Prognostic nutritional index in predicting survival of patients with gastric or gastroesophageal junction adenocarcinoma: A systematic review

Abstract

BACKGROUND

Gastric cancer is the third most common cause of cancer related death worldwide. Surgery with or without chemotherapy is the most common approach with curative intent; however, the prognosis is poor as mortality rates remain high. Several indexes have been proposed in the past few years in order to estimate the survival of patients undergoing gastrectomy. The preoperative nutritional status of gastric cancer patients has recently gained attention as a factor that could affect the postoperative course and various indexes have been developed. The aim of this systematic review was to assess the role of the prognostic nutritional index (PNI) in predicting the survival of patients with gastric or gastroesophageal adenocarcinoma who underwent gastrectomy with curative intent.

AIM

To investigate the role of PNI in predicting the survival of patients with gastric or gastroesophageal junction adenocarcinoma.

METHODS

A thorough literature search of PubMed and the Cochrane library was performed for studies comparing the overall survival (OS) of patients with gastric or gastroesophageal cancer after surgical resection depending on the preoperative PNI value. The PRISMA algorithm was used in the screening process and finally 16 studies were included in this systematic review. The review protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO).

RESULTS

Sixteen studies involving 14551 patients with gastric or esophagogastric junction adenocarcinoma undergoing open or laparoscopic or robotic gastrectomy with or without adjuvant chemotherapy were included in this systematic review. The patients were divided into high- and low-PNI groups according to cut-off values that were set according to previous reports or by using receiver operating characteristic curve analysis in each individual study. The 5-year OS of patients in the low-PNI groups ranged between 39% and 70.6%, while in the high-PNI groups, it ranged between 54.9% and 95.8%. In most of the included studies, patients with high preoperative PNI showed statistically significant better OS than the low PNI groups. In multivariate analyses, low PNI was repeatedly recognised as an independent prognostic factor for poor survival.

CONCLUSION

According to the present study, low preoperative PNI seems to be an indicator of poor OS of patients undergoing gastrectomy for gastric or gastroesophageal cancer.

Key Words: Prognostic nutritional index, Gastric adenocarcinoma, Gastroesophageal junction cancer, Prognosis, Overall survival

Core Tip: In the present systematic review, we investigated the role of prognostic nutritional index (PNI) in predicting the survival of patients with gastric or gastroesophageal junction adenocarcinoma that were submitted to surgery with or without chemotherapy. PNI is easy to calculate and provides information about the nutritional status of the patients. Low preoperative PNI seems to be associated with worse survival in patients that will undergo surgery for gastric or gastroesophageal junction adenocarcinoma and therefore could be useful for decision making in clinical practice.

INTRODUCTION

Gastric cancer (GC) is one of the most common malignancies and the third most common cause of cancer-related deaths. In 2018, there were approximately 1033701 new cases of GC reported worldwide, resulting in 783000 associated deaths[1-3]. Surgery with or without chemotherapy remains the cornerstone in the management, as it may have curative results. Despite advancements in surgical procedures, chemotherapy, radiotherapy, and targeted treatments, postoperative complications and mortality rates remain high, leading to a poor prognosis for patients[4].

Pathological characteristics such as tumor stage, nodal status, and resection margin are thought to be crucial in determining cancer patient survival[5]. However, it is now obvious that tumor pathology is not the only factor that influences cancer survival; muscle mass, nutritional profile, immunological conditions, and other variables could significantly affect surgical outcomes[6,7]. Malnutrition is particularly common among this group of patients and is attributed to inadequate oral intake, protein-losing gastropathy, ongoing bleeding due to tumors, and ineffective nutritional pathways. Numerous studies have revealed that malnutrition relates to a poor prognosis and multiple postoperative complications in cancer patients, as a result of worsening overall health and increased treatment challenges, emphasizing the need for proper perioperative management to improve the nutritional status of each individual[8-10]. Therefore, early evaluation and management of a patient’s nutritional status may enhance the prognosis and outcomes for those undergoing curative surgery for GC[11].

Historically, tumor-node-metastasis (TNM) classification has been the most prevalent and reliable indicator for patient prognosis. However, there is an increasing number of cases where patients classified at the same stage exhibit significantly different prognoses[12,13]. Recent studies have demonstrated that perioperative inflammation-based prognostic scores can predict overall survival (OS) in patients with diverse forms of cancer[14]. Albumin levels are a key indicator of a patient’s nutritional status. Several scores based on albumin levels have been developed, such as the nutritional index (NI), Glasgow Prognostic Score, Nutrient Profiling System, and Controlling Nutritional Status score[11,15]. The prognostic NI (PNI) has gained popularity as a means of predicting the surgical risk of patients with GC. It is calculated by multiplying 10 times the serum albumin value (g/dL) plus 0.005 times the lymphocytes count (/mm³). It utilizes nutritional and inflammation status instead of tumor growth, node invasion, and metastasis stage[1,16,17]. PNI has been used as a prognostic tool for patients with solid organ tumors with significant prognostic value regarding survival and postoperative complications, paving the way for individualized perioperative management[12,17,18]. In this systematic review, we aimed to assess the role of PNI in predicting the survival of patients undergoing curative intent surgery for gastric or gastroesophageal adenocarcinoma.

