Impact of Perioperative Immunonutrition on Postoperative Outcomes in Patients with Upper Gastrointestinal Cancer: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Abstract

There is no consensus on the efficacy of perioperative immunonutrition in patients with upper gastrointestinal (GI) cancer surgery. We clarified the impact of perioperative immunonutrition on postoperative outcomes in patients with upper GI cancers. We searched MEDLINE (PubMed), MEDLINE (OVID), EMBASE, Cochrane Central Register of Controlled Trials, Web of Science Core Selection, and Emcare from 1981–2022 using search terms related to immunonutrition and upper GI cancer. We included randomized controlled trials. Intervention was defined as immunonutritional therapy, including arginine, n-3 omega fatty acids, or glutamine during the perioperative period. The control was defined as standard nutritional therapy. The primary outcomes were infectious complications, defined as events with a Clavien–Dindo classification grade ≥ II that occurred within 30 days after surgery. After screening, 23 studies were included in the qualitative synthesis and in the quantitative synthesis. The meta-analysis showed that immunonutrition reduced infectious complications (relative risk ratio: 0.72; 95% confidence interval: 0.57–0.92; certainty of evidence: Moderate) compared with standard nutritional therapy. In conclusion, nutritional intervention with perioperative immunonutrition in patients with upper GI cancers significantly reduced infectious complications. The effect of immunonutrition for upper GI cancers in reducing the risk of infectious complications was about 30%.

1. Introduction

For cancer patients, surgical resection is the main treatment, and one of the risk factors for problems after surgery is preoperative malnutrition [1,2]. Patients with gastrointestinal cancers often have worsened nutritional status, which ranges from 20–70% [3,4]. The immune system and tissue repair are impacted by malnutrition [3]. Furthermore, the surgical invasion-induced catabolic processes result in the loss of vital nutrients, which might induce immune response dysregulation and increase the risk of infectious complications [1,3]. To decrease infection and total postoperative complications, nutritional interventions are necessary to reduce preoperative malnutrition.

There is no consensus on the efficacy of perioperative immunonutrition in patients with upper GI cancer after surgery. An immunonutrition therapy using either arginine, n-3 omega fatty acids, or glutamine has been developed and used clinically to reduce infectious complications and postoperative length of stay [5]. A postoperative inflammatory response is modulated by immunonutrition, which reduces immunosuppression caused by inflammatory cytokines [6]. While standard nutritional therapy is used for malnutrition, immunonutrition is used not only for malnutrition but also to reduce invasion or restore immunity. However, its efficacy and optimal timing are clinically unresolved, especially in upper GI cancer surgeries. Therefore, we planned a systematic review and meta-analysis of perioperative immunonutrition in patients with upper GI cancers.

This study aimed to clarify the effectiveness of perioperative immunonutrition in patients undergoing elective surgery for upper GI cancer. We also investigated whether the recommendations for patients who are malnourished differed from those who are not malnourished. We hypothesized that perioperative immunonutrition reduces infectious complications in patients with and without malnourishment.

2. Materials and Methods

We conducted a systematic review of the relevant literature in accordance with the Cochrane Handbook, Preferred Reporting Items for Systematic Reviews and Meta-Analysis 2020 guidelines (PRISMA-2020), and the Minds Manual for Guideline Development 2020 [7,8,9]. The protocol was published in PROSPERO (CRD42023471825).

2.1. Eligibility Criteria and Study Selection

We included randomized controlled trials (RCTs) of patients aged over 18 years who underwent elective upper GI cancer surgery with perioperative immunonutrition. Intervention was defined as perioperative immunonutritional therapy, including arginine, n-3 omega fatty acids, or glutamine was given preoperatively, postoperatively, or both. The control was defined as standard oral or intravenous nutritional therapy without immunonutrition. The intervention and control groups received the same amount of nutrition. We excluded studies in which more than 25% of patients had benign disease or cancer at other sites, review articles, case reports, crossover trials, and cluster-, quasi-, and non-randomized trials.

