Perioperative risk factors for prognosis in patients undergoing radical esophagectomy: A retrospective study

Abstract

BACKGROUND

Esophageal cancer constitutes one of the most aggressive malignant neoplasms associated with poor clinical outcomes. While surgical resection remains the cornerstone of curative intervention, optimization of perioperative care protocols has emerged as an essential strategy to reduce postoperative complications and potentially improve long-term survival rates in patients undergoing esophagectomy. However, substantial debate persists regarding the relative importance of various perioperative risk factors and their impact on post-resection outcomes.

AIM

To identify perioperative factors affecting prognosis after radical esophagectomy, aiming to improve patient outcomes through targeted interventions.

METHODS

A retrospective study analyzed 378 patients with esophageal cancer who underwent radical esophagectomy (McKeown, Sweet, or Ivor-Lewis procedures) from January 2022 through December 2023. All operations were performed by experienced surgeons following standardized perioperative protocols. The investigation gathered data on patient demographics, surgical parameters, tumor pathology (using the 8^th edition American Joint Committee on Cancer staging system), and survival outcomes. Statistical analyses utilized Kaplan-Meier estimates and Cox proportional hazards modeling, with adjustment for confounding variables.

RESULTS

Multivariate Cox proportional hazards analysis identified three independent predictors of survival: Tumor-node-metastasis staging [Hazard ratio (HR) = 2.31, 95% confidence interval (CI): 1.72-3.10, P < 0.001], tumor differentiation (moderate: HR = 1.46, 95%CI: 1.02-2.09, P = 0.038; poor: HR = 2.15, 95%CI: 1.47-3.14, P < 0.001), and extended postoperative analgesic use (> 5 days) (HR = 1.43, 95%CI: 1.08-1.89, P = 0.012). Kaplan-Meier analysis demonstrated significantly lower overall survival rates in patients requiring analgesics for > 5 days compared to ≤ 5 days (P = 0.003), with consistent patterns observed for both opioid (P = 0.019) and nonsteroidal anti-inflammatory drug use (P = 0.028). The extended analgesic group exhibited a higher proportion of elderly patients (48.47% vs 35.57%, P = 0.015), while other baseline characteristics and tumor features remained comparable between groups.

CONCLUSION

Tumor-node-metastasis staging, tumor differentiation, and duration of postoperative analgesic use independently predict survival following radical esophagectomy, underscoring the significance of optimal pain management protocols.

Key Words: Radical esophagectomy; Perioperative management; Prognostic factors; Postoperative analgesic use; Survival analysis

Core Tip: In this retrospective analysis of 378 patients who underwent radical esophagectomy for esophageal cancer, we identified tumor-node-metastasis staging, tumor differentiation, and extended postoperative analgesic use (> 5 days) as independent predictors of overall survival. Of particular significance was the adverse impact of prolonged administration of both opioids and nonsteroidal anti-inflammatory drugs on survival outcomes. These findings emphasize the critical importance of precise perioperative management, specifically regarding optimal pain control protocols, in enhancing patient outcomes. Further validation through multicenter, prospective studies and investigation of underlying mechanisms is warranted to optimize perioperative care strategies and improve long-term survival rates in this patient population.

INTRODUCTION

Esophageal cancer is a malignant neoplasm originating from esophageal epithelial cells that can develop in any segment of the esophagus, with squamous cell carcinoma being the predominant histologic type. While early-stage disease may be asymptomatic, advanced cases typically manifest with dysphagia and food obstruction[1]. Global cancer statistics from 2020 indicate that esophageal cancer ranks seventh among all malignancies worldwide, with China accounting for over half of incident cases[2]. In China, as of 2016, esophageal cancer ranked sixth in incidence and fifth in mortality among all malignancies, constituting a substantial public health burden[3]. The disease is characterized by high malignant potential and poor outcomes, with advanced-stage patients in China exhibiting only a 26% five-year survival rate[4].

Currently, surgical intervention remains the most definitive curative approach for esophageal cancer[5]. Improved surgical mortality rates can be attributed to advances in operative techniques, standardized perioperative protocols, and widespread implementation of multidisciplinary treatment strategies. However, radical esophagectomy remains a highly invasive procedure characterized by prolonged operative duration, challenging surgical exposure, and complex anatomical relationships. The high incidence of postoperative complications, often exacerbated by concurrent comorbidities, significantly impacts prognosis and quality of life[6].

