Gastric cancers missed at upper endoscopy in Central Norway 2007 to 2016 - a population-based study

Abstract

Bakgrunn Andelen intervallcancer (IC) ved øvre endoskopi (ØE) har tidligere blitt rapportert å være mellom 4.7% og 9.8% i andre vestlige land. Målet med denne populasjonsbaserte studien var å finne andelen av IC i Midt-Norge og identifisere faktorer knyttet til IC.

Metode Dette var en retrospektiv populasjonsbasert kohortstudie med 730 pasienter diagnostisert med adenokarsinom i ventrikkel i Midt-Norge 2007-2016. Pasientene ble hentet fra det Norske Kreftregisteret og det Norske Pasientregisteret, og hver pasients journal ble gjennomgått manuelt. IC ble definert som adenokarsinom i ventrikkel diagnostisert mellom 6 og 36 måneder etter en tidligere ØE. Faktorer assosiert med IC ble undersøkt. Intervallcancere ble delt inn i definitivt oversett (ØE 6-12 måneder før diagnose) og potensielt oversett (ØE 12-36 måneder før diagnose) og gruppenes egenskaper ble sammenlignet.

Resultat Av 730 pasienter med ventrikkelcancer var 67 pasienter IC og frekvensen av IC var 9.2%. IC var assosiert med cancer lokalisert i corpus (p = 0.009), Laurens histologisk type diffus (p = 0.028) og tidligere Billroth 2-anastomose (14.9% vs. 4.7%, p = 0.001). IC hadde en tendens til å bli diagnostisert på et tidligere stadium, men dette funnet var ikke signifikant (p = 0.22). Ulcus ble oftere funnet ved ØE hos pasienter med definitivt oversett versus potensielt oversett ventrikkelcancer (40.9% mot 17.8%, p = 0.041).

Konklusjon IC utgjorde 9.2% av ventrikkecancere i Midt-Norge. IC var assosiert med lokalisering i corpus, Lauren diffuse type og tidligere Billroth-2-anastomose. Intensivert oppfølging og tilstrekkelig biopsitaking av pasienter med ulcus kan gjøre at færre ventrikkelcancere blir oversett.

Background The rate of missed gastric cancer (MGC) at upper endoscopy (UE) have previously been reported to be between 4.7% and 9.8% in other Western countries. The aim of this population-based study was to calculate the rate of MGC in Central Norway and identify factors associated with MGC.

Methods This was a retrospective population-based cohort study including 730 patients diagnosed with gastric adenocarcinoma in Central Norway 2007-2016. Patients were identified from the Norwegian Cancer Registry and the Norwegian Patient Registry and each patients’ medical record was reviewed manually. MGC was defined as incident gastric adenocarcinoma diagnosed between 6 and 36 months after a previous UE. Factors associated with MGC were examined. MGCs were subdivided into definitely missed (UE 6-12 months prior) and potentially missed (UE 12-36 months prior) and their characteristics were compared.

Results Out of 730 patients with gastric cancer, 67 patients were MGCs and the rate of MGC was 9.2%. MGC was associated with localization (p=0.009) with MGC more frequently in the corpus, Laurens histological type (p=0.028) with diffuse type more frequent among MGCs and previous Billroth 2-anastomosis (14.9% vs. 4.7%, p = 0.001). MGC tended to be diagnosed at an earlier stage, but not significantly (p=0.22). An ulceration was more frequently found at the UE prior to the diagnosis in patients with definitely missed versus potentially missed gastric cancer (40.9% vs. 17.8%, p=0.041).

Conclusions MGC accounted for 9.2% of gastric cancers in Central Norway. MGC was associated with localization in the corpus, Lauren distribution diffuse and previous Billroth-2-anastomosis. Intensified follow-up and adequate biopsy sampling of patients with gastric ulcerations could reduce the rate of missed gastric cancer.