Double tract reconstruction in proximal gastric tumors: A systematic review of clinical and functional outcomes

Abstract

BACKGROUND

Proximal gastrectomy (PG) with double tract reconstruction (DTR) has recently emerged as a function-preserving alternative to total gastrectomy (TG) with Roux-en-Y (RNY) reconstruction in patients with proximally located gastric cancer.

AIM

To evaluate the current evidence comparing PG-DTR with TG-RNY in terms of perioperative outcomes, long-term survival, complication rates, nutritional status and reflux esophagitis.

METHODS

A systematic literature search was conducted using PubMed, MEDLINE, Web of Science and the Cochrane Library for studies published between 2010 and January 2025. Search terms included gastric cancer, DTR and TG. Trials comparing PG-DTR with TG-RNY or PG-esophagogastrostomy (EG) were included. Data on operative details, lymph node yield, complications (Clavien-Dindo ≥ III), nutritional markers and incidence of reflux were extracted. Nineteen trials met the inclusion criteria. The review followed the PRISMA guidelines.

RESULTS

PG-DTR was found to have comparable long-term oncological outcomes to TG-RNY, despite a lower extent of lymph node dissection. Operative time and intraoperative blood loss were generally similar, with some studies favouring PG-DTR. Rates of major postoperative complications were comparable between techniques. Notably, PG-DTR showed a significantly lower incidence of reflux esophagitis than PG-EG and was comparable or superior to TG-RNY in reflux control. Nutritionally, PG-DTR was associated with better post-operative weight maintenance and biochemical parameters such as haemoglobin, albumin and vitamin B12 levels compared to TG-RNY. No significant nutritional differences were observed between PG-DTR and PG-EG. PG-DTR appears to offer a balanced approach to the surgical treatment of proximal gastric cancer, combining oncological safety with functional and nutritional benefits.

CONCLUSION

Its superiority over TG-RNY in postoperative nutrition and reflux prevention, together with comparable complication rates and survival, supports its consideration as a preferred reconstruction method in selected patients.

Key Words: Nutritional outcomes; Reflux esophagitis; Gastric cancer; Total gastrectomy; Double tract reconstruction; Proximal gastrectomy

Core Tip: Proximal gastrectomy (PG) with double tract reconstruction (DTR) has gained attention as a function-preserving surgical technique for proximal gastric cancer. This systematic review demonstrates that PG-DTR achieves oncological outcomes comparable to total gastrectomy with Roux-en-Y reconstruction, while offering significant advantages in postoperative nutritional status. Preservation of the residual stomach and duodenum appears to play a critical role in minimizing weight loss and maintaining better haematological parameters. These findings suggest that PG-DTR may be a superior surgical option for selected patients without compromising oncological safety.

INTRODUCTION

The global incidence of gastric cancer, especially proximal stomach tumors, has shown a consistent upward trend[1]. Total gastrectomy (TG) has long been regarded the standard surgical procedure for proximal gastric cancer. However, recent treatment guidelines from East Asian countries have introduced proximal gastrectomy (PG) as a viable alternative surgical approach[2]. Compared with TG, PG offers potential advantages in preserving gastric function and improving patients’ quality of life. The Japanese Gastric Cancer Association recommends three primary reconstruction options after PG: Esophagogastrostomy (EG), double tract reconstruction (DTR), and jejunal interposition (JI)[2]. Historically, reconstruction after PG has been performed using a gastric tube in combination with either EG or JI[3]. However, these methods are linked to increased rates of anastomotic stricture and reflux esophagitis caused by severe gastroesophageal reflux[4].

DTR has been developed to address these functional complications. It involves PG with distal stomach preservation and creating a side-to-side anastomosis between the digestive tract limb of a standard Roux-en-Y (RNY) configuration and the remnant stomach. Compared to TG-RNY, PG-DTR may be better adapted to the physiological and anatomical requirements of the digestive tract. The preservation of normal digestive processes and the reservoir function of the distal stomach are thought to contribute to superior nutritional outcomes[5]. In addition to these potential nutritional benefits, several studies have shown that PG-DTR provides oncologically acceptable results[6]. In recent years, both single- and multicenter studies comparing PG-DTR with TG-RNY and PG-EG have highlighted the perioperative, oncological, and nutritional advantages of PG-DTR (Table 1)[7-25]. This systematic review aims to compare PG-DTR and TG-RNY in terms of operative characteristics, postoperative complications, long-term survival, nutritional outcomes, and rates of reflux esophagitis based on the most recent data available in the literature.

