Many economic problems require a covariance matrix estimator that is not only invertible, but also well-conditioned (that is, inverting it does not amplify estimation error). For largedimensional covariance matrices, the usual estimator -the sample covariance matrix-is typically not well-conditioned and may not even be invertible. This paper introduces an estimator that is both well-conditioned and more accurate than the sample covariance matrix asymptotically. This estimator is distribution-free and has a simple explicit formula that is easy to compute and interpret. It is the asymptotically optimal convex linear combination of the sample covariance matrix with the identity matrix. Optimality is meant with respect to a quadratic loss function, asymptotically as the number of observations and the number of variables go to infinity together. Extensive Monte-Carlo confirm that the asymptotic results tend to hold well in finite sample.
A B S T R A C T PurposePanitumumab, a fully human antibody against the epidermal growth factor receptor (EGFR), has activity in a subset of patients with metastatic colorectal cancer (mCRC). Although activating mutations in KRAS, a small G-protein downstream of EGFR, correlate with poor response to anti-EGFR antibodies in mCRC, their role as a selection marker has not been established in randomized trials.
Patients and MethodsKRAS mutations were detected using polymerase chain reaction on DNA from tumor sections collected in a phase III mCRC trial comparing panitumumab monotherapy to best supportive care (BSC). We tested whether the effect of panitumumab on progression-free survival (PFS) differed by KRAS status.
ResultsKRAS status was ascertained in 427 (92%) of 463 patients (208 panitumumab, 219 BSC). KRAS mutations were found in 43% of patients. The treatment effect on PFS in the wild-type (WT) KRAS group (hazard ratio [HR], 0.45; 95% CI: 0.34 to 0.59) was significantly greater (P Ͻ .0001) than in the mutant group (HR, 0.99; 95% CI, 0.73 to 1.36). Median PFS in the WT KRAS group was 12.3 weeks for panitumumab and 7.3 weeks for BSC. Response rates to panitumumab were 17% and 0%, for the WT and mutant groups, respectively. WT KRAS patients had longer overall survival (HR, 0.67; 95% CI, 0.55 to 0.82; treatment arms combined). Consistent with longer exposure, more grade III treatment-related toxicities occurred in the WT KRAS group. No significant differences in toxicity were observed between the WT KRAS group and the overall population.
ConclusionPanitumumab monotherapy efficacy in mCRC is confined to patients with WT KRAS tumors. KRAS status should be considered in selecting patients with mCRC as candidates for panitumumab monotherapy.
ORE PERSONS IN THE United States die from non-small cell lung cancer (NSCLC) than from breast, colorectal, and prostate cancer combined. 1 Each year, more than 60000 persons develop stages IIIB and IV NSCLC; nearly all go on to die from metastatic spread. In addition, most individuals experience symptoms caused directly by lung cancer. These symptoms are often the first manifestations of the illness and increase in frequency and severity as the disease progresses. Cough, shortness of breath, weight loss, loss of appetite, and chest tightness impair the quality of lives already cut short by NSCLC. For patients with advanced lung cancer, physical well-being and changes in quality of life correlate with survival. 2 Espe-Author Affiliations and Financial Disclosures are listed at the end of this article.
This study demonstrated that panitumumab-FOLFOX4 was well tolerated and significantly improved PFS in patients with WT KRAS tumors and underscores the importance of KRAS testing for patients with mCRC.
Gefitinib showed no added benefit in survival, TTP, or RR compared with standard chemotherapy alone. This large, placebo-controlled trial confirmed the favorable gefitinib safety profile observed in phase I and II monotherapy trials.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.