Systematic review and validation of clinical models predicting survival after oesophagectomy for adenocarcinoma

Abstract

Background: Oesophageal adenocarcinoma poses a significant global health burden, yet the staging used to predict survival has limited ability to stratify patients by outcome. This study aimed to identify published clinical models that predict survival in oesophageal adenocarcinoma and to evaluate them using an independent international multicentre dataset.

Methods: A systematic literature search (title and abstract) using the Ovid Embase and MEDLINE databases (from 1947 to 11 July 2020) was performed. Inclusion criteria were studies that developed or validated a clinical prognostication model to predict either overall or disease-specific survival in patients with oesophageal adenocarcinoma undergoing surgical treatment with curative intent. Published models were validated using an independent dataset of 2450 patients who underwent oesophagectomy for oesophageal adenocarcinoma with curative intent.

Results: Seventeen articles were eligible for inclusion in the study. Eleven models were suitable for testing in the independent validation dataset and nine of these were able to stratify patients successfully into groups with significantly different survival outcomes. Area under the receiver operating characteristic curves for individual survival prediction models ranged from 0.658 to 0.705, suggesting poor-to-fair accuracy.

Conclusion: This study highlights the need to concentrate on robust methodologies and improved, independent, validation, to increase the likelihood of clinical adoption of survival predictions models.

Fig. 1

Kaplan–Meier plots of predicted survival a Barbour et al., stratification of survival based on nodal metastasis risk grade (grades I/II were considered as a single grade; P = 0.859). b Langer et al., stratification of survival based on prognosis score (A = 3, B = 4–5, C = 6; P < 0.001). c Eil et al., predicted and actual survival of patients who underwent surgery alone (P < 0.001). d Eil et al., predicted and actual survival of patients who underwent neoadjuvant therapy and surgery (P < 0.001). e Shapiro et al., stratification of survival based on nomogram score (I = 0–4, II = 5–8, III = 9–11, IV ≥ 12; P < 0.001). f Davison et al., stratification of survival based on nodal metastasis risk grade (grades I/II were considered as a single grade; P < 0.009). g Cao et al., stratification of survival based on nomogram score (I = 0.0–6.6, II = 6.7–9.0, III = 9.1–11.0, IV = 11.1–13.8, V = 13.9–16.7, VI = 16.8–18.7, VII ≥ 18.8; P < 0.001). h Zhou et al., stratification of survival based on nomogram score (I = 0–93, II = 94–187, III = 188–280; P < 0.001). i Gabriel et al., stratification of survival based on nomogram score (I = 1, II = 2, III = 3, IV = 4–5; P < 0.001). j Xie et al., stratification of survival based on nomogram score (I = 0–88, II = 89–142, III = 143–172, IV ≥ 173; P < 0.001). k Liu et al., stratification of survival based on nomogram score (I = 0–48, II = 49–64, III = 65–88, IV ≥ 89; P < 0.001). l Du et al., stratification of survival based on nomogram score (I = 0–99, II = 100–158, III = 159–200, IV ≥ 200; P < 0.001). Box contains variables included in the model. pT-stage, pathological T-stage. pN-stage, pathological N-stage. LVI, lymphovascular invasion. NA-therapy, neoadjuvant therapy. Histo, histology. No. LNs, number of lymph nodes examined. No LNs/+, total number of lymph nodes examined and number of positive lymph nodes. CD-score, Charlson-Deyo comorbidity score

Fig. 2

Calibration plots curves for predicted survival a Shapiro et al.. b Cao et al.. c Zhou et al.. d Gabriel et al.. e Xie et al.. f Liu et al.. g Du et al.