Table 1 Single- and multicenter studies comparing proximal gastrectomy-double tract reconstruction with total gastrectomy-Roux-en-Y and proximal gastrectomy-esophagogastrostomy.

Ref.	Country	Year	Study period	Study design	Group	Sample size	Patients per group (n)	Age	Sex (male/female)	Preoperative BMI (kg/m2)	Access (% minimally invasive procedures)	Operation duration (minutes)	Bleeding (mL)	Retrieved lymph nodes (n)	Early complication (Clavien-Dindo ≥ 3)	Lenght of stay hospital (days)	TNM stage (I/II/III)	Reflux esophagitis (n)	Mean BMI difference (kg/m2)	Follow up (month)	Survival data
Kim and Kim[7]	Korea	2016	2009-2014	Retrospective	DTR	34	17	64.7 ± 9.9	14/3	24.2 ± 3.8	Laparoscopic	268.2 ± 40.9		33.4 ± 17.3	3	9.4 ± 4.1	16/1/0	2		12
					TG-RNY		17	60.9 ± 12.9	10/7	23.4 ± 5		270 ± 43.4		39.3 ± 20.4	3	10 ± 3.6	15/2/0	1		12
Jung et al[8]	Korea	2017	2008-2016	Retrospective	DTR	248	92	59.8 ± 11.4	77/15	23.5 ± 2.7	Laparoscopic	198.3 ± 38.8	84.7 ± 81.7	46.1 ± 19.6		8.5 ± 11.3		1	DTR < TGRY	26.6 ± 10.3	The 5-year OS rate; 96.1% and 95.9% for the PG-DTR and TG groups, respectively (P = 0.575)
					TG-RNY		156	58.7 ± 10.8	120/36	23.9 ± 3.3		225.4 ± 51.6	128.3 ± 112.5	60.0 ± 25.7		8.6 ± 5.7		3		43.5 ± 23.2
Ko et al[9]	Korea	2020	2008-2016	Retrospective	DTR	104	52	61.5 ± 12.3	35/17	23.7 ± 3.1	Laparoscopic	206.1 ± 38.2	133.6 ± 138.6	26.5 ± 11.1	4	13.3 ± 8.6	45/5/2	1		22.7 ± 15.4	The PG-DTR group had a significantly better 5-year OS rate than the TG group (P = 0.02)
					TG-RNY		52	63.0 ± 9.2	35/17	23.4 ± 2.9		204.2 ± 47.5	167.3 ± 94.4	36.2 ± 16.5	7	18.0 ± 15.5	40/7/5	0		36.3 ± 23.1
Li et al[10]	China	2019	2015-2017	RCT	DTR	300	103		90/13	66.10 ± 3.618 (kg)	Open	143.62 ± 2.98	140.22 ± 7.93	23 ± 3	0	10.52 ± 1.18	87/16/0	9	2.02 kg	18
					JI		98		88/10	66.33 ± 3.116 (kg)		144.37 ± 2.51	139.09 ± 8.85	22 ± 3	1	10.53 ± 1.31	81/17/0	4	1.78 kg	18
					TG-RNY		99		82/17	66.36 ± 3.824 (kg)		143.29 ± 4.51	139.78 ± 8.52	37 ± 4	1	10.68 ± 1.35	78/21/0	11	8.11 kg	18
Ji et al[11]	China	2021	2014-2019	Retrospective	DTR	64	25	64.5 ± 8.9	24/1	24.5 ± 2.6	Open	240 (210-270)	50 (50-100)	19 (16.3-23)	0	13 (11-18)	23/2/0	2		39.8	The 3-year OS rates of the EG and PG-DTR groups were 79.9% and 90.9%, respectively (P = 0.066)
					EG		39	62.1 ± 9.1	31/8	24.5 ± 3.7		195 (170-240)	100 (50-100)	22.5 (17-31)	1	11 (10-14)	21/12/6	12	N	39.8
