This multi-center cohort study systematically investigated the prognostic effect of sex disparity and prediagnosis lifestyle factors in gastric cancer, including BMI at diagnosis, usual BMI, weight loss, the history of Hp infection, and the status of smoking and drinking. To the best of our knowledge, our study is the first work targeted this same issue, with the number of 29779 gastric cancer patients in four center cohorts. In this study, we found that female patients with gastric cancer had better prognosis than male patients. Smoking history only affected the prognosis of female patients, while BMI at diagnosis, usual BMI and the amount of smoking were only independent prognostic factors for male patients.
Gastric cancer is closely related to sex. Statistically, the incidence of gastric cancer in males is almost twice as high as in females[2]. The effect of sex factor on the prognosis of gastric cancer has also been extensively studied, but several controversies about the result still remained. In most studies, female patients had a better prognosis than males[3–12]. However, several other studies showed that sex had no prognostic effect on OS[13–16]. Three studies even demonstrated that males with gastric cancer had a better prognosis than female patients[17–19]. Our study showed that female patients had a slightly better prognosis than male patients. Some studies believed the difference in prognosis between females and males could be related to the type of gastric cancer[11, 29, 30]. They found proximal and cardiac gastric cancer were more likely to be males, which was consistent with our research, and patients with these types of gastric cancer had a poorer prognosis than patients with distal and non-cardiac gastric cancer. One study suggested androgen receptors (AR) might play an important role in the prognosis of gastric cancer[12]. AR could play the role of oncoprotein when accepted the action of androgens, including regulation of cell proliferation and tumour growth. These reactions might be a major cause of poor prognosis in male patients. In addition, AR could provide conditions for the occurrence and progression of tumors by activating cyclin-dependent kinases (CDKs) and the vascular endothelial growth factor (VEGF) gene. The expression of tumour suppressor genes p53 and p27 could be inhibited by the AR, leading to differentiation and proliferation of gastric cancer cells. Another study found AR-positive cells were only absent from samples of female gastric cancer patients with T1 and T2 stage, which might be the reason why we found that females had better OS than males in patients with early pTNM stage in our study[31]. More research is needed to prove the exact mechanism.
Our study suggested that weight loss > 10% was significantly related to the prognosis of gastric cancer in both male and female patients. One previous study noted that this result might be related to the comparatively aggressive potential of the tumor in patients who lost weight, wherein the physical condition of being underweight did not cause a progression of the tumor, but rather the progression of the tumors brought about the weight loss[32]. In our study, we found underweight before developing gastric cancer was not associated with the prognosis of gastric cancer, which partly supported this opinion. Weight loss may be due to dysphagia, odynophagia, anorexia or cancer cachexia as gastric cancer progresses to more advanced stages[33, 34]. The more advanced stages the gastric cancer patients were, the greater surgical manipulation they required, and the higher risk for bleeding that requires a transfusion they had[32]. In additions, human adipose tissue may have the function of preserving nutrients and increase the chance of survival, thus it has a negative effect on survival when weight lost[24, 35, 36].
Although overweight and obesity are considered an important risk factor for the development of cancer, their presence appears to be a paradoxical protective factor for survival in patients[35]. Our study showed this phenomenon known as the obesity paradox, and was consistent with most previous research[23, 24, 33, 37–39]. To our surprise, after stratified by gender, this phenomenon was only observed in male patients. No study has yet concluded that the obesity paradox has sex disparity in gastric cancer patients. However, one study found that only male patients seemed to contribute to the obesity paradox observed in patients with acute coronary syndromes[40]. Another study showed that obesity was associated with improved prognosis of metastatic melanoma, driven by strong associations observed only in male patients treated with targeted or immune therapy[41]. And they believed it was due to different hormone levels in obese patients between the sexes. However, in the recent study, some researchers have suggested that obesity or adipose tissue is not associated with the prognosis of cancer[42, 43]. Though BMI is widely-used to measure obesity, it may misclassify body composition (fat versus muscle). The real factor associated with the prognosis was skeletal muscle mass rather than fat. Therefore, more studies are needed to confirm the effect of BMI at diagnosis on the prognosis of gastric cancer.
