Simple Postoperative Markers for Early Identification of Low-Risk Patients After Gastric or Colorectal Cancer Surgery: A Retrospective Cohort Study

Abstract

Background

Early identification of low-risk patients after curative gastric or colorectal surgery can enable safe, timely discharge. We evaluated two routinely available postoperative markers—serum prealbumin (PA) and lymphocyte percentage (L%)—as a simple rule-out strategy.

Methods

This is a single-center retrospective cohort study with a temporal external validation set. Prealbumin (mg/dL) and L% (%) were measured on postoperative day (POD)1/3/5. We prespecified a high-sensitivity threshold (sensitivity ≥90%) in the training cohort and assessed its performance in the validation cohort. Multivariable logistic models were fitted at POD3 and POD5.

Results

Postoperative day 1 L% showed limited discrimination; analyses focused on POD3/5. At both time points, PA and L% were independent protective predictors, with no effect modification by tumor site. Slim (PA + L%) and Full models had similar discrimination; both were well calibrated, and decision-curve analysis indicated clear net benefit over default strategies. Under the fixed high-sensitivity approach, the POD3 Slim model maintained high sensitivity and NPV with moderate specificity.

Conclusions

By POD3, combining PA and L% provides a low-cost, readily deployable, rule-out-oriented approach that supports safe early discharge within structured postoperative pathways. Multicenter prospective validation and impact studies are warranted.

Keywords

gastric cancer colorectal cancer postoperative complications early discharge

Key Takeaways

By POD3, serum prealbumin and lymphocyte percentage reliably identify low-risk patients with high sensitivity and negative predictive value.

A Slim model using only PA and L% performs comparably to a Full model while remaining simple and deployable.

Temporal external validation supports generalizability of the high-sensitivity, rule-out strategy for safe early discharge; decision-curve analysis shows net clinical benefit across common threshold-probability ranges (0.05-0.40).

Introduction

Gastric and colorectal cancers remain among the most challenging cancers worldwide. In 2020, there were an estimated 1.1 million new cases of gastric cancer and 770 000 related deaths. Although incidence has declined, the overall burden is expected to rise with population aging; by 2040, projections suggest 1.8 million new cases and 1.3 million deaths.¹ According to the World Cancer Research Fund (WCRF), colorectal cancer accounted for approximately 1.9 million new cases and 900 000 deaths in 2022. Surgery remains central to management, and reported postoperative complication rates range from 29.8% to 41.3% for gastric cancer and 20.7% to 35% for colorectal cancer.^2,3 In the context of constrained health care resources, early identification of high-risk patients and timely discharge of low-risk patients are key goals.

Nutritional status supports host defense and tissue repair. Although postoperative hypoalbuminemia correlates with complications, albumin is an imperfect marker with a ∼20-day half-life, whereas prealbumin has a ∼2.5-day half-life and better tracks short-term nutritional change.^4,5 Prior studies indicate that preoperative serum prealbumin not only reflects nutritional status but also predicts postoperative complications.^6-8 For example, a prospective study reported that patients with preoperative prealbumin <17.29 mg/dL were more likely to develop postoperative complications.⁹

Surgical trauma also elicits a systemic inflammatory response. Tissue injury triggers the release of inflammatory mediators, typically causing neutrophilia with a relative or absolute lymphopenia. Inflammatory markers are routinely monitored after surgery, and numerous studies have demonstrated their value for predicting postoperative complications.^10-12

Therefore, our primary aim was to develop and temporally validate a parsimonious, bedside-ready rule-out tool to inform decisions when readiness for discharge is uncertain—that is, for patients in whom continued hospitalization is unlikely to confer substantial additional benefit. The tool seeks to identify a low-risk subgroup and thereby support safe, timely discharge with structured follow-up.

Materials and Methods

Study Design and Setting

Single-center retrospective cohort study of patients undergoing curative gastric or colorectal cancer resection (Feb 2023-Jan 2025) was performed; exclusion criteria were as follows: recurrent/metastatic disease, non-radical surgery, or incomplete records. This study was conducted following the Declaration of Helsinki; ethical approval was waived per local policy for retrospective anonymized data.