MATERIALS AND METHODS

Study selection

A thorough literature search of PubMed and the Cochrane library was conducted for articles comparing the OS of patients with gastroesophageal cancer after surgical resection depending on their preoperative PNI over the past 10 years. The terms “prognostic nutritional index”, “PNI”, “gastric cancer”, “gastroesophageal cancer”, “gastric adenocarcinoma”, and “survival” were used in various combinations. The PubMed search yielded 788 results that were scrutinized against the predetermined inclusion and exclusion criteria (Table 1). After title and abstract screening and exclusion of duplicates and irrelevant articles, 71 were eligible for further assessment. After full text screening, finally 16 studies were included in our systematic review. The search and screening processes were completed by two independent reviewers using the PRISMA algorithm and any conflict was resolved through discussion (Figure 1)[19]. The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO ID CRD42023461282).

Figure 1 PRISMA flowchart.

Table 1 Inclusion and exclusion criteria.

Inclusion criteria	Exclusion criteria
Cohort studies	Metastatic or recurrent disease
Studies in English language	Patients receiving palliative care
Studies published over the past 10 yr (2013 to 2023)	Cancer other than adenocarcinoma
Adult patients (over 18 years old)	Case-control studies
Patients with histologically confirmed esophagogastric or gastric adenocarcinoma	Commentaries or letters to the editor
Open or laparoscopic gastrectomy
Effect of preoperative PNI on OS as primary outcome

PNI: Prognostic nutritional index; OS: Overall survival.

Risk of bias and quality assessment

Only cohort studies were included in our systematic review. The risk of bias and the quality of each individual study were assessed using the Cochrane Tool to assess Risk of Bias in Cohort studies and the Newcastle-Ottawa quality assessment scale (NOS), respectively[20]. The Cochrane Tool consists of seven questions and according to the answers a cohort study was categorized as of low or of high risk of bias. The NOS consists of three categories (Selection, Comparability, and Outcome) and eight items. A maximum of one star can be awarded for each item within the selection and outcome categories and a maximum of two stars for the comparability[1,12,13,15,17,21-31] (Table 2). A study with a score of over six stars is considered to be of high quality.

Table 2 Newcastle-Ottawa scale scores for the included studies.

Ref.	Selection				Comparability	Outcomes			Total
	Representativeness of the exposed cohort	Selection of the non exposed cohort	Ascertainment of exposure	Outcome of interest not present at the start of the study		Assessment of outcome	Length of follow-up	Adequacy of follow-up
Hashimoto et al[21]	*	*	*	*	*	*			6/8
Hirahara et al[31]	*	*	*	*	*	*	*	*	8/8
Hirahara et al[22]	*	*	*	*	*	*			6/8
Ishiguro et al[23]	*	*	*	*	*	*		*	7/8
Kudou et al[12]	*	*	*	*	*	*	*	*	8/8
Lee et al[13]	*	*	*	*	*	*	*	*	8/8
Lin et al[15]	*	*	*	*	*	*	*	*	8/8
Liu et al[24]	*	*	*	*	*	*		*	7/8
Murakami et al[1]	*	*	*	*	*	*		*	7/8
Saito et al[25]	*	*	*	*	*	*		*	7/8
Shen et al[26]	*	*	*	*	*	*	*	*	8/8
Takechi et al[17]	*	*	*	*	*	*			6/8
Toyokawa et al[27]	*	*	*	*	*	*	*	*	8/8
Toyokawa et al[28]	*	*	*	*	*	*	*	*	8/8
Wu et al[29]	*	*	*	*	*	*		*	7/8
Xu et al[30]	*	*	*	*	*	*	*	*	8/8

Data extraction

The following data were extracted: Year of publication, institution, study period, number of participants, patient diagnosis and operation, PNI cut-off value, patient age, sex, body mass index, albumin and lymphocyte count, tumor location and TNM stage, follow-up period, OS, and the univariate and multivariate analysis results. Data extraction was completed by four of the reviewers. Any disagreement during that phase was resolved by consulting a senior reviewer.

RESULTS

Sixteen studies involving 14551 patients with non-metastatic gastric or gastroesophageal adenocarcinoma who underwent surgery with curative intent between 1997 and 2021 were included in our systematic review. The clinical and pathological characteristics of the patients are shown in Tables 2 and 3.

Table 3 Institute, period, and patient demographics.

