Introduction

By some estimates, gastric cancer is the fifth most frequently diagnosed cancer and the third leading cause of death from cancer worldwide (Ferlay et al. 2013). Gastric cancer is often diagnosed at an advanced stage and has a poor outcome (Macdonald 2006). Postoperative chemoradiation therapy has been established as a standard of care in patients with locally advanced gastric cancer (Macdonald et al. 2001; Lee et al. 2006, 2012; Leong et al. 2011). In general, patients were excluded if they were older than 75 years of age in most clinical trials, for the reason that elderly patients are more likely to have other concomitant chronic illnesses, which may increase the risk of complications and even death during treatment. Indeed, older patients with gastric cancer are less likely to receive standard treatment compared to younger patients, even when such treatments are potentially curative (Hundahl et al. 1997). Thus, the optimal treatment strategies in elderly patients are still controversial.

We designed our study to specifically evaluate the cancer-specific survival of four subgroups of elderly patients with locally advanced gastric cancer (T3, T4 Any N or Any T, N+) according to four different treatment modalities: adjuvant radiation (RT), surgery only, RT only and no surgery/no RT by analyzing the Surveillance, Epidemiology, and End Results (SEER)-registered database.

Results

Patient characteristics

We identified 3218 patients in SEER database during the 8-year study period (between 2004 and 2011) who met inclusion criteria. There were 499 patients in adjuvant RT, 2012 patients in surgery only, 305 patients in RT only and 402 patients in no surgery/no RT. There were 1783 (55.4 %) males and 1435 (44.6 %) females. The median diagnosis age was 87 years (range 75–100), and the majority of patients were white (71.1 %). Patient demographics and pathological features are summarized in Table 1. Although the proportion fell every year from 2004 to 2011, patients aged 75 years and older still made up more than one-third of the population of gastric cancer (Fig. 1). Compared with the slight decreasing pattern of surgery only, the increasing trends of RT only and no surgery/no RT were observed in elderly patients. Moreover, the adjuvant RT accounted for <20 % of care patterns for elderly patients with locally advanced gastric cancer, of which the trend changed little from 2004 to 2011 (Fig. 2). But in patients younger than 75 years with locally advanced gastric cancer, the adjuvant RT accounted for almost 40 % of care patterns (Fig. 3).

Table 1 Patient characteristics
Full size table
Fig. 1
figure 1

Trend of the proportion of patients with gastric cancer aged 75 years and older from 2004 to 2011

Full size image
Fig. 2
figure 2

Patterns of care are illustrated for elderly patients (≥75) with locally advanced gastric cancer from 2004 to 2011 according to treatment modality. RT indicates radiation

Full size image
Fig. 3
figure 3

Patterns of care are illustrated for patients (<75) with locally advanced gastric cancer from 2004 to 2011 according to treatment modality. RT indicates radiation

Full size image

Impact of different treatment strategies on survival outcomes in elderly patients with gastric cancer

The 5-year CSS was 43.8 % in adjuvant RT, 28.5 % in surgery only, 14.9 % in RT only and 1.4 % in no surgery/no RT, which had significant difference in univariate log-rank test (P < 0.001) (Fig. 4). Furthermore, the survival analyses were stratified by different age groups (age 75–79 years, age 80–84 years, age ≥85 years, Figs. 5, 6, 7). It demonstrated that adjuvant RT significantly increase survival in all age categories, including age 75–79 years, age 80–84 years and age ≥85 years (all P < 0.001). Besides, signet ring cancer, white and black race, higher tumor grade were identified as significant risk factors for poor survival on univariate analysis (all P < 0.001) (Table 2). When multivariate analysis with Cox regression was performed, we convinced these factors as independent prognostic factors (Table 2).

