Impact of diabetes on recovery after radical gastrectomy for gastric cancer: A retrospective cohort study

Abstract

BACKGROUND

Gastric cancer remains a significant global health concern. Radical gastrectomy is the primary curative treatment. Diabetes mellitus is a common comorbidity in patients undergoing surgery for gastric cancer, including radical gastrectomy. Previous studies have suggested that diabetes can negatively affect postoperative outcomes, such as wound healing, infection rates, and overall recovery. However, the specific impact of diabetes on recovery after radical gastrectomy for gastric cancer remains poorly understood. evaluate the influence of diabetes on postoperative recovery, including hospital stay duration, complications, and readmission rates, in patients undergoing gastrectomy for gastric cancer. Understanding these effects could help optimize perioperative management and improve patient outcomes.

AIM

To investigate the impact of diabetes on recovery after radical gastrectomy for gastric cancer and associated postoperative outcomes.

METHODS

This retrospective cohort study was performed at the Endocrinology Department of Xuanwu Hospital, Capital Medical University, Beijing, China. We examined patients who underwent radical gastrectomy for cancer between January 2010 and December 2020. The patients were divided into the diabetes and non-diabetes groups. The main outcomes included length of hospital stay, postoperative complications, and 30-day readmission rate. Secondary outcomes included quality of life indicators. Propensity score matching was used to adjust for potential confounding factors.

RESULTS

A total of 1210 patients were included in the study, with 302 diabetic patients and 908 non-diabetic patients. After propensity score matching, 280 patients were included in each group. Diabetic patients demonstrated significantly longer hospital stays (mean difference 2.3 days, 95%CI: 1.7-2.9, P < 0.001) and higher rates of postoperative complications (OR 1.68, 95%CI: 1.32-2.14, P < 0.001). The 30-day readmission rate was also higher in the diabetic group as compared to the non-diabetic group (12.5% vs 7.8%, P = 0.02).

CONCLUSION

Patients with diabetes mellitus undergoing radical gastrectomy for gastric cancer experience prolonged hospital stay, increased postoperative complications, and higher readmission rates, thus requiring optimized perioperative management strategies.

Key Words: Gastric cancer; Diabetes mellitus; Radical gastrectomy; Postoperative recovery; Complications

Core Tip: This study evaluated the effect of diabetes mellitus on postoperative recovery in patients with gastric cancer who underwent radical gastrectomy. Diabetic patients experienced longer hospital stays, higher rates of postoperative complications, and increased 30-day readmission rates than non-diabetic patients. These results underscore the importance of tailored perioperative management strategies to improve the outcomes of patients with diabetes undergoing surgical interventions. Optimal care is essential to mitigate the adverse effects of diabetes on recovery following gastrectomy.

INTRODUCTION

Gastric cancer remains the fifth most common malignancy and third leading cause of cancer-related deaths worldwide, with an estimated 1089103 new cases and 768793 deaths in 2020[1]. Despite advancements in early detection and treatment modalities, the prognosis of patients with gastric cancer remains poor, with a 5-year survival rate of approximately 30% in most countries[2]. Radical gastrectomy combined with appropriate lymphadenectomy continues to be the cornerstone of curative treatment for gastric cancer[3].

The outcomes of radical gastrectomy can be significantly influenced by various factors, including comorbidities. Diabetes mellitus, a chronic metabolic disorder characterized by persistent hyperglycemia, is a comorbidity that has garnered increasing attention in recent years. The global prevalence of diabetes has been rising steadily, with an estimated 463 million adults living with diabetes as of 2019. This figure is projected to reach 700 million by 2045[4].

The relationship between diabetes and cancer has been extensively researched, with evidence suggesting that diabetes may increase the risk of various cancers, including gastric cancer[5]. Moreover, diabetes has been associated with poorer outcomes in patients with cancer, including increased mortality and reduced quality of life (QOL)[6]. However, the specific effects of diabetes on recovery after radical gastrectomy for gastric cancer remain unclear.

