Epidemiological, clinical and pathological characteristics of gastric neoplasms in the province of Cremona: the experience of the first population-based specialized gastric cancer registry in Italy

Despite its declining incidence, gastric cancer (GC) is still the fifth most common malignancy in the world and remains the third leading cause of cancer-related death in both sexes worldwide, following lung and liver cancer [1]. Except for Japan [2], where screening programs have already been implemented, the 5-year survival rate for patients (pts) with GC is generally poor. In western populations, including Europe and the United States, 5-year survival rates usually do not exceed 25–30% [3, 4]. Epidemiology of GC is characterized by a wide variation in incidence and mortality rates across different geographical areas, as reported in GLOBOCAN [1]. Most GC cases originate in Eastern Asia (58.1%), Europe (14.7%) and parts of Central and Latin America (7.8%) [1]. Differences in epidemiology have prompted and guided cancer research and cancer control strategies. For example, in Japan, due to the high incidence, active surveillance and preventive gastric resection are performed [5, 6]. In Italy, according to the latest published data [7], two geographical areas are mainly associated with high incidence and mortality rates: one in northern Italy and the other one in the middle of the country. In particular, in the northern part, the province of Cremona presents high GC mortality rates. This has been noticed since 1996 [8] but the province lacks accurate data to monitor the situation. To investigate more deeply the real epidemiological situation for inhabitants of the province of Cremona, a specialized population-based GC registry was set up at the Oncology Department of the public Hospital “ASST” of Cremona. This article aims to report the results of the first six years of such data collection. Cancer registries have a key role in cancer control. Owing to the very poor prognosis of stomach cancer, the results of epidemiological and experimental studies are fundamental to develop primary prevention strategies [9].

The project is an observational and descriptive epidemiological study, complemented by a retrospective chart review. Official collaboration with intra and extra-provincial facilities was established, in order to have access to pt data, provided that they were formally resident in the province of Cremona. A multidisciplinary team of clinicians and pathologists was involved in the project. All medical reports with diagnosis of GC were reviewed. For each incident case the following variables were registered: medical history and information about diagnosis, imaging assessments performed, details of surgery and other treatments received, genetic background and biomolecular characteristics, social and environmental factors. All data were collected, recorded, protected and processed in accordance with the cancer registration recommendations given by AIRTum (Associazione Italiana Registri Tumori) and IARC (International Agency for Research on Cancer) and in accordance with the national privacy laws [10‐12].

This work represents the first specialized GC registry in Italy. In the province of Cremona, diagnosis of stomach cancer is more frequent in M (1.4 M/F ratio) and in the elderly population (75 years old is the overall median age at diagnosis), as reported in literature [7]. In addition, in this geographical area, M are diagnosed at an earlier age than F (73 vs 78, p < 0.001). With about 360,000 people and about 150 cases per year, the AISRs (per 100,000 inhabitants) are 41.4 for M and 28.3 for F. These values clearly deviate from other Italian areas (35.9 in northern Italy, 39.3 in central Italy and 24.8 in southern Italy for M and 17.7 in northern Italy, 20.5 in central Italy and 12.8 in southern Italy for F [7]). More than one M out of 20 and one F out of 30 living in this area are at risk of developing GC during their lifetime. In the province of Cremona the GC incidence has declined by − 1.92% in M and − 3.21% in F respectively, even if a longer period of observation is needed to confirm this with a statistically significant interval of confidence. Comparing the rates of incidence to the world standard population, ASIRs amount to 20.8 for M and 12.7 for F, despite Italian average rates of 10.9 and 5.9, respectively [1]. The F incidence rate found in our registry is much closer to the rate reported by IARC (and amounting to 13.8) for the same gender of the Eastern Asian population, known to be a country characterized by a higher GC incidence [1].