Ref.	Institute	Period	Patients number	Sex	Age (yr)
Hashimoto et al[21]	Sasebo City General Hospital, Japan	2013-2020	109	68 M, 41 F	83 (80-94)
Hirahara et al[31]	Department of Digestive and General Surgery, Shimane University, Japan	2009-2016	218	145 M, 77 F	Low PNI group (n = 109): 78 (46-91). High PNI group (n = 259): 69 (36-89)
Hirahara et al[22]	Department of Digestive and General Surgery, Shimane University, Japan	2010-2016	368	254 M, 114 F	Absent postoperative complications group (n = 265): 70 (36-91). Present postoperative complications group (n = 103): 73 (41-90)
Ishiguro et al[23]	Department of Surgery in Yokohama City, Japan	2015-2021	258	183 M, 75 F	31-88
Kudou et al[12]	Department of Surgery and Science, Kyushu University; Department of Gastroenterological Surgery, National Kyushu Medical Center, Japan	2005-2016; 2010-2019 (respectively to the 2 institutes)	206	151 M, 55 F	66.3 (35-92)
Lee et al[13]	Severance Hospital, South Korea	2001-2010	7781	5150 M, 2631 F	57.1 ± 11.9
Lin et al[15]	Fujian Medical University Union Hospital, China	2009-2014	2182	1643 M, 539 F	60.8 (54-68.3)
Liu et al[24]	Sun Yat-sen University Cancer Center, China	2000-2012	1330	905 M, 425 F	59 (19-89)
Murakami et al[1]	Tottori University Faculty of Medicine, Japan	2001-2013	254	186 M, 68 F	> 70, n = 128; < 70, n = 126
Saito et al[25]	Division of Surgical Oncology, Department of Surgery, School of Medicine, Tottori University Faculty of Medicine, Japan	2005-2013	453	331 M, 122 F	Low PNI group (n = 188): 73.5. High PNI group (n = 265): 63.5
Shen et al[26]	General Surgery Department of the Jinling Hospital, China	2010-2018	525	387 M, 138 F	Training set (n = 369): 58.53 ± 10.14. Validation set (n = 156): 57.87 ± 10.28
Takechi et al[17]	Onomichi General Hospital, Hiroshima, Japan	2011-2014	182	130 M, 52 F	70 (38-91)
Toyokawa et al[27]	Osaka City University Hospital, Japan	1997-2012	240	168 M, 72 F	64.5 (58-71.3)
Toyokawa et al[28]	Osaka City University Hospital, Japan	1997-2012	225	147 M, 78 F	68 (60-75)
Wu et al[29]	Affiliated Hospital of Jiangnan University, Jiangsu Province, China	2015-2017	77	59 M, 18 F	62.58 ± 8.97
Xu et al[30]	Shantou University Medical College’s cancer hospital, China	2016-2020	236	171 M, 65 F	43.68 ± 4.62

PNI: Prognostic nutritional index; M: Male; F: Female.

The PNI cut-off values used in the studies and the survival outcomes are shown in Table 4. PNI was calculated as 10 × albumin (g/dL) + 0.005 × total lymphocyte count (/mm³) and its thresholds ranged between 44.2 and 47 in the majority of the studies; however, three studies used the cut-off values 42.3, 49.2, and 52, respectively. There were 10864 patients in the high PNI group and 3687 in the low PNI group. The patients were submitted to total or partial gastrectomy and lymphadenectomy with or without adjuvant or neoadjuvant chemotherapy. The primary endpoint of our study was OS and the results of univariate and multivariate analyses of the studies that were included are shown in Tables 5 and 6.

Table 4 Patient clinical characteristics.