Outcome

Primary endpoint: clinically relevant complications within 10 days (Clavien-Dindo ≥II) captured from medical records and discharge summaries, including index stay, readmissions, and outpatient/ED visits ≤10 days. Because most patients are discharged before POD10, we restricted outcome ascertainment to the first 10 postoperative days, emphasizing inpatient and early post-discharge risks instead of the customary 30-day window.

Predictors and Measurements

Candidate predictors were serum prealbumin (PA, mg/dL) and lymphocyte percentage (L%) measured on POD1/3/5; preoperative PA was explored to contextualize postoperative change. Variables were scaled (PA per 1 mg/dL; L% per 5%).

Temporal Validation

Training set: Feb 2023-Jan 2025; temporal external validation: Feb-Jun 2025, with no re-fitting/tuning.

Statistical Analysis

All analyses were performed in R 4.4.3; continuous data was measured as mean ± SD, and categorical data was measured by using χ2 or Fisher’s exact test; two-sided α = 0.05.

Single-Marker Performance

ROC curves in training; primary cutoff chosen as the smallest value achieving sensitivity ≥90% and evaluated in validation; Youden’s J served as a comparator. Discrimination summarized by AUC (DeLong 95% CI); at fixed thresholds, Se/Sp/PPV/NPV with exact 95% CIs were reported in both sets.

Multivariable Models

At POD3 and POD5, logistic models were prespecified: Full (age, sex, tumor site, PA, L%) and Slim (PA, L% only). Adjusted ORs (95% CI, p) were reported; tumor-site-by-marker interactions were tested via likelihood-ratio tests.

Linearity, Calibration, and Utility

Nonlinearity for PA and L%/5 was assessed with restricted cubic splines; where present, spline terms were used in sensitivity analyses and AUC differences compared by DeLong. Validation calibration used intercept (α) and slope (β) plus decile-based plots with LOESS (ideal α ≈ 0, β ≈ 1). Decision-curve analysis evaluated net benefit across threshold probabilities 0.05-0.40 vs “treat all/keep all hospitalized” and “treat none/discharge all.”

Results

During the study period, 227 patients were screened; after excluding 7 for palliative or exploratory surgery and 21 for incomplete laboratory data, 199 were analyzed. Missing POD3/5 values chiefly reflected no same-day blood draw or early discharge for nonclinical reasons; as missingness was plausibly not at random, multiple imputation was deemed inappropriate and complete-case analyses were performed. The temporal external validation cohort (February-June 2025) included 50 patients, of whom 12 (24.0%) developed postoperative complications, a rate similar to that of the training set.

Of these 199 patients, 128 (64.3%) were male and 71 (35.7%) were female, with a mean age of 61.89 ± 11.65 years. Seventy-five (37.7%) had gastric cancer and 124 (62.3%) had colorectal cancer. Postoperative complications occurred in 49 patients (24.6%). The groups with and without complications were comparable in age, BMI, sex, ASA class, tumor site, stage, and most baseline laboratories. Baseline characteristics are summarized in Table 1. Preoperative serum prealbumin levels were significantly lower in patients who developed complications than in those who did not (18.39 ± 5.38 vs 20.43 ± 5.48, P = 0.013) (Table 1).

Table 1.

Baseline Characteristics of Patients With and Without Complications

		Overall	Not complicated	Complicated	P-value
Age		61.82±11.67	61.13±11.31	63.92±12.59	0.147
BMI		24.06±3.64	24.04±3.46	24.12±4.16	0.897
Sex	Male	128	95	33	0.611
Sex	Female	71	55	16	0.611
ASA	1	35	26	9	0.931
	2	94	72	22
	3	70	52	18
Tumor location	Gastric	75	61	14	0.129
Tumor location	Colorectal	124	89	35	0.129
Pathological stage	Ⅰ	51	42	9	0.405
	Ⅱ	62	45	17
	Ⅲ	86	63	23
Preoperative hemoglobin (g/L)		128.02±22.32	129.50±21.02	123.49±25.61	0.102
Preoperative platelets (*10^9/L)		239.06±83.21	242.73±83.17	227.82±83.15	0.277
Fasting blood glucose (mmol/L)		5.65±1.36	5.61±1.28	5.77±1.57	0.461
Pre-PA (mg/dL)		19.93±5.51	20.43±5.48	18.39±5.38	0.013
Pre-L% (%)		29.54±8.52	29.69±8.82	29.08±7.57	0.663

