Early postoperative negative conversion of Helicobacter pylori stool antigen after distal gastrectomy and potential host-related factors

Abstract

BACKGROUND

The early postoperative behavior of Helicobacter pylori (H. pylori) after distal gastrectomy (DG) and the host factors influencing early negative conversion remain poorly defined. In this study, we evaluated early postoperative H. pylori dynamics using a standardized stool antigen test (SAT).

AIM

To evaluate early postoperative H. pylori dynamics and identified clinicopathological factors associated with early negative conversion.

METHODS

Among 129 patients who underwent DG for gastric cancer, 53 had active preoperative H. pylori infection, as confirmed by SAT. Postoperative status was reassessed at 3 months, and patients were categorized into the persistent infection (n = 13) or spontaneous negative conversion (n = 40) group. Clinicopathological variables were compared between groups. Multivariable logistic regression was performed, and sensitivity analyses included Firth penalized logistic regression. A post hoc power analysis was conducted to contextualize the sample size.

RESULTS

Early negative conversion on SAT occurred in 75.5% of the patients at 3 months. The early negative conversion group had a higher preoperative American Society of Anesthesiologists (ASA) score than the persistent infection group (2.10 ± 0.67 vs 1.69 ± 0.48; P = 0.024). In the primary multivariable model adjusting for age, body mass index, sex, and postoperative antibiotic use, a higher ASA score was directionally association with early negative conversion (odds ratio = 3.02; 95%CI: 0.91-10.01; P = 0.071). Sensitivity analyses using penalized likelihood and models incorporating serum albumin and hemoglobin produced directionally consistent estimates with wide confidence intervals. Post-hoc power analysis indicated limited statistical power (approximately 0.44) to detect an independent association with the current sample size.

CONCLUSION

A higher ASA score may be associated with early negative conversion after DG; however, the findings are exploratory and require validation in larger cohorts with longitudinal follow-up to confirm durable eradication.

Key Words: Helicobacter pylori; Distal gastrectomy; Gastric cancer; Stool antigen test; American Society of Anesthesiologists score; Negative conversion

Core Tip: The early postoperative course of Helicobacter pylori (H. pylori) following distal gastrectomy remains poorly characterized. Using a standardized stool antigen test protocol, this study demonstrated that spontaneous H. pylori clearance occurs in approximately three-quarters of patients within 3 months after surgery. Among evaluated clinicopathological variables, a higher preoperative American Society of Anesthesiologists score was associated with spontaneous clearance, suggesting that host-related factors may influence early postoperative H. pylori dynamics. These findings provide clinically relevant insights that may help optimize postoperative surveillance strategies and inform the timing of eradication therapy after distal gastrectomy.

INTRODUCTION

Gastric cancer (GC) remains a major global health burden and leading cause of cancer-related mortality worldwide, including in South Korea[1-3]. Helicobacter pylori (H. pylori) is a well-established carcinogen that plays a central role in gastric carcinogenesis through chronic inflammation, atrophic gastritis, and intestinal metaplasia[3]. Even after curative resection, H. pylori may contribute to persistent mucosal injury in the remnant stomach and has been implicated as a potential risk factor for remnant gastric cancer[4-6].

Distal gastrectomy (DG) induces substantial alterations gastric physiology, including reduced acid secretion, altered motility, vagotomy-related effects[7], and increased duodenogastric bile reflux[8-12]. These changes may impair the ability of H. pylori to survive in the remnant stomach, and eight studies have reported decreased infection prevalence and spontaneous clearance after DG[7-14]. The mechanisms underlying this phenomenon remain unclear but are thought to reflect the combined influence of hypoacidity, bile reflux, and alterations in the gastric mucosal environment.

Several studies have suggested that negative postoperative H. pylori status may be associated with poorer oncologic outcomes in patients undergoing curative surgery for GC[15,16], raising questions about the biological implications of bacterial persistence vs clearance. A recent multicenter prospective study confirmed a high rate of spontaneous H. pylori clearance one year after DG; however, it did not identify reliable early predictors and relied on heterogeneous diagnostic methods[17].

