Abstract
A multi-institutional squamous lung cancer consortium of investigators is developing prognostic signatures through the US NCI Lung SPECS (Strategic Partnership for Evaluation of Cancer Signatures) program. Six institutions contributed tumor specimens and published/unpublished expression-based prognostic signatures for validation using standardized sample cohorts (a primary validation cohort comprising institutional cases, and additional validation cohorts from two prospective cooperative group studies) and quality controlled assessment in independent laboratory and statistical cores. Here, we report on de novo prognostic signatures derived using the pooled institutional dataset. Highly quality-controlled cases of primary SCC from the pooled cohort (N=249) were analyzed to generate de novo prognostic signatures from among the 147 genes comprising pre-existing signatures, and from among all profiled genes. Minimax Concave Penalty (MCP) selection and Ward’s minimum variance clustering yielded survival analyses with 2 clusters that were evaluated using Cox regression and bootstrap cross validation (bCV; 500 iterations). Two significantly prognostic models were generated (see Figure): Pooled Model A (PMA) was the optimal 2-cluster model using probesets representing 6 genes selected from components of pre-existing signatures: CASP8, MDM2, SEL1L3, RILPL1, LRR1, COPZ2. Pooled Model B (PMB) was the optimal 2-cluster model using probesets representing 6 genes selected from among all those profiled: SSX1, DIAPH3, LOC619427, CASP8, EIF2S1, HSPA13. PMA and PMB each remained independently prognostic in multivariable analyses incorporating an a priori baseline model (age, sex, stage; c-index = 0.641). Two de novo prognostic signatures were derived using a pooled multi-institutional cohort of SCC assembled for validation of pre-existing signatures. PMA and PMB were each found to be independently prognostic, accounting for established clinical predictors. Both now move forward, along with validated pre-existing signatures, to additional assessment of discrimination, calibration and clinical usefulness using additional independent prospective US co-operative group cohorts of cases.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