Fig. 4
figure 4

Survival curves in elderly patients with gastric cancer according to four subgroups. The 5-year cancer-specific survival was 43.8 % in adjuvant RT, 28.5 % in surgery only, 14.9 % in RT only and 1.4 % in no surgery/no RT, which had significant difference (P < 0.001). RT indicates radiation

Full size image
Fig. 5
figure 5

Survival curves in patients with gastric cancer according to four subgroups in age 75–79 years (P < 0.001). RT indicates radiation

Full size image
Fig. 6
figure 6

Survival curves in patients with gastric cancer according to four subgroups in age 80–84 years (P < 0.001). RT indicates radiation

Full size image
Fig. 7
figure 7

Survival curves in patients with gastric cancer according to four subgroups in age ≥85 years (P < 0.001). RT indicates radiation

Full size image
Table 2 Univariate and multivariate survival analyses of patients with gastric cancer according to various clinicopathological variables
Full size table

Discussion

Gastric cancer is rampant in many countries around the world. In 2015, an estimated 24,590 people will be diagnosed and 10,720 people will eventually die of their disease in the USA (Siegel et al. 2015). The average age at diagnosis is 71 years, and almost two-thirds of those diagnosed with gastric cancer are above 65 (Howlader et al. 2015). In the present study, we found patients aged 75 years and older made up more than one-third of the population of gastric cancer, even though the proportion fell every year from 2004 to 2011. As patients aged 75 years and older are often underrepresented in the US National Cancer Institute (NCI) randomized clinical trials (Murthy et al. 2004), the external validity of clinical trials-based recommendations in older persons was still controversial. Studies have demonstrated that older patients are less likely to receive standard cancer treatment than younger patients in breast cancer and prostate cancer, even when such treatments are potentially curative (Greenfield et al. 1987; Bennett et al. 1991). As such, it is not clear whether results of gastric cancer studies are equally applicable to elderly patients. Extrapolation of results of younger patients may not be appropriate.

Older patients may be at risk of overtreatment or undertreatment because of several competing factors such as comorbidities, frailty, clinician perceptions and decreased life expectancy. Some studies showed that elderly patients have a higher incidence of postoperative complications that are caused by reduced physiological function (Audisio et al. 1997; Page et al. 2002). Additionally, rates of underlying diseases are higher among elderly patients, meaning that complications may tend to be more serious. Yamada et al. (2013) found preoperative lymphocyte count, hemoglobin level, serum albumin levels and percent vital capacity were significantly lower in the ≥85 years group than in the 75–84 years group. Patients over 85 years old are more likely to suffer postoperative pneumonia after gastrectomy than younger old patients. Takama et al. (2015) found preoperative serum albumin level and prognostic nutritional index are significant predictors of postoperative complications in patients with gastric cancer who were 85 years or older. Although the INT-0116 trial demonstrated postoperative chemoradiation improved survival for patients with locally advanced resected gastric adenocarcinoma, the acute toxicity of adjuvant chemoradiation reported was considerable: 54 % of patients experienced grade 3 or 4 hematological toxicity 33 % of patients experienced grade 3 or 4 gastrointestinal toxicity. Among the chemoradiation group, only 64 % of patients completed treatment and 17 % discontinued treatment due to toxicity (Macdonald et al. 2001). The median age of INT-0116 trial was 60 years. The rates of treatment deviation may be even higher among individuals 75 years or older. If elderly patients had a decrease in functional status due to surgery, they may no longer be good candidates for adjuvant therapy. Coburn et al. (2008) found that that age was a significant predictor of receipt of adjuvant radiation therapy after gastrectomy for nonmetastatic gastric adenocarcinoma: 44.5 % of patients received adjuvant RT in 18–59 years, 31.0 % in 60–74 years and 12.6 % in 75–85 years. Similarly, others have observed that more than 60 % of gastric cancer operations were performed in those aged 65 years and older, while only 30.8 and 23.3 % of elderly patients received adequate nodal evaluation and adjuvant RT, respectively (Dudeja et al. 2011). Specifically, age groups 80 years and older were at fivefold risk of receiving no adjuvant RT. Thus, older patients with gastric cancer are less likely to receive any type of treatment for their cancer compared to younger patients. In agreement with these findings, our study demonstrated the adjuvant RT accounted for <20 % of care patterns in elderly patients with locally advanced gastric cancer, compared with the proportion of almost 40 % in patients younger than 75 years.