Perioperative management is fundamental to esophageal cancer treatment, with surgical outcomes and patient prognosis directly linked to the quality of perioperative care[7]. Despite technological and therapeutic advances, esophageal cancer surgery continues to carry substantial morbidity and mortality risks. Key perioperative challenges include inadequate pain control, respiratory complications, delayed gastrointestinal recovery, and poor nutritional status[8,9]. Pain management is particularly crucial, as insufficient analgesia can precipitate respiratory compromise, impair early mobilization, and amplify stress response, potentially compromising both immediate recovery and long-term outcomes.

While numerous studies have investigated perioperative risk factors[10-12], including age, nutritional status, comorbidities, surgical approach, operative duration, blood loss, and extent of lymphadenectomy, their prognostic significance remains controversial with inconsistent findings across studies. The emergence of enhanced recovery after surgery protocols has transformed perioperative management strategies[13], with optimal pain control as a cornerstone component. Although optimized perioperative care can reduce surgical complications and potentially enhance long-term survival, implementing comprehensive care protocols presents practical challenges due to patient heterogeneity, varying pain thresholds, and resource limitations.

Our study aims to comprehensively evaluate perioperative factors influencing outcomes following radical esophagectomy. Through retrospective analysis of institutional clinical data, we seek to identify key perioperative risk factors associated with adverse postoperative outcomes and decreased survival, assess their relative prognostic impact, determine potentially modifiable factors amenable to intervention, and establish evidence-based recommendations for perioperative management. These findings will guide clinical decision-making and ultimately enhance treatment outcomes and quality of life for patients with esophageal cancer.

MATERIALS AND METHODS

Study design and patients

This retrospective study analyzed clinical data from 378 patients who underwent radical esophagectomy for esophageal cancer at the Department of Thoracic Surgery, Fourth Hospital of Hebei Medical University between January 2022 and December 2023. Patients were eligible for inclusion if they met the following criteria: (1) Histologically confirmed esophageal cancer by preoperative endoscopic biopsy; (2) Absence of distant metastasis on preoperative imaging studies, including contrast-enhanced chest and abdominal computed tomography, cervical ultrasonography, and whole-body positron emission tomography-computed tomography; (3) Medical fitness for surgery with no contraindications; (4) Completion of standard radical esophagectomy; and (5) Availability of complete clinical records. Exclusion criteria comprised: (1) Prior neoadjuvant therapy; (2) History of other malignancies; (3) Presence of severe cardiac, pulmonary, hepatic, or renal dysfunction; (4) Intraoperative discovery of distant metastases precluding complete resection; (5) Perioperative mortality; and (6) Loss to follow-up or incomplete follow-up data.

Surgical methods

All procedures were performed by experienced surgeons (associate chief physician or above) with a minimum of 15 years of expertise in esophageal cancer surgery. The surgical approach (McKeown three-incision, Sweet left thoracic, or Ivor-Lewis two-incision) was selected based on tumor location, patient characteristics, and surgical expertise. All operations adhered to standard oncological principles of radical esophagectomy, ensuring complete resection of the diseased esophageal segment and regional lymph nodes with negative margins.

The surgical procedure was performed under general anesthesia with double-lumen endotracheal intubation. Patient positioning varied by approach: Supine with contralateral head rotation for cervical access, lateral decubitus with 30° elevation for thoracic access, and supine for abdominal access. The thoracic phase involved esophageal mobilization and systematic lymphadenectomy of the periesophageal, subcarinal, paratracheal, and supradiaphragmatic nodes. The abdominal phase included gastric mobilization along greater and lesser curvatures while preserving the right gastroepiploic artery, accompanied by lymph node dissection around the cardia, left gastric artery, and celiac axis.

Gastrointestinal continuity was restored using a 3-4 cm wide gastric tube, created with a linear stapler along the greater curvature. The gastric conduit was routed through either the posterior mediastinum or retrosternal space for esophagogastric anastomosis, performed either mechanically or manually, based on anatomical considerations and surgeon preference. Standard placement of thoracic drainage and nasogastric tubes was performed. Strict adherence to oncological principles included “no-touch” technique to minimize tumor manipulation. All specimens underwent pathological examination for margin status and lymph node involvement.

