This study investigated CCND1 amplification in a cohort of patients with early breast cancer, focusing on HR-positive, HER2-negative patients, and the correlation of amplification with outcome and clinicopathologic parameters. The amplification rate in this study was found among unselected breast cancer patients, a figure that is consistent with previously published amplification frequencies ranging between 10 and 15% [29,30,31]. Although it was a recurring alteration, no statistically significant correlation was observed between CCND1 amplification status in HR-positive, HER2-negative breast cancer patients and their prognosis. Various studies have reported this lack of correlation [32,33,34,35]. Like others, we have shown that CCND1 amplification status does not add further information about DFS or OS to the established biomarkers of tumor size — lymph-node involvement, and grading [36,37,38].

Other authors, however, have reported a prognostic value for CCND1 status. Lundgren et al. noted a higher rate of recurrence in CCND1-amplified tumors independent of T, N, grading, and Ki-67 expression (hazard ratio 1.60; 95% CI, 1.08 to 2.41; P = 0.03), while Bostner et al. (hazard ratio 1.59; 95% CI, 0.96 to 2.64; P = 0.0093), reported an even clearer connection between CCND1 amplification and a higher risk of recurrence in an ER-positive subset of patients [33, 39]. Lundberg et al. reported a poorer 15-year survival in ER-positive breast cancer patients with CCND1 amplifications in the Cox proportional hazards model [40]. Other studies have suggested that CCDN1 amplification may be of greater significance among ER-positive patients [37, 41].

In the present study, CCND1 amplification was lower among TNBC patients and higher in HER2-positive and HR-positive patients. These figures correspond to the literature findings. A greater frequency of CCND1 amplifications is observed in ER-positive subsets of breast cancer patients, and a shorter DFS for the ER-positive subset of CCND1-amplified patients has been reported [31, 34, 38, 39, 41, 42]. In 2019, Lundberg et al. reported that CCND1 amplifications—especially in ER-positive, HER2-negative patients without lymph-node involvement—correlated with poor OS [40]. The present study did not find any statistically significant difference in the DFS or OS over a 10-year follow-up period for HR-positive, HER2-negative patients. There may be an impact in different follow-up length, as Lundberg et al. reported a poorer survival after 15 years [40]. But their Kaplan–Meier estimates suggest that survival differences were evident even at much earlier time points. The existing literature indeed presents a heterogeneous landscape regarding the prognostic impact of CCND1 amplification and at what timepoint it occures. For example, Holm et al. reported worse 10-year overall survival in CCND1-amplified luminal A tumors, consistent with the findings of Chin et al., who observed a similar trend with a median follow-up of 6.6 years [16, 43]. Additionally, Bieche et al. demonstrated decreased relapse-free survival within a 10-year period [35]. Conversely, Valla et al. and others found no significant difference in survival at 10 years [32]. Given these mixed findings, further investigation is necessary, particularly in low-risk, HR-positive, node-negative cohorts, to determine whether specific factors can reliably predict late tumor recurrence.

Since CCND1 appears to play a more important role among HR-positive patients, there has been discussion on whether earlier relapse may be caused by a diminished therapeutic response to tamoxifen. It is still not clear whether cyclin D1 protein overexpression, rather than CCND1 status, could be used to predict the therapeutic response to tamoxifen, and conflicting results on this have been published [29, 36, 44]. While Quintayo et al. reported that the prognosis is poorer in the presence of CCND1 amplifications, irrespective of cyclin D1 protein levels, Jirström and coworkers found that the tamoxifen response is influenced by both amplification status and the level of cyclin D1 expression [36].

Endocrine resistance and relapse during treatment with CDK 4/6 inhibitor therapy is the main focus of research for HR-positive patients [45]. Different mechanisms of resistance have been discussed, such as RB1 mutations and loss of function [46], up-regulation of phosphorylated 3-phosphoinositide-dependent protein kinase 1 (PDK1), which was observed in ribociclib-resistant cell lines [47], or hyperactivity of cyclin A/CDK2, in which cells were found to be able to escape CDK4 loss [48]. Molecular mechanisms would also explain the poorer outcome in CCND1-amplified/cyclin D1-overexpressing tumors, leading to increased cyclin D1 expression and inappropriate cyclin D-CDK 4/6 activity, and suggesting that CCND1 amplification could serve as a predictive biomarker [49, 50]. Data from the PALOMA-1 trial, in cohort two of which only patients with amplifications of cyclin D1 or loss of p16 or both were included, showed no additional benefit for palbociclib in patients with CCND1-amplified tumors. Recruitment of this cohort was therefore stopped after an interim analysis [51].

CCND1 amplification rate in this trial was 12.9% which found by protein expression of cyclin D1 assessed bei FISH analysis. This is according to the literature where in one of the largest examined patient cohorts in the Arimidex, Tamoxifen, Alone or in Combination (ATAC) trial where over 1000 patients where assessed by chromogenic in situ hybridisation (CISH) and immunohistochemistry the ampflification rate was 8.7% [33]. The largest study using SNPs to find CCND1 amplification reported 22–35% amplification rate [40]. It seems that the method does not correlated to survival as positive and negative results were reported [33, 40, 52, 53]. Nevertheless mRNA levels might provide a more accurate prediction of prognosis in breast cancer patients concerning CCND1 amplification and this should be further examined.

However, further trials are needed in this field, since patients with luminal B-like tumors in the present study showed an amplification rate of 17.3%. Some data suggest that resistance to CDK 4/6 inhibitors is associated with strong up-regulation of CCND1 protein levels [54]. A phase 1/phase 2 clinical trial (NCT03901469) examined the combination of a small molecule inhibitor, ZEN-3694, in combination with either palbociclib or abemaciclib to overcome endocrine resistance. The combination was shown to reverse CDK6 and CCND1 protein levels [54]. No further biomarker analyses have yet been published from CDK 4/6 inhibitor trials; for example, the Natalee trial (NCT03701334) examining ribociclib for adjuvant breast cancer patients will assess CCND1 as one of the biomarkers. The present study found that in the HER2-positive cohort, 22% of tumors showed CCND1 amplification, so it may be useful to examine CCND1 in further trials with triple-positive patients in whom HER2 antibodies are administered in combination with CDK4/6 inhibitors, and to evaluate whether HER2-low tumors also show high CCND1 amplification rates.

There are no evaluated suggested cutoff values for examining CCND1 amplification using FISH analysis. CCND1 amplification was therefore defined as a CCND1 probe to CEP11 ratio of 2.0 or greater—the same ratio recommended by the U.S. Food and Drug Administration for HER2 amplification diagnosis for breast cancer [26]. Lundgren and coworkers used single nucleotide polymorphism (SNP) arrays to identify amplified versus nonamplified tumors [40]. Jirström et al. also used FISH to evaluate the CCND1 gene, and it was scored as amplified when the ratio was greater than 1 in at least 20% of tumor cells [44]. Further research is necessary to standardize cutoff values for distinguishing between amplified and nonamplified tumors. Differences in preanalytical and analytical methods with regard to the antibody clones and reading of the immunohistochemistry stains may lead to different results. There are also biological limitations resulting from the use of a tissue microarray, and limitations due to the small sample size of the HR-positive, HER2-negative cohort, in which 378 patients were examined. In addition, CCND1, which can be expressed heterogeneously, may be differently interpreted, leading to false-negative results. In addition, the study was conducted on archived tumor tissue and was performed retrospectively.