Data in the literature regarding elderly patients with gastric cancer seem to be limited and sometimes conflicting. Hoffman et al. (2012) examined the survival benefit of postoperative chemoradiation therapy for elderly patients with resected gastric adenocarcinoma. They identified 1023 patients aged 65 years and older (median, 76) who underwent gastrectomy for non-metastatic stage IB–IV gastric adenocarcinoma diagnosed between 2000 and 2002 in the linked Surveillance, Epidemiology, and End Results–Medicare database. In this study, thirty percent of patients received adjuvant chemoradiation and receipt of adjuvant chemoradiation therapy did not significantly increase survival. In contrast, Strauss et al. (2010) analyzed patients aged 65 years or older with stage IB through stage IV (M0) gastric cancer, from 1991 to 2002, who underwent gastric resection in the Surveillance, Epidemiology, and End Results–Medicare database and found adjuvant therapy significantly reduced the mortality rate for stages III and IV (M0), trended toward improved survival for stage II, and showed no benefit for stage IB. Besides, they observed trends toward improved survival in all age categories except 80–85 years. In our opinion, these studies identified patients based on the old edition of the AJCC staging manual and from old times of diagnosis. In our study, all cases were identified from diagnosed time between 2004 and 2011 and restaged according to the criteria described in the seventh edition AJCC staging manual released in 2010. We found elderly patients in adjuvant RT had the significantly best CSS and detected survival benefits from the administration of adjuvant RT in all age categories, including age 75–79 years, age 80–84 years and age ≥85 years.

Although this is a large population-based study, it has several potential limitations. First, the SEER registry does not collect information on the comorbidities, nutritional status or performance status of the patients. One reason that the elderly patients may be undergoing less aggressive treatment may be due to comorbidities and poor performance status. Second, our study is the lack of data in the SEER registry on the use of chemotherapy and curability of surgery, resulting in a potentially significant confounder in the current study. It is possible that patients may have received adjuvant chemotherapy or preoperative chemotherapy. As for patients who are older than 75 years of age, they may receive less chemotherapy. Finally, the current analysis of the nonrandomized patient population could not exclude the possibility of selection bias. However, our study has its convincing power for its larger population-based study.

In conclusion, surgery and adjuvant RT may be the recommended treatment strategy in elderly patients with locally advanced gastric cancer, especially to patients medically fit for the combined modality therapy. Ideally, randomized studies for elderly patients should be performed.

Materials and methods

Patient selection in the SEER database

The SEER, a population-based reporting system, was surveyed for the retrospective collection of data used in the analysis. The SEER program collects and publishes cancer incidence and survival data from 18 population-based cancer registries, covering approximately 28 % of the population in the USA. The SEER data contain no identifiers and are publicly available for studies of cancer-based epidemiology and survival analysis.

Cases of gastric carcinoma (C16.0–16.9) diagnosed from 2004 to 2011 were extracted from the SEER database (SEER*Stat 8.2.1) according to the site recode classifications. All cases were restaged according to the criteria described in the American Joint Committee on Cancer (AJCC) staging manual (seventh edition, 2010). Histological type were limited to adenocarcinoma (ICD-03, 8140/3, 8144/3, 8210/3, 8211/3, 8221/3, 8255/3, 8260/3, 8261/3, 8262/3, 8263/3, 8310/3, 8323/3), mucinous adenocarcinoma (ICD-03, 8480/3, 8481/3) and signet ring cell carcinoma (ICD-03, 8490/3). Only patients aged 75 years and older with locally advanced gastric cancer (T3, T4 Any N or Any T, N+) and whose gastric cancer was a single primary tumor were included into the current study. We selected this range because AJCC TMN stage was available since 2004 and patients diagnosed after 2011 were excluded to ensure an adequate follow-up time. Patients were excluded if they were treated with either preoperative or intraoperative radiation, or were treated with both preoperative and postoperative radiation, and the sequence of radiation therapy and surgery was unknown. Other exclusion criteria were as follows: synchronous distance metastases, unknown TNM stage, unknown survival months and unknown treatment modality.

This study was based on the publicly available data from the SEER database, and we had got the permission to access these research data (Reference number: 10963-Nov 2014).

Statistical analysis

Age, sex, race, histological grade, histotype and cancer-specific survival (CSS) were extracted from SEER database. CSS was calculated from the date of diagnosis to the date of cancer-specific death. Deaths attributed to the rectal cancer were treated as events and deaths from other causes were treated as censored observations. The intergroup comparison of clinicopathologic variables was performed with the Chi-square test. Survival was analyzed using the Kaplan–Meier method (Kaplan 1958). The association between each of the potential prognostic factors and the estimated CSS was tested with the log–rank test (Mantel 1966). Multivariate analysis was performed using the Cox regression model (Gill 1992). The statistical test was two-sided, and P < 0.05 was considered statistically significant. PASW Statistics 13 (SPSS Inc., Chicago, USA) was used for the statistical analysis.