Values are mean ± SD or n (%). Complications are events within 10 postoperative days (Clavien-Dindo ≥ II). Group comparisons used t test or Mann-Whitney U test for continuous variables and χ² or Fisher’s exact test for categorical variables (two-sided).

Abbreviations: Pre-PA, preoperative prealbumin; Pre-L%, preoperative lymphocyte percentage.

Postoperative complications are detailed in Table 2. Grade II (Clavien-Dindo) events were most frequent. Anastomotic leakage or intestinal fistula occurred in 3.52% of patients, all successfully managed with conservative treatment. The most severe events were grade IIIb complications in two patients (1.0%), both due to postoperative intra-abdominal hemorrhage. No postoperative mortality was observed. Detailed data for the analyzed indicators are provided in Table 3. Notably, most indicators are closely associated with postoperative complications(Tables 2 and 3).

Table 2.

Spectrum of Clinically Relevant Postoperative Complications Within 10 days

Clavien-Dindo grade	Type of complication	Number of cases	Percentage (%)
Ⅱ	Respiratory complications (eg, respiratory tract infections and pleural effusion)	14	7.04%
	Incisional SSI	9	4.52%
	Organ/space SSI	8	4.02%
	Anastomotic leakage/enterocutaneous fistula	7	3.52%
	Infections at other sites (eg, urinary tract infection)	4	2.01%
	Postoperative ileus	5	2.51%
Ⅲ	Intra-abdominal hemorrhage	2	1.01%

Complications are listed by Clavien-Dindo grade and type with case counts and percentages for the whole cohort.

Abbreviations: SSI, surgical site infection.

Table 3.

Lymphocyte Percentage and Prealbumin on POD1, POD3, and POD5 by Complication Status

Markers		Overall	Not complicated	Complicated	P-value
PA (mg/dL)	POD1	15.55 ± 4.59	16.44 ± 4.23	12.82 ± 4.61	<0.001
	POD3	13.18 ± 4.70	14.74 ± 4.00	8.43 ± 3.32	<0.001
	POD5	15.37 ± 5.56	17.23 ± 4.11	9.70 ± 5.61	<0.001
L% (%)	POD1	8.58 ± 4.64	8.93 ± 5.06	7.53 ± 2.83	0.067
	POD3	15.85 ± 7.02	17.61 ± 6.77	10.47 ± 4.68	<0.001
	POD5	19.32 ± 7.88	21.94 ± 6.84	11.30 ± 4.86	<0.001

Values are mean ± SD. P-values from t test or Mann-Whitney U test, two-sided.

Abbreviations: PA, prealbumin; L%, lymphocyte percentage; POD, postoperative day.

To evaluate the discriminatory capacity of each indicator, ROC curves were first constructed. We adopted a high-sensitivity strategy as the primary thresholding approach: in the training set, the operating cutoff was prespecified as the smallest value achieving sensitivity ≥90%, and its performance was evaluated in the temporal validation cohort. For comparison, we additionally derived an “optimal” cutoff using Youden’s J index. The resulting thresholds and their performance metrics are summarized in Table 4 and Additional file 1 (Supplemental Figure 1). Because POD1 L% did not reach statistical significance and exhibited only modest discrimination on ROC analysis, subsequent analyses focused on POD3 and POD5 (Table 4).

Table 4.

Single-Marker AUCs and Operating Cutoffs at POD1, POD3, and POD5 (Training Set)

Marker	AUC	Cut_Se90	Cut_Youden
POD1PA	0.727	18.90 mg/dL	13.56 mg/dL
POD1L%	0.565	11.40%	10.60%
POD3PA	0.884	13.63 mg/dL	10.61 mg/dL
POD3L%	0.820	16.15%	12.85%
POD5PA	0.845	19.60 mg/dL	11.36 mg/dL
POD5L%	0.898	18.50%	14.65%

For each marker, AUC is shown with the high-sensitivity cutoff (Cut_Se90, the smallest value achieving sensitivity ≥90%) and the Youden’s J cutoff. Units: PA in mg/dL; L% in percentage points.