Ref.	Tumor location	TNM stage	Operation	Chemotherapy
Hashimoto et al[21]	NS	I, n = 53 (48.6%). II, n = 31 (28.4%). III, n = 25 (22.9%)	Open surgery 54 (49.5%), laparoscopic 55 (50.5%). Distal/total/proximal gastrectomy: 70/37/2. D2 lymphadenectomy, n = 38 (34.9%)	Adjuvant 13 (11.9%). Neoadjuvant NS
Hirahara et al[31]	EGJ, n = 6. Upper, n = 41. Middle, n = 91. Lower, n = 80	Ia-Ib, n = 92, IIa-IIb, n = 51, IIIa-IIIc, n = 75. T1/2/3/4: 80/27/45/66. N0/1/2/3: 120/30/33/35	Laparoscopic total/laparoscopic partial/laparoscopy assisted distal gastrectomy: 60/14/144	Adjuvant: Yes n = 79, no n = 139. Neoadjuvant chemotherapy in the exclusion criteria
Hirahara et al[22]	EGJ, n = 11. U, n = 70. M, n = 162. L, n = 125	IA-IB, n = 217. IIA-IIB, n = 65. IIIC-IIIC, n = 86. T1/2/3/4: 192/48/54/74. N0/1/2/3: 244/40/42/42	Laparoscopic total/laparoscopic partial/laparoscopy assisted distal gastrectomy: 82/37/249	Adjuvant: Yes n = 100, no n = 268
Ishiguro et al[23]	Upper, n = 63 (24.4%). Middle, n = 113 (43.8%). Lower, n = 82 (31.8%)	T1, n = 138 (53.5%). T2 or T3, n = 120 (46.5%). Lymphatic invasion positive/negative: 90 (34.9%)/168 (65.1%)	Total/distal/partial gastrectomy: 66/180/11. D1+/D2 lymphadenectomy: 139/112	77% of the patients in the high PNI group and 47% in the low PNI group (amongst stages II and III patients)
Kudou et al[12]	EGJ = 96, UGC = 110	T1 97 (47.1%), T2 25 (12.1%), T3 55 (26.7%), T4 29 (14.1%). N0 136 (66.0%), N1 33 (16.0%), N2 13 (6.3%), N3 24 (11.7%). I/II/III: 113 (54.9%)/52 (25.2%)/41 (19.9%)	Total/proximal gastrectomy: 161/45. D1 lymphadenectomy (for T1 tumors), n = 97. D2 lymphadenectomy (for T2-4 tumors), n = 64	Adjuvant: Yes, n = 51 (24.8%), no, n = 155 (75.2%). Neoadjuvant chemotherapy in the exclusion criteria
Lee et al[13]	NS	T1 4182 (53.8%), T2 944 (12.1%), T3 913 (11.7%), T4a 1700 (21.9%), T4b 42 (0.5%). N0 4967 (63.8%), N1 941 (12.1%), N2 798 (10.3%), N3 1075 (13.8%). Stage I 4608 (59.2%), II 1286 (16.5%), III 1887 (24.3%)	Subtotal gastrectomy 5895 (75.8%). Total gastrectomy 1886 (24.2%)	Patients with stage II or higher disease were recommended for adjuvant chemotherapy (numbers not mentioned). Neoadjuvant chemotherapy in the exclusion criteria
Lin et al[15]	Upper 521 (23.9%). Middle 465 (21.3%). Lower 923 (42.3%). Mixed 273 (12.5%)	TNM stage: I 632 (29.0), II 526 (24.1), III 1024 (46.9)	Total gastrectomy 1134 (52.0%). Distal gastrectomy 998 (45.7%). Proximal gastrectomy 50 (2.3%)	1223 patients (56%): Adjuvant chemotherapy n = 1223 (56%). Neoadjuvant chemotherapy NS
Liu et al[24]	Upper third 511 (38.4%). Middle third 278 (20.9%). Lower third 541 (40.7%)	I 220 (16.5%). II 334 (25.1%). III 776 (58.3%)	D2 gastrectomy with R0 resection	Adjuvant chemotherapy n = 817. Neoadjuvant chemotherapy in the exclusion criteria
Murakami et al[1]	NS	T1 n = 147, T2/3/4 n = 107. N0 n = 181, N1/2/3 n = 73. Stage I n = 161, II/III n = 93	Distal/proximal gastrectomy n = 181, total gastrectomy n = 73. D0/1/1+ lymphadenectomy n = 171, D2 lymphadenectomy n = 83	NS
Saito et al[25]	NS	T1 n = 284, T2/3/4 n = 169. Lymph node metastasis absent/present: 343/110	Curative gastrectomy (R0 resection) with regional dissection of lymph nodes. Partial/proximal/total gastrectomy: 311/42/100	Adjuvant chemotherapy n = 64, neoadjuvant chemotherapy n = 5, perioperative chemotherapy n = 10
Shen et al[26]	Upper 158. Middle 202. Lower 165	Training/validation set: I 138 (37.40%)/64 (41.03%), II 84 (22.76%)/39 (25.00%), III 147 (39.84%)/53 (33.97%)	Robotic gastrectomy proximal/distal/total: 110/272/143	Neoadjuvant chemotherapy n = 116, adjuvant n = 267
Takechi et al[17]	NS	Stage: I n = 114 (62.6%), II n = 38 (20.9%), III n = 30 (16.5%)	Distal/total/proximal gastrectomy: 124 (68.1%)/51 (28%)/7 (3.8%). D1/D1+/D2 lymphadenectomy: 32 (17.6%)/74 (40.7%)/76 (41.8%)	Postoperative patients with stages II and III GC n = 33 (18.1%). Neoadjuvant NS
Toyokawa et al[27]	Upper n = 57 (23.8%). Middle n = 98 (40.8%). Lower n = 83 (34.6%). Whole n = 2 (0.8%)	Only stage II patients: IIA n = 111 (46.3%), IIB n = 129 (53.7%)	Total/proximal/distal gastrectomy: 72/1/167	Adjuvant chemotherapy: Yes 62/no 178. Neoadjuvant in the exclusion criteria
Toyokawa et al[28]	Upper/middle/lower n = 209 (92.9%). Whole 16 (7.1%)	IIIA 80 (35.6%), IIIB 72 (32.0%), IIIC 73 (32.4%)	Total/distal gastrectomy: 108 (48%)/117 (52%)	Adjuvant chemotherapy: Yes 41 (18.2%)/no 184 (81.8%)
Wu et al[29]	NS	Only stage III: n = 77 (100%)	Partial gastrectomy (n = 15), total gastrectomy (n = 62)	The average number of chemotherapy cycles was 6.77 ± 4.14, and all patients completed > 2 chemotherapy cycles. Neoadjuvant chemotherapy in the exclusion criteria
Xu et al[30]	EGJ	I 48 (20.3%), II 53 (22.4%), III 135 (57.2%)	Curative gastro-esophageal resection with R0 resection	NS

NS: Not stated; TNM: Tumor-node-metastasis; EGJ: Esophagogastric junction; U: Upper; M: Middle; L: Lower; UGC: Upper gastric cancer; GC: Gastric cancer.

Table 5 Prognostic nutritional index cut-off value calculation method, threshold value and range, follow-up, and survival of the patients.

