KIBRA; a novel biomarker predicting recurrence free survival of breast cancer patients receiving adjuvant therapy

Tissue blocks were first examined for its physical suitability. The histopathologists reviewed the Haematoxylin and Eosin (H&E) stained slides of each case and made a circle on the slide localizing a representative area in the invasive tumour-front with minimum fixation artifacts. The H&E slides were superimposed on the donor block to identify the area in the tissue block to be punched. From each of these donor blocks, a core of 2 mm diameter breast cancer tissue was extracted using TMA builder™ (Thermo Fisher). The cores were transposed into the recipient TMA wax mold prepared previously which contained 24 pits to hold 23 breast cancer cores. A core of brain tissue was transposed into the 24th pit in the mold as a guide to identify the rows and the columns of the TMA. A map for each TMA block was designed as a grid to link the biomarker score to the clinico-pathological data of each case. All IHC stained slides were assessed by the first author eliminating inter-observer variation and she was blinded to the prognostic features of each breast cancer.

ER α clone 1D5 (Dako-M7047), PR (Dako-M3569) and Her2 (Dako-A0485) had been used with the secondary antibody (Dako Real EnVision™) for IHC staining of all breast cancers. Anti KIBRA antibody (Abcam-ab107637) was used in 1 in 150 dilution as per manufacturer’s instructions with the same secondary antibody to detect the expression of KIBRA. Different dilutions were tried before selecting 1 in 150 dilution for KIBRA. Ki67 (Dako M7240) in 1/75 dilution, anti Claudin3 (Abcam-ab15102) in 1/150 dilution, CK 5/6(Dako M7237) in 1/50 dilution and EGFR (Dako M3563) in 1/100 dilution were used for the corresponding markers. EGFR antigen retrieval was done using proteinase. For Claudin3 and CK5/6, microwave antigen retrieval with pH 9 Tris buffered saline was done at 1100 W for 20 min following preheating. Citrate buffer at pH 6 was used for antigen retrieval by pressure cooking for 7 min after pre-heating, for the rest of the antibodies. IHC staining was done manually with a positive control.

The Allred score for ER and PgR and UK recommendations for HER2 were used in the assessment of staining, on the original whole sections [6]. A score of ≤ 2 for ER and PgR and a score of 0 or + 1 for Her2 were considered the criterion for categorizing triple negative breast cancer (TNBC). ER and PgR were considered positive when the Allred score for each was ≥ 3. HER2 was considered positive when the score was 3+. Patients with HER2 equivocal expression (2+) were excluded when insitu hybridization results were not available.

KIBRA expression was scored as; no staining = 0, weak staining = 1, moderate staining = 2 and strong staining = 3, separately for cytoplasmic and nuclear expression. Intensity of staining in normal breast epithelial cells were considered score 2 and other scores were assigned accordingly. A score < 2 was considered KIBRAN-low or KIBRAC-low depending on the cellular location; nucleus or cytoplasm respectively. Breast cancers with a score of < 2 for both cytoplasmic and nuclear staining were considered overall Kibra-low. Claudin3 was also assessed with same scoring but the membrane staining also was assessed similarly. Therefore Claudin3–low breast cancer was defined as < 2 score for all three cellular levels; nuclear, cytoplasmic and membrane. All breast cancers were classified into the molecular subtypes using the IHC surrogates for molecular classification [7]. TNBCs with positivity for at least one of the basal markers (CK5/6 and EGFR) were considered basal-like breast cancers. Since there is no consensus on the cut off for Ki67, we analysed the recurrence free survival (RFS) of the cohort against different cut off levels of Ki67 (≥ 14%, 20 and 25%) published in the literature before selecting the cutoff.

Patients whose breast cancer tissue could be included in TMA construction were enrolled for the study. Mean follow-up time was 38.3 (SD ± 23.4) months. The actual minimum follow-up period was 12 months (81.1% -24 months, 56.4%-36 months, 35.1%-48 months and 24%- ≥ 5 years).

Recurrence free survival (RFS) time was calculated from the date of surgery/first therapeutic intervention to one of the following events; first loco-regional and/or first distant recurrence. [8] Radiological and histopathological evidence were used to confirm the recurrence. The date on which the said investigation done was considered the date of recurrence. Patients who did not experience a recurrence or death were censored at the last follow-up [8].

The Pearson chi-square test was used to determine the association between the expression of KIBRA at each of the two cellular locations with the clinico-pathological features. Kaplan-Meier model was used to estimate the RFS and the log-rank test was used to compare the survival of different groups. The Kaplan-Meier model for univariate and Cox-regression model with backward stepwise factor retention method for multivariate analyses were used to estimate the predictors of survival.

IHC for KIBRA expression was first done on routine breast cancer tissue sections and on normal breast glandular tissue, before TMAs were stained. Both normal and lactating breast tissue showed nuclear as well as cytoplasmic staining. [Fig. 1a, b] KIBRA expression in breast cancers detected by IHC is given in Fig. 2. Details of KIBRA expression in the study cohort (909/1124 patients) is given in the Table 1.