Abbreviations: AUC, area under the curve; PA, prealbumin; L%, lymphocyte percentage; POD, postoperative day.

To control potential confounding, multivariable logistic regression models were fitted at POD3 and POD5, adjusting for age, sex, and tumor site. Prealbumin and L% were independent protective predictors of complications (see Table 5). Furthermore, the model revealed that patients with colorectal cancer had a higher risk of postoperative complications on POD3 and POD5 compared with those with gastric cancer, with odds ratios of 0.19 (95% CI, 0.06-0.57) and 0.14 (95% CI, 0.03-0.53), respectively. The wide confidence intervals indicate a certain degree of uncertainty in these estimates, which may be attributable to the limited number of events within subgroups. Interaction analyses detected no significant effect modification by tumor site for either marker (POD3 P = 0.348; POD5 P = 0.791). Although complication rates differed by tumor site, marker effects were consistent across gastric cancer and colorectal cancer, supporting pooled models and shared thresholds (Table 5).

Table 5.

Adjusted Associations of PA and L% With Complications at POD3 and POD5

Markers	aOR	95% CI	P-value
POD3PA	0.613	0.518-0.725	<0.001
POD3L%/5	0.424	0.257-0.698	<0.001
POD5PA	0.697	0.604-0.804	<0.001
POD5L%/5	0.244	0.137-0.435	<0.001

Adjusted odds ratios (aOR) from multivariable logistic models are reported per prescaled unit: PA = per 1 mg/dL increase; L%/5 = per 5 percentage-point increase. Lower aOR (<1) indicates a protective association.

Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; PA, prealbumin; L%, lymphocyte percentage.

At each time point, we specified two logistic models: a Full model (age, sex, tumor site, PA, and L%) and a Slim model (PA and L% only) (Figure 1). In temporal external validation, AUCs were similar and DeLong tests showed no significant differences at either time point (Table 6). Calibration was acceptable at POD3 for both models (α ˜ 0, β ˜ 1), whereas at POD5 the Full model overestimated risk (α < 0 with a 95% CI not crossing 0) and the Slim model remained well calibrated (α ˜ 0, β ˜ 1) (Figure 2). Decision-curve analysis showed consistent net benefit for both models across threshold probabilities of 0.05-0.40, with nearly overlapping curves and superiority to “treat none/discharge all” and “treat all/keep all hospitalized” strategies (Figure 3).

Table 6.

External Discrimination of Full Versus Slim Models in the Validation Cohort

Timepoint	Model	AUC	95% CI	ΔAUC	P-value
POD3	Full	0.943	0.878-1.000	0.002	0.880
POD3	Slim	0.941	0.870-1.000	0.002	0.880
POD5	Full	0.950	0.892-1.000	0.011	0.413
POD5	Slim	0.939	0.869-1.000	0.011	0.413

AUCs with 95% CIs are shown for POD3 and POD5. ΔAUC and P-values are from DeLong tests comparing Full vs Slim at each timepoint.

Abbreviations: AUC, area under the curve; CI, confidence interval; POD, postoperative day.

Restricted cubic spline analyses tested linearity for PA and L%/5. At POD3, neither marker showed nonlinearity; at POD5, PA was nonlinear, so a natural spline with 2 degrees of freedom replaced the linear term in sensitivity analyses. In the training set, spline models yielded slightly higher AUCs without statistical significance (Full: P = 0.108; Slim: P = 0.072), and validation results were likewise nonsignificant (Full: P = 0.525; Slim: P = 0.615) (Table 7). Spline models also exhibited calibration slopes β < 0.8, so the linear specification was retained.

Table 7.

