Clinicopathological Characteristics and Prognosis of Gastric Cancer With Old Tuberculosis: A Propensity Score Matching Analysis


 Background: It is known that Bacillus Calmette–Guérin (BCG), which serves as an adjuvant immunotherapy, has certain beneficial effects on the prognosis for gastric cancer. However, no existing reports have probed into the impact of old tuberculosis (TB) lesion on patients with gastric cancer. Methods: Using the database from a high-volume center, 2649 consecutive gastric cancer patients admitted to West China Hospital undergoing curative-intent gastrectomy were identified between 2009 and 2016. Patients were divided into TB group and non-tuberculosis (NTB) group based on the presence or absence of a coexistent old pulmonary TB lesion. Then, multinomial logistic regression models tested the risk factors for the postoperative pulmonary complications (PPCs). Kaplan-Meier analyses and Cox regression models tested the effect of old TB lesion on overall survival (OS) rate. The propensity score-matching (PSM) method was used to assemble a well-balanced cohort.Results: Altogether 81 of the enrolled patients had old pulmonary TB lesions. The presence of an old TB lesion was associated with superior overall survival (OS). Besides, the 3-year OS rates were 84.7% in TB group and 69.2% in NTB group, and the difference between two groups was statistically significant (p=0.005). Similarly, in the following analysis on 73 pairs of matched patients, significant differences were also noted between the two cohorts (p=0.001), with the 3-year OS rates of 90.4% and 69.2%, respectively. Further, multivariable Cox regression analysis showed that old TB was an independent positive factor for OS before and after propensity score-matching (PSM) (p=0.002/p<0.001, HR=0.471/0.220, respectively). Additionally, the old TB lesion appeared to be an independent risk factor for postoperative pulmonary complications (PPCs) (both p<0.001, HR=3.375 and HR=3.415 in univariate and multivariate logistic regression, respectively).Conclusions: Old TB lesion is a risk factor for PPCs. Despite that, it represented as an independent positive factor for the prognosis of gastric cancer in general. This study may shed light on further immune therapy for gastric cancer.


Introduction
Gastric cancer and tuberculosis (TB) are the two major health problems worldwide. TB remains a major cause of ill health despite the technological and therapeutic advances, and the number of TB cases is estimated to be 10.0 million in 2018 [1]. Typically, gastric cancer is the fth most common cancer and the third leading cause of cancer death worldwide [2]. Given that both gastric cancer and TB possess a substantial impact on public health, the association between these two epidemics deserves further investigation.
The relationship between gastric cancer and TB has always been a hot topic for decades. In the 20th century, several clinical trials report that using the Bacillus Calmette-Guérin (BCG), a vaccine derived from mycobacterial tuberculosis, as an adjuvant non-speci c immune potentiator can prolong the survival of gastric cancer patients [3][4][5]. Particularly, in recently years, immunotherapy seems promising for gastric cancer in monotherapy or in combining strategies, which has revolutionized the oncology landscape by targeting the host immune system [6]. For instance, blocking immune checkpoints such as cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4), programmed cell death-1 (PD-1) and its ligand (PD-L1 or B7-H1), has proven e cacy in several solid cancers [7]. Whether old healed TB play a role in the outcomes of gastric cancer patients via some immunotherapy ways like BCG remains unclear.
Considering the fact that no existing study has investigated the impact of old TB lesion on the clinicopathological characteristics and prognosis of gastric cancer, this study aimed to explore the impact of old healed TB lesion on the survival and clinicopathological features of gastric cancer patients based on the surgical outcomes from a high-volume center in Southwest China.

Materials And Methods
Study design and patient population Informed consent was waived due to the retrospective nature of this study. However, patient records were deidenti ed and anonymized before statistical analysis.
A total of 2649 patients who underwent radical cystectomy for gastric cancer were retrospectively collected from West China Hospital of China from January 1st, 2009 to December 31st, 2016. The survival outcomes of these patients were followed up for at least 3 years after surgery. The study owchart is presented in Fig. 1. The patient inclusion criteria were as follows: (1) those with histologically con rmed adenocarcinoma of the stomach; and (2) those who underwent curative-intent surgery. Meanwhile, patients conforming to any one of the following criteria were excluded from this study: (1) those with carcinoma of the remnant gastric; (2) those who received R1 or R2 surgery; (3) those with other malignancies in stomach; (4) those received previous chemotherapy and/or radiation therapy; (5) those with other malignancies prior to the operation; and (6)  There were 2568 patients in NTB group and 81 in TB group, respectively. As discovered from Table 1 and Table 2, there were some heterogeneities in baseline characteristics between the two groups. To be speci c, TB group was associated with a higher proportion of male patients, prior pulmonary comorbidities, ex-smoker or current smoker, PPCs and a smaller tumor size and shorter PODs. In addition, signi cant differences were also observed in the distributions of pathological stages and Bormann types. respectively. There were signi cant differences in OS between the two cohorts (Fig.2a). The estimated 3year OS rate in TB group was 84.7%, which was signi cantly higher than that of 69.2% in NTB group (p=0.005; Fig.2a). The median follow-up period in TB group was 57 months, about 7 months longer than that in NTB group (50 months, p = 0.011).
After one-to-two PSM, 73 pairs of patients were quali ed for further analysis. Good balance was demonstrated for both of the matched patients, with no clinically signi cant differences in the covariates, as depicted in Table 1 and Table 2. Besides, signi cant differences in OS rates were also detected in the PSM cohorts (Fig.2b). The 3-year OS rate in TB group was 90.4%, which was higher than that of 69.2% in NTB group (p=0.001; Fig.2b). After adjusting for potential confounders by multivariate Cox regression analysis, the coexistence of old TB lesion remained signi cantly and strongly associated with the decreased tumor-related mortality (adjusted HR: 0.220, [95% CI: 0.107-0.454], p<0.001), as shown in Table  5.

