Risk Prediction Model Created to Monitor 90-Day Mortality Following Esophagectomy in Esophageal Cancer

Oncologists can now use the International Esodata Study Group’s risk predictor model to identify patients with esophageal cancer who might be at a very high-risk of death following an esophagectomy.

The International Esodata Study Group (IESG) risk prediction model helped to determine a patient’s risk of death within 90 days of undergoing an esophagectomy and could help to inform the decision-making process, according to a study published in JAMA Surgery.

The 30-day mortality rate post esophagectomy was 2.0% (n = 164) and the 90-day mortality rate was 4.2% (n= 353); this corresponded to a 2-fold increase in mortality rate. The IESG allowed for patients to be identified into different risk groups within the 90 days. Among those who died within 90 days of esophagectomy, 54.1% (n = 191) died in the hospital before being discharged, 9.6% (n = 34) died after re-admission related to their procedure, and 1.7% (n = 6) died following an unrelated re-admission within 30 days of discharge. An additional 32.9% (n = 116) died within 90 days of hospital discharge without being readmitted.

“For everyday practice, the data sets that are incorporated are readily accessible before surgery. The application of the IESG score at the first patient visit or at the multidisciplinary tumor board would add granularity to patient selection and would aid practitioners and patients in the decision-making process and in providing informed consent,” the study’s investigators wrote.

The study spanned over 5 years and included 8403 patients. Almost all patients (97.9%) had solid tumors present upon enrolling on the trial. Additionally, 12.6% of patients had uncomplicated diabetes, 10.1% had chronic pulmonary disease, and 5.2% had peripheral vascular disease. The majority of patients had an ECOG performance status of 0 (50.0%), with a minority who had a status of 1 (44.7%), 2 (4.5%), and 3 (0.8%). Additionally, 46.6% of patients had received neoadjuvant chemoradiotherapy and 28.5% underwent neoadjuvant chemotherapy.

Patients were assigned to either the development cohort (n = 4172), or the validation cohort (n = 4231), both of which were found to be not significantly different in key domains. The only difference between cohorts was the rate of congestive heart failure, which was notably higher in the validation cohort.

Rates of 30-day mortality in the development and validation cohorts were 2.1% and 1.8% (P =.23), and the 90-day mortality rates were 4.4% and 4.0% (P = .34), respectively.

Investigators used the univariate regression model to compare patients who were alive after 90-days post operation (n = 3988) with patients who died within 90 days (n = 184). They used 10 variables with P< .10 to consider potential factors of risk. These factors included age, body mass index, sex, World Health Organization classification, history of myocardial infraction, connective tissue disease, peripheral vascular disease, moderate to severe liver disease,

neoadjuvant treatment, and hospital volume.

The final risk score ranged from -10 to +5 and helped to identify 16 risk groups of patients. If patients had a positive score, they had a lower risk of death while those with a negative score had a higher risk of death. The development cohort was divided into 5 different homogeneous risk groups according to their score, including very low risk (1 or less), low risk (0), medium risk (-1 to -2), high risk (-3 to -4), and very high risk (-5 or greater).

Investigators used a distribution to track the deaths within 90-days between each cohort. The areas under the curve for the development cohort was 0.68 (95% CI, 0.64-0.72), while the validation cohort was 0.64 (95% CI, 0.60-0.69).

When comparing patients who had a very low risk score with those who had a very high-risk score, investigators identified an 11-fold increase in 90-day mortality in the development cohort (1.8% vs 18.2%). Moreover, when comparing the very low risk group with the very high-risk group in the validation cohort, investigators identified a 7-fold increase in 90-day mortality (2.1% vs 14.1%).

“On the basis of a prospectively collected, large, contemporary data set and on preoperative variables combining clinical, demographic, and hospital volume data, the proposed

IESG risk prediction model allows for stratification of individual patient risk of death within 90 days after esophagectomy for cancer. The IESG risk score is easily accessible and

provides an evidence-based schema for assigning risk to 1 component of the multimodality treatment of patients with esophageal cancer and for allocation of the most appropriate treatment,” the investigators concluded.


D'Journo XB, Boulate D, Fourdrain A, et al. Risk prediction model of 90-day mortality after esophagectomy for cancer. JAMA Surg. Published online ahead of print, June 23, 2021;e212376. doi:10.1001/jamasurg.2021.2376