Ma et al[12]	China	2020	2010-2018	PSM	DTR	192	86	60.5 ± 8.2	79/7		Laparoscopic/open: 61/25	177 ± 30	91 ± 32	23 ± 6	2	11.3 ± 3.4	31/27/28		8.8 kg	48	5 year OS rates in pathological stage II; PG-DTR and TG-RNY were 71.5% and 65.3%, respectively (P = 0.365)
					TG-RNY		86	62,7 ± 10,5	76/10		Laparoscopic/open: 57/29	164 ± 17	106 ± 62	25 ± 7	3	10.5 ± 2.1	35/22/29		12.7 kg	48	5-year OS rates in pathological stage III; PG-DTR and TG-RNY were 51.9% and 42.1%, respectively (P = 0.272)
Wang et al[13]	China	2020	2016-2017	Retrospective	DTR	36	12	55.8 ± 4.1	6/6	23.29 ± 0.76	Laparoscopic	215.8 ± 14.69	104.2 ± 14.38	28.50 ± 3.76	0	10.8 ± 1.19	10/2/0	1	DTR < TGRY	12
					TG-RNY		24	58.38 ± 1.82	15/9	23.5 ± 0.68		209.3 ± 12.87	111.7 ± 12.2	36.29 ± 1.35	0	13.0 ± 2.1	22/2/0	11		12
Sato et al[14]	Japan	2021	2013-2019	Retrospective	DTR	150	75	68.8 ± 10.2	59/16	22.4 (15.8-32.2)	Laparoscopic/robot: 65/10	312 (199-552)	23 (4-192)		9	9 (7-58)	60/11/4	6		36.3 (22.2-51.0)
					TG-RNY		75	66.2 ± 11.8	51/24	22.8 (16.0-34.8)	Laparoscopic/robot: 64/11	311 (206-495)	25 (3-182)		4	9 (7-49)	44/16/12	4		36.3 (22.2-51.0)
Kimuro et al[15]	Japan	2021	2011-2018	Retrospective	DTR	16	8	70.4 (55-77)	8/0	22.9 (19.3-24.9)	Laparoscopic	294.1 ± 33.9	8.2 ± 4.3		2	15.4 ± 3.3	8/0/0		92 ± 5.7 (weight 24th month)	24
					TG-RNY		8	68.3 (45-83)	8/0	23.7 (20.8-29.8)		383.4 ± 20.7	48.8 ± 16.8		1	14.6 ± 2.2	8/0/0		82.4 ± 3.7 (weight 24th month)	24
Eom et al[16]	Korea	2021	2013-2017	Retrospective	DTR	103	58	65.5 (52.8-72.0)	47/11	23.5 ± 3.1	Laparoscopic			34.0 (24.6-40.3)				2	There is no significant difference	12
					EG		45	63.0 (53.5-72.5)	13/12	25.0 ± 3.3				26.0 (20.5-36.5)				8		12
Hwang et al[17]	Korea	2022	2016-2018	RCT	DTR	128	63	58.6 ± 10.2	35/28	24.5 ± 2.9	Laparoscopic	217.1 ± 67.5	75.5 ± 74.7	40.8 ± 18.3	3	7.4 ± 3.2	60/3/0	9		0.5
					TG-RNY		65	61.3 ± 11.5	47/18	24.5 ± 2.9		200.8 ± 51.9	65.5 ± 62.7	55.1 ± 24.9	2	7.3 ± 2.9	58/7/0	9		0.5
Li et al[18]	China	2023	2012-2015	Retrospective	DTR	100	50	65.8 ± 4.5	40/10	24.1 ± 1.8	Laparoscopic	192.1 ± 7.7	122.4 ± 45.8	25.7 ± 4.6	3	14.1 ± 4.4	50/0/0	1	22.9 ± 2.0 (BMI: 12nd month)	60	5-year OS rates in the PG-DTR and TG-RNY were 98% and 90%, respectively (P = 0.08)
					TG-RNY		50	66.9 ± 3.6	40/10	23.8 ± 2.0		187.4 ± 6.7	117.0 ± 51.3	34.4 ± 5.2	2	14.6 ± 6.0	50/0/0	2	20.6 ± 1.3	60