2.2. Search Strategy

Appendix A provides the search formulae. We searched MEDLINE (PubMed), MEDLINE (OVID), Embase (OVID), Cochrane Central Register of Controlled Trials (CENTRAL), Web of Science Core Selection, and Emcare (OVID). The period covered by the RCT was 2000–2022.

2.3. Study Selection and Data Collection

Two independent reviewers screened the titles and abstracts, assessing their eligibility based on the full texts. The same reviewers performed independent data extraction from the included studies using a standardized data collection form. Reviewer disagreements were resolved through discussion or with a mediating third reviewer. The original authors were contacted for missing data.

2.4. Risk-of-Bias Assessment

Two of the three researchers carried out risk-of-bias (ROB) assessments using the Cochrane Collaboration ROB tool, which has five domains: randomization, deviation from intervention, missing data, measurement of outcome, and selective reporting [7]. The ratings “high risk”, “some concerns”, and “low risk” were assigned to each domain and overall. The resolution of the disagreement was decided by a third person.

2.5. Outcomes

The primary outcomes were the total postoperative and infectious complications. The secondary outcomes were severe complications, anastomotic leakage, postoperative pneumonia, postoperative mortality, nutritional intervention adverse events, and postoperative hospitalization. Postoperative complications were defined as events with a Clavien–Dindo (CD) classification grade of ≥II that occurred within 30 days after surgery. Severe complications were defined as those with a CD grade of ≥III.

2.6. Synthesis of Results

We pooled the relative risk ratios (RRs) and 95% confidence intervals (CIs) for postoperative complications, postoperative mortality, and nutritional intervention adverse events, and the mean differences (MDs) and 95% CIs for postoperative hospitalization in patients with upper GI cancer. An intention-to-treat analysis was performed for dichotomous data where possible. We used Review Manager software 5.4.2 and performed meta-analyses with a random-effects model, assuming that the true effect would be low owing to many unmeasured or unknown factors and individual differences between studies in accordance with the Cochrane Handbook [7].

Statistical heterogeneity was evaluated by visually inspecting forest plots and calculating the I² statistic (I² values of 0–40% may not be important; 30–60% may represent moderate heterogeneity; 50–90% may represent substantial heterogeneity; 75–100%, considerable heterogeneity) [7]. When there was substantial heterogeneity (I² > 50%), we assessed the reason.

To elucidate the influence of effect modifiers, subgroup analyses according to malnutrition status (with or without malnutrition), intervention timing (preoperative, postoperative, or perioperative), cancer site (esophageal or gastric), and difference in ingredient (arginine absent or arginine present) were performed when sufficient data were available. We also performed a sensitivity analysis for the frequency of malnourishment (>50% or >75%). In one of these analyses, studies using imputed statistics were excluded, while the other included only participants who completed the study with complete data [7]. Potential publication bias was assessed by visual inspection of the funnel plots for outcomes in more than 10 studies [7].

2.7. Certainty Assessment

Based on the Cochrane Handbook [7], we summarized the findings for total postoperative complications, infectious complications, severe complications, anastomotic leakage, postoperative pneumonia, postoperative mortality, nutritional intervention adverse events, and postoperative hospitalization. The summary included grading of certainty of evidence (COE) according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach [10]. We started with “high” COE [10]. If there were any serious concerns in any domain, we lowered the grade from “high” COE. The effect estimates displayed in the Summary of Findings table were created using RRs and MD. To determine the inconsistency domain of the GRADE ratings, we examined the consistency of the RR and MD.

3. Results

3.1. Study Selection

The PRISMA flowchart is shown in Figure 1. A total of 391 records were searched on 16 October 2023. After screening, 23 studies (2249 patients) were included in the qualitative synthesis [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33] and in the quantitative synthesis [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33]. No unpublished data or ongoing studies were identified. The reasons for exclusion were incorrect population (n = 40), incorrect control (n = 4), protocol without results (n = 1), insufficient outcome data (n = 9), duplicate records (n = 2), and other reasons (n = 16).

3.2. Study and Patient Characteristics

Table 1. Summary of included studies.

