Secondary Primary Cancer after Primary Gastric Cancer: Literature Review and Big Data Analysis Using the Health Insurance Review and Assessment Service (HIRA) Database of Republic of Korea

Simple Summary With the development of new chemotherapeutic agents and advances in surgical techniques, the overall survival of gastric cancer patients continues to rise. Long-term follow-up of gastric cancer patients has resulted in an increase in the number of reports of secondary primary cancer (SPC) in these patients. This review analyzes the SPCs of gastric cancer survivors using the nationwide claims data of Republic of Korea, and summarizes previous studies of the incidence of SPC in national and institutional databases. Abstract Advances in cancer screening and early detection, as well as improvements in surgical techniques and therapeutics, have contributed to decreasing gastric cancer mortality. The number of gastric cancer survivors continues to rise; however, long-term follow-up has revealed an increase in the risk of post-gastrectomy symptoms or other health problems, such as extra-gastric secondary primary cancer (SPC), in these survivors. Therefore, evidence-based screening for new primary cancer is needed in these populations; however, the incidence of SPC varies by country or continent and its characteristics have not been clearly reported. The characteristics of SPC are of increasing interest to both treatment providers and gastric cancer survivors; thus, this literature review explores not only the epidemiology and biology of SPC but also clinical and biological factors that influence its prognosis.


Introduction
Cancer research and advances in cancer treatment modalities have increased the rates of overall survival in cancer patients. After completion of cancer treatment, cancer survivors return to pre-treatment life, but their quality of life is not the same as before treatment. Patients face financial burden, psychological difficulty, work-related problems, and physiological changes in the body after organ resection [1]. Along with providing emotional support and explaining the disease to patients, health care providers should also be aware that secondary primary cancers (SPCs) can develop in cancer survivors.
In gastric cancer, the development of new chemotherapeutic agents and advances in surgical techniques have resulted in more survivors in recent decades [2][3][4][5][6]. In East Asian countries, the advent of national screening programs has resulted in more than half of all gastric cancer patients being diagnosed with early stage tumors, and their 5-year survival rate exceeds 90% [2,7,8]. SPCs affect the long-term survival of cancer survivors, and several national and institutional database studies have been published previously. Therefore, the purpose of this study is to review the literature describing the occurrence of SPCs in gastric Table 1. List of single institutional-based studies on the incidence of secondary primary cancer (SPC) after gastric cancer.

Authors
Country Publication Year

Synchronous/ Metachronous Common SPCs
Takekuni K et al. [ Nationwide population-based studies in Japan, Taiwan, United States of America, Portugal, and Sweden are available and listed in Table 2 [20][21][22][23][24]. The incidence of SPCs in patients with gastric cancer ranges from 4.4% to 5.5%. In most studies, the risk of SPC in patients with gastric cancer was estimated using standardized incidence ratio (SIR), which is defined as the number of observed cancer occurrences divided by the expected number. The expected number of cancers was calculated by multiplying the first primary cancer incidence by the person-years after first primary cancer diagnosis in the general population. The SIR for all SPCs in gastric cancer patients ranges from 0.91 to 1.46. The most common SPC site after gastric cancer is the gastrointestinal tract, such as esophagus, small intestine, and colon. Some studies showed increased incidences of SPCs in thyroid, pancreas, bone and soft tissues, ovaries, bladder, kidney, liver, and intrahepatic bile duct, and non-Hodgkin's lymphoma. By contrast, the incidence of SPCs in lung, breast and prostate, melanoma, and myeloma decreased after a diagnosis of gastric cancer. In both men and women, the incidences of SPCs in the gastrointestinal tract and non-Hodgkin's lymphoma after diagnosis of gastric cancer were high. In men, the incidence of SPCs were elevated in prostate and bladder, whereas in women, the incidences of SPCs were elevated in bone and soft tissue, ovaries, kidneys, liver and intrahepatic bile ducts. The incidence of SPCs in gastric cancer survivors was higher in men than in women [12,19,21,23].