The effect of smoking on the prognosis of gastric cancer has long been investigated. Most previous studies demonstrated that cigarette consumption might bring an adverse effect on the prognosis[20, 22, 44–49], but there were also some studies that showed no relationship between smoking and the prognosis[50–52]. One study even found an unexpected association between smoking and better survival in gastric cancer patients[21]. Consistent with most previous studies, our study suggested that smoking was associated with a poor prognosis for gastric cancer. In one study, cigarette smoke exposure could increase TxA2 release[20]. TxA2 mediated diverse biologic effects, such as platelet activation, cell contraction and angiogenesis, which might facilitate tumor growth and metastasis in smokers. In addition, nicotine is one of the main ingredients in cigarettes. Although nicotine was not carcinogenic by itself, it induced proliferation and angiogenesis in several preclinical models[53–55]. Moreover, smoking had an adverse impact on the pulmonary, circulatory, and immunologic systems, and on wound healing, which might reduce the effect of gastrectomy or other treatments[56, 57]. However, the impact of cigarette consumption was different between the sexes. Female patients with smoking history had a worse prognosis than non-smokers, regardless of how much they smoked in our study. Among male patients, smoking only had adverse effects if they smoked more than 40 cigarettes per day. Although the exact mechanism was unclear, the reason could be that females were more susceptible to the adverse effects of smoking as a result of biological differences, such as immunological or hormonal determinants[58–65].
There are many available studies regarding the prognostic effect of drinking on gastric cancer, but the results were inconsistent. Some studies believed alcohol consumption was associated with poor prognosis[21, 44, 66, 67], but others showed drinking had few impacts in the survival of gastric cancer patients[22, 48, 50, 51, 68]. Our study showed there was no significant association between drinking and long-term prognosis, regardless of the amount and duration of alcohol consumption. As such, the association between alcohol consumption and the prognosis of gastric cancer still needs further study.
The history of Hp infection has always been a high-risk factor for gastric cancer[2]. However, its effect on prognosis has not yet been established. There were many studies that believed Hp infection would develop the progress or lead to the recurrence of gastric cancer, which brought poor prognosis for gastric cancer patients[28, 69–72]. One study found Hp infection was significantly associated with heparanase expression, which might promote the invasion and metastasis of gastric cancer and cause a poorer prognosis in the end[73]. In another study, the effects of Hp infection on gastric cancer were as follows: it caused DNA damage and affected the repair of the tumor microenvironment (TME) to the damage, activated the oncogenic signaling pathways to promote cancer growth, and modulated the immune environment within the TME[74]. However, more studies have found the opposite that Hp positive status was associated with better outcome[26, 27, 75–86]. Those studies suggested that although Hp infection increased the risk of gastric cancer, the histological type of gastric cancer induced by Hp was associated with a better prognosis[75, 77, 86]. And they believed the immune change caused by Hp infection was beneficial to the prognosis of gastric cancer patients[26, 78, 82]. In addition, some studies found that there was no relationship between Hp infection and the prognosis of gastric cancer[87–89], which were consistent with our research. Further investigations are needed to clarify the paradoxical effects of Hp infection in future studies.
One of the main strengths of this study is the large number of patients included, which allows the identification of sex differences with clinical relevance and statistical significance. Another strength is we analyzed three groups included total patients, gastrectomy, and no surgery groups, which avoids the risk of error in survival outcomes due to different treatment methods. However, when interpreting the findings of this study, several potential limitations need to be considered. Firstly, many patients may change their lifestyle after the diagnosis of gastric cancer, such as stopping or limiting their tobacco and alcohol consumption. This may lead to differences in survival due to lifestyle changes after diagnosis, which has a potential confounding effect on the result of our study. Secondly, the incorrect collection of smoking and drinking history in patients with gastric cancer may occur when lifestyle factors were collected in an indirect manner due to recall bias. Thirdly, the sample of no surgery patients was relatively small. Therefore, the results might not be used as a reference for patients who cannot undergo surgery.