Ma et al[19]	China	2022	2013-2018	Retrospective	DTR	66	33	23 patients (> 60 years)	28/5		Laparoscopic	204 ± 37	171 ± 71	21.5 ± 9.5	3	15.4 ± 2.3	14/9/10	8	DTR < TGRY	54	5-year OS rates in the PG-DTR and TG groups were 61% and 60%, respectively (P = 0.61)
					TG-RNY		33	21 patients (> 60 years)	28/5			205 ± 50	176 ± 96	29.0 ± 10.9	3	17.2 ± 3.8	10/12/11	11		61
Park et al[20]	Korea	2023	2016-2020	RCT	DTR	128	63	58.6 ± 10.3	35/28	24.5 ± 2.9	Laparoscopic	217.1 ± 67.5	75.5 ± 74.6	40.8 ± 18.3		7.4 ± 3.2	60/3/0	1		24	2-year OS rates; PG-DTR and TG-RNY were 98.5% and 100%, respectively (P = 0.33)
					TG-RNY		65	61.3 ± 11.5	47/18	24.5 ± 2.9		200.8 ± 51.9	65.5 ± 62.7	55.1 ± 24.9		7.3 ± 2.9	58/7/0	2		24
Hasegawa et al[21]	Japan	2024	2011-2022	Retrospective	DTR	73	39	66.6 ± 14.0	27/12	22.3 ± 3.0	Laparoscopic/robot/open: 21/11/2	404.5 ± 88.8	114.8 ± 143.9	23.6 ± 12.5	6	16.1 ± 10.9	30/6/3	0	There is no significant difference	36
					EG		34	69.4 ± 9.9	27/7	23.0 ± 3.6	Laparoscopic/robot/open: 24/15/0	365.1 ± 99.5	122.3 ± 195.5	23.9 ± 13.9	2	15.8 ± 12.6	29/4/1	5		36
Shoda et al[22]	Japan	2024	2020-2022	Prospective	DTR	20	10	72.3 ± 6.3	9/1	24.0 ± 2.9	Laparoscopic	554.6 ± 163.2	123.8 ± 118.3		0	14.6 ± 3.8	5/0/5		0.17 ± 0.08	12
					EG		10	68.8 ± 8.9	6/4	23.0 ± 4.2		413.5 ± 91.2	78.7 ± 75.8		0	12.0 ± 6.7	8/1/1		0.14 ± 0.08	12
Wang et al[23]	China	2024	2020-2022	Retrospective	DTR	295	87	64.8 ± 6.67	69/18	24.24 ± 3.27	Laparoscopic	230 (210-255)	85 (50-100)	22 (19-31)	2	7 (7-9)	30/22/35	8	2.51 ± 1.38	12
					EG		96	64.8 ± 6.62	74/22	23.79 ± 2.79		195 (180-250)	100 (50-100)	21 (18-26)	1	8 (7-9)	26/21/49	42	4.37 ± 1.81	12
					TG-RNY		112	63.15 ± 9.64	81/31	24.75 ± 3.36		200 (180-240)	100 (50-100)	28 (22-35)	4	8 (7-9)	23/26/63	26	3.96 ± 1.59	12
Sun et al[24]	China	2024	2019-2023	Retrospective	DTR	156	93	63.16 ± 9.75	74/19	23.08 ± 3.64	Laparoscopic	166.38	113.23	18.10	5	13.83		33	1.40 ± 0.655	12
					EG		63	61.70 ± 8.30	47/16	23.20 ± 3.39		174.06	150.16	16.81	5	13.11		49	1.81 ± 0.759	12
Wu et al[25]	China	2024	2020-2022	Retrospective	DTR	43	23	70 (66-76)	19/4	22.1 ± 2.8	Laparoscopic	187 (177-198)	21.2 ± 4.1	20.4 ± 10.3	0	8 (8-9)	15/8/0	9	22.3 ± 2.6 (BMI: 12nd month)	12
					SOA		20	65.5 (60.8-71.8)	17/3	21.9 ± 3.3		197.5 (190-207)	21.9 ± 3.0	24.8 ± 11.0	0	8.0 (7.0-8.8)	16/4/0	2	22.0 ± 2.5	12

