Introduction

Gastric remnant cancers (GRC) are adenocarcinomas arising in the residual stomach after partial gastrectomy and has been considered a distinct clinical entity. GRC has been reported to account for 1-8% of all gastric adenocarcinomas. The advent of proton pump inhibitors and H. pylori eradication therapy have led to marked decline in distal gastrectomies for benign ulcer disease. However, there is a long latency period between surgery for benign disease and development of GRC. The clinical relevance of GRC as a separate entity has been questioned since the prognosis of GRC does not seem to differ from that of primary gastric cancers overall. In the current study we aimed to characterize patients with GRC and their prognosis in a large Western population-based cohort.

Aims & Methods

Patients with gastric adenocarcinoma in Central Norway between 2001 and 2016 (n=1217) were identified from combined searches in the Norwegian Cancer Registry (NCR) and Norwegian Patient Registry (NPR) with subsequent manual assessment of all patient records. GRCs were defined as gastric adenocarcinomas in patients previously operated with a distal gastrectomy for either benign or malignant disease (n=78). GRC patients were compared to non-GRC overall (n=1139) as well as to patients with proximal non-GRC (n=586). Patient, tumour and treatment characteristics, including indication for previous distal gastrectomy and reconstruction method, TNM stage, Lauren histological type and treatment were recorded. Minimum follow-up time was 6 years and 7 months. Survival was analysed by the Kaplan-Meier method. The Cox proportional hazard method was used in a multivariable analysis to identify factors independently associated with survival.

Results

The proportion of GRC was 6.4% for the entire cohort (78 of 1217 patients). The indication for primary distal gastrectomy was benign disease in 76 (97.4%) patients and median time to GRC was 37.6 (range 15.7-68.0) years. The annual number of GRC (p=0.002), number of non-GRC (p= 0.032) and GRC/non-GRC ratio (p=0.003) all declined during the study period. GRC patients were older at diagnosis (78.7 years) compared to non-GRC (74.9 years) and proximal non-GRC (72.7 years) patients. GRC patient were more frequently males (83.3%), compared to non-GRC (62.9%) and proximal non-GRC (71.6%) patients. A higher proportion of GRC (23.1%) were diagnosed in stages 0+I compared to non-GRC (12.8%) and proximal non-GRC (10.9%). Lauren histological classification did not differ significantly between groups. R0/R1 resection was performed in 41.6%, 41.7% and 39.7% of GRC, non-GRC and proximal non-GRC patients, respectively. The overall survival of GRC patients did not differ from non-GRC and proximal non-GRC, with median survival of 10.4 (9.3-11.5) months, 7.4 (5.1-9.9) months and 10.3 (8.6-12.1) months, and 5-year survival of 23.1%, 18.5% and 24.9%, respectively.In a multivariable Cox analysis, age (HR 1.01, 95%CI 1.00-1.02, p<0.001), Lauren diffuse histology (HR 1.25 95%CI 1.07-1.45, p=0.006), TNM stage (HR 1.45, 95%CI 1.40-1.50, p<0.001) and neoadjuvant chemotherapy (HR 0.68, 95%CI 0.54-0.85, p=0.002) were independently associated with mortality, whereas sex, proximal tumour location and GRC were not.

Conclusion

GRCs declined during the study period, but the latency between distal gastrectomy and GRC diagnosis was long. GRC patients were more often male and older than other gastric cancer patients, but GRC per se was not independently associated with survival after adjusting for TNM stage and tumour location.

Introduction

Early gastric cancers (EGC) are confined to the gastric mucosa and submucosa irrespective of lymph node metastases. The proportion of gastric cancers diagnosed at an early stage varies considerably between Western countries but is low compared to countries with a screening program for gastric cancer. The majority of EGC patients have been treated with surgical resection including standard lymphadenectomy in Western countries. In contrast, Asian countries have treated larger proportions of EGC with endoscopic submucosal dissection (ESD) followed by estimation of risk of lymph node metastases based on histopathological characteristics. The aim of the present study was to characterize EGCs in Central Norway before ESD was an available treatment option. Furthermore, we aimed to assess histopathological characteristics associated with lymph node metastases and long-term survival in our cohort of EGC patients.