Several molecular and cellular mechanisms underlie the adverse effects of diabetes on postoperative recovery. At the molecular level, chronic hyperglycemia leads to the formation of advanced glycation end products, which impair collagen synthesis and crosslinking, compromising wound healing[7]. Diabetes-induced microvascular dysfunction is characterized by reduced endothelial nitric oxide production and increased oxidative stress, which compromises tissue perfusion and oxygenation. Furthermore, diabetes impairs immune function through multiple pathways, including decreased neutrophil chemotaxis and phagocytosis, impaired T-cell responses, and dysregulated cytokine production[8].

Previous studies investigating the effects of diabetes on the outcomes after gastric cancer surgery have yielded conflicting results. Some studies have reported increased postoperative complications and mortality in patients[9,10], whereas other studies have found no significant differences[11,12]. These inconsistencies may be attributed to variations in study design, sample size, and definition of diabetes-related outcomes.

Given the high prevalence of both gastric cancer and diabetes, particularly in aging populations, understanding the interplay between these conditions is crucial for optimizing patient care and improving patient outcomes. This knowledge can inform preoperative risk assessment, guide perioperative management strategies, and assist in postoperative care planning for patients with diabetes undergoing radical gastrectomy.

This study aimed to address this knowledge gap by conducting a comprehensive analysis of the effects of diabetes on recovery after radical gastrectomy for gastric cancer.

MATERIALS AND METHODS

Study design and patient population

We performed a retrospective cohort study using data from Xuanwu Hospital, Capital Medical University, China, a large tertiary medical center with a high volume of cancer surgeries. This study was approved by the Ethics Committee of the Xuanwu Hospital, Capital Medical University, and informed consent was obtained from all participants. All patients who underwent radical gastrectomy for histologically confirmed gastric adenocarcinoma between January 1, 2010 and December 31, 2020 were eligible for inclusion. Exclusion criteria were: Age < 18 years, palliative or non-curative intent surgery, concurrent malignancies, incomplete medical records, and loss to follow-up within 30 days of surgery.

Data collection

Patient data were extracted from electronic medical records and the hospital's gastric cancer database following a standardized protocol. Two trained research assistants independently extracted the data and any discrepancies were resolved through discussion with a senior researcher. Data quality was ensured through regular audits of 10% of the randomly selected records. For missing data, defined as data present in < 5% of cases, multiple imputations using chained equations were performed with 20 iterations, including all baseline characteristics, treatment variables, and outcomes in the imputation model. Information collected from each patient included: (1) Demographic characteristics of age, sex, body mass index (BMI), smoking status, and alcohol consumption; (2) Comorbidities including hypertension, cardiovascular disease, chronic obstructive pulmonary disease, and chronic kidney disease; (3) Diabetes-related data including duration of diabetes, type of diabetes (type 1 or 2), glycated hemoglobin (HbA1c) levels, and diabetes medication; (4) Tumor characteristics including TNM stage (according to the 8th edition of the American Joint Committee on Cancer staging system), tumor location, and histological type; (5) Surgical details including type of gastrectomy (total or subtotal), extent of lymphadenectomy, operative time, and estimated blood loss; and (6) Perioperative management including use of neoadjuvant or adjuvant therapy, antibiotic prophylaxis, and venous thromboembolism prophylaxis.

Definition of diabetes mellitus

Patients were classified as having diabetes mellitus if they met one or more of the following criteria: (1) Documented diagnosis of diabetes in medical records; (2) Use of antidiabetic medications (oral agents or insulin) at the time of hospital admission; (3) HbA1c ≥ 6.5% within 3 months prior to surgery; and (4) Fasting plasma glucose ≥ 126 mg/dL (7.0 mmol/L) on two separate occasions during the preoperative evaluation.