Despite these findings, important gaps remain. Most prior studies have focused on long-term postoperative outcomes, and few have examined early clearance dynamics using standardized diagnostic modalities. Furthermore, the potential contribution of systemic host factors - rather than local gastric changes to early H. pylori persistence or clearance has not been adequately investigated.

Therefore, this study aimed to evaluate early postoperative changes in the H. pylori status and to identify clinicopathological factors associated with spontaneous negative conversion in patients undergoing DG for GC.

MATERIALS AND METHODS

Study population

Between June 2022 and August 2025, 262 patients underwent DG for histologically confirmed gastric adenocarcinoma at a single tertiary institution. Of these, 129 patients met the following criteria and were enrolled in the study: (1) Availability of both preoperative and postoperative serum anti-H. pylori immunoglobulin G (IgG) result; (2) Availability of both preoperative and postoperative stool antigen test (SAT) results; and (3) No history of H. pylori eradication therapy before surgery or during the postoperative follow-up.

The final study cohort comprised 129 patients. Among them, patients with active preoperative H. pylori infection confirmed by SAT were further analyzed to assess postoperative persistence vs spontaneous negative conversion (Figure 1).

Figure 1 Flow diagram of patient inclusion and selection. A total of 262 patients underwent distal gastrectomy for gastric adenocarcinoma between June 2022 and August 2025. Among them, 129 patients had both preoperative and postoperative stool antigen test for Helicobacter pylori (H. pylori) and serum anti-H. pylori immunoglobulin G available. After excluding patients who received H. pylori eradication therapy before surgery or during follow-up, and those without complete paired testing, 53 patients with active preoperative infection constituted the analytic cohort for evaluating postoperative persistence vs spontaneous clearance. DG: Distal gastrectomy; SAT: Stool antigen test.

Preoperative and postoperative assessment

Preoperative evaluation included measurement of serum anti-H. pylori IgG and H. pylori SAT using enzyme immunoassay techniques. All patients underwent upper endoscopy before surgery. Postoperative follow-up was conducted three months after DG and included clinical evaluation, physical examinations, laboratory testing, upper endoscopy, and repeat SAT to reassess the H. pylori status.

Surgical procedure

All procedures were performed by a single experienced gastric surgeon at a single tertiary institution. The surgical approach followed the Korean Practice Guidelines for Gastric Cancer established by the Korean Gastric Cancer Association[18]. All patients underwent DG with Billroth-II reconstruction combined with Braun enteroenterostomy, ensuring a uniform postoperative anatomical configuration across the cohort. A standard D2 lymph node dissection was performed in all cases; in accordance with the accepted oncologic principles for GC surgery.

Perioperative antibiotic use

All patients received prophylactic intravenous first-generation cephalosporins on the day of surgery in accordance with the institutional protocol. No routine postoperative antibiotics beyond the prophylactic regimen were administered. In this study, postoperative antibiotic exposure was defined as administration of additional broad-spectrum antibiotics exclusively in patients who developed Clavien-Dindo grade ≥ II complications (e.g., infectious complications requiring therapeutic antibiotics). This definition enabled evaluation of postoperative antibiotics independent of standard surgical prophylaxis.

Proton pump inhibitor administration

Proton pump inhibitor (PPI) use was verified through review of the electronic medical records. Patients discontinued PPIs at least two weeks before the postoperative H. pylori SAT, in accordance with established diagnostic guidelines, to minimized the risk of false-negative results. The using timing of PPI discontinuation relative to SAT was confirmed to reduce confounding effects on H. pylori conversion.

H. pylori status definition

Patients who tested positive for H. pylori on preoperative SAT were categorized at three months postoperatively as follows: (1) Persistent infection - positive SAT result both before and after surgery; and (2) Negative conversion - positive SAT preoperatively and negative postoperatively without eradication therapy.

Patients with negative preoperative SAT results were excluded from comparative analyses of clearance patterns.

