A nomogram with a risk-stratification system can accurately predict cancer-specific survival (CSS) in patients with primary long bone osteosarcoma (PLBOS), according to research findings published in Translational Oncology

To create the nomogram, researchers evaluated data from 1199 patients with PLBOS enrolled in the Surveillance, Epidemiology, and End Results (SEER) database from 2004 to 2015.

The data set was randomly divided into a training cohort (840 patients) and a validation cohort (359 patients). The median follow-up was 64 months (range, 1 to 179 months) in the training cohort and 65 months (range, 3 to 178 months) in the validation cohort.


Continue Reading

There was no significant difference in baseline characteristics between the training and validation cohorts (P >.05). In the overall cohort, the median age was 17 years (IQR, 12-28 years), and the median tumor size was 95 mm (IQR, 70-130 mm). In most patients (60.8%), the tumor extended beyond the periosteum.

On multivariate analysis, age, histological type, surgery of primary site, tumor size, local extension, regional lymph node invasion, and distant metastasis were significantly associated with CSS (P <.05).

Based on these 7 independent prognostic factors, the researchers constructed a nomogram with a risk-stratification system for predicting the 3-year and 5-year CSS rates. Patients were stratified into low-risk (0-162), medium-risk (162-215.8), and high-risk (>215.8) prognostic groups according to the nomogram-predicted scores. 

Kaplan-Meier analyses showed a significant difference in CSS between the 3 risk groups. In the training cohort, patients in the high-risk group had a significantly lower CSS rate compared with those in the medium-risk and low-risk groups (P <.0001). These findings were confirmed in the validation cohort (P <.0001).

When compared with the low-risk group, the medium- and high-risk groups were older and had larger tumor sizes, fewer peripheral types, more local extension, more lymph node invasion and distant metastasis, and a lower rate of surgery.

The predictive accuracy of the nomogram was evaluated using the concordance index (C-index), calibration curves, and decision curve analysis (DCA).

The C-index of the nomogram in the training cohort (0.767 [0.74-0.795]) and the validation cohort (0.715 [0.665-0.764]) was higher when compared with the TNM staging system in the training cohort (0.676 [0.645–0.707]) and the validation cohort (0.644 [0.596-0.692]), indicating that the nomogram was more accurate than the TNM system.

The calibration curves of the 3-year and 5-year CSS rates in the training and validation cohorts showed that the predictive value of the nomogram was consistent with the observed value. In addition, the DCA plot for 3-year and 5-year CSS in the training and the validation cohorts showed more positive net benefits compared with the TNM-related DCA in both cohorts. 

Overall, the nomogram had good predictive performance and showed noninferior performance compared with the TNM staging system, according to the researchers.

“This model can help clinicians evaluate prognoses, identify high-risk individuals, and give individualized treatment recommendation[s] [for] PLBOS patients,” the researchers concluded.

Reference

Tian S, Liu S, Qing X, et al. A predictive model with a risk-classification system for cancer-specific survival in patients with primary osteosarcoma of long bone. Transl Oncol. Published online February 5, 2022. doi:10.1016/j.tranon.2022.101349