Age at diagnosis turned out to be the decisive prognostic factor in our material of TNBCs. We observed the most favorable outcome among patients aged 49–57 years. As compared to this age group, the risk of breast cancer-specific mortality was increased 3-fold among the younger patients (Table 3). Overrepresentation of aggressive course of disease among young TNBC patients is also reported in previous literature [26–29], especially among very young patients aged < 35 years at diagnosis [30–31]. The prognostic role of age in TNBC has not been settled [32–38], and a considerable part of the research is controversial [39–42]. Among patients older than 57 years at diagnosis, our results indicated a 4-fold increased risk of breast cancer-specific mortality (Table 3), and the risk was further accentuated among the eldest patients (aged > 73) (Fig. 3). Likewise, the majority of previous research associates increasing age with decreasing survival in TNBC [37, 43–45], with the highest reported mortality among the oldest old [46]. The survival deficit of elderly patients has been accounted for the lack of organized screening in the age group [47], as well as comorbidities [27, 48], and the consequent less intensive treatments [49]. Also, it has been suggested that the variation of outcomes in different age groups may partly be explained by the distribution of biological TNBC subtypes [37, 44, 50–51].
Basal-like subtype is overrepresented among young TNBC patients and characterized by BRCA mutations, drug resistance, and a more aggressive course of disease than the other TNBC subtypes [37, 52–54]. Our findings (Fig. 3) are in line with the majority of previous research reporting higher survival and lower recurrence rates for basal-like than non-basal-like TNBC [42, 50]. However, the prognostic role of basal differentiation in TNBC is not completely settled [52, 55–56]. In our material, basal differentiation of TNBC was a statistically significant prognostic feature only among the youngest patient subgroup (women aged < 49), predicting more than 4-fold increased risk of TNBC-specific mortality (Table 3). Apart from previous literature [50], in our material basal differentiation did not show prognostic significance among the older patients (Fig. 3). Heterogeneity of TNBC has been presented as an explanation for the observed survival differences between age groups [3, 57–58]. Among the young, aggressive course of disease has been explained by the lower proliferation rates in basal-like TNBC predisposing chemotherapy insensitivity [50], whereas among the older patients, alterations in tumor microenvironment have been addressed [32].
Regarding the established prognostic factors of breast cancer, in our material tumor size indicated up to 2.2-fold mortality risk in age-dependent multivariate analysis (Table 3, Fig. 3). Ample evidence from previous literature indicates the independent prognostic role of tumor size in TNBC [30, 59–60], but the impact of tumor size in different age groups has not been settled. In our material, tumor size was a significant predictor of survival only among the subgroups of patients > 57, but not > 73 years at diagnosis. This finding may be explained by the Finnish national breast cancer screening program [61]. As a consequence, a significant improvement of breast cancer survival, particularly among TNBC patients, has been reported [62].
According to general understanding, TNBC has a reduced probability of lymph node involvement, due to its tendency to spread hematogenously [63]. In our material, nodal status alone did not predict disease survival (Table 3). However, after standardizing for age at diagnosis, axillary metastasis proved to be an independent prognostic factor in the whole material and among patients aged > 57 years, indicating 2.0- and 2.5-fold increased mortality, respectively. Among younger patients, instead, age was such a strong prognostic factor that it overrode the impact of nodal status. In previous literature, no consensus reigns on the prognostic impact of nodal status in TNBC. Part of previous research considers axillary metastases as an adverse clinical feature [64–65], especially for short-term survival [66]. The association between nodal status and outcome of TNBC, however, appears not to be straightforward [66–67], while even extensive metastatic disease has not been observed to impair survival [68]. The discrepant and confusing association between nodal status and disease outcome have been linked to the genetic and clinical heterogeneity of TNBC [66].
Defects in cell cycle regulation and DNA repair mechanisms are common characteristics of TNBC, and enhanced cell cycle is recognized as a specific cellular feature of young patients [64, 69]. This is also reflected in proliferation indices, such as immunohistochemically detected Ki-67 labeling index, which has been reported to decrease with increasing age [69]. Similar to our present findings, the majority of researchers have associated high Ki-67 with unfavorable disease outcome among TNBC patients [9, 60, 70–71]. In our material, however, Ki-67 was an independent prognosticator only among patients aged > 57 years for whom ech 10 % increase in Ki-67 indicated 2.9-fold increased risk of mortality. As previously reported [72], Ki-67 was not a clinically applicable prognostic factor among the youngest patient subgroup in our material, possibly reflecting the age-dependent distribution of different genetic and transcriptional phenotypes in TNBC [27, 32, 69]. However, the clinical applicability of Ki-67 in predicting the outcome of TNBC is impaired by the varying criteria for high vs low labeling index with cut-points rangig from 14% [1] to 25 %, and een up to 60 % [11, 72–73]. Moreover, poor understanding of the functions and dynamics of Ki-67 protein during the cell cycle hamper detailed prognostic conclusions [74].
Previous literature has identified various cell cycle-regulating proteins with expression peaks at specific steps of cell division. Due to their distinct expression patterns in cell cycle, these proteins have been suggested with advantages over Ki-67 in achieving more detailed prognostic information [6]. Previously, based on invasive breast carcinomas of all intrinsic subtypes, we have reported 8.4-fold increased risk of mortality associated with the combination of separase, securin and cdk1, each a crucial regulator of cell cycle progression [75]. Contradictory to our previous results, apart from geminin, all the tested novel biomarkers for cell proliferation failed to show statistically significant prognostic associations in our present material of TNBCs.
Geminin, a DNA replication inhibitor, is activated during cell cycle progression by the anaphase-promotion complex (APC) leading to initiation of sister chromatid separation [76]. The specific role of geminin is to limit DNA replication and trigger sister chromatid separation and, therefore, it is considered to qualify exceptionally well as a cell cycle-specific proliferation marker [77–79]. In previous literature, loss of geminin expression has been observed to predict poor survival in breast cancer [80–82] and aggressive course of disease in the subgroup of TNBCs [83–85]. In our material, geminin identified TNBC patients with increased mortality among subgroups generally associated with favorable outcome, i.e. patients aged > 57 with node-negative disease and small tumor size. Consistent with previous literature, adverse outcome in TNBC was associated with increased geminin expression but the association was not linear. Emerging evidence emphasizes the multifaceted role of geminin in malignancy [77, 86]. Based on the current understanding, geminin overexpression arrests cell proliferation whereas loss of geminin may generate re-replication, predisposing cancer cells to aneuploidy and drug-resistance [80, 87].
In summary, our results emphasize the nature of TNBC as a heterogeneous disease with varying courses of disease in different age groups. Traditional clinical features do not provide optimal prognostic characterization for all TNBC patients, and young age should be considered as an additional adverse prognostic feature. Contrary to other breast cancer subtypes, proliferation does not explain for the survival differences in all age-subgroups of TNBC patients. Our findings provide additional support for the hypothesis that young TNBC patients comprise a unique disease entity, while elderly patients represent a more coherent subgroup which, in this therapeutically challenging disease, may indicate potential regarding future targeted treatments. While the explanation for the survival deficit between different age groups in TNBC still remains unsettled, it is evident that TNBC diagnosed at a young age warrants individual therapeutic considerations, especially concerning the increasing number of patients receiving neoadjuvant treatment.