Ref.	PNI calculation	PNI cut-off value and groups	PNI range	Follow up (months)	Outcome
Hashimoto et al[21]	ROC curve analysis	44.2. PNI > 44.2 (n = 72), PNI < 44.2 (n = 37)	NS	23.9 (0.4-81.9)	The 30-d, 180-d, 1-yr, and 3-yr cumulative OS rates were 100%, 97.0%, 91.6%, and 74.7%, respectively
Hirahara et al[31]	ROC curve analysis	44.3. PNI < 44.3 (n = 109), PNI > 44.3, n = 109	NS	Observation period from date of surgery till day of death	5-yr OS: Low PNI, 50.9%; high PNI, 73.6% (P < 0.001). In elderly patients (age > 70) 5-yr OS: Low PNI 52.5%, high PNI 82.5%. Non elderly patients (age < 70), 5-yr OS: Low PNI 46.6%, high PNI 54.7%
Hirahara et al[22]	ROC curve analysis	44.5. PNI < 44.5 n = 114, PNI > 44.5, n = 254	NS	NS	NS
Ishiguro et al[23]	Set according to previous reports	47. PNI < 47 (n = 75), PNI > 47 (n = 183)	NS	NS	5-yr OS: A: 44.7%; B: 77.2% (P < 0.001)
Kudou et al[12]	ROC curve analysis	44.7. PNI < 44.7 (n = 167, 81.1%), PNI > 44.7 (n = 39, 18.9%)	NS	60	Worse 5-year OS rates were associated with PNI < 44.7 (vs > 44.7) (OS: 41.7% vs 84.5%, HR = 5.460, P < 0.0001). In subgroup analysis PNI < 44.7 (vs > 44.7) was significantly associated with poor prognosis in patients with stages II and III disease
Lee et al[13]	ROC curve analysis	46.7. PNI < 46,7 (n = 779), PNI > 46,7, n = 7002	54.2 ± 5.9	60	The low PNI group had a poor prognosis for all stages of disease (for all stages and stages I, II, and III: P < 0.001)
Lin et al[15]	Set according to previous reports	46. PNI ≤ 46 (n = 1348, (61.8%), PNI > 46 (n = 834, 38.2%)	NS	52 (1-118)	Low PNI 5-yr OS = 55.5%, high PNI 5-yr OS = 75.4%
Liu et al[24]	Set according to previous reports	45. Low PNI group PNI < 45. Number of patients NS		35 (range 1-179). Final follow-up June 2015, 806 patients were alive by then	NS
Murakami et al[1]	ROC curve analysis	Preoperative PNI of ≥ 52 (pre-PNIhigh) n = 82, preoperative PNI < 52 (pre-PNIlow) n = 172, postoperative PNI ≥ 49 (post-PNIhigh) n = 95, postoperative PNI < 49 (pre-PNIlow) n = 159. Group A, patients with pre-PNIhigh and post-PNIhigh; group B, patients with either pre-PNIhigh and post-PNIlow or pre-PNIlow and post-PNIhigh; group C, patients with pre-PNIlow and post-PNIlow	Preoperative PNI range 30.6-63.6. Postoperative range 24.2-61.7	NS	5-yr OS prePNIhigh 95.8%, prePNIlow 70% (P < 0.0001). 5-yr OS postPNIhigh 91.4%, postPNIlow 70.1% (P < 0.0001). 5-yr OS prePNIlow and postPNIhigh 80.1%, prePNIlow and postPNIlow 67.1% (P = 0.031). 5-yr OS prePNIhigh and postPNIhigh 100%, prePNIhigh and postPNIlow 83.4% (P = 0.0021)
Saito et al[25]	ROC curve analysis	46.7. PNI ≥ 46.7 (n = 265, 58.5%) and PNI < 46.7 (PNIlow, n = 188, 41.5%)	Range 27.7-63.6	NS	5-yr OS PNIlow 59.5%, PNIhigh 88.2% (P < 0.0001)
Shen et al[26]	X-tile 3.6.1 software1 (Yale University, New Haven, CT, United States)	45.39. Training set low PNI n = 48 (13.01%), high PNI 321 (86.99%), validation set low PNI n = 29 (18.59%), high PNI n = 127 (81.41%). Patients were randomly divided into the training set and the validation set at a 7:3 ratio	NS	41 (range 2-102) training set and 38 (range 1-101) validation set	3-yr and 5-yr OS rates were 80.9% and 74.8% in the training set, and 81.6% and 73.5% in the validation set
Takechi et al[17]	Set according to previous reports	45. PNI < 45 (n = 97), PNI ≥ 45 (n = 85)	NS	39 (range, 1-72)	NS
Toyokawa et al[27]	ROC curve analysis	49.2. PNI ≤ 49.2 (n = 136), PNI > 49.2) (n = 104)	NS	100.5 (70.0-136.8)	The 5-yr OS rate for the entire study population was 78.8%
Toyokawa et al[28]	ROC curve analysis	45.6. PNI ≤ 45.6 (n = 90, 40%), PNI > 45.6 (n = 135, 60%)	46.8 (IQR: 42.5-49.9)	Median 80 (69-124)	The 5-yr OS rate for the entire study population was 48.7%.
Wu et al[29]	ROC curve analysis	42.3. Low PNI group PNI < 42.3. Number of patients NS	NS	Shortest 30, longest 64	3-yr OS low PNI group < 40%, high PNI group > 60%. Exact number NS (only survival curves available)
Xu et al[30]	ROC curve analysis	45.6. Propensity matching patients. PNI < 45.6 (n = 58), PNI > 45.6 (n = 85)	NS	Every 3 months first 2 yr, every 12 months for 3rd-5th yr, once per year after that. Final follow-up December 2022	Low PNI group had a 5-yr OS rate of 46.9%, high PNI group had a 5-yr OS rate of 71.30%

¹A bioinformatics tool for biomarker assessment and outcome-based cut-point optimization designed by Yale University.