Perioperative management

Preoperative preparation: All patients underwent comprehensive preoperative evaluation upon admission. Standard assessment included complete blood count, liver and kidney function tests, coagulation profile, and cardiopulmonary function evaluation. The American Society of Anesthesiologists (ASA) classification was applied to patients aged 65 years or older or those with comorbidities. Nutritional status was evaluated using the Nutritional Risk Screening 2002 scoring system; patients scoring ≥ 3 or with serum albumin < 35 g/L received nutritional supplementation. Anemia (hemoglobin < 120 g/L in men, < 110 g/L in women) was corrected preoperatively. Pulmonary function was assessed through detailed history, physical examination, spirometry, and arterial blood gas analysis. Patients with impaired pulmonary function received respiratory physiotherapy and nebulization treatment. Medical conditions including hypertension and diabetes were optimized to achieve target parameters. Bowel preparation was conducted one day preoperatively using oral polyethylene glycol electrolyte solution or mannitol, with 8-hour fasting and 4-hour clear liquid restrictions prior to surgery. Intravenous access was established 24 hours preoperatively, and prophylactic second-generation cephalosporins were administered 30 minutes before incision. All patients received psychological support and detailed operative instructions.

Anesthesia method: Combined general and thoracic epidural anesthesia was utilized. Standard monitoring included electrocardiogram, blood pressure, and pulse oximetry, complemented by invasive arterial pressure monitoring. Anesthesia induction was performed with midazolam (0.05 mg/kg), sufentanil (0.5 μg/kg), propofol (1.5-2.0 mg/kg), and cisatracurium (0.2 mg/kg). A double-lumen endobronchial tube was positioned using fiberoptic bronchoscopic guidance. Anesthesia maintenance was achieved through continuous propofol (4-12 mg/kg/h) and remifentanil (0.1-0.3 μg/kg/minute) infusion, with intermittent cisatracurium administration. Volume-controlled ventilation was sustained with 6-8 mL/kg tidal volume, 12-14 respirations/min, and end-tidal CO₂ maintained at 35-45 mmHg. Single-lung ventilation parameters were modified accordingly, with positive end-expiratory pressure applied as needed. Mean arterial pressure was maintained at or above 65 mmHg using vasoactive medications when required. Temperature, urine output, and blood gas values were monitored to preserve physiological homeostasis.

Postoperative management: Patients were transferred to the intensive care unit for immediate postoperative monitoring. Analgesia was administered through a patient-controlled analgesia pump (sufentanil + tropisetron) in conjunction with thoracic epidural analgesia. Continuous monitoring included vital parameters, drainage characteristics, and urinary output. Early mobilization, pulmonary exercises, and nebulization therapy were implemented to prevent respiratory complications. Endotracheal suctioning was performed at 2-4 hours intervals. Nutritional support followed a systematic approach: Enteral nutrition via nasogastric tube was initiated 24 hours postoperatively, supplemented with parenteral nutrition. Oral intake commenced between postoperative days 3-5 following verification of anastomotic integrity, progressing from clear liquids to soft diet. Prophylactic antibiotic therapy was typically discontinued within 48-72 hours. Deep venous thrombosis prophylaxis included early ambulation, intermittent pneumatic compression, and subcutaneous low-molecular-weight heparin administration. Thoracic drainage tubes were removed when output was < 200 mL/day and serous. Patients were transferred to the general ward upon clinical stabilization, with immediate intervention for any complications.

Data collection

Baseline characteristics collected included demographic variables (age, sex, body mass index), medical history (hypertension, diabetes mellitus, chronic gastritis), lifestyle factors (smoking status, alcohol consumption), and clinical parameters [ASA classification, New York Heart Association (NYHA) functional classification, preoperative glomerular filtration rate]. The ASA physical status classification system was defined as follows: (1) Class I: Normal healthy patient; (2) Class II: Patient with mild systemic disease; (3) Class III: Patient with severe systemic disease limiting activity; (4) Class IV: Patient with severe systemic disease that poses constant threat to life; (5) Class V: Moribund patient not expected to survive without surgery; and (6) The NYHA functional classification was categorized as: Class I: No limitation of physical activity; class II: Mild limitation with symptoms during ordinary activity; class III: Marked limitation with symptoms during less than ordinary activity; and class IV: Symptoms at rest with any physical activity causing discomfort.

Operative data included anesthetic technique, surgical approach, anastomotic location, operative duration, length of postoperative analgesia, adjuvant therapy (chemotherapy or radiotherapy), and postoperative complications (pneumonia, pleural effusion). Tumor characteristics were evaluated according to the 8^th edition of the American Joint Committee on Cancer staging system, including tumor-node-metastasis (TNM) stage, tumor location and maximum diameter, histologic grade, depth of invasion, perineural invasion, lymphovascular invasion, and lymph node status.

