On the prognostic value of survival models with application to gene expression signatures.

As part of the validation of any statistical model, it is a good statistical practice to quantify the prediction accuracy and the amount of prognostic information represented by the model; this includes gene expression signatures derived from high-dimensional microarray data. Several approaches exist for right-censored survival data measuring the gain in prognostic information compared with established clinical parameters or biomarkers in terms of explained variation or explained randomness. They are either model-based or use estimates of prediction accuracy.As these measures differ in their underlying mechanisms, they vary in their interpretation, assumptions and properties, in particular in how they deal with the presence of censoring. It remains unclear, under what conditions and to what extent they are comparable. We present a comparison of several common measures and illustrate their behaviour in high-dimensional situations in simulation examples as well as in applications to real gene expression microarray data sets. An overview of available software implementations in R is given.

FGFR4 Arg388 genotype is associated with pathological complete response to neoadjuvant chemotherapy for primary breast cancer.

BACKGROUND: A single-nucleotide polymorphism (SNP) in the FGFR4 gene is associated with poor prognosis in solid tumors. A recent study presented the first evidence that FGFR4 Arg388 could predict resistance to adjuvant chemotherapy in breast cancer. The present study evaluates the potential of this SNP to predict response to neoadjuvant chemotherapy (NCT) for primary breast cancer (PBC). METHODS: As part of a randomized phase II trial, 257 patients received either doxorubicin-cyclophosphamide (AC) or doxorubicin-pemetrexed (AP) followed by docetaxel (Doc; Taxotere) as NCT for T2-4/N0-2/M0 PBC. FGFR4 genotype analyzed on germline DNA was correlated with clinicopathologic variables, clinical response, and pathological complete response (pCR) using univariate and multivariate analyses. RESULTS: Only axillary lymph node status was associated with FGFR4 Arg388 [odds ratio (OR) 1.82, P = 0.03]. Joint analysis of both treatment arms revealed a correlation of FGFR4 Arg388 with clinical response (OR 2.14, P = 0.03) but not with pCR. In the AC-Doc arm, however, FGFR4 Arg388 was a strong predictor of pCR in the multivariate analysis (OR 3.79, P = 0.03). A significant interaction between FGFR4 genotype and treatment (P = 0.01) was found, indicating a therapy-specific effect. CONCLUSION: We provide the evidence that FGFR4 388Arg is an independent predictor of pCR following AC-Doc as NCT in PBC.