BMI: Body mass index; TNM: Tumor-node-metastasis; RCT: Randomized controlled trial; PSM: Propensity score-matched; PG: Proximal gastrectomy; DTR: Double tract reconstruction; TG: Total gastrectomy; RNY: Roux-en-Y; EG: Esophagogastrostomy; JI: Jejunal interposition; SOA: Stomach outlet anastomosis; OS: Overall survival.

MATERIALS AND METHODS

A systematic literature search of international databases (PubMed, Web of Science, MEDLINE, and Cochrane Library) was conducted for trials published between 2010 and January 2025. The search terms included (“stomach” and “gastric”) and (“neoplasm” and “cancer” and “tumor” and “carcinoma”) and (“double tract reconstruction” and “double tract anastomosis”) and (“total gastrectomy”). A total of 1106 articles were initially identified based on these criteria. The analysis focused on studies comparing various reconstruction methods following PG, including DTR. Case reports and letters to the editor were excluded from analysis. Ultimately, 19 studies were included. Data extracted from these articles encompassed demographic characteristics, surgery type, operative time, intraoperative blood loss, number of lymph nodes dissected, tumor staging, postoperative complications of Clavien-Dindo grade III or higher, body mass index (BMI), and incidence of reflux esophagitis. All included studies were reported as PRISMA complain (Figure 1)[26].

Figure 1 PRISMA complain of this study. RCT: Randomized controlled trial; PSM: Propensity score-matched.

Risk of bias assessment

The methodological quality of the included studies was assessed using both the Cochrane Risk of Bias 2.0 (RoB 2.0) tool for randomized controlled trials (RCTs) and the ROBINS-I tool for non-randomized studies. Among the 19 included studies, four were RCTs, one was a prospective cohort study with a defined intervention protocol, one was a propensity score-matched study, and 13 were retrospective observational studies.

RCTs (RoB 2.0): According to the RoB 2.0 assessment, two of the four RCTs (Hwang et al[17]; Li et al[10]) were judged to have a low overall risk of bias across all five domains, including the randomization process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selection of reported results. The study by Park et al[20] raised some concerns due to limitations in outcome measurement reporting, while Shoda et al[22], although prospective, exhibited some concerns in randomization transparency and selective reporting. A detailed summary is presented in the supplementary risk of bias table (Supplementary Table 1).

Non-randomized studies (ROBINS-I): Among the 15 non-randomized studies, the overall risk of bias was high in 14 studies, and moderate in 1 (Ma et al[12]; propensity score-matched; Supplementary Table 2). The predominant sources of bias include: (1) Confounding and selection bias: Most retrospective studies lacked sufficient control for confounding factors, such as baseline nutritional status, tumor stage, and institutional surgical expertise; (2) Classification of interventions: Given the surgical nature of the interventions, the consistency in defining PG-DTR vs TG-RNY or PG-EG was not always clear, especially in multicenter data or studies spanning long recruitment periods; (3) Outcome assessment: Although primary outcomes like reflux esophagitis, BMI change, and lymph node retrieval were reported, assessment methods (objective vs subjective) were variably described, leading to some concerns in the outcome measurement domain; and (4) Selective reporting: Several studies reported only statistically significant outcomes, raising concerns about reporting bias.

Impact of study design heterogeneity: Significant methodological and clinical heterogeneity was observed among the included studies, particularly in: (1) Study design and sample size: Study designs ranged from single-center retrospective cohorts with small sample sizes to multicenter trials with larger populations; (2) Surgical techniques and access: Laparoscopic vs open vs robotic approaches were variably used across and even within studies; and (3) Outcome definitions and follow-up durations varied considerably (from 0.5 to 60 months), affecting comparability. This heterogeneity limited the feasibility and reliability of quantitative synthesis (meta-analysis) for specific endpoints, especially functional and nutritional outcomes. Although pooled estimates were calculated for a subset of comparable studies (n = 13), the certainty of evidence remained moderate to low, primarily due to the risk of bias and inconsistency across studies.

RESULTS

The included studies generally compared PG-DTR, performed either by open or laparoscopic approach, with TG-RNY or PG-EG.

Oncological safety and overall survival

Standard D2 Lymphadenectomy is typically performed during TG-RNY, whereas PG-DTR and PG-EG are usually accompanied by D1+ lymphadenectomy involving stations 1-2-3a-4sa-4sb-7-8a-9-11p. Regarding the number of lymph nodes retrieved, TG is superior to both PG-DTR and PG-EG[8,9,13,17,19,20,23]. No significant difference was found between PG-DTR and PG-EG[11,16,21,22,24]. The advantage of TG in lymph node yield is attributed to the broader range of lymph node stations resected. For Siewert type II and III adenocarcinoma, studies comparing PG followed by DTR with TG-RNY have shown comparable results in terms of tumor recurrence, metastasis, and long-term survival[8,11,12,18-21]. Ko et al[9] reported that DTR resulted in superior overall survival compared to TG-RNY. This difference in overall survival has been attributed to the older age of patients who experienced mortality in the TG-RNY group and the nutritional disadvantages associated with TG-RNY than PG-DTR.