Author (Year)	Country	Cancer Type	Sample Size	Age (Mean)	BMI (Mean)	Male (%)	Malnutrition (%)	Intervention Timing
Li XK (2021) [11]	China	Esophageal	103	62.1	23.6	68.9	NA	Pre + Post
Kanekiyo S (2019) [12]	Japan	Esophageal	40	65.0	21.9	80.0	25.0	Pre + Post
Mudge LA (2018) [13]	Denmark	Esophageal	276	62.5	26.9	80.8	16.0	Pre + Post
Ma C (2018) [14]	Taiwan	Gastric	34	60.2	24.2	64.7	NA	Pre + Post
Scislo L (2018) [15]	Poland	Gastric	98	62.9	24.7	72.5	NA	Post
Kitagawa H (2017) [16]	Japan	Esophageal	29	67.1	21.0	75.9	24.1	Pre
Matsuda Y (2017) [17]	Japan	Esophageal	72	64.1	21.3	75.0	34.7	Post
Ida S (2017) [18]	Japan	Gastric	123	65.0	22.7	72.4	NA	Pre + Post
Marano L (2013) [19]	Italy	Gastric	109	66.6	23.1	65.1	57.8	Post
Fujitani K (2012) [20]	Japan	Gastric	231	64.0	22.7	72.4	2.2	Pre
Sultan J (2012) [21]	UK	Esophageal	132	67.0	25.8	75.8	8.3	Pre + Post
Jiang ZM (2010) [22]	China	Upper GI	203	58.2	23.1	64.5	NA	Post
Okamoto Y (2009) [23]	Japan	Gastric	60	66.9	NA	70.0	NA	Pre
Gunerhan Y (2009) [24]	Turkey	Upper GI	29	64.6	24.1	41.4	55.2	Pre
Kłek S (2005) [25]	Poland	Upper GI	90	61.9	NA	56.7	54.4	Post
Farreras N (2005) [26]	Spain	Gastric	60	68.0	NA	53.3	21.7	Post
Adiamah A (2021) [27]	UK	Upper GI	108	66.6	NA	76.9	NA	Post
Yildiz SY (2016) [28]	Turkey	Upper GI	41	64.1	22.0	70.7	NA	Pre + Post
Sakurai Y (2007) [29]	Japan	Esophageal	30	63.0	NA	43.3	NA	Pre + Post
Lobo DN (2006) [30]	UK	Upper GI	108	66.2	NA	76.9	8.3	Post
Healy LA (2017) [31]	Ireland	Esophageal	191	62.0	28.1	79.1	4.2	Pre + Post
Ryan AM (2009) [32]	Ireland	Esophageal	53	63.9	25.9	71.7	4.0	Pre + Post
Aiko S (2009) [33]	Japan	Esophageal	29	62.1	NA	90.0	NA	Post

BMI—body mass index, GI—gastrointestinal, NA—not applicable, UK—United Kingdom.

summarizes the characteristics of the 23 studies included in the quantitative synthesis. Of these studies, 10 were for esophageal cancer [11,12,13,16,17,21,29,31,32,33], 7 were for gastric cancer [14,15,18,19,20,23], and 6 were for mixed upper GI [22,24,25,27,28,30]. Regarding the nutritional intervention timing, 5 studies were conducted preoperatively [16,20,23,24,33], 8 were postoperatively [15,17,19,22,25,26,27,31], and 10 were preoperatively and postoperatively [11,12,13,14,18,21,28,29,31,32]. We did not find any literature that included only patients with malnutrition. Table 1 summarizes the findings using the GRADE approach.

3.3. Risk of Bias

Figure 2 summarizes the ROB in the included studies. Regarding postoperative complications, there was a low ROB for incomplete outcome data and selective reporting and a low ROB or “some concerns” for random sequence generation and allocation concealment. The ROB for the participant and personnel blinding and outcome assessment were low”, some concerns” or high.

3.4. Meta-Analysis Results

Table 2. Summary of findings.

