Tumour-specific HMG-CoAR is an independent predictor of recurrence free survival in epithelial ovarian cancer

Abstract

Background: Our group previously reported that tumour-specific expression of the rate-limiting enzyme in the mevalonate pathway, 3-hydroxy-3-methylglutharyl-coenzyme A reductase (HMG-CoAR) is associated with more favourable tumour parameters and a good prognosis in breast cancer. In the present study, the prognostic value of HMG-CoAR expression was examined in tumours from a cohort of patients with primary epithelial ovarian cancer.

Methods: HMG-CoAR expression was assessed using immunohistochemistry (IHC) on tissue microarrays (TMA) consisting of 76 ovarian cancer cases, analysed using automated algorithms to develop a quantitative scoring model. Kaplan Meier analysis and Cox proportional hazards modelling were used to estimate the risk of recurrence free survival (RFS).

Results: Seventy-two tumours were suitable for analysis. Cytoplasmic HMG-CoAR expression was present in 65% (n = 46) of tumours. No relationship was seen between HMG-CoAR and age, histological subtype, grade, disease stage, estrogen receptor or Ki-67 status. Patients with tumours expressing HMG-CoAR had a significantly prolonged RFS (p = 0.012). Multivariate Cox regression analysis revealed that HMG-CoAR expression was an independent predictor of improved RFS (RR = 0.49, 95% CI (0.25-0.93); p = 0.03) when adjusted for established prognostic factors such as residual disease, tumour stage and grade.

Conclusion: HMG-CoAR expression is an independent predictor of prolonged RFS in primary ovarian cancer. As HMG-CoAR inhibitors, also known as statins, have demonstrated anti-neoplastic effects in vitro, further studies are required to evaluate HMG-CoAR expression as a surrogate marker of response to statin treatment, especially in conjunction with current chemotherapeutic regimens.

Figure 1

HMG-CoAR Expression in EOC. Examples of immunohistochemical HMG-CoAR staining in EOC with negative, intermediate and strong cytoplasmic expression (5× and 20× magnification) (A). Areas of membranous expression (B) and granular staining (C) were also seen (20× magnification). HMG-CoAR expression was also evident in normal fallopian tube (D) and normal ovarian surface epithelium (E) (20× magnification).

Figure 2

HMG-CoAR is Associated with Prolonged RFS in EOC. Kaplan Meier analysis of manually assessed HMG-CoAR cytoplasmic intensity revealed a trend towards an improved RFS (A) and OS (B). Dichotomization of data as positive versus negative revealed that HMG-CoAR was associated with an improved RFS (C) but not an improved OS (D).

Figure 3

Automated Analysis of HMG-CoAR Protein Expression. Using Genie pattern recognitiion software, tumour and stroma were identified and tumour specific HMG-CoAR was quantified using a colour deconvolution algorithm. The images shown are IHC and mark-up images, markups show different levels of HMG-CoAR as described by the colour coded legend.

Figure 4

HMG-CoAR Autoscore is Associated with an Improved RFS. There was an excellent correlation between automated and manual cytoplasmic intensity (A). A HMG-CoAR autoscore was calculated by combining cytoplasmic intensity and the percentage of positive tumour cells. The distribution of the HMG-CoAR autoscore is illustrated in the histogram (B). Using a threshold of the 25th percentile, an increased HMG-CoAR autoscore was associated with a prolonged RFS (C) but not OS (D).