POD5 Linear Versus Spline Specifications: Discrimination and Calibration in Training and Validation

Model		AUC	95% CI	α	β	ΔAUC	P-value
Train	Full-Linear	0.951	0.921-0.981	0.000	1.000	0.012	0.108
	Full-Spline	0.963	0.940-0.986	0.000	1.000	0.012	0.108
	Slim-Linear	0.934	0.896-0.972	0.000	1.000	0.015	0.072
	Slim-Spline	0.949	0.921-0.978	0.000	1.000	0.015	0.072
Valid	Full-Linear	0.950	0.892-1.000	−1.486	0.968	−0.007	0.525
	Full-Spline	0.943	0.883-1.000	−1.325	0.710	−0.007	0.525
	Slim-Linear	0.939	0.869-1.000	−0.979	0.977	0.007	0.615
	Slim-Spline	0.945	0.887-1.000	−0.732	0.757	0.007	0.615

Comparison of linear and restricted cubic spline forms for PA with L% kept linear, for both Full and Slim models. AUCs with 95% CIs are shown alongside calibration intercept (α) and slope (β). ΔAUC and P-values are from DeLong tests (spline minus linear).

Abbreviations: AUC, area under the curve; CI, confidence interval; α, calibration intercept; β, calibration slope.

Accordingly, we designated the POD3 Slim model as the primary model. It enables earlier decision making using low-cost, widely available tests and imposes no additional operational burden. Under the fixed “training-set sensitivity ≥90%” strategy, the validation cohort achieved sensitivity 100% and NPV 100% with specificity 47.3% at POD3, indicating no missed complications while classifying approximately half of complication-free patients as low risk. These findings support safe early discharge and warrant multicenter validation and integration into discharge pathways.

Discussion

Both albumin and prealbumin were associated with postoperative complications, consistent with their roles as negative acute-phase proteins. Under acute stress, hepatic protein synthesis is reprioritized away from negative acute-phase proteins (eg, prealbumin) toward positive acute-phase proteins,¹³ contributing to early declines in prealbumin. Postoperative lymphopenia is well recognized and reflects neuroendocrine-inflammatory stress, attributed to endogenous glucocorticoids, sympathoadrenal activation (catecholamine-induced apoptosis), pro-inflammatory cytokines, lymphocyte redistribution, and possible opioid effects. It is associated with adverse outcomes and serves as a pragmatic marker of postoperative stress; in our cohort, lymphocyte percentage complemented prealbumin, and together they improved prediction for early discharge.^14-20

In our cohort, patients undergoing gastrectomy exhibited a lower risk of postoperative complications than those undergoing colorectal surgery. Although this may appear counterintuitive, the pattern is explicable. On the one hand, unlike studies that report 30-day outcomes, we restricted outcome ascertainment to the first 10 postoperative days—the typical length of stay in our center and the interval most relevant to discharge planning; consequently, several gastrectomy-related complications that tend to manifest later (eg, nutritional deficiencies, delayed gastric emptying, and anastomotic stricture) fell outside the endpoint. On the other hand, colorectal procedures entail higher risks of surgical-field contamination and a higher propensity to develop functional ileus postoperatively, which plausibly translate into higher early complication rates within this 10-day window.

Although POD5 prealbumin and lymphocyte percentage showed excellent discrimination and high NPV, relying on POD5 may be overly conservative under contemporary ERAS pathways and minimally invasive surgery. In our setting, many patients—especially after colon resection—ambulate and tolerate semiliquid diets by POD2-3; accordingly, we prioritized POD3 as an earlier decision point.

From a clinical standpoint, we position the model as a decision-support, rule-out tool to help identify low-risk patients in the early postoperative period; it is not intended to function as a rigid discharge rule. For patients who have met conventional discharge milestones—independent ambulation, return of bowel function, and adequate pain control—we do not recommend prolonging hospitalization solely because a laboratory value (eg, PA or L%) falls marginally below the suggested threshold. With a structured follow-up plan in place (eg, laboratory reassessment and symptom review within 1-3 days at our clinic or a local provider), such patients can be discharged safely. By contrast, the model is most informative when readiness for discharge is uncertain—for example, in patients who restrict oral intake or mobilization out of caution, or who have low pain tolerance with higher analgesic requirements—for whom continued hospitalization is unlikely to confer substantial additional benefit. In these scenarios, the model provides a quantitative, auditable risk estimate to be integrated with clinical judgment (vital signs, physical findings, drain output, imaging) in a multidisciplinary decision. Ultimately, discharge decisions are multifactorial and should align with ERAS pathways, clinician and nursing assessments, patient preferences, and available home support.