Primary and secondary outcome measures

There were three primary outcomes. First, length of hospital stay was defined as the number of days from the date of surgery to the date of discharge. Second, postoperative complications were assessed using the Clavien-Dindo classification system[13], which is a standardized method for grading surgical complications based on the type of treatment required to correct the complication. Complications of grade II or higher were considered significant, which helps standardize the reporting of surgical outcomes and allows for meaningful comparisons across different studies. Third, the 30-day readmission rate was defined as any unplanned readmission to the hospital within 30 days of discharge.

The secondary outcome of QOL was assessed using the European Organization for Research and Treatment of Cancer QOL Questionnaire (EORTC QLQ-C30) and gastric cancer-specific module (QLQ-STO22)[14-16].

Statistical analyses

All statistical analyses were performed using R-software, version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was defined as a two-sided P-value of < 0.05.

Continuous variables are presented as mean ± SD or median with interquartile range (IQR), depending on the distribution of data. Categorical variables are expressed as frequencies and percentages. Normality was assessed using the Shapiro-Wilk test and visual inspection of the Q-Q plots.

To minimize the impact of potential confounding factors, we employed propensity score matching. The propensity score was calculated using a logistic regression model that included the following variables: Age, sex, BMI, smoking status, alcohol consumption, comorbidities, tumor stage, and type of gastrectomy. Diabetic and non-diabetic patients were matched 1: 1 using the nearest neighbor method with a caliper width of 0.2 SDs of the logit of the propensity score. The balance of covariates before and after matching was assessed using standardized mean differences, with a value < 0.1 considered indicative of good balance.

For the matched cohort, continuous outcomes were compared using paired t-tests or Wilcoxon signed-rank tests as appropriate. McNemar's test was used to assess categorical outcomes. The length of hospital stay was also analyzed using a linear regression model, adjusting for potential residual confounders. Postoperative complications were compared using logistic regression and the results were presented as ORs with 95%CIs. The 30-day readmission rates were analyzed using the Cox proportional hazards model, and the results were expressed as HRs with 95%CIs.

QOL scores were compared between diabetic and non-diabetic patients using linear mixed-effects models to account for repeated measurements over time. The models included fixed effects for diabetes status, time point, and their interaction as well as random effects for patients.

Sample size calculation

Based on previous studies, we estimated that the presence of diabetes is associated with a 20% increase in the length of hospital stay. Assuming an SD of 5 days in both groups, a power of 80%, and a two-sided alpha of 0.05, we calculated that a minimum of 250 patients per group would be required to detect this difference. To account for potential exclusions and increase the power of the secondary analyses, we aimed to include at least 300 patients in each group after propensity score matching.

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki, and adhered to the ethical guidelines for medical and health research involving human subjects. Patient confidentiality was maintained throughout the study and all data were de-identified prior to analysis. The study results will be reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) Guidelines for cohort studies[17].

RESULTS

Patient characteristics

In total, 1427 patients underwent radical gastrectomy for gastric cancer during the study period. After applying the exclusion criteria, 1210 patients were included in the final analysis, comprising 302 patients with diabetes mellitus (25%) and 908 patients without diabetes (75%). Propensity score matching created two comparable groups of 280 patients (560 patients each) for the primary analysis.

Table 1 shows the baseline characteristics of the study population before and after propensity score matching. Before matching, diabetic patients were significantly older (mean age 68.5 ± 9.2 vs 63.7 ± 11.3 years, P < 0.001) and had a higher BMI (26.3 ± 4.1 vs 24.8 ± 3.7 kg/m², P < 0.001) compared to non-diabetic patients. They also had a higher prevalence of hypertension (62% vs 41%, P < 0.001) and cardiovascular diseases (28% vs 17%, P < 0.001). After propensity score matching, these differences were substantially reduced, with all standardized mean differences of < 0.1, indicating a good balance between the groups.

Table 1 Baseline characteristics of patients before and after propensity score matching, n (%).

