Surgical treatment outcomes of patients with T1-T2 gastric cancer: does the age matter when excellent treatment results are expected?

Even though the incidence of gastric cancer (GC) has decreased over the past decades, it continues to be a major healthcare problem being the fourth most common malignancy and the second cause of death among all cancers [1]. Treatment strategy of gastric cancer has dramatically changed during the last decades mainly as a result of advances in chemotherapy. However, surgery remains the main and only curative treatment option for GC. Survival rates after curative surgery vary depending on several factors, but the stage of disease and the quality of surgery are the two most important predictors [2]. Best results of surgical GC treatment are achieved when patients undergo surgery while the tumor has only limited invasion. Reports from Asian and Western countries show excellent overall survival (OS) results of patients who underwent surgery for early gastric cancer (pT1) with a 5-year OS rate up to 99% [3‐6]. Results for patients with pT2 gastric cancer were also high with a 5-year OS reaching up to 66% [7]. Despite these advancements, such favorable outcomes will be difficult to maintain in the increasing aging population. The increasing lifetime expectancy and improved treatment of age-related chronic diseases have led to a greater number of older patients suffering from GC, who can be potentially cured by surgical resection. While surgery is standard in patients with a stable clinical condition, the indication for operations in elderly patients, especially in those with comorbidities, remains unclear [8‐10]. Therefore, identification of factors affecting short-term and long-term surgical treatment outcomes in elderly is essential for treatment personalization and optimization. This is especially significant for a cohort of patients in which excellent outcomes can be expected. Moreover, because early GC stages are relatively rare in the West, only limited data are available to guide treatment decisions for such a population.

The aim of our study was to analyze the differences in surgical treatment outcomes between elderly and non-elderly patients with early invasion (pT1-T2) gastric cancer.

With the population in Europe and Lithuania aging, the proportion of patients with gastric cancer who are elderly at the time of diagnosis is increasing [11]. Naturally, the definition of elderly patients varies in different studies with cutoff values ranging from 65 to 70 years to 75 or even 80 years [11‐13]. However, average lifetime expectancy in Lithuania is lower compared to most developed Western or Asian countries. According to WHO report, the average lifetime expectancy in Lithuania in 2015 was 73.6 years. Therefore, we used 70 years as the threshold. With this selection, the elderly group amounted to 41.6% of the total cohort in this study.

Several aspects should be discussed with respect to the clinicopathological differences between the groups of our study. We found two distinct histological features. First, the intestinal type of tumor was more often found in the E group. These results support previously published data [14‐17]. Higher frequency of intestinal gastric cancer with increasing age has been explained by studies, which showed intestinal GC development in areas of the stomach where intestinal metaplasia occurs. With chronic atrophic gastritis, a gradual change from normal mucosa to intestinal-type mucosa takes many years to develop. The process peaks in elderly age contributing to an increased risk of intestinal gastric cancer [18, 19]. The second difference that we noted was that the incidence of differentiated cancer (well or moderately differentiated adenocarcinoma) was significantly higher in the E group. Higher incidence of differentiated tumors supports the hypothesis that gastric cancer in an elderly age develops first as a differentiated type of tumor and only later, probably due to long-lasting chronic gastritis, these tumors progress to an undifferentiated one. Both of these unique findings emphasize the major role of chronic gastritis in the GC development and progression in elderly individuals.

There was no significant difference in the overall and surgical postoperative complication rate between the NE and E groups. These results oppose studies published in the early 2000s, where age was shown to be a major predictor of postoperative complications or mortality [20, 21]. Instead, it supports results of more recent studies, demonstrating that advances in surgical and anesthesiological techniques have reduced surgical complications and consequently improved short-term surgical outcomes in elderly patients [22, 23]. One of the differences between gastric cancer surgeries between the different age groups is lymph node dissection, which is usually limited in the elderly. Although the rate of D1 or insufficient (≤ 15 lymph nodes) lymphanodectomy was not significantly higher in the E group, we did find a significant difference when the average number of retrieved lymph nodes was compared between the groups (22.18 vs. 19.50, p = 0.005). However, in contrast to most Western and Japanese reports [24‐26], we failed to show the association between limited lymphanodectomy and lower postoperative morbidity or mortality. Perhaps, this discrepancy can be explained by the retrospective design of our study and the fact that limited lymphanodectomies were performed for patients with a high morbidity risk.

