Etiology and prognosis of gastroesophageal cancers

Sammanfattning: Gastric and esophageal cancer – gastroesophageal cancers, belong to the most fatal malignancies in the world. Understanding the cause of these diseases is key to interventions such as primary prevention and/or surveillance with the potential of lowering the disease burden. Several important exposures have been identified or suggested but the causes of gastroesophageal cancers are still largely unknown. Further, only one in three gastric cancer patients and one in six esophageal cancer patients are still alive five years after their diagnosis. The prognosis is considerably improved if the tumor can be surgically removed, but 70-80% of the patients are not eligible for surgery due to advanced spread of the disease at diagnosis. Both curable and non-curable gastroesophageal cancer patients may be treated with chemotherapy – but reports about the survival outcome from real-world patients based on the treatment they receive, are very scarce. In this thesis we explored important biological determinants for gastroesophageal cancers to encourage further research on their etiology. Furthermore, we did a follow-up study on the survival of gastroesophageal cancer patients who were treated with various chemotherapy regimens to determine their effectiveness in a real-world setting. We used population-based registers in Sweden and a case-control study in a high-risk region to explore the association between biological risk markers and gastroesophageal cancers. In study I we employed the nationwide Stomach Biopsy Cohort (SBC) study in Sweden to estimate the risk for gastric cancer associated with family history of gastric mucosal lesions. In study II we used the Epidemiology Strengthened by histoPathology Reports in Sweden (ESPRESSO) study to explore the risk for gastroesophageal cancers associated with esophageal lesions. In study III, we performed a case-control study in a high-risk area in China to study the association between gastric atrophy and esophageal squamous cell carcinoma, and examine its interaction with poor oral health, which might further increase the risk of esophageal squamous cell carcinoma. In study IV, we explored the influence of various chemotherapy regimens on the survival of gastroesophageal cancer patients based on a regional study on cancer chemotherapy at the Regional Cancer Center in Stockholm-Gotland. In study I, we found that the excess risk was 50-60% higher for gastric non-cardia cancer among individuals who had a first-degree relative with gastric mucosal lesions (atrophic gastritis/intestinal metaplasia/dysplasia) compared to the general Swedish population. In study II, we demonstrated that non-dysplastic gastric and glandular metaplasia patients had a similar excess risk (Standardized Incidence Ratio, SIR 11.9; 95% 95% Confidence Interval, CI, 9.9-14.1) for esophageal adenocarcinoma as intestinal metaplasia patients (SIR (10.8; 95% CI 7.8-14.6). In study III, we confirmed an association between gastric atrophy and esophageal squamous cell carcinoma (Odds Ratio 1.61; 95% CI 1.33-1.96), which was further increased in the presence of poor oral health (Relative Excess Risk due to Interaction 1.28; 95% CI 0.39-2.18). In study IV, we discovered that among patients who were intended to have curative treatment, esophageal cancer patients who received cisplatin-fluorouracil had better survival than those with carboplatin-fluorouracil (Hazard Ratio, HR, for carboplatin-fluorouracil vs cisplatin-fluorouracil 2.18; 95% CI 1.09-4.37), but gastroesophageal junction cancer patients treated with cisplatin-fluorouracil had worse survival than patients with fluorouracil-oxaliplatin (HR for fluorouracil-oxaliplatin vs cisplatin-fluorouracil 0.28; 95% CI 0.08-0.96). We conclude that family history of gastric mucosal lesions can be employed for further risk stratification for non-cardia gastric cancer but needs to be evaluated regarding cost-effectiveness. Further, non-dysplastic columnar metaplasia patients may benefit from strengthened surveillance, but further validation studies are required. Moreover, gastric atrophy and its interaction with poor oral health are associated with esophageal squamous cell carcinoma in a high-risk region in China, thus future studies of the microbial alterations associated with gastric atrophy and poor oral health in the development of esophageal squamous cell carcinoma are warranted. Last, the choice of cisplatin-fluorouracil was associated with better survival outcome in esophageal but worse outcome in gastroesophageal junction cancer patients in Sweden. This finding needs to be further explored on a national level in Sweden.