TNM Staging was classified as: Stage 0: Carcinoma in situ; stage I: Local tumor invasion without nodal involvement; stage II: (1) Muscular invasion without nodal involvement; and (2) With regional lymph node metastasis; stage III: Advanced local invasion with regional lymph node involvement; and stage IV: Distant metastasis. The primary endpoint was all-cause mortality, defined as death from any cause during the follow-up period. Mortality was determined through active follow-up with participants or their next of kin. Date and cause of death were documented. Participants were followed from the index date until death, loss to follow-up, or study completion, whichever occurred first.

Statistical analysis

Statistical analyses were conducted using SPSS software (version 26.0, IBM Corporation, Armonk, NY, United States). The normality of continuous variables’ distribution was assessed using the Shapiro-Wilk test. Normally distributed continuous variables were expressed as mean ± SD and compared using independent-samples t-tests, while non-normally distributed variables were presented as median (interquartile range) and analyzed using the Mann-Whitney U test. Variance homogeneity was evaluated using Levene's test. Categorical variables were expressed as frequencies (percentages) and analyzed using Pearson’s χ² test or Fisher's exact test when expected cell counts were less than 5.

Overall survival was calculated using the Kaplan-Meier method, with intergroup differences assessed using the log-rank test. The proportional hazards assumption for Cox regression analysis was verified through examination of Schoenfeld residuals. Prognostic factors were identified through univariate and multivariate Cox proportional hazards models. To control for potential confounding factors, the following variables were selected a priori based on clinical significance and published literature: Age, sex, body mass index, ASA physical status classification, comorbidities, tumor stage, surgical approach, and operative time. Variables showing P ≤ 0.1 in univariate analysis were included in the multivariate Cox regression model. The magnitude and precision of associations were quantified using hazard ratios (HRs) with corresponding 95% confidence intervals (CIs). All statistical tests were two-sided, with statistical significance defined as P < 0.05.

RESULTS

Univariate analysis of perioperative Cox proportional hazards model for overall survival

Baseline characteristics: Advanced age (≥ 65 years; HR = 1.52, 95%CI: 1.18-1.96, P = 0.001), ASA physical status class IV (HR = 1.45, 95%CI: 1.08-1.95, P = 0.01), and NYHA functional class III (HR = 1.68, 95%CI: 1.12-2.52, P = 0.01) demonstrated significant associations with decreased overall survival. Hypertension and diabetes exhibited borderline significance (P < 0.10) (Table 1).

Table 1 Univariate Cox analysis of baseline characteristics for overall survival, n (%).

Baseline characteristic	n = 378	Hazard ratio	95%CI	P value
Age
<65 years	215 (56.88)	1.00	Reference	-
≥ 65 years	163 (43.12)	1.52	1.18-1.96	0.001
Gender
Male	292 (77.25)	1.00	Reference	-
Female	86 (22.75)	0.89	0.67-1.19	0.44
BMI (kg/m²), mean ± SD	23.10 ± 3.20	0.97	0.93-1.02	0.24
Hypertension
No	246 (65.08)	1.00	Reference	-
Yes	132 (34.92)	1.28	0.99-1.65	0.06
Diabetes
No	298 (78.84)	1.00	Reference	-
Yes	80 (21.16)	1.31	0.98-1.75	0.07
Smoking history
No	156 (41.27)	1.00	Reference	-
Yes	222 (58.73)	1.22	0.94-1.58	0.13
History of drinking
No	169 (44.70)	1.00	Reference	-
Yes	209 (55.29)	1.15	0.89-1.48	0.28
Chronic gastritis
No	271 (71.69)	1.00	Reference	-
Yes	107 (28.31)	1.12	0.85-1.47	0.43
ASA classification
III	312 (82.54)	1.00	Reference	-
IV	66 (17.46)	1.45	1.08-1.95	0.01
NYHA classification
I	256 (67.72)	1.00	Reference	-
II	89 (23.54)	1.33	0.99-1.78	0.06
III	33 (8.74)	1.68	1.12-2.52	0.01
Glomerular filtration rate
< 90 mL/minute/1.73 m²	142 (37.57)	1.00	Reference	-
≥ 90 mL/minute/1.73 m²	236 (62.43)	0.81	0.63-1.05	0.11

CI: Confidence interval; BMI: Body mass index; ASA: American Society of Anesthesiologists; NYHA: New York Heart Association.