Univariate and Multivariate Cox Regression
In univariate analysis, old TB, age, ECOG score, pulmonary dysfunctions, perioperative blood transfusions, tumor size, surgical procedure, surgical approach, lower third and diffused locations, histological grading, neural/vascular invasion, Bormann's type, pathological T-N-M category, pathological TNM stage, POCs and adjuvant chemotherapy were associated with the patient's OS (Table 4). Based on univariate analysis results, factors of TB history, age, ECOG score, comorbidities, blood transfusions, tumor size, surgical procedure, approach, location, histological grading, Bormann's type, neural/vascular invasion, T-N-M stage, POCs and adjuvant chemotherapy were included for stepwise multivariate Cox regression analysis. It was discovered that old TB lesion and adjuvant chemotherapy were the independent factors to determine good survival, while older age, pulmonary dysfunctions, blood transfusion, bigger tumor size, diffused location, higher TNM classi cations, PPCs predicted poor prognosis (Table 4).
After PSM, old TB, tumor size, blood transfusion, surgical procedure, location, Bormann type, T-N-M category, P-stage, PPCs and adjuvant chemotherapy were identi ed to be associated with OS in the univariate analysis. Stepwise analysis of the multivariate Cox regression included the same covariates as in the multivariate analysis before PSM. And nally, several independent positive prognosticators for OSold TB, adjuvant chemotherapy, proximal gastrectomy and laparoscopic or robot-assisted gastrectomy, and several negative prognosticators including middle third or lower third locations and G3 and PPCs were identi ed after PSM. Detailed univariate and multivariate Cox regression analyses of OS after PSM are shown in Table 5.

