The Target for Statins,HMG-CoA Reductase,Is Expressed in Ductal Carcinoma-In Situ and May Predict Patient Response to Radiotherapy

Background

Patients with ductal carcinoma-in-situ (DCIS) are currently not prescribed adjuvant systemic treatment after surgery and radiotherapy. Prediction of DCIS patients who would benefit from radiotherapy is warranted. Statins have been suggested to exert radio-sensitizing effects. The target for cholesterol-lowering statins is HMG-CoA reductase (HMGCR), the rate-limiting enzyme in the mevalonate pathway. The aim of this study was to examine HMGCR expression in DCIS and study its treatment predictive value.

Methods

A population-based cohort including 458 women diagnosed with primary DCIS between 1986 and 2004 were followed until November 2011 to study long-term survival. Tumor tissue microarrays were constructed, and immunohistochemical analyses were performed to detect cytoplasmic protein expression of HMGCR. The association between DCIS HMGCR expression and invasive breast cancer recurrence-free survival (RFS_inv) and overall survival (OS) was analyzed by Kaplan–Meier curves, log rank test, and Cox proportional hazard analysis.

Results

HMGCR was strongly expressed in 24 % of the assessed DCIS samples, moderately expressed in 46 %, and weakly expressed in 23 %; no expression was detected in 7 % of the samples. During the follow-up time (median 13.8 years), 61 patients were diagnosed with an invasive breast cancer recurrence, and 80 patients died. A crude analysis showed no survival benefit from radiotherapy. However, patients with strong HMGCR expression showed an improved RFS_inv (log rank, p = 0.03) and OS (log rank, p = 0.04) after radiotherapy. No statistically significant interaction was observed for HMGCR and radiotherapy (RFS_inv p = 0.69 and OS p = 0.29).

Conclusions

This study demonstrates HMGCR expression in DCIS and suggests HMGCR as a predictive marker of response to postoperative radiotherapy in DCIS, although the test for interaction was nonsignificant. Future DCIS studies addressing the potential of statin treatment targeting HMGCR are warranted.     

Response of acromegaly to long term bromocriptine therapy: a biochemical and clinical assessment.

The long term effects of bromocriptine in 12 acromegalics treated for a mean duration of 10.2 months are reported. Seven showed a significant (P less than 0.05) and sustained fall in serum immunoreactive growth hormone (GH) levels throughout 24 h, 6 of whom had a 50% or greater reduction in mean circulating GH during glucose tolerance testing. Only one patient had mean serum GH levels throughout the day suppressed to normal (less than 5 mIU/l) but 3 had suppression of mean serum GH during GTT to normal or very near normal (less than 10 mIU/l). The effective dose was 20 mg daily. Only 4 patients reported any improvement in soft tissue swelling and acral features, which was unrelated to the GH response. Possible reasons for the discrepancy between clinical and biochemical responses are discussed. In 9 of the 12 patients bromocriptine was discontinued and pituitary ablative therapy offered. Three out of 4 patients who underwent trans-sphenoidal hypophysectomy had mean GH levels during GTT reduced to less than 7 mIU/l. In the three who continued bromocriptine treatment GH suppression was maintained at less than 10 mIU/l for up to 3 years but with little change in acral features. Although bromocriptine is safe and was well tolerated it is not as effective as existing forms of pituitary ablative therapy and should be reserved for those cases where ablation is contraindicated or unsuccessful.     

Novel Biologically Inspired Approaches to Extracting Online Information from Temporal Data

In this paper, we aim to develop novel learning approaches for extracting invariant features from time series. Specifically, we implement an existing method of solving the generalized eigenproblem and use this to firstly implement the biologically inspired technique of slow feature analysis (SFA) originally developed by Wiskott and Sejnowski (Neural Comput 14:715–770, 2002) and a rival method derived earlier by Stone (Neural Comput 8(7):1463–1492, 1996). Secondly, we investigate preprocessing the data using echo state networks (ESNs) (Lukosevicius and Jaeger in Comput Sci Rev 3(3):127–149, 2009) and show that the combination of generalized eigensolver and ESN is very powerful as a more biologically plausible implementation of SFA. Thirdly, we also investigate the effect of higher-order derivatives as a smoothing constraint and show the overall smoothness in the output signal. We demonstrate the potential of our proposed techniques, benchmarked against state-of-the-art approaches, using datasets comprising artificial, MNIST digits and hand-written character trajectories.