Oesophageal cancer : Studies on mortality trends and the impact of surgical and oncologic strategies on postoperative complications

Oesophageal cancer carries a poor prognosis, and curative treatment relies on multimodality therapy combining surgical resection with neoadjuvant chemo(radio)therapy. While survival has been investigated in epidemiological studies, few of them have examined trends separately for the main histological subtypes—oesophageal adenocarcinoma (OAC) and squamous cell carcinoma (OSCC)—and even less for patients undergoing surgery and those not. Oesophagectomy is considered a high-risk procedure characterized by substantial postoperative morbidity. A lack of standardized definitions has led to heterogenous reporting, and the true incidence of postoperative complications in routine practice is not well established. Minimally invasive oesophagectomy (MIO) and neoadjuvant treatment (nT) represent two of the most important advances in recent decades, with clinical trials demonstrating safety and efficacy. However, evidence from nationwide, unselected populations remains limited and inconclusive. This research aimed to investigate mortality and postoperative morbidity related to oesophageal cancer and oesophagectomy. The specific objectives were: (I) to examine survival trends in oesophageal cancer; (II) to evaluate and describe trends in postoperative complications following oesophagectomy; (III) to compare postoperative outcomes after MIO versus open oesophagectomy in nationwide practice; and (IV) to assess the impact of nT compared to upfront surgery, and of neoadjuvant chemoradiotherapy (nCRT) compared to neoadjuvant chemotherapy (nCT), on postoperative complications and mortality. This thesis is comprised of four original studies. All studies were conducted as nationwide, population-based, retrospective cohort studies in Finland. In the first study, national government-maintained registries were used as data sources. In the other three studies, the Finnish National Esophago-Gastric Cancer Cohort was used as data source. Postoperative complications were recorded according to the Esophagectomy Complications Consensus Group (ECCG) definitions. All studies employed standard hypothesis testing methods and multivariable models used in epidemiological research. Study I investigated survival trends during a 30-year period in Finland. 1-, 3- and 5-year survivals were tabulated for oesophageal cancer, stratified in 5-year periods, and further by histology and surgical treatment status. 5-year survivals in these groups were presented as Kaplan Meier curves. Cox regression was used to assess risk-factors for mortality. The study found overall improved survival trends in the 9102 patients included. In patients undergoing oesophagectomy, 5-year survival increased from 20% to 49% in OAC and from 11% to 54% in OSCC over the study period. In patients not undergoing surgery, 5-year survival increased from 5% to 8% in OAC and 5% to 7% in OSCC. The proportion of patients undergoing surgery decreased from 42% to 32% in OAC and from 35% to 12% in OSCC. In Cox regression analysis, female sex was a protective factor for 5-year all-cause mortality in OSCC (HR male vs female 1.43 (1.33-1.54) in all patients, 1.56 (1.33-1.83) in operated patients, 1.28 (1.18-1.39) in non-operated patients), which was not observed in OAC. Study II described trends in all postoperative complications in the ECCG framework in Finland from 1987 to 2016. Postoperative outcomes were assessed in ten-year periods, marking changes in practice or epidemiology of oesophageal cancer. This study included 1493 patients undergoing oesophagectomy. The approximately 65% overall complication rate remained stagnant during the study period. Key outcomes such as pneumonia (15.5%, 19.1%, 22.5%) and anastomotic leakage (8.2%, 8.9%, 12.1%) increased in incidence, however major complications, lengths of hospital- and ICU-stay and severe outcomes such as sepsis or multiple organ dysfunction syndrome decreased during the period. The most striking finding was a significantly decreased 90-day mortality rate, from 18% to 5.4%. Study III assessed the effect of MIO on postoperative surgical complications compared to open oesophagectomy. Postoperative complications and mortality were compared using logistic regression, adjusted for relevant confounding. This study included 699 patients from 2007 to 2016 and found MIO associated with a lower risk of major complications (35% vs 47%, OR 0.39 (0.25–0.60)), reoperations (18% vs 26%, OR 0.43 (0.26–0.73)), anastomotic leak (10% vs 14%, OR 0.47 (0.25–0.89)), intrathoracic abscess (3% vs 9%, OR 0.13 (0.05–0.36)), pulmonary complications (36% vs 46%, OR 0.63 (0.41–0.96)) and infectious complications (11% vs 22%, OR 0.41 (0.23–0.73)). No adverse outcomes were increased in patients undergoing MIO compared to open oesophagectomy. Study IV compared the risk of postoperative complications and mortality after nT compared to upfront surgery, after nCRT compared to nCT and in patients with aborted or dose-reduced nT regimens compared to those with full intended nT regimens. Outcomes were compared using logistic regression, adjusted for relevant confounding. The study included 774 patients undergoing oesophagectomy in 2005-2016. In multivariable analysis, the risk of overall postoperative complications, major complications, pneumonia, anastomotic leak or any of the complication categories in the ECCG framework were not affected by treatment strategy in any of the comparisons. In conclusion, the prognosis of oesophageal cancer has improved in recent decades, with the greatest survival benefits seen among patients undergoing surgery, whereas the prognosis of patients not undergoing surgery remains poor. Oesophagectomy continues to be a procedure of high postoperative morbidity. Even though the overall incidence of complications has remained unchanged, postoperative mortality has significantly declined, and the severity of complications has decreased. Minimally invasive oesophagectomy was associated with improved postoperative outcomes and can be safely and effectively implemented into routine nationwide practice. Neoadjuvant treatment was not associated with increased risk of postoperative complications or mortality compared to upfront surgery. The same was observed comparing neoadjuvant chemoradiotherapy to neoadjuvant chemotherapy. Adverse events during neoadjuvant treatment leading to dose reduction or treatment discontinuation do not seem to worsen postoperative outcomes.