Health-related quality of life trajectories up to 15 years after curative treatment for esophageal cancer - a prospective cohort study

Abstract

Background: The differentiation of specific, long-term health-related quality of life (HRQL) trajectories among esophageal cancer survivors remains unclear. We aimed to identify potentially distinctly different HRQL-trajectories and uncover the underlying factors of such trajectories in patients having undergone surgery (esophagectomy) for esophageal cancer. Materials and Methods: This nationwide, prospective, and longitudinal cohort study included 420 patients who underwent curative treatment for esophageal cancer, including esophageal cancer surgery, in Sweden from 2001-2005. The main outcome was HRQL summary score trajectories, measured by the well-validated EORTC QLQ-C30 questionnaire at 6 months, 3, 5, 10, and 15 years after esophagectomy, and analyzed using growth mixture models. Potentially underlying factors for these trajectories (age, sex, education, proxy baseline HRQL, comorbidity, tumor histology, chemo(radio)therapy, pathological tumor stage, and postoperative complications) were analyzed using weighted logistic regression providing odds ratios (OR) with 95% confidence intervals (CI). Results: Four distinct HRQL summary score trajectories were identified: Persistently good, improving, deteriorating, and persistently poor. The odds of belonging to a persistently poor trajectory were decreased by longer education (>12 y versus <9 y: OR 0.18, 95% CI 0.05-0.66) and adenocarcinoma histology (adenocarcinoma versus squamous cell carcinoma: OR 0.37, 95% CI 0.16-0.85), and increased by more advanced pathological tumor stage (III-IV versus 0-I: OR 2.82, 95% CI 1.08-7.41) and postoperative complications (OR 2.94, 95% CI 1.36-6.36). Conclusion: Distinct trajectories with persistently poor or deteriorating HRQL were identified after curative treatment for esophageal cancer. Education, tumor histology, pathological tumor stage, and postoperative complications might influence HRQL trajectories. The results may contribute to a more tailored follow-up with timely and targeted interventions. Future research remains to confirm these findings.