Characteristic	Before matching			After matching
	Diabetic (n = 300)	Non-diabetic (n = 900)	P value	Diabetic (n = 280)	Non-diabetic (n = 280)	P value
Age (years)	68.5 ± 9.2	63.7 ± 11.3	< 0.001	67.8 ± 9.5	67.3 ± 10.1	0.54
Male sex	190 (62.9)	536 (59.0)	0.22	172 (61.4)	169 (60.4)	0.80
BMI (kg/m²)	26.3 ± 4.1	24.8 ± 3.7	< 0.001	25.9 ± 3.9	25.7 ± 3.8	0.53
Hypertension	187 (61.9)	372 (41.0)	< 0.001	168 (60.0)	165 (58.9)	0.79
CVD	85 (28.1)	154 (17.0)	< 0.001	75 (26.8)	72 (25.7)	0.77
Tumor stage			0.18			0.95
I	70 (23.2)	236 (26.0)		66 (23.6)	68 (24.3)
II	93 (30.8)	300 (33.0)		89 (31.8)	87 (31.1)
III	139 (46.0)	372 (41.0)		125 (44.6)	125 (44.6)

BMI: Body mass index; CVD: Cardiovascular disease.

Among patients with diabetes, 267 (88.4%) had type 2 diabetes and 35 (11.6%) had type 1 diabetes. The median duration of diabetes was 8.5 years (IQR: 4.2-13.7 years). The mean HbA1c level was 7.4% ± 1.2%. Regarding diabetes management, 143 patients (47.4 %) were on oral antidiabetic medications alone, 84 (27.8%) were on insulin therapy, and 75 (24.8%) were on a combination of oral medication and insulin.

Primary outcomes

In the matched cohort, patients with diabetes had a significantly longer mean length of hospital stay compared to non-diabetic patients (12.7 ± 5.3 days vs 10.4 ± 4.1 days, P < 0.001). The mean difference was 2.3 days (95%CI: 1.7-2.9 days). After adjusting for potential residual confounders in a linear regression model, diabetes remained significantly associated with increased length of stay (β = 2.1 days, 95%CI: 1.5-2.7 days, P < 0.001).

Diabetic patients experienced a higher rate of overall postoperative complications (Clavien-Dindo grade II or higher) than non-diabetic patients (38.2% vs 26.8%, P = 0.004). The odds ratio for experiencing a complication in diabetic patients was 1.68 (95%CI: 1.32-2.14, P < 0.001) after adjusting for age, BMI, and tumor stage.

As shown in Table 2, patients with diabetes had significantly higher rates of surgical site infections (15.7% vs 8.9%, P = 0.015), pneumonia (12.5% vs 7.1%, P = 0.037), and urinary tract infections (9.3% vs 4.6%, P = 0.029). Additional complications listed in Table 2 included anastomotic leakage (6.1% vs 3.9%, P = 0.241) and deep vein thrombosis (3.2% vs 1.8%, P = 0.281), though these differences were not statistically significant.

Table 2 Incidence of specific postoperative complications in diabetic and non-diabetic patients, n (%).

Complication	Diabetic (n = 280)	Non-diabetic (n = 280)	P value
Surgical site infection	44 (15.7)	25 (8.9)	0.015
Pneumonia	35 (12.5)	20 (7.1)	0.037
Urinary tract infection	26 (9.3)	13 (4.6)	0.029
Anastomotic leakage	17 (6.1)	11 (3.9)	0.241
Deep vein thrombosis	9 (3.2)	5 (1.8)	0.281

The 30-day readmission rate was significantly higher in the diabetes group than in the non-diabetes group (12.5% vs 7.8%, P = 0.020). The HR for readmission in diabetic patients was 1.64 (95%CI: 1.18-2.28, P = 0.003) after adjusting for age, comorbidities, and postoperative complications.