Surgery-related factors: Extended postoperative analgesic administration (> 5 days; HR = 1.62, 95%CI: 1.24-2.12, P < 0.001), postoperative pneumonia (HR = 1.58, 95%CI: 1.21-2.06, P = 0.001), pleural effusion (HR = 1.42, 95%CI: 1.07-1.88, P = 0.02), and adjuvant chemoradiotherapy (HR = 1.45, 95%CI: 1.12-1.88, P = 0.01) were associated with inferior survival outcomes. Prolonged operative time showed marginal correlation with adverse outcomes (HR = 1.00, P = 0.05) (Table 2).

Table 2 Univariate Cox analysis of surgery-related factors for overall survival, n (%).

Surgery-related factors	n = 378	Hazard ratio	95%CI	P value
Type of anesthesia
General anesthesia	203 (53.70)	1.00	Reference	-
Epidural anesthesia	175 (46.30)	0.92	0.71-1.19	0.52
Surgical methods
Open surgery	245 (64.81)	1.00	Reference	-
Minimally invasive surgery	133 (35.19)	0.78	0.60-1.02	0.07
Location of anastomotic site
Cervical	228 (60.32)	1.00	Reference	-
Thoracic	150 (39.68)	1.25	0.97-1.62	0.09
Operative time, minutes, mean ± SD	266.01 ± 83.13	1.00	0.99-1.01	0.05
Duration of analgesic use
≤ 5 days	149 (39.42)	1.00	Reference	-
> 5 days	229 (60.58)	1.62	1.24-2.12	< 0.001
Postoperative radiotherapy or chemotherapy
No	198 (52.38)	1.00	Reference	-
Yes	180 (47.62)	1.45	1.12-1.88	0.01
Postoperative pneumonia
No	265 (70.11)	1.00	Reference	-
Yes	113 (29.89)	1.58	1.21-2.06	0.001
Postoperative pleural effusion
No	284 (75.13)	1.00	Reference	-
Yes	94 (24.87)	1.42	1.07-1.88	0.02

CI: Confidence interval.

Tumor pathological characteristics: Advanced TNM stage (III/IV; HR = 2.85, 95%CI: 2.15-3.78, P < 0.001), increased tumor size (≥ 3 cm; HR = 1.68, 95%CI: 1.29-2.19, P < 0.001), submucosal invasion (HR = 1.92, 95%CI: 1.46-2.52, P < 0.001), poor differentiation (HR = 2.42, 95%CI: 1.67-3.51, P < 0.001), perineural invasion (HR = 1.75, 95%CI: 1.33-2.30, P < 0.001), vascular invasion (HR = 1.89, 95%CI: 1.42-2.52, P < 0.001), and lymph node metastasis (HR = 2.32, 95%CI: 1.77-3.04, P < 0.001) demonstrated significant correlation with decreased survival (Table 3).

Table 3 Univariate Cox analysis of pathological characteristics for overall survival, n (%).

Pathological characteristic	n = 378	Hazard ratio	95%CI	P value
TNM stage
I or II stage	167 (44.18)	1.00	Reference	-
III or IV stage	211 (55.82)	2.85	2.15-3.78	< 0.001
Tumor location
Upper segment	72 (19.05)	1.00	Reference	-
Middle segment	198 (52.38)	0.92	0.66-1.28	0.62
Lower segment	108 (28.57)	0.88	0.61-1.27	0.50
Tumor diameter
< 3 cm	156 (41.27)	1.00	Reference	-
≥ 3 cm	222 (58.73)	1.68	1.29-2.19	< 0.001
Depth of tumor infiltration
On the muscle layer	143 (37.83)	1.00	Reference	-
Subcutaneous muscle layer	235 (62.17)	1.92	1.46-2.52	< 0.001
Histological differentiation
Well differentiated	89 (23.54)	1.00	Reference	-
Moderately differentiated	196 (51.85)	1.58	1.12-2.23	0.01
Poorly differentiated	93 (24.61)	2.42	1.67-3.51	< 0.001
Perineural invasion
No	286 (75.66)	1.00	Reference	-
Yes	92 (24.34)	1.75	1.33-2.30	< 0.001
Tumor thrombus
No	301 (79.63)	1.00	Reference	-
Yes	77 (20.37)	1.89	1.42-2.52	< 0.001
Lymph node metastasis
No	184 (48.68)	1.00	Reference	-
Yes	194 (51.32)	2.32	1.77-3.04	< 0.001

CI: Confidence interval; TNM: Tumor-node-metastasis.