PNI: Prognostic nutritional index; OS: Overall survival; ROC: Receiver operating characteristic; NS: Not stated; IQR: Interquartile range; HR: Hazard ratio.

Table 6 Univariate and multivariate analysis results.

Ref.	Univariate analysis	Multivariate analysis
Hashimoto et al[21]	Low PNI associated with poor OS (P = 0.049)	Low PNI was an independent prognostic factor for poor OS (P = 0.044)
Hirahara et al[31]	Low PNI value was a significant risk factor for shorter OS (P < 0.001)	PNI was confirmed as an independent prognostic factor for OS (P < 0.001)
Hirahara et al[22]	PNI was significantly associated with OS (HR = 3.316, 95%CI: 2.133-5.196, P < 0.001)	In patients with high PNI, only CEA was was independently associated with OS (P = 0.002)
Ishiguro et al[23]	PNI was significantly associated with OS (P < 0.001)	PNI was an independent predictor of OS (HR = 3.452, 95%CI: 2.042-5.836, P = 0.007)
Kudou et al[12]	PNI < 44.7 (vs > 44.7) was associated with worse OS (P < 0.0001)	PNI (P < 0.0001, HR = 8.946) was independently associated with OS
Lee et al[13]	Low PNI was significantly associated with worse OS (HR = 2.864, 95%CI: 2.544-3.223, P < 0.001)	Low PNI was independently associated with OS (HR = 1.383, 95%CI: 1.221-1.568, P < 0.001)
Lin et al[15]	PNI was significantly associated with OS (P < 0.001)	PNI was independently associated with OS (P = 0.004) and the 5-yr OS rate in the low PNI group was significantly lower than that in the normal PNI group (55.5% vs 75.4%, P < 0.05)
Liu et al[24]	PNI was associated with OS (HR = 1.627, 95%CI: 1.274-2.078, P < 0.001)	PNI (HR = 1.356, 95%CI: 1.051-1.748, P = 0.019) was independently associated with OS. In stage stratified analysis PNI was not significantly associated with OS
Murakami et al[1]	5-yr survival rates were 100.0, 83.0, and 67.1% for groups A, B, and C, respectively	5-yr OS 100%, 92.4%, and 78.3% for groups A, B, and C, respectively, in non-elderly patients (age < 70) (P = 0.017). 5-yr OS 100%, 75.1%, and 59% for groups A, B, and C, respectively, for elderly patients (age > 70) (P = 0.0029). Group stratification mentioned in Table 4
Saito et al[25]	NS	5-yr OS PNI low group 59.5%, PNI high group 88.2% (P < 0.0001). Median age of the PNI high group (63.5 yr) was significantly younger than of the PNI low group (73.5 yr)
Shen et al[26]	PNI was an independent prognostic factor for OS. PNI (≤ 45.39 vs > 45.39) (HR = 0.439, 95%CI: 0.236-0.734, P = 0.002)	PNI was an independent prognostic factor for OS. PNI (≤ 45.39 vs > 45.39) (HR = 0.553, 95%CI: 0.306-0.993, P = 0.048)
Takechi et al[17]	Low PNI was significantly associated with worse OS (HR = 4.261, 95%CI: 1.734-10.47, P = 0.002). Stage I GC patients in the high PNI group showed significantly better OS than patients in the low PNI group (P < 0.001). No significant difference in OS between PNI groups in stage II and III GC patients	Only PNI score was an independent prognostic factor for OS (HR = 2.889, 95%CI: 1.104-7.563, P = 0.031)
Toyokawa et al[27]	PNI was significantly associated with OS (HR = 0.381, 95%CI: 0.219-0.662, P = 0.001)	PNI was an independent prognostic factor for OS (HR = 0.415, 95%CI: 0.234-0.736, P = 0.003)
Toyokawa et al[28]	PNI was not significantly associated with OS (P = 0.073)	PNI was not significantly associated with OS (P = 0.676)
Wu et al[29]	-	The group with high pre-chemotherapy PNI values had significantly better overall survival than the group with low pre-chemotherapy PNI values (HR = 0.485, 95%CI: 0.255-0.920; P = 0.027)
Xu et al[30]	Lower PNI was a significant predictor of shorter OS (P = 0.004)	In comparison to the high PNI group, the hazard of endpoint mortality was 2.442 times greater in the low PNI group (P = 0.003)

PNI: Prognostic nutritional index; OS: Overall survival; HR: Hazard ratio; CI: Confidence interval; GC: Gastric cancer; CEA: Carcinoembryonic antigen.