Outcomes	Relative Effect (95% CI)	Sample Size (Studies)	Certainty of the Evidence (GRADE)	Comments
Total postoperative complications	RR 0.94 (0.76 to 1.16)	1062 (9 RCTs)	⊕⊕⊕⊝ Moderate a	Immunonutrition probably does not reduce total postoperative complications.
Infectious complications	RR 0.72 (0.57 to 0.92)	1775 (18 RCTs)	⊕⊕⊕⊝ Moderate b	Immunonutrition probably reduces infectious complications.
Severe complications	RR 1.06 (0.67 to 1.67)	446 (4 RCTs)	⊕⊕⊕⊝ Moderate a	Immunonutrition probably does not reduce severe complications.
Anastomotic leakage	RR 0.72 (0.51 to 1.03)	1704 (17 RCTs)	⊕⊕⊕⊝ Moderate a	Immunonutrition probably reduces anastomotic leakage.
Postoperative pneumonia	RR 1.00 (0.79 to 1.25)	2087 (21 RCTs)	⊕⊕⊕⊝ Moderate a	Immunonutrition probably does not reduce postoperative pneumonia.
Mortality	RR 0.79 (0.44 to 1.42)	1822 (16 RCTs)	⊕⊕⊕⊝ Moderate a	Immunonutrition probably does not reduce mortality.
Postoperative hospital stay	MD −1.45 (−2.43 to −0.46)	1847 (19 RCTs)	⊕⊕⊕⊕ High	Immunonutrition reduces postoperative hospital stays.
Adverse events	RR 0.82 (0.63 to 1.06)	387 (6 RCTs)	⊕⊕⊕⊝ Moderate a	Immunonutrition probably does not increase adverse events.

CI—confidence interval; MD—mean difference; RCT—randomized control trials; RR—risk ratio. ^a Downgraded one point because of the inconsistency of the forest plot. ^b Downgraded one point because of inconsistency due to substantial heterogeneity.

shows a summary of the findings of this study. The results of the meta-analysis are shown in Figure 3.

Nine studies reported total postoperative complications. Immunonutrition does not reduce total postoperative complications compared to standard nutritional therapy (RR: 0.94, 95% CI: 0.76–1.16, I² = 47%, n = 9, COE: moderate; Figure 3a).

Eighteen studies reported infectious complications. Immunonutrition reduces infectious complications compared to standard nutritional therapy (RR: 0.72, 95% CI: 0.57–0.92, I² = 50%, n = 18, COE: moderate, Figure 3b).

Four studies reported severe complications. Immunonutrition is unlikely to reduce severe complications compared to standard nutritional therapy (RR: 1.06, 95% CI: 0.67–1.67, I² = 0%, n = 4; COE: moderate; Figure 3c).

Seventeen studies reported anastomotic leakage. Immunonutrition probably reduces anastomotic leakage compared to standard nutritional therapy (RR: 0.72, 95% CI: 0.51–1.03, I² = 0%, n = 17, COE: moderate, Figure 3d).

Twenty-one studies reported postoperative pneumonia. Immunonutrition is unlikely to reduce postoperative pneumonia compared to standard nutritional therapy (RR: 1.00, 95% CI: 0.79–1.25, I² = 0%, n = 21, COE: moderate; Figure 3e).

Sixteen studies reported mortality rates. Immunonutrition is unlikely to reduce mortality compared to standard nutritional therapy (RR: 0.79, 95% CI: 0.44–1.42, I² = 0%, n = 16, COE: moderate; Figure 3f).

Nineteen studies reported postoperative hospital stays. Immunonutrition reduces postoperative hospital stay compared to standard nutritional therapy (MD: −1.45, 95% CI: −2.43, −0.46, I² = 57%, n = 19; COE: high, Figure 3g).

Six studies reported nutritional intervention adverse events. Immunonutrition is unlikely to increase nutritional intervention adverse events compared to standard nutritional therapy (RR: 0.82, 95% CI: 0.63–1.06, I² = 0%, n = 6, COE: moderate, Figure 3h).