Operative time and blood loss

Several studies have reported no statistically significant difference in operative time and intraoperative bleeding between PG-DTR and TG-RNY[7,9-11,13,14,17-20,23,25]. However, studies by Kimura et al[15] and Jung et al[8] found that PG-DTR was associated with shorter operative times and less intraoperative blood loss than TG-RNY. These results were thought to be influenced by anastomotic technique and surgeon experience. EG, another reconstruction method following PG involves only a single anastomosis and is considered technically simpler than DTR. Several studies have shown that EG can be performed with a shorter operative time and less blood loss than DTR[11,21,24].

Early postoperative complications and reflux esophagitis

Compared to TG-RNY and EG, DTR has shown similar rates of early postoperative complications[7-15,17-25]. Reflux esophagitis remains one of the most discussed long-term complications. Evaluation for reflux was made according to the Visick scoring or the Los Angeles classification. Several studies have shown that PG-DTR has comparable or even lower rates of reflux esophagitis than TG-RNY[7,9,12,13,17-20], with some suggesting a slight advantage[10,23]. Given that one of the traditional justifications for TG over PG is a reduced risk of reflux, the comparable results achieved with PG-DTR are noteworthy. A greater distance between the esophagojejunostomy and gastrojejunostomy appears to correlate with a reduced incidence of reflux, although a distance greater than 30 cm may compromise endoscopic surveillance and delay the detection of secondary malignancies[20].

Among PG reconstruction techniques, EG has consistently been associated with higher rates of reflux esophagitis. Modifications such as fixation of the remnant stomach to the diaphragm to create a neo-His angle or pseudo-fundus[16,23], fundoplication with conventional anastomosis[21,22], and the side overlap with fundoplication by Yamashita technique[21,27] have been introduced to mitigate this problem. However, EG remains less favorable than DTR in terms of reflux outcomes[11,21,23,24]. These findings underscore the importance of placing intestinal segments between the esophagus and the remaining stomach. Nevertheless, it is critical to interpret these findings with caution, as the majority of the underlying studies are retrospective in design and subject to selection and measurement biases. The overall quality of evidence supporting the superiority of PG-DTR in terms of reflux control is therefore considered moderate, rather than high. This highlights the need for further high-quality RCTs directly comparing PG-DTR with other reconstruction techniques to draw more definitive conclusions regarding long-term functional outcomes.

Nutritional outcomes

Body weight is a surrogate marker of nutritional status. Changes in patient body weight are given in mean BMI or kilograms. Studies have shown that PG-DTR is more effective in maintaining postoperative body weight than TG-RNY[8,10,12,18,19]. This advantage is likely due to the preservation of the gastric reservoir and maintaining the duodenal passage, which supports enzyme flow and nutrient absorption. EG is also associated with better weight maintenance than TG-RNY, but is not significantly different from DTR. Although some studies have reported better weight maintenance with DTR after the first postoperative year[24], the overall consensus is that weight outcomes are comparable between DTR and EG[16,21-23]. The preserved reservoir function of the remnant stomach and the intact duodenopancreatic axis is thought to mitigate excessive weight loss.

The stomach plays a central role in vitamin B12 metabolism. While evidence on long-term postoperative B12 and iron levels is inconsistent, patients who undergo PG-DTR appear to require less supplementation than those who undergo TG-RNY[7,8,10,18-20]. PG-DTR has also been associated with more favorable changes in laboratory parameters such as hemoglobin, total protein, and albumin[10,13,15,18,19]. No significant differences were observed between PG-DTR and PG-EG in postoperative levels of B12, total protein, albumin, or haemoglobin[11,21,24], probably due to the preserved functional capacity of the remnant stomach.