OS

PNI was significantly associated with the OS in all of the studies included except for Toyokawa et al[27] enrolled 225 patients with stage III only gastric adenocarcinoma; 184 of them were submitted to adjuvant chemotherapy and PNI was not associated with OS. Hirahara et al[22] included 368 patients that were submitted to laparoscopic or laparoscopy assisted gastrectomy and 100 of them were also submitted to adjuvant chemotherapy. The authors demonstrated in univariate analysis that PNI was significantly associated with OS; however, the same result was not reached in multivariate analysis which showed that only carcinoembryonic antigen was significantly associated with OS.

Of 258 patients who underwent curative resection for GC and were included in Ishiguro et al[23]’s study, adjuvant chemotherapy was not administered to patients with stage I GC but only to patients with stage II or III. The authors demonstrated that PNI was independently associated with OS.

Lin et al[15] included 632 patients with stage I GC, 526 with stage II, and 1024 with stage III who underwent curative gastrectomy; 56% of them received adjuvant chemotherapy. The authors showed that PNI was independently associated with OS. Saito et al[24] included 111 GC patients with and 343 without lymphatic invasion; 64 patients received adjuvant and 5 neoadjuvant chemotherapy. The authors demonstrated that high PNI was significantly associated with better OS.

Hashimoto et al[21] included only elderly patients between 80 and 94 years of age. Fifty-four of them were submitted to open surgery and fifty-five to laparoscopic surgery; however, it was not stated whether neoadjuvant chemotherapy was administered. The authors demonstrated that PNI was an independent prognostic factor for OS and reported a cumulative 3-year OS rate of 74.7%. Xu et al[30] included younger patients (mean age 43.68 ± 4.62) and they also showed that low PNI was significantly associated with lower OS.

Shen et al[26] included 525 patients with stages I-III GC in their study who were submitted to robotic gastrectomy, 116 of them to neoadjuvant chemotherapy and 267 of them to adjuvant chemotherapy, and they randomly divided the patients to a training and a validation set. The authors showed that PNI was significantly associated with OS in both sets and that PNI was an independent prognostic factor for OS.

Wu et al[29], Liu et al[24], and Toyokawa et al[27] also showed that PNI was significantly associated with OS. Of note, Wu et al[29] included only patients with stage III adenocarcinoma in their study. Toyokawa et al[28], who also included only stage III patients, demonstrated that PNI was not significantly associated with OS in those patients. In another study, which included stage II only GC patients, Toyokawa et al[27] showed that PNI was an independent prognostic factor for OS. Takechi et al[17] confirmed a significant association of PNI and OS only for stage I patients but not for stage II or III patients. Finally, Lee et al[13] and Kudou et al[12] showed that PNI was significantly associated with OS in stage-stratified analysis for stage I, stage II, and stage III gastric adenocarcinoma patients.

Murakami et al[1] reported a 5-year OS rate of 70% and 95.8% in preoperatively low- and high PNI groups, respectively (P < 0.0001). In this study, postoperative PNI values were also recorded at 1 month after surgery. The authors demonstrated that the patients with preoperative and postoperative PNI higher than the threshold had a better 5-year OS (100%) compared to those who had preoperative or postoperative PNI lower than the threshold (5-year OS 83%) and compared to the patients with preoperative and postoperative low PNI (5-year OS 67.1%). The authors also compared the 5-year OS rates of these groups in elderly (age over 70 years old) and non-elderly patients (age under 70 years old) and they reported lower OS rates for elderly patients (100% vs 75.1% vs 59% for elderly patients and 100% vs 92.4% vs 78.3% for non-elderly patients). Hirahara et al[31] also examined the 5-year OS in low- and high-PNI groups in elderly and non-elderly patients. A greater difference in 5-year OS was reported between the low- and high-PNI groups in elderly patients (52.5% vs 82.5%) compared to non-elderly patients (46.6% vs 54.7%).

DISCUSSION

In the present study, we assessed the role of PNI in the prognosis of patients with gastric or gastroesophageal junction cancer. However, there are discrepancies in the literature depending on cancer stage. For instance, according to Migita et al[32], a low PNI is a significant predictor of poor OS in patients with GC at stages I and III, but not at stages II and IV. In the study of Sakurai et al[33], a low PNI was found to be a negative prognostic factor in stages I and II, but not in stage III. This disparity could be explained by the fact that, in addition to cancer stage, other clinicopathological factors, including but not limited to patient’s age, nutritional status, or lymphatic or vessel invasion, could influence the survival of patients with different stages of GC[34-36]. Of note, the data are still not conclusive as to whether PNI has better prognostic value in early or advanced GC stage[36-38].

Undoubtedly, there is a link between the nutritional status and the prognosis of patients with GC. Many studies have shown that malnutrition has a negative impact on the prognosis of cancer patients due to its effects on the immune system function resulting in impaired general health and increased treatment complications[9,36]. PNI is a systemic inflammatory marker that has been shown to be useful in cancer prognosis. However, there has been little research into the impact of inflammation on the tumor microenvironment[15]. It is unclear whether a low preoperative PNI is a cause or a result of tumor progression. A low preoperative PNI, according to a meta-analysis of Li et al[39], was significantly associated with poor OS, as well as increased postoperative mortality. Therefore, assessing nutritional status is critical because it allows for the identification of malnourished patients at high risk and the implementation of appropriate nutritional interventions that could possibly improve prognosis and reduce complications. However, Migita et al[40] showed that preoperative oral nutritional supplementation did not improve low preoperative PNI, therefore more research needs to be done regarding the optimal means of preoperative nutritional support.