3.5. Subgroup Analyses

We performed subgroup analyses according to intervention timing (preoperative, postoperative, or perioperative), cancer site (esophageal or gastric), and difference in ingredient (arginine absent or arginine present). Subgroup and sensitivity analysis for malnutrition were not performed because of the paucity of studies that included patients with malnourishment.

3.5.1. Subgroup Analyses According to Intervention Timing

The results of the subgroup analyses according to intervention timing are shown in Figure 4. In the subgroup analysis, total postoperative complications were reduced for postoperative administration compared to preoperative plus postoperative administration (p = 0.05, Figure 4a). The subgroup analyses showed no efficacy differences between the groups according to the intervention timing for infectious complications (p = 0.42, Figure 4b), anastomotic leakage (p = 0.65, Figure 4c), postoperative pneumonia (p = 0.66, Figure 4d), and mortality (p = 0.81, Figure 4e). Postoperative hospital stay was shorter with postoperative administration compared to preoperative or preoperative plus postoperative administration (p < 0.001, Figure 4f). The subgroup analysis for adverse events showed no efficacy difference between the groups (p = 0.89, Figure 4g). The subgroup analysis for severe complications was not performed due to the small number of studies.

3.5.2. Subgroup Analyses According to Cancer Site

The results of the subgroup analyses according to cancer site are shown in Figure 5. The subgroup analyses showed no efficacy differences between the groups according to the cancer site for total postoperative complications (p = 0.33, Figure 5a), infectious complications (p = 0.17, Figure 5b), anastomotic leakage (p = 0.39, Figure 5c), postoperative pneumonia (p = 0.68, Figure 5d), and mortality (p = 0.48, Figure 5e). Postoperative hospital stay was shorter in the gastric cancer group than in the esophageal cancer group (p < 0.001, Figure 5f). The subgroup analysis for adverse events showed no efficacy difference between the groups (p = 0.16, Figure 5g). The subgroup analysis for severe complications was not performed due to the small number of studies.

3.5.3. Subgroup Analyses for Ingredient Difference

The results of the subgroup analyses according to ingredient differences are shown in Figure 6. The subgroup analyses showed no efficacy differences between the groups according to the ingredient difference for total postoperative complications (p = 0.73, Figure 6a), infectious complications (p = 0.70, Figure 6b), severe complications (p = 0.88, Figure 6c), anastomotic leakage (p = 0.88, Figure 6d), postoperative pneumonia (p = 0.88, Figure 6e), mortality (p = 0.81, Figure 6f), and postoperative hospital stay (p = 0.49, Figure 6g). The subgroup analysis for adverse events was not performed due to the small number of studies.

3.5.4. Subgroup Analyses for Total Duration

The results of the subgroup analyses according to total duration are shown in Figure 7. The subgroup analyses showed no efficacy differences between the groups according to the total duration for total postoperative complications (p = 0.23, Figure 7a), infectious complications (p = 0.43, Figure 7b), anastomotic leakage (p = 0.91, Figure 7c), postoperative pneumonia (p = 0.09, Figure 7d), and mortality (p = 0.31, Figure 7e). Postoperative hospital stay was shorter in the <10 days group than in the ≥10 days group (p < 0.001, Figure 7f). The subgroup analysis for adverse events showed no efficacy difference between the groups (p = 0.13, Figure 7g). The subgroup analysis for severe complications was not performed due to the small number of studies.

3.5.5. Subgroup Analyses for Malnutrition

The results of the subgroup analyses according to malnutrition are shown in Figure 8. The subgroup analyses showed no efficacy differences between the groups according to the malnutrition for infectious complications (p = 0.51, Figure 8a), anastomotic leakage (p = 0.94, Figure 8b), postoperative pneumonia (p = 0.09, Figure 8c), and mortality (p = 0.94, Figure 8d). Postoperative hospital stay was shorter in the malnutrition group than in the no-malnutrition group (p < 0.001, Figure 8e). The subgroup analyses for total complications, severe complications, and adverse events were not performed due to the small number of studies.