Aims & Methods

The study was a retrospective population-based study using patients identified from the Norwegian Cancer Registry and Norwegian Patient Registry (NPR). The annual number of gastroscopies was retrieved from NPR. 1205 patients were diagnosed with gastric cancer from 2001 to 2016, a period when gastrectomy with a modified D2 resection was standard treatment. In accordance with Western diagnostic criteria, patients with carcinoma in situ were not included. Patients who received neoadjuvant chemotherapy and histologically ypT0 or ypT1 were excluded to study only chemo naïve (true) EGC. Histopathological specimens were re-examined and eCura score risk stratification was calculated. Finally, long-term survival and cause of death were assessed.

Results

Of the 1205 patients with gastric cancer 88 (7.3%) had EGC. The number of gastric cancers declined (p=0.010), but the relative proportion of EGC was unchanged during the study period despite a doubling of annual gastroscopies at a national level during the study period. Patients with EGC had median age 77 (43-91) years vs. 74 (21-99) years for non-EGC controls. EGCs were more often of the Lauren intestinal type (p<0.001) and were more often localized in the antrum or corpus. Eighty-three patients underwent surgical resection and eight (9.0%) of these had N+ disease. An eCura score could be calculated for 79 EGCs. A significant proportion (9.4%, n=5) of the patients with a low risk eCura had lymph node metastases, whereas further exclusion of tumours with histological ulceration or SM2 invasion in a post-hoc analysis identified a N0 cohort (n=36). Recurrence of cancer was diagnosed in seven (8.1%) patients after a median time of 29 (14-124) months from the EGC diagnosis. Fifty-five (62.5%) EGC patients died during follow-up, whereof 43 (78.2%) deaths were unrelated to EGC, seven (12.7%) died of cancer recurrence and five (9.1%) died from postoperative complications. Median survival for EGC patients was 117.1 months (95% CI 99.8-134.3) and 5-year overall survival 75%. Approximately half of deaths related to cancer recurrence (3 of 7) occurred more than five years after the EGC diagnosis.

Conclusion

EGCs constituted a minor but constant proportion of gastric cancers despite a considerable increase in gastroscopies. In this historic cohort with formal gastric resection for EGC, 9.0% had pN+ disease. Low risk eCura score alone was insufficient to predict pN0 disease whereas additional exclusion of patients with histological ulceration or SM2 infiltration identified a pN0 cohort. Long-term follow-up of more than five years is needed to assess EGC related mortality.

Introduction

Gastric adenocarcinomas associated with Epstein-Barr virus (EBV) infection account for 5-10% of gastric cancers and is recognized as a distinct molecular subtype. Adenocarcinomas arising in the residual stomach after partial gastrectomy are named gastric remnant cancers (GRC) and usually develop decades after distal gastrectomy for benign disease. The relevance of GRC as a separate entity has been questioned since many tumour characteristics as well as prognosis of GRC does not seem to differ from other primary gastric cancers. However, it has been reported that a 5-fold higher proportion of GRC were EBV positive compared to other gastric adenocarcinomas in a meta-analysis of studies on Eastern patient cohorts. Symptoms and endoscopic findings at diagnosis may also differ between GRC and other gastric cancers. The aim of this study was to assess EBV infection as well as clinico-endoscopic characteristics in patients with GRC in a Western population-based patient cohort.

Aims & Methods

All incident gastric adenocarcinoma in Central Norway between 2001 and 2016 were identified by a combined search in the Norwegian Cancer Registry and Norwegian Patient Registry followed by manual assessment of all patient records. Patient and tumour characteristics were extracted. GRCs was defined as gastric adenocarcinomas in patients previously operated with a distal gastrectomy for either benign or malignant disease. Among 1217 patients with adenocarcinoma there were 78 with GRCs. All tumours were histologically classified according to Lauren and EBV-status was assessed by in situ hybridization (ISH) in GRC patients (n=78) and in proximal (localized in the gastric corpus) non-GRC (n=56) controls. Symptoms and findings at upper endoscopy at the time of diagnosis of GRC and proximal non-GRC patients were recorded.

Results

Seventy-eight patients had GRC, of whom 76 (97.4%) patients had benign disease as indication for primary distal gastrectomy. The median latency time from primary distal gastrectomy to GRC was 37.6 (range 15.7-68.0) years. Fourteen (18.7%) GRC and 6 (10.9%) proximal non-GRCs were EBV-ISH positive, but the proportions did not differ significantly (p=0.23). EBV-status was not associated with patient age, sex or Lauren histological type, neither for GRC-patients, nor for the study population overall.
The indication for upper endoscopy when the cancer was diagnosed was more frequently anaemia in GRC compared to proximal non-GRC patients (32.1% vs. 16.1%, p=0.036) and less frequently upper abdominal pain (12.8% vs. 35.7%, p=0.002). At the upper endoscopy, ulceration was a less frequent finding (17.9% vs. 57.1%, p<0.001) in GRC compared to proximal non-GRC patients. Median overall survival in EBV positive GRCs was 22.2 (0-64.9) months compared to proximal non-GRC 6.7 (4.0-9.5) months, p = 0.40, with 5-year survival of 28.6% and 22.6%, respectively.

