The first competing risk survival nomogram in patients with papillary renal cell carcinoma

Abstract

There is still a lack of competing risk analysis of patients with papillary renal cell carcinoma (pRCC) following surgery. We performed the cumulative incidence function (CIF) to estimate the absolute risks of cancer-specific mortality (CSM) and other-cause mortality (OCM) of pRCC over time, and constructed a nomogram predicting the probability of 2-, 3- and 5-year CSM based on competing risk regression. A total of 5993 pRCC patients who underwent nephrectomy between 2010 and 2016 were identified from the Surveillance, Epidemiology, and End Results (SEER) database. The 2-, 3-, 5-year CSM rates were 3.2%, 4.4% and 6.5%, respectively, and that of OCM were 3.2%, 5.0% and 9.3%, respectively. The estimates of 5-year cumulative mortality were most pronounced among patients aged > 75 years in OCM (17.0%). On multivariable analyses, age, tumor grade, T stage, N stage, and with or without bone, liver and lung metastases were identified as independent predictors of CSM following surgery and were integrated to generate the nomogram. The nomogram achieved a satisfactory discrimination with the AUC_t of 0.730 at 5-year, and the calibration curves presented impressive agreements. Taken together, age-related OCM is a significant portion of all-cause mortality in elderly patients and our nomogram can be used for decision-making and patient counselling.

Figure 1

The flowchart of patient selection in the SEER database. pRCC papillary renal cell carcinoma.

Figure 2

Cumulative incidences curves of cancer and noncancer death in the total cohort. The figure was performed using R software (version 4.0.3, R Core Team 2020, https://www.r-project.org/).

Figure 3

Cumulative incidences curves of CSM according to age (A), tumor grade (B), T stage (C) and N stage (D). The differences between groups were assessed by Gray’s test. CSM cancer-specific mortality. The figure was performed using R software (version 4.0.3, R Core Team 2020, https://www.r-project.org/).

Figure 4

Cumulative incidences curves of OCM according to age (A) and sex (B). The differences between groups were assessed by Gray’s test. OCM other-cause mortality. The figure was performed using R software (version 4.0.3, R Core Team 2020, https://www.r-project.org/).

Figure 5

Competing risk nomogram predicting 2-, 3- and 5-year probabilities of CSM. CSM cancer-specific mortality. The figure was performed using R software (version 4.0.3, R Core Team 2020, https://www.r-project.org/).

Figure 6

The calibration curves of the predicted probabilities and observed CSM rates. The 45-degree diagonal represents perfect agreements between the nomogram-predicted probabilities (X-axes) and the observed CSM rates (Y-axes). CSM cancer-specific mortality. The figure was performed using R software (version 4.0.3, R Core Team 2020, https://www.r-project.org/).