Statistical analysis

Continuous variables are presented as mean ± SD and were compared using Welch’s t test. Categorical variables are presented as n (%) and were compared using Fisher’s exact test. Factors associated with early postoperative negative conversion were evaluated using multivariable logistic regression with maximum likelihood estimation. Given the limited number of outcome events, Firth penalized logistic regression was additionally performed as a sensitivity analysis to reduce small-sample bias (Supplementary Table 1). To assess the risk of type II error, a post-hoc power analysis was conducted using a Wald-based normal approximation derived from the observed American Society of Anesthesiologists (ASA) effect size (Supplementary Table 2). The approximate sample size required to achieve 80% power under a similar effect size was also estimated.

RESULTS

Patient characteristics

The analysis included 129 patients who underwent DG. Among them, 53 (41.1%) had active H. pylori infection preoperatively, whereas 76 (58.9%) showed no active infection. Baseline clinicopathological features were generally comparable between two groups (Table 1). However, patients with active infection were significantly younger than those without infection (62.4 ± 11.7 years vs 66.8 ± 11.7 years; P = 0.036). The sex, body mass index (BMI), ASA score, tumor depth (early gastric cancer vs advanced gastric cancer), nodal status, Clavien-Dindo grade, PPI use, antibiotic use, and serum H. pylori IgG results did not differ significantly between groups (all P > 0.05).

Table 1 Baseline clinicopathological characteristics of the overall cohort according to preoperative Helicobacter pylori infection status, n (%)/mean ± SD.

Variables	Total (n = 129)	Preoperative status of helicobacter infection		P value
		No active infection	Active infection
Overall	129 (100.0)	76 (58.9)	53 (41.1)
Age (years)	65.0 ± 11.9	66.8 ± 11.7	62.4 ± 11.7	0.036
Sex				0.634
Male	92 (71.3)	53 (69.7)	39 (73.6)
Female	37 (28.7)	23 (30.3)	14 (26.4)
Body mass index (kg/m2)	24.7 ± 3.3	24.8 ± 3.3	24.5 ± 3.3	0.652
American Society of Anesthesiologists score	2.08 ± 0.65	2.13 ± 0.66	2.00 ± 0.65	0.264
Tumor depth (EGC vs AGC)				0.834
EGC	106 (82.2)	62 (81.6)	44 (83.0)
AGC	23 (17.8)	14 (18.4)	9 (17.0)
Nodal status				0.3031
Negative	111 (86.0)	63 (82.9)	48 (90.6)
Positive	18 (14.0)	13 (17.1)	5 (9.4)
Clavien-Dindo Grade				0.453
0-I	103 (79.8)	59 (77.6)	44 (83.0)
≥ II	26 (20.2)	17 (22.4)	9 (17.0)
Use of proton pump inhibitors				0.486
Negative	120 (93.0)	72 (94.7)	48 (90.6)
Positive	9 (7.0)	4 (5.3)	5 (9.4)
Postoperative antibiotics use				0.565
Negative	104 (80.6)	60 (78.9)	44 (83.0)
Positive	25 (19.4)	16 (21.1)	9 (17.0)
Serum helicobacter immunoglobulin G				< 0.001
Negative (%)	25 (19.4)	24 (31.6)	1 (1.9)
Positive (%)	104 (80.6)	52 (68.4)	52 (98.1)

¹Fisher’s exact test, or Welch’s t-test as appropriate.

Comparisons between groups were performed using χ² test. AGC: Advanced gastric cancer; EGC: Early gastric cancer.

Negative conversion of H. pylori

Among the 53 patients with preoperative active infection, 40 (75.5%) exhibited spontaneous negative conversion three months postoperatively, whereas 13 (24.5%) had persistent infection. In univariate analyses, preoperative ASA score was significantly higher in the early negative conversion group than in the persistent infection group (P = 0.024). No other demographic, laboratory, and perioperative factors showed statistically significant differences (Table 2).

Table 2 Comparison between persistent infection and early negative conversion groups, n (%)/mean ± SD.

