High tumor expression of CTLA4 identifies lymph node-negative basal-like breast cancer patients with excellent prognosis

Abstract

Background: Tumor immune cell infiltration is a favorable prognostic factor in triple-negative breast cancer. Most triple-negative tumors belong to the aggressive basal-like subtype. We hypothesized that immune gene expression may identify low-risk patients for whom adjuvant chemotherapy can be de-escalated.

Methods: The expression of 753 immune-related genes was analyzed in tumor biopsies from 45 patients with basal-like disease and no lymph node metastases (Oslo1 cohort) and evaluated for prognostic value. Findings were validated in two independent cohorts. Oslo1 biopsies were also analyzed for tumor-infiltrating lymphocytes (TIL) and tertiary lymphoid structures (TLS).

Results: Here we show that a high expression of CTLA4 (above 63^rd percentile) is associated with an excellent prognosis in the Oslo1 cohort. None of the patients in the CTLA4^high group suffered disease recurrence (median follow-up 7.4 years) or breast cancer-related death (median follow-up 17.7 years). Analysis of the SCAN-B (n = 233; 97% without distant recurrence in CTLA4^high group) and METABRIC cohorts (n = 155; 93% disease-specific survival in CTLA4^high group) validates this finding, which also applies to patients who did not receive chemotherapy. CTLA4 expression correlates with TIL score and TLS levels (Oslo1 cohort), but no TIL^low/CTLA4^high patients died from breast cancer, suggesting that the CTLA4 readout identifies low-risk patients not captured by TIL assessment.

Conclusions: A high primary tumor expression of CTLA4 identifies patients with an excellent prognosis, for whom standard chemotherapy may be de-escalated or omitted.

Fig. 1. Patient disposition.

Selection criteria for patients/samples included in the study for each cohort; Oslo1 (a), SCAN-B (b), and METABRIC (c).

Fig. 2. Oslo1: differential gene expression by systemic recurrence.

Boxplot of scaled log₂-transformed normalized gene expression in patients (n = 45) from the Oslo1 cohort with or without subsequent systemic recurrence of breast cancer. P values were calculated by Wilcoxon’s rank sum test. In the boxplots, the center line represents the median value, the hinges represent the interquartile range (IQR). Whiskers extend to extreme values, omitting outliers >1.5 × IQR from the box limits.

Fig. 3. Kaplan-Meier analysis of disease-specific survival in patients with gene expression above and below median.

Data from the Oslo1 cohort (n = 45). P values were calculated using the log-rank method.

Fig. 4. Oslo1: Survival and disease recurrence by CTLA4 expression.

Overall survival, recurrence-free interval, and disease-specific survival in Oslo1 patients with CTLA4 expression above and below the cutoff at the 63rd percentile.

Fig. 5. SCAN-B: survival and disease recurrence by CTLA4 expression.

5-year overall survival, recurrence-free interval, and distant recurrence-free interval in patients with CTLA4 expression above and below the cutoffs at the 63rd ((a–c); derived from Oslo1) and the 60th ((d–f); derived from SCAN-B) percentile in the SCAN-B cohort.

Fig. 6. METABRIC: Survival by CTLA4 expression.

Disease-specific and overall survival with CTLA4 expression above and below the cutoff value at the 63rd (a, b) and 60th (c, d) percentile in the METABRIC dataset.

Fig. 7. Outcomes in patients with and without (neo-) adjuvant chemotherapy.

5-year outcomes by CTLA4 expression in patients with (a, c, e) and without (b, d, f) chemotherapy in the Oslo1, SCAN-B, and METABRIC cohorts. The cutoff for high CTLA4 expression was set at the 63rd percentile within each cohort. Disease-specific survival (DSS) was used as outcome measure for Oslo1 and METABRIC, while distant recurrence-free survival (DRFI) was used for SCAN-B, as cause of death was not available for this cohort.

Fig. 8. CTLA4 expression versus TIL, TLS, and GC score.

a Correlation plot of tumor-infiltrating lymphocytes (TIL) vs CTLA4. b CTLA4 expression in samples grouped by prevalence of tertiary lymphoid structures (TLS). c CTLA4 expression in samples grouped by prevalence of germinal centers (GC). Data from the Oslo1 cohort (basal-like N0). All patients with biopsies available for evaluation were included in this analysis. n = 44 for TIL and TLS scores, n = 42 for GC score. Normalized CTLA4 counts were log₂ transformed and scaled to a mean of 0 and a standard deviation of 1. TILs were scored according to recommendations by the International TILs Working Group (2014). In the box plots, the center lines represent median values, hinges the IQR, and whiskers the extreme values, omitting outliers extending >1.5 × IQR from the hinges. In boxplots, P values were calculated using the Wilcoxon rank-sum test. CTLA4, cytotoxic T-lymphocyte-associated protein 4; R, Pearson correlation coefficient.

Fig. 9. Outcomes by CTLA4 expression in triple-negative breast cancer.

Survival/recurrence rates in patients with high and low CTLA4 expression and triple-negative breast cancer (regardless of molecular subtype) in Oslo1 (a–c) and SCAN-B (d–f). For each cohort, the cutoff for high CTLA4 expression was set at the 63rd percentile of gene expression in basal-like samples in the same cohort.