Discussion
To the best of our knowledge, this study is by far the rst cancer series that demonstrates the in uence of a pulmonary TB history on the clinicopathological characteristics and prognosis of gastric cancer patients. According to our results, Old TB lesion is a risk factor for PPCs. Despite that, it represented as an independent positive factor for the prognosis of gastric cancer in general.
Our ndings were consistent with a previous work reporting that post-TB lesion was an independent risk factor for PPCs. The impacts of old TB on pulmonary impairment, which usually manifest as restrictive dysfunction, obstructive dysfunction or both, have been extensively acknowledged in previous literature [14] [15]. Furthermore, a prospective study by Eun Sun Kim et al [16]. on patients with non-small-cell lung cancer (NSCLC) revealed that, the presence of old TB lesion on chest X-ray radiograph was an independent risk factor for PPCs. Our ndings were similar to the previous results. The possible explanation may be the destructive and brosing properties of pulmonary TB, which cause pulmonary impairment [17], thus contributing to the increased incidence of PPCs compared with patients without a prior TB. Additionally, multivariable Cox regression analysis suggested that PPCs was the independent risk factor signi cantly associated with the poor outcomes of gastric cancer, consistent with a previous study [18]. These results suggest that special attentions should be paid to the perioperative respiratory management of the special group.
As for the positive effect of old TB on the survival of gastric cancer, the possible explanations are elaborated below. Firstly, old TB possibly improves the survival of gastric cancer patients by the same way as BCG. BCG has once been used as a nonspeci c immunotherapy to treat cancer, but it is largely abandoned later due to a lack of persistent effects. Even so, evidence from several clinical trials has indicated that BCG immunotherapy prolongs the survival time of gastric cancer patients. Although the precise mechanisms underlying the bene cial effects of BCG remain unclear so far, it is possible that the following factors may be involved, including the production of tumor necrosis factor (TNF)-alpha and other antitumor cytokines, as observed in bladder cancer [19], the enhancement of monocyte-and lymphocyte-mediated tumor cell-killing synergism [20], or the tolerance to chemotherapy [21]. Secondly, one research on mycobacterial TB product heparin-binding hemagglutinin (HBHA) unravels that HBHA, as an immunostimulant, can lyse tumor cells and suppress tumor growth by stimulating dendritic cells (DCs), which thereby further activate the T cell-mediated tumor cell killing [22]. Additionally, another study documents the anti-cancer activity of Mycobacterium indicus pranii (Mw) in melanoma, which shares the same antigen with mycobacterial TB [23]. Mw modi es MMP-9 expression in murine macrophages in vitro [24], thus defending against the migration and invasion of melanoma. It is reported that after TB infection, granulomas that contain quiet M. TB antigen and soluble pathogen contribute to the persistent memory of adaptive immune response [25], which may play its antitumor effects [26]. Therefore, it is reasonable to assume that adaptive immune response may exert the anti-cancer effects via TNF, DCs, T cells (especially the activated CTLs and Th1 cells), and some other potential ways, as implicated in the above studies.
In this study, it was noticed that several similar studies investigating the impacts of TB on the survival of lung cancer patients reported contradictory ndings. For instance, Kuo and coworkers [27] suggested that NSCLC patients with active TB showed better survival outcomes, and the effective T lymphocyte in ltration in tumor might be the underlying mechanism. Chia-Hao Chang et al. [28] reported that previous TB had a gender-dependent impact on lung cancer, with better 1-year progression-free survival (PFS) in female patients but worse PFS in male patients. In this case, the differences in carcinogenesis and in ammation between male and female patients might account for the related mechanisms. Contrarily, several studies show that old TB is negatively associated with the survival of lung cancer patients [29] [30]. For example, a previous study has indicated that old pulmonary TB is an independent risk factor for PPCs [16], which is also related to a poorer long-term outcome, with a 6-month reduction in the mean OS of lung cancer [31]. Thus, it was hypothesized that the prognostic effect of old TB on lung cancer patients might be confounded by PPCs, which were not included as a confounding factor in the analysis of all the above studies. In our study, it was found that old pulmonary TB was a protective factor for the OS of gastric cancer, but a risk factor for PPCs. Despite the higher incidence of PPCs in TB group and the negative effect on prognosis brought by PPCs, this study showed that old pulmonary TB still represented a favorable factor for the OS of gastric cancer. It is hypothesized that the overwhelming protective effect, which is contributed by old TB, may interpret the improved prognosis of gastric cancer patients.
In particular, 11 patients in our cohort had concurrent perigastric lymph node TB, of them, 6 cases without old pulmonary TB were assigned to NTB group. In view of the small sample size and the minimal heterogeneity in TB group, the 11 cases with lymph node TB were not analyzed separately in earlier analysis. These 11 patients had relatively early N stages (including 6 at N0, 4 at N1 and 1 at N2) and pathological TNM stages (including 5 at stage I, 4 at stage II, 1 at stage III and 1 at stage IV). The median OS for these 11 patients was 71.2 (range, 34.7-129.7) months, and the 3-year OS rate was 81.8% (95% CI: 44.7-95.1). Our ndings were consistent with a previous study reporting that gastric cancer patients with tumor-related sarcoid granuloma reactions in regional lymph nodes had favorable prognosis [32]. As far as we know, this is the rst study to illustrate the positive effect of perigastric lymph node TB on pathological patterns and prognosis. Typically, the protective effect of perigastric lymph node TB may be best explained by the effective T lymphocyte in ltration in tumor, as mentioned previously. Moreover, earlier studies have discovered that latent pulmonary TB can transport to regional lymph node and disseminate via the bloodstream [33]. No matter which body part they reach, the microorganisms may gain entry into epithelial cells and broblasts, or are engulfed by the local tissue macrophages. This usually elicits a low-grade in ammatory response and attracts the antigen-responding T-and B-cells. Hence, it is reasonable to believe that the in ammatory response possibly inhibits the migration of gastric cancer.
Certain limitations should be noted in this study. First, as a retrospective cohort study, the inherent nature determines its diminishing ability to determine causality. However, this study is a relatively large series of gastric cancer patients with a prior history of TB, and it is unlikely that large-scale prospective randomized trials will be organized, since gastric cancer concurrent with old healed TB is rare. Second, it is notable that some misclassi cation of exposure can not be exclude due to the self-report of TB. Not all subjects may remember their TB, or they may be unaware that they have had TB in the past. Thirdly, despite the relatively considerable sample size in a single center, it may still be too small to generalize to the entire population. Thereby, multi-center studies and/or prospective studies are warranted to verify the putative positive effect of old TB.
In conclusion, old TB lesion is a favorable prognostic factor for the survival of gastric cancer patients, but a risk factor for PPCs as well. Our study shed light on further immune therapy study for gastric cancer. Of course, attentions should also be paid to this special group. Due to the small sample size in our study, replication of our ndings in other cohorts is warranted. Apart from epidemiological studies, more research is needed to elicit the potential immunological links between TB and gastric cancer.

Availability of data and materials
The data that support the results of this research is available on request from the corresponding author.
Considering privacy or ethical restrictions, the data is not publicly available.
Ethics approval and consent to participate Patient records were de-identi ed and anonymized prior to the analysis. The Research Ethics Committee of West China Hospital approved this retrospective study, and the Surgical Gastric Cancer Patient Registry number was No. WCH-SGCPR-2020-04.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no con ict of interest.     Figure 1 Flowchart for the cohort study. NTB group: Non-Tuberculosis group; TB group: Tuberculosis group Kaplan-Meier curves demonstrating overall survival (OS) of (a) all the patients and (b) the matched patients between the TB group and NTB group TB, tuberculosis; NTB, non-tuberculosis