Variable	Total (n = 53)	Persistent (n = 13)	Negative conversion (n = 40)	P value
Age (years)	62.4 ± 11.7	61.5 ± 9.9	62.7 ± 12.4	0.711
Sex, male	39 (73.6)	8 (61.5)	31 (77.5)	0.292
Body mass index (kg/m2)	24.5 ± 3.3	24.3 ± 3.9	24.6 ± 3.1	0.809
American Society of Anesthesiologists score (continuous)	2.0 ± 0.7	1.7 ± 0.5	2.1 ± 0.7	0.024
Diabetes mellitus	15 (28.3)	1 (7.7)	14 (35.0)	0.08
Hypertension	17 (32.1)	2 (15.4)	15 (37.5)	0.183
Chronic kidney disease	2 (3.8)	0 (0.0)	2 (5.0)	1
Serum albumin (g/dL)	4.2 ± 0.4	4.2 ± 0.5	4.2 ± 0.4	0.804
Hemoglobin (g/dL)	13.2 ± 2.0	13.8 ± 1.5	13.0 ± 2.1	0.168
Follow-up time (days after surgery)	110.8 ± 19.4	106.5 ± 7.8	112.2 ± 21.8	0.169
Postoperative proton pump inhibitor use	2 (3.8)	0 (0.0)	2 (5.0)	1
Postoperative antibiotics use	8 (15.1)	1 (7.7)	7 (17.5)	0.662
Operation time (minutes)	165.8 ± 50.3	169.6 ± 41.8	164.5 ± 53.2	0.724
Estimated blood loss (mL)	97.2 ± 74.0	75.1 ± 29.7	104.5 ± 82.6	0.063
Postoperative complications (Clavien-Dindo ≥ II)	9 (17.0)	1 (7.7)	8 (20.0)	0.424

Continuous variables compared using Welch’s t test. Categorical variables compared using Fisher’s exact test.

Multivariate analysis for predictors in negative conversion of H. pylori

In the primary multivariate logistic regression model (age, BMI, sex, and postoperative antibiotic use), ASA score showed a directional association with early negative conversion [odds ratio (OR) = 3.02; 95%CI: 0.91-10.01; P = 0.071; Table 3]. Given the limited number of outcome events, this finding should be interpreted as a hypothesis-generating. In sensitivity analyses using Firth penalized logistic regression, the ASA association remained directionally consistent but was attenuated (OR = 2.55; 95%CI: 0.83-7.85; P = 0.104).

Table 3 Primary multivariable logistic regression for early negative conversion (maximum likelihood estimation vs Firth).

Variable	MLE OR (95%CI)	MLE P value	Firth OR (95%CI)	Firth P value
American Society of Anesthesiologists score (per 1-point increase)	3.02 (0.91-10.01)	0.071	2.55 (0.83-7.85)	0.104
Age (per 1-year increase)	1.02 (0.96-1.07)	0.592	1.01 (0.96-1.07)	0.629
Body mass index (per 1 kg/m2 increase)	1.09 (0.87-1.36)	0.46	1.07 (0.87-1.33)	0.507
Sex (female vs male)	0.48 (0.11-2.03)	0.319	0.51 (0.13-2.07)	0.346
Postoperative antibiotics use (yes vs no)	1.85 (0.19-18.07)	0.596	1.46 (0.19-11.01)	0.716

Firth: Firth penalized logistic regression (sensitivity analysis to reduce small-sample bias); MLE: Maximum likelihood estimation; OR: Odds ratio.

The postoperative use of antibiotics also showed an elevation in odds but did not reach statistical significance (OR = 1.85, 95%CI: 0.19-18.07; P = 0.596). The age, BMI, and sex were not significantly associated with negative conversion (all P > 0.05). In the forest plot summarizing the multivariate estimates, the ASA score demonstrated the most prominent directional effect among all variables examined (Figure 2).