Secondary outcomes

Analysis of the EORTC QLQ-C30 and QLQ-STO22 questionnaires revealed that diabetic patients had lower QOL than non-diabetic patients in several domains. As shown in Table 3, significant differences were observed in physical functioning (mean difference: -8.3 points, 95%CI: -11.7 to -4.9, P < 0.001), fatigue (+7.2 points, 95%CI: 4.1 to 10.3, P < 0.001), and pain (+6.5 points, 95%CI: 3.2 to 9.8, P < 0.001) at 3 months post-surgery. These differences persisted, albeit to a lesser extent, at 6 and 12 months postoperatively, with data presented in Table 3 showing physical functioning differences decreasing to -5.6 points (95%CI: -8.5 to -2.7, P < 0.001) at 12 months.

Table 3 Mean quality of life scores at 3, 6, and 12 months following surgery, mean ± SD.

Domain	Time point	Diabetic	Non-diabetic	Mean difference (95%CI)	P value
Physical functioning	3 months	65.2 ± 18.7	73.5 ± 16.9	-8.3 (-11.7 to -4.9)	< 0.001
	6 months	71.8 ± 17.3	78.9 ± 15.6	-7.1 (-10.2 to -4.0)	< 0.001
	12 months	75.6 ± 16.1	81.2 ± 14.8	-5.6 (-8.5 to -2.7)	< 0.001
Fatigue	3 months	43.7 ± 22.4	36.5 ± 20.1	7.2 (4.1 to 10.3)	< 0.001
	6 months	38.9 ± 20.8	33.1 ± 18.7	5.8 (2.9 to 8.7)	< 0.001
	12 months	35.2 ± 19.5	30.8 ± 17.9	4.4 (1.7 to 7.1)	0.002
Pain	3 months	32.8 ± 24.6	26.3 ± 22.1	6.5 (3.2 to 9.8)	< 0.001
	6 months	28.5 ± 22.9	23.7 ± 20.8	4.8 (1.7 to 7.9)	0.003
	12 months	25.1 ± 21.3	21.4 ± 19.5	3.7 (0.8 to 6.6)	0.013
Reflux symptoms	3 months	26.9 ± 22.7	21.1 ± 19.8	5.8 (2.7 to 8.9)	0.002
	6 months	24.3 ± 21.1	19.8 ± 18.5	4.5 (1.6 to 7.4)	0.004
	12 months	22.7 ± 20.3	18.9 ± 17.9	3.8 (1.1 to 6.5)	0.006
Eating restrictions	3 months	38.6 ± 24.9	31.5 ± 22.3	7.1 (3.9 to 10.3)	< 0.001
	6 months	34.2 ± 23.1	28.7 ± 20.8	5.5 (2.5 to 8.5)	< 0.001
	12 months	31.3 ± 21.7	26.9 ± 19.5	4.4 (1.6 to 7.2)	0.002

Higher scores indicate better physical functioning and worse symptoms/problems in other domains.

Gastric cancer-specific symptoms, as measured by the QLQ-STO22 and detailed in Table 3, showed that diabetic patients experienced more problems with reflux symptoms (+5.8 points, 95%CI: 2.7 to 8.9, P = 0.002) and eating restrictions (+7.1 points, 95%CI: 3.9 to 10.3, P < 0.001) throughout the first year after surgery. The data demonstrated that these differences remained significant at 12 months post-surgery, with eating restrictions showing a difference of +4.4 points (95%CI: 1.6 to 7.2, P = 0.002).

DISCUSSION

This large retrospective cohort study provides comprehensive evidence that diabetes mellitus is associated with poor outcomes in patients undergoing radical gastrectomy for gastric cancer. Our findings demonstrate that patients with diabetes experience longer hospital stays, higher rates of postoperative complications, increased 30-day readmission rates, reduced long-term survival, and a lower QOL than patients without diabetes.