Positive KIBRA nuclear expression (KIBRAN) (score 2 or 3) was significantly associated with low recurrence rate (p = 0.014), low lymph node stage (0 or 1) (p = 0.006), positive ER expression (p < 0.001), positive PgR expression (p < 0.001) and Ki67 positivity in ≥14% of cancer cells (p < 0.001). Distribution of KIBRAN positive tumours within each molecular subtype was also statistically significant (p = 0.002). Cytoplasmic expression of KIBRA (KIBRAC) was associated with high grade tumours (p = 0.003) and HER2 over-expression (p = 0.004). KIBRAC was also associated with Ki67 positivity in ≥ 14% of cancer cells (p < 0.001). Prevalence of overall KIBRA-low tumours was most significantly present in TNBC (50%) and basal-like subgroups (39.3%) (p = 0.028).

The overall KIBRA-low expression was significantly associated with negative ER expression (p = 0.005), negative PgR expression (p = 0.001) and Ki67 positivity in < 14% of cancer cells (p < 0.001). Overall KIBRA-low expression was present in 63.8% of Claudin-low breast cancers (p < 0.001). Recurrences developed in 32.3% of the overall KIBRA-low breast cancers which were Claudin-low as well (p = 0.019). KIBRA expression had no association with age, TNM stage [10] and lympho-vascular invasion.

Out of all patients who had chemotherapy, 44.2% had overall KIBRA-low breast cancers while 41.3% of those who had hormone therapy also had overall KIBRA-low expression. Recurrences were seen in 20.7% of those who had chemotherapy (p = 0.038) and 16.7% of patients who had endocrine therapy (p = 0.012). Therefore, survival analysis of patients who had these adjuvant therapies was also done, in relation to the low KIBRA expression.

The survival analysis included 525 patients who had follow-up details on recurrence and KIBRA expression data. These 525 patients had 124 events (loco-regional recurrence or distant metastasis). The second and subsequent recurrences of a patient were not included. The estimated median RFS time of the study cohort was 102.00 months (SE 11.45; 95% CI 79.56–124.44). Five year RFS of the cohort was 70.5%. There was no statistically significant RFS difference due to the expression of KIBRAN (p = 0.052) or KIBRAC (p = 0.937) alone. However, the overall Kibra-low breast cancer patients had a RFS disadvantage compared to KIBRA positive breast cancers (p = 0.037) (Fig. 3a). Since KIBRAN expression associated with good prognostic features and KIBRAC associated with poor prognostic features as described above, the study cohort was divided into four groups according to the possible combinations of staining; score of 2 or 3 for KIBRAN alone, KIBRAC alone and KIBRAN with KIBRAC were the three positive groups. (Fig. 2a, b) A score of 0 or 1 for both cytoplasm and nucleus was included in the 4th group (Table 1). When RFS of patients according to the four possible staining patterns were compared, KIBRAN alone had the best prognosis and overall low staining for KIBRA had the worst prognosis. (Fig. 3b) However, the difference between the survival curves of the four types did not reach statistical significance (p = 0.05). Therefore, we considered it was best to proceed with the overall low KIBRA expression for the rest of the analysis.

To find out whether the disadvantage of overall low KIBRA expression on RFS is retained within the molecular subgroups, RFS was analysed according to the molecular subtypes. The molecular subtype composition of the cohort is given in the Table 1. Since basal-like breast cancers comprised only 8.1% of the cohort, it was combined with the rest of TNBCs for the univariate and multivariate analysis. Similarly the two luminal B subtypes were also grouped together as luminal B. Overall low KIBRA expression had no effect on the RFS of patients in luminal A (p = 0.069), luminal B (p = 0.073), TNBC (p = 0.420) and HER2-enriched (p = 0.531) subtypes. (Table 2).

Although there was no significant effect on the RFS of the above subgroups, RFS of all luminal breast cancers (all ER positive) was affected by overall low KIBRA expression as it imparted a RFS disadvantage. (Fig. 4a) No such effect on RFS was observed in the group of ER negative patients (p > 0.05). RFS analysis was repeated on the intermediate and high risk categories separately, but there was no significant effect on the RFS of either category due to the overall low expression of KIBRA (p > 0.05).

Ki67 expression at cut off of 14% (p = 0.04), 20% (p = 0.042) and 25% (p = 0.042) of the cells; all had a significant RFS difference and therefore the lowest, 14% was selected as the cut off for molecular classification for the current study as there is no consensus on this in the literature.

Out of the 525 patients who had ER expression, there were 183 who fulfilled the following three criteria; had ER positive breast cancer, had endocrine therapy and had follow-up details on recurrences.

Out of the 183 patients, 25 experienced an event (loco-regional recurrences and distant metastasis). Patients who had overall KIBRA-low expression demonstrated a RFS disadvantage compared to those who had overall positive KIBRA expression (p = 0.020) (Fig. 4b). Patients who did not receive endocrine therapy, did not show a survival difference due to the expression of KIBRA (p = 0.645).

There were 484 patients who had chemotherapy and whose follow up details were available. Out of them, 119 had events (loco-regional recurrences and distant metastasis). The survival effect of the overall expression of KIBRA was analysed. Patients who received chemotherapy and had overall KIBRA-low expression demonstrated a RFS disadvantage compared to those who had overall positive KIBRA expression (p = 0.018) (Fig. 4c). Patients who did not receive chemotherapy, probably not indicated, did not show a survival difference due to the expression of KIBRA (p = 0.211).

Univariate and subsequently multivariate analyses were done to identify the factors with an independent effect on the RFS of patients in different molecular subtypes and patients who received the two treatment modalities (Tables 2, 3, 4 and 5).