly, Roche, daiichi. S. Michiels: Advisory / Consultancy, punctual statistical advice: IDDI, Belgium; Advisory / Consultancy, advice on surrogacy: Janssen Cilag France; Advisory / Consultancy, Data and safety monitoring member: Hexal, J&J, Ipsen, Neovacs, Genticel, Mabxience, Steba, IQVIA, Roche, Sensorion, Biophytis. All other authors have declared no conflicts of interest.