However, these observations should be interpreted with caution, as most of the supporting studies are retrospective and subject to selection bias, inconsistent outcome reporting, and short-to-intermediate follow-up durations. Consequently, the evidence supporting PG-DTR’s nutritional superiority is considered moderate. High-quality RCTs directly comparing PG-DTR with TG-RNY and PG-EG are crucial to validate these preliminary findings and guide evidence-based surgical decision-making. To facilitate a comprehensive comparison of the perioperative, functional, and oncologic outcomes across studies, a summary forest plot was constructed (Figure 2). This figure visually synthesizes the pooled effect estimates for, including early complications, reflux esophagitis, intraoperative bleeding, number of retrieved lymph nodes, postoperative of length hospital stay, and BMI difference. Collectively, these data underscore the comparative advantages and trade-offs associated with PG-DTR than TG-RNY, reinforcing the overall interpretative framework of this review.

Figure 2 Summary forest plot comparing proximal gastrectomy with double-tract reconstruction to total gastrectomy with Roux-en-Y across binary and continuous outcomes. Effect sizes represent odds ratios (log scale) for binary outcomes and mean differences for continuous outcomes. A value less than 0 favors proximal gastrectomy with double-tract reconstruction. PG: Proximal gastrectomy; DTR: Double tract reconstruction; TG: Total gastrectomy; RNY: Roux-en-Y; BMI: Body mass index; CD: Clavien-Dindo.

DISCUSSION

Clinical implications and real-world applicability of PG-DTR

While accumulating evidence supports the functional and oncological safety of PG with DTR, limited guidance exists regarding the clinical criteria for its optimal implementation. The included studies in this study provide a heterogeneous mix of patient populations, surgical approaches, and institutional protocols, reflecting the current variability in adopting this technique. PG-DTR appears to be particularly suitable for patients with early to locally advanced proximal gastric cancer (cT1-T3, N0-N1) in whom the tumor is located in the upper third of the stomach but without esophageal invasion or need for TG due to margin concerns[8,12,13,20]. Patients with relatively preserved performance status and adequate nutritional reserves may benefit more from function-preserving strategies, such as PG-DTR. Additionally, carefully selected elderly patients may experience better postoperative quality of life and nutritional outcomes than those undergoing TG-RNY[10,19].

Despite its potential benefits, PG-DTR faces persistent real-world barriers to widespread adoption. First, PG-DTR demands considerable technical expertise, especially in constructing a tension-free esophagojejunostomy and accurately positioning the jejunal limbs to ensure uninterrupted food passage through the remnant stomach and jejunum. Studies, such as those by Wang et al[13], Ma et al[12], and Li et al[10], were conducted in high-volume centers with extensive experience in laparoscopic or open gastric cancer surgery, suggesting that institutional learning curves may considerably impact outcomes.

Second, anatomical limitations, including patient BMI, remnant stomach size, and mesenteric length, may limit the feasibility of DTR. Sato et al[14] and Hwang et al[17] noted the importance of sufficient jejunal mobility to prevent anastomotic tension and maintain adequate blood supply. In patients with short mesentery or intra-abdominal adhesions, alternative reconstruction techniques such as EG or JI may be preferred. Third, the reconstruction is more complex and time-consuming than standard RNY or EG anastomoses, which may limit its application in minimally invasive surgery, especially in centers with limited access to robotic platforms. However, some included studies (e.g., Sato et al[14], Eom et al[16], and Shoda et al[22]) have demonstrated the feasibility of PG-DTR in laparoscopic and robotic settings, provided that surgical teams have sufficient expertise and institutional support.

Finally, there is a lack of uniform consensus or guidelines on when PG-DTR should be favored over other techniques. Although it shows promise in preserving postoperative function, the decision to employ PG-DTR should be individualized, considering tumor characteristics, patient comorbidities, anatomical feasibility, and surgical expertise. In summary, while PG-DTR is a function-preserving and oncologically acceptable alternative to TG-RNY in selected patients with proximal gastric cancer, its broader adoption in routine clinical practice requires addressing technical challenges, standardizing surgical protocols, and training surgeons in both open and minimally invasive techniques.

CONCLUSION

In recent years, PG-DTR has emerged as a promising surgical technique for proximal gastric tumors. Although PG-DTR may involve a slightly longer operative time than TG-RNY and PG-EG, it yields comparable outcomes regarding major postoperative complications (Clavien-Dindo grade ≥ III), hospital stay, and overall survival. Its nutritional superiority over TG-RNY and the lower incidence of reflux esophagitis compared to PG-EG further support its clinical utility. PG-DTR is widely regarded as a rational and effective surgical approach for patients with proximal gastric cancer.