3.6. Funnel Plots

The funnel plots are shown in Figure 9. Funnel plots were visualized as symmetrical, indicating minimal publication bias in the reporting of infectious complications (Figure 9a), anastomotic leakage (Figure 9b), postoperative pneumonia (Figure 9c), mortality (Figure 9d), and postoperative hospital stay (Figure 9e). For total postoperative complications, severe complications, and adverse events, funnel plots were not prepared because there were fewer than 10 references, as per the Cochrane Handbook [7].

4. Discussion

The results of the present systematic review and meta-analysis of 23 studies and 2249 patients revealed that immunonutrition probably reduces infectious complications in patients with upper GI cancers after surgery. In addition, immunonutrition probably reduces the rates of anastomotic leakage and the postoperative hospital stay. However, immunonutrition is unlikely to reduce the rate of total complications, severe complications, postoperative pneumonia, and postoperative mortality. Compared to standard nutrition, immunonutrition is unlikely to increase nutritional intervention adverse events. In a subgroup analysis, postoperative administration decreased the total number of postoperative complications and shortened the length of hospital stay. The subgroup analyses showed no efficacy differences between the groups according to the ingredient difference. We did not perform sensitivity analysis because of the paucity of studies that included patients with malnutrition. This study shows that compared to standard nutritional therapy, nutritional intervention with immunonutrition can reduce postoperative complications, especially infectious complications, without increasing nutrition-related adverse events.

There was no consensus on the impact of immunonutrition on postoperative outcomes in patients with upper GI cancers because different results have been reported; however, this may be due to the different numbers of studies included in the meta-analysis. A summary of the results of previous meta-analyses is presented in

Table 3. Summary of meta-analyses.

Author (Year)	Type of Surgery	Number of Studies	Number of Patients	Intervention	Control	Outcome and Effect
Wong, C.S. (2016) [38]	Upper GI Surgery	19 RCTs	2016	Enteral IMN	Standard EN	Wound infection (RR 0.69, 95%CI: 0.50, 0.94) Length of hospital stay        (MD: −2.51, 95%CI: −3.47, −1.55) Anastomotic leakage (RR: 0.68, 95%CI: 0.42, 1.10) Pneumonia (RR: 0.81, 95%CI: 0.65, 1.01) Mortality (RR: 0.77, 95%CI: 0.43, 1.37)
Mingliang, W. (2020) [34]	Esophagectomy	7 RCTs	606	Enteral IMN	Standard EN	Infectious complications        (RR: 0.97, 95%CI: 0.78, 1.20) Pneumonia (RR: 0.97, 95%CI: 0.71, 1.33) Wound infection (RR 0.80, 95%CI: 0.46, 1.40) Anastomotic leakage (RR: 0.59, 95%CI: 0.33, 1.04)
Fu, H. (2022) [39]	Total gastrectomy	10 RCTs	1056	Enteral IMN	Standard EN	Wound infection (OR 0.77, 95%CI: 0.50, 1.19) Infectious complications        (OR: 0.72, 95%CI: 0.48, 1.09)
Zhuo, Z.G. (2021) [35]	Esophagectomy	6 RCTs	638	Enteral IMN	Standard EN	Anastomotic leakage (OR: 0.75, 95%CI: 0.43, 1.31) Wound infection (OR 0.99, 95%CI: 0.49, 2.00) Pneumonia (OR: 1.04, 95%CI: 0.65, 1.65)
Song, G.M. (2017) [37]	Gastrectomy	11 RCTs	840	Enteral IMN	Standard EN	Infectious complications        (OR: 0.56, 95%CI: 0.36, 0.86) Non-infectious complications        (OR: 1.04, 95%CI: 0.52, 2.10) Length of hospital stay        (MD: −0.42, 95%CI: −0.74, −0.10)
Tian, X. (2022) [36]	Esophagectomy	14 RCTs	1071	Enteral IMN	Standard EN	Infectious complications        (OR: 0.44, 95%CI: 0.20, 0.98) Length of hospital stay        (MD: −1.06, 95%CI: −2.56, −0.10) Anastomotic leakage (OR: 0.69, 95%CI: 0.20, 2.13) Pneumonia (OR: 0.45, 95%CI: 0.19, 0.99) Wound infection (OR 0.82, 95%CI: 0.17, 3.61)

EN—enteral nutrition; IMN—immunonutrition; MD—mean difference; OR—odds ratio; RCT—randomized controlled trials; RR—relative risk ratio.

