Tumor-Infiltrating Lymphocytes in Patients Receiving Trastuzumab/Pertuzumab-Based Chemotherapy


Michail Ignatiadis; Gert Van den Eynden; Salgado Roberto; Marco Fornili; Yacine Bareche; Christine Desmedt; Françoise Rothé; Marion Maetens; David Venet; Esther Holgado; Virginia McNally; Astrid Kiermaier; Heidi M. Savage; Timothy R. Wilson; Javier Cortes; Andreas Schneeweiss; Karen Willard-Gallo; Elia Biganzoli; Christos Sotiriou


J Natl Cancer Inst. 2019;111(1):69-77. 

In This Article


TRYPHAENA Study Design

The TRYPHAENA study (NCT00976989) design and results have been previously reported[8] and are described in the Supplementary Methods (available online). Written informed consent was obtained from each participant. Approval for the protocol and for any modifications was obtained from independent ethics committees (committees are listed in the Supplementary Methods, available online).

Stromal Tumor-infiltrating Lymphocyte Evaluation

The percentages of stromal coverage by TILs (hereafter referred to as TIL percentage) before (at diagnosis) and after (for patients with residual invasive breast cancer) dual HER2 blockade were evaluated in hematoxylin and eosin (H&E) slides from diagnostic biopsies and surgical samples from patients enrolled in the TRYPHAENA study. TIL percentage at surgery was evaluated in patients with residual invasive breast cancer. In the analyses, the geometric means of the percentage of TILs present were determined by two independent pathologists;[9] the agreement between the two pathologists was evaluated as explained in the Supplementary Methods (available online). TIL percentage at diagnosis was assessed according to the 2014 guidelines developed by the International Immuno-Oncology Biomarker Working Group.[10] TIL percentage for patients with residual invasive disease was scored according to guidelines developed by the International Immuno-Oncology Biomarker Working Group.[11] In cases of pCR or if no tumor bed could be identified, TIL assessment was not performed.

Gene Expression Using the NanoString nCounter Assay

Sample processing and RNA extraction for gene expression profiling are described in the Supplementary Methods (available online). Gene expression data from the NanoString nCounter assay (previously reported custom designed 800-gene panel tailored for breast cancer[12]) were analyzed after quality control filtering and normalization using the R package NanoStringQCPro (version As recommended, the mean of the positive probes and the housekeeping genes were used in order to account for differences in sample input. Moreover, negative control probes were used for background correction, and samples that did not pass quality control were removed. Raw and normalized NanoString data have been deposited in the Gene Expression Omnibus database under accession number GSE109710.

We calculated the expression of the ERBB2 and ESR1, genes that have been shown to be associated with pCR after neoadjuvant chemotherapy and anti-HER2 agents,[4,13] and the expression of well-known genes associated with CD8+ T-cell cytolytic activity: granzyme A and B (GZMA and GZMB) and perforin 1 (PRF1), with T-regulatory cells: forkhead box P3 (FOXP3), with immune checkpoint inhibition: cytotoxic T-lymphocyte-associated protein 4 (CTLA4), programmed death–ligand 1 (PDL1) and programmed cell death protein 1 (PD1), with T-helper phenotype 1: interferon-gamma (IFNG) and 2: interleukin 4 (IL4). Moreover, we computed three immune-gene signatures, a STAT1-related gene signature (immune_STAT1),[14] a signature related to the presence of tertiary lymphoid structures (immune_TLS),[15] and an immune-gene signature associated with benefit from adjuvant trastuzumab in the North Central Cancer Treatment Group N9831 adjuvant trastuzumab trial (immune_trastuzumab).[16] Details about single gene/signature scores calculations and PAM50 subtype definition are provided in the Supplementary Methods (available online).

Statistical Analyses

The association between TIL percentage, at baseline and at residual disease or surgery, with clinicopathologic variables was evaluated using linear regression. Clinicopathological variables included age (<50 and ≥50 years), histology grade (III vs I/II), estrogen receptor (ER) status (positive vs negative), clinical stage (III vs II), and chemotherapy (anthracyclines vs nonanthracyclines). The dependence of pCR, defined as absence of invasive tumor in the breast and lymph nodes (ypT0/is ypN0), on TIL percentage at baseline and clinicopathologic variables was assessed using logistic regression models.

The presence of nonlinear effects of TIL percentage at baseline on pCR probability on the logit scale, modeled by restricted cubic splines, was evaluated by Wald test. pCR probabilities according to TIL quartiles at baseline were also estimated, with 95% confidence intervals (CIs) computed by the Wilson method. EFS was defined as the time from the date of surgery to the first documentation of disease progression or death. Any evidence of contralateral disease in situ was not considered disease progression. EFS was described separately in patients who achieved a pCR from those who did not. Patients who were withdrawn from the study without documented disease progression were censored at the date of the last assessment when the patient was known to be disease-free. Median follow-up was obtained by the reverse Kaplan-Meier method. Association of EFS with TIL percentage and clinicopathological variables was examined using the Cox regression model. All tests of statistical significance were two-sided, and a P value of less than .05 was considered statistically significant.

Statistical analysis of gene expression data generated using the NanoString nCounter Assay is described in the Supplementary Methods (available online). This study has been reported according to the Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK) criteria.[17]