Conclusion

A high proportion of GRC were EBV positive, but the proportion did not differ significantly from other proximal gastric cancers. At the time of diagnosis GRC patients did more often have anaemia, but less frequently upper abdominal pain or gastric ulceration.

Introduction

It is unclear whether post-operative complications reduce the long-term survival following gastrectomy for gastric cancer, despite research on the subject. We aimed to help fill this knowledge gap in a study that took the main methodological concerns of the existing literature into account.

Aims & Methods

This population-based cohort study included patients who underwent gastrectomy for gastric adenocarcinoma in Sweden between 2006 and 2020, with follow-up until 2024. Data came from medical records and national health data registries. The exposure was in-hospital complications defined by the Clavien-Dindo classification, categorised into three groups: no complications, mild complications (grade I-II), and severe complications (grade IIIa-IVb). The outcomes were all-cause and disease-specific mortality between 91 days and 5 years following gastrectomy. Multivariable Cox regression provided hazard ratios (HR) with 95% confidence intervals (CI), adjusted for age, sex, education, comorbidity, neoadjuvant therapy, extent of the gastrectomy (total or subtotal), microscopically radical surgery, and pathological tumour stage.

Results

The study included 2102 patients. Compared to patients without complications (n=1219), those with severe complications (n=383, 18%) had an adjusted HR of 1.10 (95% CI 0.94-1.29) for all-cause mortality and HR of 1.15 (95% CI 0.97-1.36) for disease-specific mortality. The corresponding unadjusted HRs were statistically significant (HR 1.22 [95% CI 1.04-1.42] for all-cause mortality and HR 1.24 [95% CI 1.04-1.46] for disease-specific mortality). Patients with mild complications (n=500, 24%) had no increased risk of mortality outcomes.

Table 2. In-hospital complications (according to the Clavien-Dindo classification) and risk of mortality within 91 days to 5 years of gastrectomy for gastric adenocarcinoma, presented as hazard ratios (HR) with 95% confidence intervals.

Conclusion

Severe in-hospital complications might increase the 5-year mortality following gastric cancer surgery also after excluding the initial 90 days of surgery.

Introduction

Esophageal cancer (EC) is the eight most common cancer worldwide and causes the sixth highest number of deaths per year. The incidence of EC in Norway has increased by over 20 % over the past two decades, and a similar increase is seen in other Western countries. The overall five-year survival is approximately 20 %, which makes it one of the most fatal malignancies. Tumour stage at diagnosis is the main prognostic factor, and an early diagnosis is of most importance to improve survival. Upper endoscopy (UE) with biopsies is the standard procedure for diagnosing EC. However, several studies have questioned the efficacy of UE for diagnosis of esophageal neoplasms. The rate of missed esophageal cancers (MEC) at UE is of concern in many Western countries which have been reported to account for 6.4 to 12.7 % of all ECs. We aimed to calculate the rate of MEC and identify factors associated with MEC in Central Norway.

Aims & Methods

Retrospective population-based cohort study including 625 patients diagnosed with esophageal cancer (EC) in Central Norway 2004-2021. EC cases were identified by a search in the Cancer Registry of Norway, using ICD-10 codes C15x, and contained information about histological diagnosis and tumour localization. The dataset was verified by manual review of electronic patient records. MEC was defined as EC diagnosed 6-36 months after a previous non-diagnostic upper endoscopy. Clinical and disease variables were extracted from patient records. The rate of MEC in Central Norway was calculated annually and for the study period overall. Factors possibly associated with MEC, including tumour localization, histological type, as well as symptoms and findings were recorded. The median time interval from prior UE to date of diagnosis was calculated, as well as biopsy rates of patients with Barrett´s esophagus (BE) and ulcers at the prior UE.