Figure 2 Forest plot of multivariate logistic regression analysis evaluating predictors of spontaneous. Helicobacter pylori clearance after distal gastrectomy. Odds ratios and 95%CI are shown for age, body mass index, sex, postoperative antibiotic use, and American Society of Anesthesiologists score. American Society of Anesthesiologists score demonstrated the strongest directional association with spontaneous clearance, although statistical significance was not reached. ASA: American Society of Anesthesiologists; BMI: Body mass index; H. pylori: Helicobacter pylori.

DISCUSSION

In this cohort, the ASA score demonstrated a directional association with early postoperative negative conversion of H. pylori on SAT. Although the multivariable estimates were imprecise and did not consistently reach statistical significance across models, the observed effect size suggests a hypothesis-generating signal that warrants validation in larger cohorts with rigorous control of confounding and adequate statistical power. Importantly, the ASA signal remained directionally consistent in sensitivity analyses incorporating laboratory markers of systemic status (serum albumin and hemoglobin), suggesting that ASA may capture broader host vulnerability not fully reflected by single laboratory parameters.

These findings should be interpreted alongside postoperative physiological changes after DG and Billroth-II reconstruction with Braun anastomosis, including reduced acid secretion and increased duodenogastric bile reflux, which may alter the gastric microenvironment and affect H. pylori viability. Perioperative antibiotics and PPI may also influence early testing results; therefore, persistent longitudinal testing is required to determine whether early negative conversion reflects durable eradication or transient negativity.

This single-center cohort study demonstrated a high rate of spontaneous H. pylori clearance (75.5%) three months following DG. This finding is consistent with eleven studies reporting reduced H. pylori prevalence after gastric resection[7-14,19,20], including the recent multicenter prospective study by Omori et al[17], which reported a 76.9% clearance rate at 1 year postoperatively. Collectively, these findings indicate that spontaneous postoperative clearance is a reproducible phenomenon across diverse populations and surgical approaches; and that a substantial proportion of patients achieve eradication without antibiotic therapy. Importantly, our shorter follow-up interval suggests that most of the spontaneous eradication likely occurred during the early postoperative period, with later time points reflecting stabilization rather than ongoing clearance.

A key novel finding of our study was the directional association between higher preoperative ASA scores and an increased likelihood of spontaneous H. pylori negative conversion. Although the multivariate analysis yielded a borderline P value (OR = 3.02; 95%CI: 0.91-10.01; P = 0.071), the magnitude of effect suggests a potentially meaningful host-related contribution. In the multivariate forest plot, in which the ASA score shows the most pronounced directional effect among all variables examined (Figure 2). Six studies have primarily emphasized local gastric mechanisms, such as bile reflux, postoperative hypoacidity, mucosal inflammation[7-12], and altered luminal conditions, to explain the reduced H. pylori colonization after DG. Our findings extend this conceptual framework by suggesting that the systemic physiological status, reflected by the ASA score, may also play substantial role in shaping the postoperative microbial ecology of the remnant stomach.

This concept is supported by the growing evidence that H. pylori persistence depends on its capacity to induce immune tolerance in immunocompetent hosts[21]. The bacterium influences both the innate and adaptive pathways, generating a microenvironment favorable for long-term colonization. Patients with poorer systemic health or impaired recovery; more common among those with higher ASA scores; may experience the disruption of this immunologic equilibrium. Such disruption of these host microbe interactions could render the remnant stomach less favorable for bacterial persistence, thereby facilitating spontaneous clearance.

Local physiological changes after DG may further amplify this effect. Surgery markedly changes remnant gastric physiology, including acid suppression[7], bile reflux, and motility patterns[8-12]. Experimental studies have demonstrated nitric oxide-mediated relaxation and functional remodeling of gastric smooth muscles after gastrectomy[22]. These motility changes may destabilize the colonization niche, particularly when combined with systemic frailty, thereby synergistically reducing the H. pylori viability within the postoperative stomach.