To facilitate bedside adoption, we provide a two-dimensional rule-out map (Figure 4) that depicts the decision boundary at a prespecified high-sensitivity operating point (sensitivity ≥90%). Given the model’s strong performance at this target, we also evaluated a less stringent operating point (sensitivity ≥80%) in the temporal validation cohort; the negative predictive value remained 1.00 and specificity increased to 0.842. Accordingly, we supply two companion maps (Se ≥ 90% and Se ≥ 80%) so clinicians can select the threshold that best balances an acceptable miss rate against the aim of earlier discharge. Use is straightforward: plot the patient’s POD3 values on the axes and read the color-coded zone—green denotes the low-risk region defined by the rule-out threshold. Unlike traditional nomograms, this map requires neither point summation nor a calculator or probability readout; classification is available at a glance. This simplicity is by design and reflects the rationale for our parsimonious modeling approach.

Limitations include the single-center retrospective design, modest sample size (with 12 events in the temporal external validation set), and complete-case analyses driven by nonclinical missingness (eg, lack of same-day blood draws or early self-discharge), which may introduce selection bias. Although the diagnosis term suggested a lower 10-day complication risk after gastrectomy than after colorectal surgery, the current sample size did not support site-stratified model development or thresholding. Heterogeneity was assessed by interaction testing only, with no effect modification detected. We plan to address this in subsequent work by conducting site-stratified model development, calibration, and threshold determination with expanded samples. These findings warrant multicenter, prospective external validation and impact studies to evaluate effects on discharge timing, readmissions, and patient-centered outcomes.

Supplemental Material

Supplemental Material - Simple Postoperative Markers for Early Identification of Low-Risk Patients After Gastric or Colorectal Cancer Surgery: A Retrospective Cohort Study

Supplemental Material for Simple Postoperative Markers for Early Identification of Low-Risk Patients After Gastric or Colorectal Cancer Surgery: A Retrospective Cohort Study by Duchen Li, Xuezheng Jiang, Xianpu Zhu, Hongyuan Chen, and Xiaoqiao Zhang in The American Surgeon™

Footnotes

ORCID iDs

Duchen Li

Hongyuan Chen

Xiaoqiao Zhang

Ethical Considerations

This study is retrospective in nature and utilizes anonymized data. It complies with the relevant regulations issued by the National Health Commission of the People’s Republic of China and is therefore exempt from ethical review.

Consent to Participate

No individual informed consent was required.

Authors’ contributions

All authors made substantial contributions to this study: participated in data collection, formal analysis, and interpretation; drafted and critically revised the manuscript; had full access to the data; approved the final version; and agree to be accountable for all aspects of the work. In addition, XZJ, XPZ, and DCL collected and analyzed the data; DCL, HYC, and XQZ drafted the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

De-identified data may be shared with qualified researchers upon reasonable request to the corresponding author, subject to Institutional Review Board approval and a data use agreement. Data will be available for 36 months after publication.*

Supplemental Material

Supplemental material for this article is available online.

Appendix

References

Morgan

Arnold

Camargo

, et al. The current and future incidence and mortality of gastric cancer in 185 countries, 2020–40: a population-based modelling study. eClinicalMedicine. 2022;47:101404. doi:10.1016/j.eclinm.2022.101404.

Baiocchi

Giacopuzzi

Reim

, et al. Incidence and grading of complications after gastrectomy for cancer using the GASTRODATA registry: a European retrospective observational study. Ann Surg. 2020;272(5):807-813. doi:10.1097/SLA.0000000000004341.

Sun

Zhang

, et al. Risk factors for postoperative complications of laparoscopic right colectomy: a post hoc analysis of the RELARC trial. Dis Colon Rectum. 2024;67(9):1194-1200. doi:10.1097/DCR.0000000000003331.