The PNI cut-off values varied between the studies that were included in this systematic review and the methods that were used to calculate the PNI cut-off value are mentioned in Table 4. An optimal cut-off value for predicting long-term outcomes has not been established in the literature[36]. Future research should focus on standardizing the PNI thresholds by performing receiver operating characteristic curve analysis in prospective studies that include patients with minimum clinicopathological characteristics heterogeneity in order to identify the PNI cut-off value with the maximum sensitivity and specificity, as a standardized PNI cut-off value may have significant impact in daily clinical practice and decision-making. For instance, in the study of Kosuga et al[41], it was shown that preoperative PNI may help clarify the extent of lymphadenectomy in patients undergoing gastrectomy for GC.

In our study, there are some limitations that need to be addressed. First, all of the included studies were retrospective cohort studies, which are prone to selection or recall bias. Furthermore, not all patients included in this study had the same stage of cancer. Also, some of them received neoadjuvant therapy prior to surgery, which possibly affected the overall course of the disease. Patients were not divided according to specific tumor characteristics, such as the TNM stage, the size or depth or the tumor, the Siewert type, or the tumor differentiation, therefore a correlation between the PNI and the survival depending on the multiple tumor characteristics could not be established. It would be of interest if future studies would stratify patients and assess the prognostic significance of the PNI based on those characteristics. Due to high heterogeneity of the recorded data, a meta-analysis could not be performed. Of note, the fact that not all deaths were confirmed cancer-related deaths is something to take into account. Furthermore, the studies that we included in our study were all performed in Eastern Asia countries and there were no studies performed in Western countries that matched our inclusion criteria. This could be attributed to the fact that East Asia has the highest prevalence of GC (20-25 patients per 100000, less than 5 patients per 100000 in Northern America)[42]. Finally, patients underwent operations in different institutions by different surgical teams of variable experience, which may have had an impact on the postoperative events.

CONCLUSION

In conclusion, all of the studies that we included showed that patients with higher preoperative PNI demonstrate better survival than those with lower PNI after surgery for gastric or gastroesophageal cancer with or without chemotherapy regardless of the tumor stage, patients age, total or partial gastrectomy, and open or laparoscopic gastrectomy, except for one study that included stage III GC patients. Future studies should focus on stratifying patients based on tumor stage, as well as on standardizing the PNI cut-offs. Moreover, more research needs to be done in terms of preoperative nutritional support as it could increase PNI and therefore improve short- and long-term outcomes. Moreover, more studies should be performed in Western countries in order to examine whether the association between PNI and survival persists in those patients who undoubtedly present different genetic factors. Finally, PNI could be a useful clinical tool, as it is easy to calculate with standard everyday labs and may lead to individualized patient care and clinical decisions with optimal results.

ARTICLE HIGHLIGHTS

Research background

Gastric cancer (GC) is a major health problem worldwide. Patients with GC that are eligible for surgery are submitted to gastrectomy with lymphadenectomy followed or not by adjuvant chemotherapy. It is important to identify prognostic factors that could predict the survival of those patients. The prognostic nutritional index (PNI) is an indicator of the nutritional and immune status of GC patients that could assist in identifying patients that will benefit the most from being submitted to surgery and that will present better survival rates.

Research motivation

GC patients with high preoperative PNI seem to present higher survival rates than those with lower PNI. PNI is easy to calculate and low-cost but in order to be used in everyday clinical practice, future research should be conducted to establish a standardized PNI threshold for GC patients that could be submitted to surgery.

Research objectives

To identify whether the PNI could be used in predicting survival outcomes in patients with GC that are submitted to surgery.

Research methods

We performed a thorough literature search of PubMed, the Cochrane library, and Reference Citation Analysis for cohort studies that included patients with gastric adenocarcinoma who were submitted to gastrectomy. The keywords that we used for our search were “Prognostic nutritional index”, “survival”, and “gastric cancer” in combinations, which lead to the retrieval of 16 studies that matched our inclusion criteria. We performed risk of bias assessment and quality assessment of each individual study and our study was prospectively registered in PROSPERO.

Research results

Our systematic review showed that the PNI could be an important prognostic marker in patients undergoing surgery for gastric adenocarcinoma. All of the studies that we included demonstrated that preoperative PNI is significantly associated with survival in GC patients except for one study which included stage III only gastric adenocarcinoma patients. However, two of the studies that we included showed that PNI is significantly associated with survival in patients with stage III gastric adenocarcinoma. Further studies should aim to identify a standardized PNI threshold for gastric adenocarcinoma patients. Moreover, future studies should analyze the correlation between PNI and the stage of disease and whether PNI is associated with survival regardless of the disease stage.

Research conclusions

PNI could be an important prognostic marker that could assist in predicting the survival of patients submitted to gastrectomy.

Research perspectives

Future research should aim at identifying a standardized PNI cut-off value. Furthermore, the correlation between PNI and tumor stage, Lauren classification, and patients’ clinicopathological characteristics should be analyzed.