Comparison between extended (> 5 days) and short-term (≤ 5 days) analgesic administration

The extended analgesic group demonstrated a higher proportion of elderly patients (≥ 65 years; 48.47% vs 35.57%, P = 0.015) (Table 4). Other baseline characteristics, surgery-related parameters (Table 5), and tumor pathological features (Table 6) revealed no significant differences between groups (all P > 0.05).

Table 4 Patient characteristics by analgesic use duration after esophagectomy, n (%).

Baseline characteristic	Short-term analgesic use (n = 149)	Long-term analgesic use (n = 229)	t value or χ2	P value
Age			5.87	0.02
< 65 years	96 (64.43)	118 (51.53)
≥ 65 years	53 (35.57)	111 (48.47)
Gender			0.22	0.64
Male	117 (78.52)	175 (76.42)
Female	32 (21.48)	54 (23.58)
BMI (kg/m²), mean ± SD	23.25 ± 3.15	23.01 ± 3.24	0.72	0.47
Hypertension			0.64	0.42
No	107 (71.81)	139 (60.70)
Yes	42 (28.19)	90 (39.30)
Diabetes			0.55	0.46
No	126 (84.56)	172 (75.11)
Yes	23 (15.44)	57 (24.89)
Smoking history			0.74	0.39
No	66 (44.30)	90 (39.30)
Yes	83 (55.70)	139 (60.70)
History of drinking			0.89	0.35
No	71 (47.65)	98 (42.79)
Yes	78 (52.35)	131 (57.21)
Chronic gastritis			0.52	0.47
No	104 (69.80)	167 (72.93)
Yes	45 (30.20)	62 (27.07)
ASA classification			0.65	0.42
III	133 (89.26)	179 (78.17)
IV	16 (10.74)	50 (21.83)
NYHA classification			1.24	0.54
I	108 (72.48)	148 (64.63)
II	34 (22.82)	55 (24.02)
III	7 (4.70)	26 (11.35)
Glomerular filtration rate			2.24	0.14
< 90 mL/minute/1.73 m²	50 (33.56)	92 (40.17)
≥ 90 mL/minute/1.73 m²	99 (66.44)	137 (59.83)

CI: Confidence interval; BMI: Body mass index; ASA: American Society of Anesthesiologists; NYHA: New York Heart Association.

Table 5 Surgery-related factors by analgesic use duration after esophagectomy, n (%).

Surgery-related factors	Short-term analgesic use (n = 149)	Long-term analgesic use (n = 229)	t value or χ2	P value
Type of anesthesia			0.33	0.57
General anesthesia	77 (51.68)	126 (55.02)
Epidural anesthesia	72 (48.32)	103 (44.98)
Surgical methods			1.96	0.17
Open surgery	83 (55.70)	162 (70.74)
Minimally invasive surgery	66 (44.30)	67 (29.26)
Location of anastomotic site			1.34	0.25
Cervical	78 (52.35)	150 (65.50)
Thoracic	71 (47.65)	79 (34.50)
Operative time, minutes, mean ± SD	265.83 ± 76.28	266.35 ± 85.42	1.73	0.09
Duration of analgesic use			2.74	0.10
≤ 5 days	85 (57.05)	113 (49.34)
> 5 days	64 (42.95)	116 (50.66)
Postoperative radiotherapy or chemotherapy			1.84	0.18
No	118 (79.19)	147 (64.19)
Yes	31 (20.81)	82 (35.81)
Postoperative pneumonia			1.95	0.16
No	122 (81.88)	162 (70.74)
Yes	27 (18.12)	67 (29.26)

CI: Confidence interval.

Table 6 Pathological characteristics by analgesic use duration after esophagectomy, n (%).