Biology of SPCs
Several risk factors for gastric cancer have been documented, namely Helicobacter pylori infection, male sex, old age, low socioeconomic status, smoking, alcohol consumption, salted food intake, preserved food, genetic factors, and familial predisposition [25]. These factors often overlap with risk factors for other cancers.
Chemo/radiotherapy has been proposed to contribute to the development of SPCs. Treatments for the first malignancy may contribute to the development of SPCs by causing damage to specific regions of DNA, chromosomal rearrangements, or chromosome loss [43,44]. Some genetic factors have been found to be associated with the development of gastric and other cancers. Microsatellite instability is closely associated with gastric and sporadic colorectal carcinoma [45][46][47]. Lynch syndrome, an autosomal dominant inherited cancer predisposition syndrome, is a risk factor for tumors of gastric, colorectal, small bowel, gallbladder, and biliary tract, and is also associated with microsatellite instability [48][49][50]. In addition, several hereditary diseases, such as Li-Fraumeni syndrome, breast-ovarian cancer syndrome, multiple endocrine neoplasia, Peutz-Jegher syndrome, and Cowden syndrome, increase the risk of gastric and other cancers [40,51].

Predisposing Factors for SPCs
A Taiwanese population-based study employing multivariate logistic regression analysis demonstrated that age over 70 years, male sex, and comorbidities, such as diabetes mellitus, chronic obstructive pulmonary disease, and liver cirrhosis are risk factors for developing SPCs [21]. Several single institution-based studies found several risk factors for developing SPCs after gastric cancer, such as age over 60 years, differentiated histology, earlier stage gastric cancer, and multiplicity of lesions [14,19].

Prognosis of Patients with SPCs
The long-term outcomes of gastric cancer patients with SPCs have not been reported by national population-based studies. However, several single-institutional retrospective studies have reported the long-term survival outcomes of gastric cancer patients with SPCs. In a Korean study, the 5-year overall survival rate according to gastric cancer stage regardless of SPC site was as follows; 61% for stage I, 39% for stage II, 30% for stage III, and 0% for stage IV [16]. In studies comparing the survival outcomes of patients with and without SPCs, the 5-year survival rates of patients with SPCs were statistically poorer than those of patients without SPCs [18,19]. In addition, the 5-year survival rate of patients without SPCs was 76.5%, that of patients with metachronous SPC was 67.5%, and that of patients with synchronous SPC was 34.1%, which were statistically significant (p < 0.001 for metachronous SPC vs. no SPC and p < 0.001 for synchronous SPC vs. no SPC). In a Japanese study analyzing 10-year survival rates, the survival rate of patients with metachronous SPC was higher than that of patients without SPC or with synchronous SPC (75.2% for metachronous SPC; 69.3% for no SPC; and 40.1% for synchronous; p < 0.01) [11]. The 5-year survival of metachronous SPC patients was significantly superior to that of synchronous SPC patients in a Korean retrospective study (77.9% for metachronous SPC and 55.2% for synchronous SPC; p = 0.002) [17]. However, in a Spanish study, no significant differences in survival rates were found between patients with synchronous and those with metachronous SPC [15].

Materials and Methods
Nationwide population and cancer data were obtained from the public medical insurance system of Republic of Korea, the National Health Insurance (NHI) system. The Health Insurance Review & Assessment (HIRA) database is an involuntary government-operated organization that builds accurate review and quality assessments for the NHI system. The number of people registered in the NHI was 52,272,755 in 2016, representing about 97% of the total population of Republic of Korea. The HIRA database provides information on age, sex, diagnostic codes, prescribed drugs, and treatment, including surgical history and procedures.
The nationwide cohort was abstracted from the HIRA claims data and included patients who were diagnosed with gastric cancer from January 2009 to December 2016. Firstly, patients with gastric cancer (C16) were confirmed by the Korean Classification of Disease, sixth edition (KCD-6), a modified version of the International Classification of Disease 10 (ICD-10) for the Korean health care system. SPC was defined as another 'first' diagnosis of primary cancer with an interval of 6 months or more after diagnosis of gastric cancer in the HIRA dataset. Patients already diagnosed with other cancers prior to gastric cancer were excluded from the study. This study was approved by the institutional review board of Ajou University Hospital (AJIRB-MED-EXP-17-101).
Descriptive statistics were calculated to estimate the frequency of SPC using the diagnostic code assigned on the first hospital visit. Additionally, SIRs and the corresponding 95% confidence interval (CI) were analyzed to estimate the relative risk of SPC in patients with gastric cancer, as compared with the general population. The CI was calculated using normal approximation, and the mid-p exact test was used when the observed value was below 5. All statistical analyses were performed using R-statistics, version 3.0.2.