The observed increase in the length of hospital stay (mean difference of 2.3 days) for diabetic patients was clinically significant and consistent with previous studies in other surgical populations[18,19]. This prolonged hospitalization may be attributed to several factors, including a higher incidence of postoperative complications, need for more intensive glycemic control, and management of diabetes-related comorbidities, and has significant economic impacts due to the high costs associated with inpatient care[20].

The higher rate of postoperative complications in patients with diabetes (OR, 1.68) was a critical finding in our study. The increased incidence of surgical site infections, pneumonia, and urinary tract infections in patients with diabetes is consistent with the known effects of diabetes on immune function and wound healing[21]. These complications contribute to prolonged hospital stays and may also have long-term implications for patient recovery and QOL.

The elevated 30-day readmission rate in patients with diabetes (12.5% vs 7.8%) is particularly concerning, as it suggests that the impact of diabetes on recovery extends beyond the initial hospital stay. This finding highlights the need for enhanced postdischarge care and close follow-up of patients with diabetes to prevent and promptly address any complications that may arise after leaving the hospital.

Analysis of the EORTC QLQ-C30 and QLQ-STO22 questionnaires revealed a dynamic pattern in QOL scores over time. While both groups showed improvements across all domains during the 12-month follow-up period, patients with diabetes consistently scored lower than non-diabetic patients. The gap was most pronounced at 3 months post-surgery, particularly in terms of physical functioning (8.3-point difference) but gradually narrowed by 12 months (5.6-point difference). This suggests that patients with diabetes may require longer recovery periods to achieve optimal functioning. Similarly, fatigue scores showed the greatest between-group difference at 3 months (7.2 points) with gradual convergence over time (4.4 points at 12 months), indicating that patients with diabetes experience more prolonged recovery-related fatigue. The persistent differences in eating restrictions and reflux symptoms (7.1 and 5.8 points respectively, at 3 months) highlight the need for targeted nutritional support and symptom management strategies in patients with diabetes.

The robustness of our findings, as demonstrated by the sensitivity analyses, strengthens the validity of our conclusions. The consistent results obtained from the unmatched cohort analysis and the stability of the findings after multiple imputations for missing data support the reliability of our primary analysis.

This study has several strengths, including its large sample size, use of propensity score matching to minimize confounding factors, and comprehensive assessment of both short- and long-term outcomes. The inclusion of QOL measures provides a patient-centered perspective that is often lacking in similar studies.

However, this study has some limitations. First, the retrospective nature of the study introduces the potential for unmeasured confounding, although our E-value calculations suggest that any unmeasured confounder would need to have a substantial effect on negating our findings. Second, the single-center design may have limited the generalizability of our results to other healthcare settings and populations. Third, although we adjusted for HbA1c levels, we did not have detailed data on the quality of glycemic control throughout the perioperative period, which may have provided additional insights into the mechanisms underlying the observed associations.

CONCLUSION

This study provides robust evidence that diabetes mellitus is associated with poor short- and long-term outcomes in patients undergoing radical gastrectomy for gastric cancer. These findings have several important clinical implications and should inform the perioperative care protocols for patients with diabetes who undergo radical gastrectomy. First, preoperative optimization of glycemic control should be prioritized with consideration of endocrinology consultations for patients with HbA1c > 7.5%. Second, implementing enhanced recovery after surgery protocols specifically modified for patients with diabetes may help mitigate complications. Such protocols should include strict glycemic control (target range 140-180 mg/dL), early mobilization to prevent muscle deconditioning, specialized wound care protocols, and prophylactic measures against common infections. Third, given the high readmission rates, establishing a structured follow-up program with regular monitoring of glycemic control and early detection of complications is crucial. Finally, persistent QOL impairments suggest the need for extended rehabilitation support and nutritional counseling, particularly focusing on managing eating restrictions and reflux symptoms. By addressing the specific needs of this high-risk population, we may improve not only immediate postoperative outcomes but also long-term survival and QOL in diabetic patients with gastric cancer.