Ingenbleek

Young

. Significance of transthyretin in protein metabolism. Clin Chem Lab Med. 2002;40(12):1281-1291. doi:10.1515/CCLM.2002.222.

Ranasinghe

Biswas

Vincent

. Prealbumin: the clinical utility and analytical methodologies. Ann Clin Biochem. 2022;59(1):7-14. doi:10.1177/0004563220931885.

Zhang

Tang

Zhang

Jiang

Gong

Fan

. Serum prealbumin level as a biomarker of survival outcomes in patients with gastric cancer: a meta-analysis. Biomarkers. 2024;29(6):410-417. doi:10.1080/1354750X.2024.2402419.

Bae

Lee

Han

, et al. Prealbumin levels as a useful marker for predicting infectious complications after gastric surgery. J Gastrointest Surg. 2011;15(12):2136-2144. doi:10.1007/s11605-011-1719-z.

Zhou

Hiki

Mine

, et al. Role of prealbumin as a powerful and simple index for predicting postoperative complications after gastric cancer surgery. Ann Surg Oncol. 2017;24(2):510-517. doi:10.1245/s10434-016-5548-x.

Raghuraman

Gurushankari

Laya

, et al. Role of specific nutritional biomarkers in predicting post-operative complications among patients undergoing elective abdominal surgery. Langenbecks Arch Surg. 2023;408(1):453. doi:10.1007/s00423-023-03186-8.

10.

Fuss

Voloboyeva

Polovyj

Yaremkevych

. Neutrophil to lymphocyte ratio in predicting postoperative complications and prognosis in patients with colorectal cancer. Pol Przegl Chir. 2022;94(6):33-37. doi:10.5604/01.3001.0015.8385.

11.

Dos Santos

BDN

Beruti

Azevedo

, et al. Using inflammatory parameters for safe and early discharge after minimally invasive colorectal surgery for colorectal cancer. Tech Coloproctol. 2025;29(1):97. doi:10.1007/s10151-025-03134-2.

12.

Cheng

Zhang

Zhuang

Sun

. Prognostic value of lymphocyte-to-C-reactive protein ratio in patients with gastric cancer after surgery: a multicentre study. Jpn J Clin Oncol. 2020;50(10):1141-1149. doi:10.1093/jjco/hyaa099.

13.

Evans

Corkins

Malone

, et al. The use of visceral proteins as nutrition markers: an ASPEN position paper. Nutr Clin Pract. 2021;36(1):22-28. doi:10.1002/ncp.10588.

14.

Hamid

Bancewicz

Brown

Ward

Irving

Ford

. The significance of changes in blood lymphocyte populations following surgical operations. Clin Exp Immunol. 1984;56(1):49-57.

15.

Cioca

Watanabe

Isobe

. Apoptosis of peripheral blood lymphocytes is induced by catecholamines. Jpn Heart J. 2000;41(3):385-398. doi:10.1536/jhj.41.385.

16.

Cuenca

Delano

Kelly-Scumpia

, et al. A paradoxical role for myeloid-derived suppressor cells in sepsis and trauma. Mol Med. 2011;17(3-4):281-292. doi:10.2119/molmed.2010.00178.

17.

Dąbrowska

Słotwiński

. The immune response to surgery and infection. Cent Eur J Immunol. 2014;39(4):532-537. doi:10.5114/ceji.2014.47741.

18.

Balakin

Yurku

Ivanov

Izotov

Nakhod

Pustovoyt

. Regulation of stress-induced immunosuppression in the context of neuroendocrine, cytokine, and cellular processes. Biology. 2025;14(1):76. doi:10.3390/biology14010076.

19.

Guo

Yang

Zhong

, et al. The relationship between anesthesia, surgery and postoperative immune function in cancer patients: a review. Front Immunol. 2024;15:1441020. doi:10.3389/fimmu.2024.1441020.

20.

Wang

Zhang

Chen

. Lymphopenia in sepsis: a narrative review. Crit Care. 2024;28(1):315. doi:10.1186/s13054-024-05099-4.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.36 MB

0.00 MB