Discussion
This study includes the largest data set ever reported for evaluating the national incidence of SPCs among gastric cancer survivors using claims data from the HIRA database in Republic of Korea. The major strengths of this study are the large sample size (10,324

Discussion
This study includes the largest data set ever reported for evaluating the national incidence of SPCs among gastric cancer survivors using claims data from the HIRA database in Republic of Korea. The major strengths of this study are the large sample size (10,324 SPCs among 225,973 gastric cancer patients) and the accuracy of the reported data, which can be attributed to the mandatory requirement for cancer registration in the health insurance system of Korea. Therefore, we suggest that the results may be more reliable than those of previous studies.
Several previous studies of SPCs in gastric cancer survivors have revealed that they share several common features. The incidence of SPCs in gastric cancer survivors is clearly higher than that in the general population. Due to the nature of the therapies used to treat primary cancers patients, the genetic predispositions of these patients, as well as the common environmental factors that lead to cancer, cancer survivors are at a higher risk of developing SPCs than the general population [1]. In addition, patients with malignancies tend to be more health-conscious or have greater access to hospitals than the general population. SPC increases with age and occurs most often within the 2-year period following a gastric cancer diagnosis. Although the overall incidence increases with age, the SIR for SPC is the highest in relatively young gastric cancer patients [21,52,53]. Young cancer patients are considered to be genetically predisposed to cancer, and careful followup and screening strategies are required. Furthermore, SPCs in gastric cancer survivors develop more frequently in males than in females [11,14,16,23].
The most common types of SPCs in gastric cancer patients are tumors of the digestive tract, such as those of the esophagus, small bowel, and colon. The pathophysiology underlying the high incidence of digestive organ cancers is unclear but may be due to factors affecting the gastrointestinal tract, such as N-nitroso compounds, drinking alcohol, and smoking, as well as common carcinogenic processes [54][55][56]. Another possibility for the high incidence is that these cancers are readly detected to regular gastric cancer follow-up examinations, such as endoscopy and abdominopelvic computed tomography [20,57,58]. The high incidence of intra-abdominal SPCs, such as hepatocellular carcinoma, pancreas cancer, gallbladder cancer, and bile duct cancer, in gastric cancer patients is also thought to be due to regular follow-up examinations for recurrence after gastric cancer treatments.
The exact mechanism involved in the development of SPC after gastric cancer has not been fully elucidated. Genetic vulnerability, immunologic factors, and exposure to carcinogens including those used in gastric cancer treatments, as well as common risk factors between gastric cancer and SPC, are considered important [13,44]. Further studies are needed to verify the causal relationship between gastric cancer and SPCs.
Few studies have analyzed survival outcomes according SPC type because SPCs are rare. Survival analysis performed irrespective of SPC type is the main method. However, each type of cancer has a different prognosis. The prognosis of patients with early stage gastric cancer is determined by the types of SPC they acquire, not by the gastric cancer itself. To analyze the true prognosis of gastric cancer patients with an SPC, detailed analysis according to SPC type and stages of gastric cancer and SPC are needed.
This study has several limitations. First, the HIRA dataset does not include histologic types or germline mutations in mismatch repair genes of primary gastric cancer or SPCs; therefore, we were unable to evaluate their relationships with various cancers (e.g., Lynch syndrome). Additionally, we were unable to elucidate the causal relationship and mechanisms between gastric cancer treatment (surgery, endoscopic resection, and chemotherapy) and SPC type. Third, detailed individual data and cancer stages were not considered in this study. This is a common limitation of studies using the HIRA dataset. Therefore, we were unable to evaluate whether there was difference in the incidence of SPC according to the stage of gastric cancer. However, based on the observation that about 88% of SPCs are detected within 5 years after a gastric cancer diagnosis, this study suggests that follow-up protocols for the early detection of SPCs could improve the survival of gastric cancer patients.

Conclusions
This study showed that gastric cancer survivors had an increased likelihood of being identified with SPC. The occurrence of SPCs after gastric cancer treatment increased with age and was common in the digestive tract. Close surveillance after gastric cancer treatment should be considered during follow-up not only to detect recurrence but also to ensure the early detection of SPCs. Further research will be needed to elucidate the mechanisms of SPC development and the prognosis of patients with SPCs.

Data Availability Statement:
The data used in the study are potentially identifiable and are not publicly available. The raw claims datasets generated and/or analyzed during the study are not publicly available due ethical restrictions by the HIRA.

Conflicts of Interest:
The authors declare no conflict of interest.