Pathological characteristic	Short-term analgesic use (n = 149)	Long-term analgesic use (n = 229)	t value or χ²	P value
TNM stage			1.90	0.17
I or II stage	78 (52.35)	88 (38.43)
III or IV stage	71 (47.65)	141 (61.57)
Tumor location			0.16	0.92
Upper segment	29 (19.46)	43 (18.78)
Middle segment	77 (51.68)	121 (52.84)
Lower segment	43 (28.86)	65 (28.38)
Tumor diameter			1.40	0.24
< 3 cm	74 (49.66)	81 (35.37)
≥ 3 cm	75 (50.34)	148 (64.63)
Depth of tumor infiltration			1.13	0.29
On the muscle layer	65 (43.62)	78 (34.06)
Subcutaneous muscle layer	84 (56.38)	151 (65.94)
Histological differentiation			2.74	0.26
Well differentiated	42 (28.19)	47 (20.52)
Moderately differentiated	81 (54.36)	115 (50.22)
Poorly differentiated	26 (17.45)	67 (29.26)
Perineural invasion			1.49	0.22
No	124 (83.22)	162 (70.74)
Yes	25 (16.78)	67 (29.26)
Tumor thrombus			1.22	0.27
No	125 (83.89)	176 (76.86)
Yes	24 (16.11)	53 (23.14)
Lymph node metastasis			1.90	0.17
No	86 (57.72)	98 (42.79)
Yes	63 (42.28)	131 (57.21)

CI: Confidence interval; TNM: Tumor-node-metastasis.

Survival analysis by duration of analgesic administration

Kaplan-Meier analyses revealed significantly improved survival in patients with ≤ 5 days of analgesic administration compared to > 5 days (P = 0.003) (Figure 1A). Consistent patterns were observed for both opioid (P = 0.019) (Figure 1B) and nonsteroidal anti-inflammatory drug (NSAID) administration (P = 0.028) (Figure 1C).

Figure 1 Duration of analgesic use after esophageal cancer surgery. A: Kaplan-Meier analysis of different analgesic use duration after radical esophagectomy; B: Kaplan-Meier analysis of different opioid analgesics use duration after radical esophagectomy; C: Kaplan-Meier Analysis of different nonsteroidal anti-inflammatory drug analgesics use duration after radical esophagectomy.

Multivariate analysis using perioperative Cox proportional hazards model for overall survival

After adjusting for potential confounding variables (P ≤ 0.1 in univariate analysis), TNM stage (HR = 2.31, 95%CI: 1.72-3.10, P < 0.001), tumor differentiation (moderate: HR = 1.46, 95%CI: 1.02-2.09, P = 0.038; poor: HR = 2.15, 95%CI: 1.47-3.14, P < 0.001), and extended analgesic administration (> 5 days) (HR = 1.43, 95%CI: 1.08-1.89, P = 0.012) emerged as independent predictors of decreased survival (Table 7).

Table 7 Multivariate analysis of survival-related factors in radically resected esophageal squamous cell carcinoma, n (%).

Variables	Hazard ratio	95%CI	P value
TNM staging	-	-	-
Phase I and II	1.00	Reference	-
Phase III and IV	2.31	1.72-3.10	< 0.001
Degree of tumor differentiation	-	-	-
Well differentiated	1.00	Reference	-
Moderately differentiated	1.46	1.02-2.09	0.038
Poorly differentiated	2.15	1.47-3.14	< 0.001
Postoperative analgesics used for more than 5 days	-	-	-
≤ 5 days	1.00	Reference	-
> 5 days	1.43	1.08-1.89	0.012

CI: Confidence interval; TNM: Tumor-node-metastasis.

DISCUSSION

Our multivariate Cox proportional hazards analysis identified TNM stage, tumor differentiation, and extended postoperative analgesic administration as independent prognostic factors in patients with esophageal cancer. Advanced TNM stage (III/IV) significantly increased mortality risk (HR = 2.31, 95%CI: 1.72-3.10, P < 0.001) compared to stage I/II, aligning with previous findings[14]. The prognostic value of TNM staging was further validated by univariate analysis, which demonstrated that its components - tumor size, invasion depth, and lymph node status - all correlated with survival. The depth of tumor invasion represents a critical prognostic determinant, particularly given the distinct anatomical structure of the esophagus[15]. The absence of serosa and the relatively thin esophageal wall facilitate rapid transmural invasion; consequently, once tumors penetrate the muscularis propria, the risk of both locoregional infiltration and distant dissemination increases markedly. Lymph node metastasis significantly impacts prognosis by reflecting tumor invasiveness and metastatic potential, facilitating tumor spread through disruption of local immune barriers, and necessitating more extensive surgery with increased risk of surgical trauma and complications[16-18].