Results

There were 625 patients in the entire cohort. The majority (n=404, 64.6%) were localized in the lower third of the esophagus and were adenocarcinomas (n=356, 57.0%). Fifty-one (8.2%) cancers were MEC. There was a significant increase in annual numbers of EC (p=0.001) and MEC (p=0.007) over the study period, but no increase in MEC rate (p=0.393). MEC patients were older than controls (74.0 years vs. 70.2 years, p=0.038), whereas sex did not differ significantly between MEC and controls. MEC was not associated with localization, sex or histological type. The most common findings at the prior UE were hiatus hernia (n=27, 52.9%), BE (n=26, 51.0%) and esophagitis (n=10, 19.6%), while the most common alarm symptom was gastrointestinal bleeding (n=11, 21.6%). The median time from prior UE to MEC diagnosis was 23.0 (6.2-35.6) months. In all patients with histological dysplasia (n=10) at the prior UE the follow-up interval was longer than recommended in current guidelines. A total of 24% of patients with either BE or an ulceration were not biopsied at UE preceding the MEC diagnosis.

Conclusion

MEC accounted for 8.2% of esophageal cancer in Central Norway. MEC was associated with higher age at diagnosis. There was no association between MEC and histological type or localization and a thorough examination of the entire esophagus is necessary to reduce MEC rates. Inappropriate follow-up of endoscopic BE, histological dysplasia or ulcerations were identified as preventable causes. Awareness of MEC, and more consistent biopsies of ulcerations and BE could reduce the MEC rate.

Disclosure

No conflict of interest.

Introduction

Gastric cancers (GCs) are histologically classified into differentiated and undifferentiated types. Mixed histological type is considered more aggressive clinical behavior compared to pure histological type. However, no reports of large-scale prospective studies regarding the long-term prognosis of mixed histological type cured in patients by endoscopic resection (ER).

Aims & Methods

We aimed to determine the long-term prognosis of mixed histological-type GCs in patients cured by ER. We used data from the J-WEB EGC study, a nationwide, multicenter, prospective cohort that enrolled 9,715 patients with 10,926 lesions treated by ER between 2010 and 2012 across 41 institutions in Japan. For this analysis, we included patients classified as eCura A or B, according to the Japanese gastric cancer treatment guidelines, represent pathological criteria for curative resection following ER. eCura A included: differentiated-type, pT1a, without ulceration (UL0), negative horizontal and vertical margins (HM0, VM0) (A-1), and no lymphovascular invasion (Ly0, V0); differentiated-type, ≤30 mm, pT1a, with ulceration (UL1), HM0, VM0, Ly0, V0 (A-2); and undifferentiated-type, ≤20 mm, pT1a, UL0, HM0, VM0, Ly0, V0 (A-3). eCura B included differentiated-type, pT1b with submucosal invasion <500 μm, HM0, VM0, Ly0, and V0. We excluded patients with multiple lesion and those with lesions containing unusual type. Each patient was followed up for 5 years. The primary endpoints were 5-year overall survival (OS) and 5-year primary disease-specific survival (pDSS) to primary GC. Patients were divided into pure and histological types. OS and pDSS were assessed using the Kaplan–Meier method, and hazard ratios (HRs) for all-cause mortality was estimated using the Cox proportional hazards regression model with death as the event. After stratification into eCuraA-1, eCuraA-2, eCuraA-3, and eCura B, the 5-year OS and pDSS were compared between the pure and mixed histological types.

Results

A total of 7,242 patients with 8,187 GC lesions were classified as eCura A or B. After excluding 646 patients with multiple GC lesions and one patient with an GC lesion containing a neuroendocrine tumor component, 6,595 patients with a single GC lesion were included in the analysis. Of these, 6,434 were pure histological type and 161 mixed histological type. There was no difference in OS between pure and mixed histological types in all patients (92.5%, 95% confidence interval [CI]: 87.4–95.7 vs. 90.5%, 95% CI: 89.7–91.2; p = 0.415, respectively). Additionally, mixed histological type was not an independent factor in adjusted HRs for OS all-cause mortality (0.93, 95% CI: 0.52–1.64). The 5-year pDSS of the mixed histological-type GCs was significantly lower than that of the pure histological-type GCs (99.4%, 95% CI: 96.6–99.9 vs 99.96%, 95% CI: 99.9–100, p = 0.003, respectively). Among eCura A-2 patients, pDSS of the mixed histological-type GCs was significantly lower than that of the pure histological-type GCs (94.4%, 95% CI: 74.2–99.0 vs. 100%, 95% CI: 99.3–100; p < 0.001, respectively).

Conclusion

The long-term prognosis of curative ER patients, including those with mixed histological-type GCs, was favorable. However, the stratification analysis showed that mixed histological-type GCs in eCura A with ulcers was more likely to cause primary GC death than pure histological-type and might need careful surveillance.