The immunologic constant of rejection classification refines the prognostic value of conventional prognostic signatures in breast cancer

Abstract

Background: The immunologic constant of rejection (ICR) is a broad phenomenon of Th-1 immunity-mediated, tissue-specific destruction.

Methods: We tested the prognostic value of a 20-gene ICR expression signature in 8766 early breast cancers.

Results: Thirty-three percent of tumours were ICR1, 29% ICR2, 23% ICR3, and 15% ICR4. In univariate analysis, ICR4 was associated with a 36% reduction in risk of metastatic relapse when compared with ICR1-3 (p = 2.30E-03). In multivariate analysis including notably the three major prognostic signatures (Recurrence score, 70-gene signature, ROR-P), ICR was the strongest predictive variable (p = 9.80E-04). ICR showed no prognostic value in the HR+/HER2- subtype, but prognostic value in the HER2+ and TN subtypes. Furthermore, in each molecular subtype and among the tumours defined as high risk by the three prognostic signatures, ICR4 patients had a 41-75% reduction in risk of relapse as compared with ICR1-3 patients. ICR added significant prognostic information to that provided by the clinico-genomic models in the overall population and in each molecular subtype. ICR4 was independently associated with achievement of pathological complete response to neoadjuvant chemotherapy (p = 2.97E-04).

Conclusion: ICR signature adds prognostic information to that of current proliferation-based signatures, with which it could be integrated to improve patients' stratification and guide adjuvant treatment.

Fig. 1

Correlations of ICR classes with immunity-related parameters and prognostic gene expression signatures of breast cancer. For each ICR class, are indicated the percentage of samples with high lymphocyte infiltrate score samples (a), metagene expression scores of T cells (b), CD8+ cells (c), and B cells (d), activation score of IFNα (e), IFNγ (f), and TNFα (g) pathways, percentage of high-risk samples according to the 70-gene signature (h), the Recurrence score (i), and the Risk of Relapse (ROR-P) score (j), activation score of P53 pathway (k), and percentage of high-risk samples according to the chromosomal instability signature (l). The p-values are indicated (Fisher’s exact test or ANOVA test when appropriate)

Fig. 2

Metastasis-free survival according to the ICR classification in breast cancer. a Kaplan–Meier MFS curves in all patients according to the four ICR classes. b Similar to (a), but after pooling the ICR1, 2 and 3 classes in the ICR1-3 class. c-e Similar to (b), but in the HR+/HER2− subtype (c), the HER2+ subtype (d, and the TN subtype (e). The p-values are indicated (log-rank test)

Fig. 3

Metastasis-free survival according to the ICR classification within the different risk groups defined by three proliferation-based prognostic signatures in the 2131 HR+/HER2− patients. Kaplan–Meier MFS curves in patients according to the ICR1-3 and the ICR4 classes in the HR+/HER2− subtype and according to the relapse risk groups defined by the 70-gene signature (a), the Recurrence score (b), and the ROR-P score (c). d– f Similar to (a–c), respectively, but the low-risk and intermediate-risk curves are not stratified according to ICR and are pooled for (e) and (f). The p-values are indicated (log-rank test)