Moreover there seems to be a lower incidence in the northern district of the province, with a statistically significant result, but more appropriate spatial analysis is needed to better investigate whether a geographical spread of incidence did exist across this area. Registry data show a significant association between sex and primary cancer location (p < 0.001), with a prevalence of proximal cancer (including GEJ) in M, as well as a significant association between sex and ADK histotype (p < 0.05), with a prevalence of DGC in F. HER-2 gene amplification was detected in 26% of cases (data consistent with literature). According to available evidence from literature, the HP is present in more than 90% of non-cardia GC [36]. The percentage of HP infection in the inhabitants of the province of Cremona, equal to 18%, is inconsistent with this known data but it should be considered that these percentages cannot be directly compared each other because 18% is the percentage based on the confirmation of the presence of HP in the histological report and moreover includes all GC tumours. HP infection has been shown to be significantly associated with district of residence (p < 0.001), with a lower presence of bacteria in pts living in the northern district. However, the rate still remains far below the percentage of infection usually reported in literature and this needs more accurate investigations. Seven point six percent of resected GCs received neoadj treatment and only 40% had postoperative CT. Sixteen point eight percent of resected GCs received adj therapies and overall their OS was improved (p < 0.001), regardless of sex or tumour histotype. Overall, 58.3% of pts were diagnosed in advanced stage (III-IV) and 41.3% underwent surgery. Median OS was 16.3 mm (95% CI, 13.8–18.8) and age standardized 5-years survival was 31.44% for M and 40.50% for F. In 49.2% of cases, pts developed PD in a mean time of 24.13 mm since diagnosis. For these pts, median OS decreased to 7.87 mm (CI95% 5.78–9.95) after PD confirmation. As expected, a strong association was detected in OS by disease stage (p < 0.001) in M as well as in F. In resected DGC, F pts seemed to fare better compared to M. This consideration is clearly hypothesis-generating and needs further investigation, taking into account several possible factors including different hormonal settings, for example. Multivariable regression analysis in resected GCs, showed type of ADK (HR_{DGC vs intestinal ADK} = 2.19, p = 0.001) stage of disease (p < 0.001, 3 d.f.) and adj-CT (HR_{yes vs no} = 0.29; p < 0.001) as prognostic factors on OS.

The majority of GCs are sporadic, but 1–3% are a form of HDGC [37]. In small group of South Korean, Japanese and Portuguese people, pathogenic germline CDH-1 mutations in HDGC and early-onset GC ranged between 8 and 15% [38‐41]. Previous studies reported that the frequency of CDH-1 mutation is inversely proportional to the rate of incidence [42, 43]. Analyzing data according to IGCLC criteria [19‐21], we can assume that hereditary predisposition does not seem one of the main reasons for high GC incidence in the province of Cremona, but it is still premature to draw definitive conclusions. At the same time, the low rate of HP infection warrants a deeper analysis by considering the ten years before diagnosis [44] and the investigation of the real prevalence of HP infection in the inhabitants of this area. The latest IARC Working Group did not consider GC as a public health priority, despite its high morbidity and mortality, and very few countries have made efforts to control its incidence up to now [45]. The only exception is the Republic of Korea, with an established nationwide screening program performed by CT scan every two years in people over 40 years of age [46]. In Japan in 2013, the government established a national health program of antibiotic treatment against HP in pts with chronic gastritis confirmed by endoscopy [47]. In Taiwan and China, during screening for colorectal cancer the evaluation of HP infection is included. Subjects with infection are included in a free endoscopy screening and free antibiotic treatment [48]. In Latin America, Chile has set up a screening program to detect HP infection by endoscopy and preventive treatment in symptomatic adults over 40 years of age [49]. Beside these countries, in the rest of the world few public health programs have been set up in order to prevent GC [50, 51]. However, it is advisable that all countries characterized by a high GC incidence should also include GC in their national cancer control plans, in order to lessen the human and economic impact of this cancer. Other plausible factors to take into account for such a high incidence rate can be found in local dietary factors or environmental causes, such as contaminated water and pesticides. The area of Cremona is characterized by a plain surrounded by rivers. Literature suggests that one important source of exposure to potential carcinogens in agro ecosystems is through water contamination by agrichemicals [52]. Agriculture represents the principal economic resource of the area, which mainly produces corn [53].

The high rate of GC incidence and differences across districts, low rate of HP infection and high percentage of pts diagnosed at an advanced stage highlight the importance of early cancer diagnosis and deeper investigation of causes in the province of Cremona.

Our findings confirm and support the IARC position, which suggests and advises a deep investigation into the primary causes of GC in order to improve preventive interventional health strategies and screening procedures. High GC incidence regions, such as the province of Cremona, should consider the development of screening programs in order to increase the rate of early diagnosis and to prolong survival.