Tumor differentiation emerged as another crucial prognostic indicator, with both moderate (HR = 1.46, 95%CI: 1.02-2.09, P = 0.038) and poor differentiation (HR = 2.15, 95%CI: 1.47-3.14, P < 0.001) associated with inferior outcomes. This gradient relationship likely reflects the enhanced invasiveness, complex molecular alterations, and treatment resistance of poorly differentiated tumors[19,20]. The independent prognostic value of TNM stage and differentiation suggests they capture distinct aspects of tumor biology-anatomical progression and biological behavior, respectively. The integration of both TNM staging and tumor differentiation may provide more comprehensive prognostic stratification and guide therapeutic decision-making. Specifically, patients with early-stage disease but poor differentiation may benefit from intensified multimodal therapy, whereas selected patients with advanced-stage but well-differentiated tumors might remain candidates for surgical intervention, provided they meet appropriate physiological criteria for operability.

Notably, extended postoperative analgesic administration (> 5 days) independently predicted decreased survival (HR = 1.43, 95%CI: 1.08-1.89, P = 0.012), validating previous findings that reported a comparable hazard ratio (HR = 1.38) for extended analgesic use in gastrointestinal cancer patients[21]. Kaplan-Meier analyses demonstrated significantly decreased survival with prolonged administration of both opioids (P = 0.019) and NSAIDs (P = 0.028). Several mechanisms may explain this association:

Firstly, extended analgesic requirements may indicate complicated recovery, particularly in elderly patients who constituted a larger proportion of the > 5 days group (48.47% vs 35.57%, P = 0.015). Secondly, regarding opioid administration, our findings support the expanding evidence suggesting that opioids may influence tumor progression[22]. Recent molecular studies have demonstrated that opioids enhance tumor angiogenesis and suppress apoptosis through μ-opioid receptor activation[23]. The immunosuppressive effects of opioids, specifically on natural killer cell activity and T lymphocyte function[24], may explain the increased early recurrence risk observed in our study. Thirdly, our results indicate that extended postoperative administration may be detrimental. This aligns with Zhao-Fleming et al’s research[25], which demonstrated that prolonged NSAID use can result in complications affecting overall recovery. Furthermore, our observations support the findings of Arron et al[26], who documented increased surgical site complications with extended NSAID use, potentially compromising long-term outcomes.

These findings collectively suggest that optimizing perioperative pain management requires precise balance between adequate pain control and minimizing the potential adverse impacts of extended analgesic administration. This perspective aligns with recent guidelines recommending multimodal analgesic approaches to minimize the duration of both opioid and NSAID exposure while maintaining effective pain control[27].

Several limitations warrant consideration in this investigation. Primarily, the single-center retrospective design introduces potential selection bias, despite stringent inclusion and exclusion criteria. The institutional source limitation may affect the external validity and generalizability of our findings. Additionally, the retrospective methodology inherently presents challenges in data collection, particularly regarding critical perioperative parameters such as intraoperative blood loss and precise analgesic dosing, potentially affecting our analytical accuracy. Furthermore, while we identified an association between extended postoperative analgesic administration and adverse outcomes, the causal relationship remains uncertain due to potential unmeasured confounding variables, including underlying complications or compromised physical status that might independently influence prognosis. Lastly, the relatively brief follow-up period may inadequately capture long-term survival outcomes.

Future research directions should address these limitations through: (1) Implementation of multicenter, prospective trials with expanded cohorts and comprehensive perioperative data collection, particularly regarding pharmacological interventions; (2) Mechanistic studies investigating analgesic-mediated effects on prognosis, emphasizing molecular biological pathways; (3) Development of individualized perioperative analgesic protocols that optimize pain management while minimizing adverse prognostic implications; and (4) Extended follow-up periods to evaluate the long-term impact of perioperative factors on survival outcomes. Moreover, integration of genomic and immunological profiling may provide deeper insights into prognostic determinants, facilitating the advancement of precision medicine approaches.

CONCLUSION

In conclusion, this comprehensive analysis demonstrates that TNM staging, tumor differentiation, and duration of postoperative analgesic administration serve as independent prognostic indicators in patients undergoing radical esophagectomy for esophageal carcinoma. These findings emphasize the critical importance of judicious perioperative analgesic management and the optimization of comprehensive perioperative care protocols. The implementation of evidence-based strategies in perioperative management may significantly influence clinical outcomes in this patient population. Further investigation into these prognostic determinants may facilitate the development of more individualized therapeutic approaches and ultimately enhance patient survival outcomes.