The age-related differences observed in our cohort further highlights host-dependent influences. Patients with persistent active infection were significantly younger, consistent with epidemiological data showing a higher H. pylori prevalence in younger populations[2,3]. Omori et al[17] also identified age and sex host characteristics influencing postoperative infection patterns. However, although age was associated with preoperative infection prevalence, it did not independently predict postoperative clearance in multivariate analysis. This suggests that systemic physiological vulnerability, as reflected by the ASA score, may outweigh the chronological age in determining the early postoperative ecological stability of H. pylori. Collectively, these findings indicate that postoperative H. pylori dynamics are shaped by both local gastric factors and systemic physiology and host immunity.

Our findings, in comparisons with the Omori et al’s cohort[17], also suggest that reconstruction method may play a less dominant role than previously assumed. The prospective study by Omori et al[17]. included Billroth-I and Roux-en-Y reconstructions (Billroth-II in only, 0.5%), yet their clearance rates (76.9%) were virtually identical to that observed in our cohort (75.5%), despite our cohort consisting entirely of Billroth-II with Braun anastomosis reconstructions. Previous studies comparing reconstruction types have shown that bile reflux and chemical gastritis are more frequent and severe after Billroth-II reconstruction, whereas H. pylori-associated inflammatory cell infiltration is more prominent when bile reflux is limited, suggesting a reciprocal relationship between bile exposure and H. pylori-driven inflammation in the remnant stomach[10-12]. Nevertheless, the almost identical clearance rates across fundamentally different reconstruction methods indicate that reconstruction-driven factors, particularly bile reflux, may not be the dominant determinants of postoperative H. pylori clearance. Despite the well-documented differences in bile exposure between Billroth-II and Billroth-I/Roux-en-Y reconstructions, the clearance rates in our cohort and those of Omori et al[17] argue that bile reflux related mucosal changes alone are insufficient to explain spontaneous eradication. Rather, the similarity in clearance rates, regardless of the reconstruction method, suggests that host-driven physiological and immunologic factors exert a stronger influence over the remnant gastric ecology, overshadowing anatomic and luminal variations.

Several studies have described a paradox in which a negative H. pylori e status after gastrectomy is associated with poorer oncologic outcomes[15,16]. Our findings offer a possible explanation: If spontaneous clearance is more common among physiologically frail patients (those with higher ASA scores), then postoperative negativity may serve as an indirect marker of underlying vulnerability rather than a casual driver of worse prognosis. This interpretation distinguishes host factors from the bacterial status; and may help conflicting observations in the literature.

This study has several limitations. First, the sample size was small (n = 53, H. pylori-positive patients), and the number of persistent infections (n = 13) limited the statistical power and increased the risk of model overfitting; therefore, the observed ASA association should be regarded as exploratory rather than confirmatory. Second, H. pylori status was assessed at a single postoperative time point (3 months). Early negative conversion on SAT cannot distinguish transient negativity from durable eradication, and false-negative SAT results may occur in the early postoperative period due to altered gastrointestinal physiology, bile reflux, mucosal injury, or medication effects. Third, the findings derive from a single-center East Asian gastric cancer cohort undergoing D2 lymphadenectomy followed by Billroth-II reconstruction with Braun anastomosis, and may not be generalizable to other populations, surgical extents, or reconstruction methods.

Future studies should validate the role of systemic host factors in postoperative H. pylori clearance through adequately powered prospective multicenter investigations. Incorporating detailed comorbidity assessments, medication profiles, nutritional and immunological markers, remnant gastric motility evaluations, and standardized SATs may provide deeper insights into the interaction between the host physiology and postoperative microbial ecology. These findings may support individualized postoperative surveillance strategies and guide evidence-based decisions regarding the need for H. pylori eradication therapy after DG.

CONCLUSION

A substantial proportion of patients exhibited early postoperative negative conversion of H. pylori on SAT at 3 months after DG. Higher ASA score showed a directional association with early negative conversion. However, given the limited statistical power and single-time-point assessment, these findings should be regarded as hypothesis-generating. Larger multicenter studies with longitudinal follow-up are required to confirm durable eradication and to clarify the clinical implications for postoperative surveillance and eradication strategies.