Validated Competing Event Model for the Stage I-II Endometrial Cancer Population

Abstract

Comorbidity and other factors influence the risk of competing mortality among patients with early-stage endometrial cancer. Competing event models could improve our ability to identify patients likely to benefit from treatment intensification.

Selection of Head and Neck Cancer Patients for Intensive Therapy

“…GCE models were used (gcerisk package, R) 31 to estimate effects of covariates on the relative hazard for primary versus competing events (ie, l 1 l 2 ), based on a proportional relative hazards model. 14,18 The quantity l 1 l 2 is termed the u þ ratio and the quantity l 1 l 1 þl 2 is termed the u ratio. The 2 quantities are related in that u Z u þ /(1 þ u þ ), so saying a risk factor increases or decreases the u þ ratio is equivalent to saying the factor increases or decreases the u ratio.…”

“…[11][12][13][14][15][16][17] The u ratio represents the proportion of the overall event hazard that is attributable to cancer. 14,15,[17][18][19] Heuristically, the u ratio can be thought of as quantifying "the degree to which cancer is the patient's problem." It also has a direct clinical interpretation: an u ratio of 0.50 means a patient has an equal risk for a primary event as for a competing event.…”

“…20 Nomograms to predict u ratios for individual patients (ie, u score) have been developed using supervised classification models based on generalized competing event (GCE) regression. 14,15,17,18 In particular, recent studies of Radiation Therapy Oncology Group trial data and 2 large meta-analyses have found that patients with HNC with a higher u score (0.8) selectively benefit from intensive therapy. 15,17 This scoring method is being used prospectively to define eligibility for the ongoing NRG HN004 trial.…”

Undertreatment of women with locoregionally advanced head and neck cancer

“…Some studies have demonstrated no difference in survival between women and men, 11,[19][20][21] whereas other studies have found differences in survival. 24,26 The GCE model differentiates patients according to their hazard for an event of interest, such as cancer recurrence or mortality, with respect to competing events, such as death from noncancer causes, and this makes it useful for identifying individual patients who may benefit from intensive treatment Cancer September 1, 2019 and groups of patients who may be undertreated. 22,23 Generalized competing event (GCE) modeling is a useful and relatively novel method for analyzing effects of treatment and risk factors on outcomes.…”

“…21,24,25 This approach has been validated for several cancers, including head and neck, endometrial, prostate, and breast cancer. 24,26 The GCE model differentiates patients according to their hazard for an event of interest, such as cancer recurrence or mortality, with respect to competing events, such as death from noncancer causes, and this makes it useful for identifying individual patients who may benefit from intensive treatment Cancer September 1, 2019 and groups of patients who may be undertreated. 24 The advantage of GCE models over standard prognostic models lies in assessing the extent to which a factor affects the ratio of a primary event to a competing event (ie, ω + ) or the proportion of an overall event hazard attributable to a primary event (ie, ω).…”