. Mingliang et al. [34] and Zhuo et al. [35] reported that immunonutrition does not reduce infectious complications for esophageal cancer. On the other hand, Tian et al. [36] reported a decrease in infectious complications (odds ratio: 0.48, 95% CI: 0.20–0.98). The difference between them is the number of RCTs included in the meta-analysis; the former may have been underpowered. Also, Song et al. [37] reported a reduction in infectious complications in patients with gastric cancer (OR: 0.56, 95% CI: 0.36–0.86). This study has the strength of summarizing a much larger number of studies compared to previous reports and showing a possible reduction in infectious complications.

Subgroup analysis by the timing of immunonutrition administration showed the differences in total postoperative complications and postoperative hospital stay for postoperative administration. Tian et al. [36] reported that postoperative administration in patients with esophageal cancer is associated with lower rates of infectious complications and postoperative pneumonia. Osland et al. [40] reported that perioperative and postoperative dosing in patients with GI cancer can reduce postoperative complications. Similarly, this present SR showed that postoperative administration may have been effective in patients with upper GI cancers.

The mechanisms by which immunonutrition improves postoperative outcomes include improved nutritional status for malnutrition and resistance to infection by modulating immune function. High levels of postoperative inflammation from invasive surgery cause immunodeficiency, although immunonutrition has the opposite consequence of excessive inflammation suppression [6]. According to one study, preoperative immunonutrition decreased inflammatory cytokines following pancreaticoduodenectomy, which may be responsible for the decrease in postoperative complications [41]. Ates et al. [42] reported that postoperative administration of immunonutrition resulted in lower postoperative CRP. These results suggest that pre- and postoperative administration may reduce inflammation. Additionally, it functions as an immunostimulation against inflammation following surgical invasion, which decreases immunity to infection and induces immunosuppression [3]. The former should be used preoperatively as modulating and the latter postoperatively as stimulating. In this study, subgroup analysis showed that immunonutrition with arginine was not effective in preventing postoperative complications and reducing hospital stays. This may mean that immune-enhancing nutrients are not effective in patients with upper GI cancer. These may have different results depending on the degree of surgical invasiveness and the extent of resection.

We performed subgroup analysis with a cutoff value of >50% malnutrition, but there was no difference in postoperative complications between patients with and without malnutrition, and only postoperative hospital stay was significantly different. There were no RCTs in which all patients were malnourished. Riso et al. [43] compared postoperative outcomes in patients with malnutrition in head and neck cancer and reported no severe complications in the immunonutrition group. In this subgroup analysis, the reason for the absence of significant differences in postoperative complications was due to the small number of RCTs that included malnutrition. Immunomodulation by administration of immunonutrition may be necessary in patients with malnutrition because of the possibility of immune impairment, but this was not evident in this study. Further studies on patients with malnutrition are required.

This study had several limitations. First, there are few RCTs in patients with malnutrition. Further studies are required in patients with malnutrition. Second, the mechanism by which immunonutrition may improve postoperative outcomes in upper GI cancers remains unclear. One hypothesis is that it suppresses postoperative inflammation. However, further studies are required. Third, the optimal immunonutritional dose remains unknown. Further RCTs with different dosage designs are required. Despite these limitations, the meta-analysis includes large numbers of RCTs, which addresses the problem of previous meta-analyses that were unable to show statistical differences despite clinical differences owing to the small number of RCTs. Our findings have several significant clinical implications.

5. Conclusions

Nutritional intervention with perioperative immunonutrition in patients with upper GI cancer significantly reduced infectious complications without increasing nutritional intervention adverse events. The effect of immunonutrition for upper GI cancers in reducing the risk of infectious complications was about 30%. Further studies with different dosage designs are required in patients with malnutrition.