Other well-known surgical risk factors for postoperative morbidity are the extent of gastric resection (total or subtotal) and multivisceral resection [20, 25, 26]. Our results confirmed the association between these two factors and higher postoperative complication rates in the E group. Generally, according to the principles of surgical oncology, subtotal gastrectomy may be carried out if an adequate proximal margin can be achieved. However, the requirements for proximal margin vary between different guidelines. Our study cohort included only patients with early invasion (pT1-T2) GC in which the recommended proximal margin is 5–8 cm according to the European Society for Medical Oncology guidelines. On the other hand, the Japanese gastric cancer treatment guidelines indicate only 3–5 cm in cases of T2 tumors and 2 cm in cases of T1 tumors. In our opinion, less radical resection margins are preferred to avoid a total gastrectomy, especially for patients with high risk of postoperative complications.

Even though the rate of total complications was comparable between the study groups, we found significantly higher number of medical complications in the E group. Furthermore, 5.7% of elderly patients died after postoperative complications occurred, while there were no deaths in the NE group. Hayashi et al. reported similar results [27], in which elderly patients with complicated postoperative courses after the gastrectomy had higher mortality rates. Consequentially, elderly patients should receive special attention if postoperative complications occur. Precise surgical risk assessment prior to surgery is crucial when an optimal treatment strategy is being determined for individual patients. As expected, the number of patients with an ASA score of 3 or 4 was significantly higher in the E group (31.7 vs. 61.4%). In contrast to other reports [24, 28‐30], we found that higher ASA score is an independent risk factor for postoperative complications in the E group. Moreover, we correlated a higher ASA score to an increase in 90-day mortality rates in the NE group. Therefore, evaluating the physical status of patients using the ASA classification is a reliable tool in predicting the short-term outcomes in both the E and NE groups.

In determining the appropriate treatment for each individual, patient prognosis should also be considered before turning to surgery. Many studies have specifically compared the long term-outcome of GC in elderly patients with non-elderly. Most of these have confirmed that the prognosis for elderly patients was poorer [19, 31, 32]. Our results are consistent with such reports as we found a significantly higher 5-year overall survival rate in the NE group. Additionally, after performing a multivariate Cox regression analysis, we determined different prognostic factors for poor OS results in the two groups. These results indicate different pathways of poor long-term outcomes in non-elderly and elderly patients. In the NE group, two subgroups of determinants were significant. First, patient characteristics—male gender and lower BMI—were correlated to a poorer prognosis. Male gender and worse prognosis were previously shown by Sato et al. [33] and were most likely linked to a shorter lifetime expectancy in the male population. Lower BMI has been known to be associated with specific respiratory postoperative complications and respiratory causes of death, complicating survival rates [11]. The second subgroup corresponds to tumor-related factors, specifically deeper tumor invasion and diffuse tumor type. Similarly, Ikoma and colleagues [7] have published data in which T2 invasion was found to be an independent risk factor for shorter survival. Deeper invasion of the tumor was associated with more advanced disease and to its link of significantly higher rates of lymph node metastasis. Histological tumor type according to Lauren classification also influences OS results. Series of reports have shown that diffuse-type cancer has a worse prognosis [34, 35] as was confirmed by our results.

Contrary to the predictive factors determined in the NE group, we failed to find any patient or tumor-related links to poor OS results in the E group. Although total gastrectomy and postoperative complications were two independent risk factors of decreased OS in the E group, both of them were related to the surgery itself. We hypothesize that with a total gastrectomy, poor OS results are due to the prolonged duration of surgery and more extensive intraabdominal manipulations, which can result in higher risks of postoperative complications and deaths in early postoperative period. Furthermore, Mantovani et al. [36] suggested that prolonged inflammatory responses could promote the proliferation and survival of cancer cells. Based on this report, the association of poor long-term outcomes and postoperative complications can be explained by the idea that residual cancer cells stimulated by inflammatory responses, caused by postoperative complications, result in proliferation and metastasis of cancer cells. Additionally, postoperative complications are associated with adjuvant chemotherapy omission and treatment delays resulting in postpone of residual cancer treatment. Therefore, postoperative complication prevention may play a major role in improving not only short-term but also long-term results in elderly patients with GC.

Elderly patients with early invasion of GC have a similar risk of postoperative complications as the non-elderly population. However, they should receive special attention in cases of complicated postoperative courses because mortality rates in the elderly are significantly higher.

Our study suggests that the ASA score may be useful in predicting postoperative complications in elderly patients undergoing surgery for GC. It offers clinicians another tool in optimizing treatment strategies. Finally, our results suggest that radical surgery with at least limited lymph node dissection should be considered, even for elderly patients.