The value of lactate dehydrogenase to albumin ratio and immune inflammation biomarkers in colorectal cancer

Background Colorectal cancer (CRC) is one of the most prevalent gastrointestinal cancers. Evidence for the importance of inflammation and immunology in the development and progression of CRC is growing steadily. The purpose of this study was to determine the clinical importance of Lactic Dehydrogenase (LDH) to Albumin (ALB) Ratio (LAR) and immune-inflammation biomarkers (IIBs) in patients with CRC. Methods This study enrolled 382 CRC patients. The LAR was determined as the serum LDH(U/l) to ALB(g/l) ratio. We compared the levels of LAR and IIBs in different TNM stages and tumor differentiation. The relationship between LAR and IIBs and overall survival (OS) of CRC was determined by Cox regression models. A prognostic nomogram was created using the results of the multivariate analysis and the effectiveness of the nomogram was assessed using the ROC, calibration, and decision curves. We evaluated the relationship between LAR and IIBs and clinical features of CRC. Results The levels of LAR, SII, NLR and PLR in TNM IV stage group (LAR:5.92 (5.23–8.24); SII: 1040.02 (499.51–1683.54); NLR: 2.87 (2.07–5.3); PLR:187.08 (125.31–276.63)) were significantly higher than those in other groups. LAR and NLR showed no significant difference in different tumor differentiation groups, while SII and PLR in undifferentiated groups (SII:543.72 (372.63–1110.20); PLR: 147.06 (106.04–203.92)) were significantly higher than those in well and moderate groups (SII: 474.29 (323.75–716.01); PLR: 126.28 (104.31–167.88)). LAR (HR = 1.317, 95% CI = 1.019–1.454), TNM stage (HR = 2.895, 95% CI = 1.838–4.559), age (HR = 1.766, 95% CI = 1.069–2.922) and lymphocytes (HR = 0.663, 95% CI = 0.456–0.963) were predictors of OS. IIBs, including SII, NLR, and PLR are independent of OS. The LAR-based nomogram AUCs of 1-year, 3-year and 5-year survival probabilities in the training cohort were 0.86, 0.72, and 0.71, respectively, and the AUCs of the validation cohort were 0.85, 0.71, and 0.69 respectively. The LAR-based nomogram's ROC curves and calibration curves demonstrated higher OS discriminative performance. The decision curves demonstrated greater net benefit in the survival prediction. Conclusion Preoperative LAR is a potential prognostic marker in CRC patients, while SII, NLR, and PLR are independent of OS. LAR was associated with tumor stage in CRC patients, but not with tumor differentiation.


Introduction
Colorectal cancer (CRC) is one of the most frequent gastrointestinal cancers globally. The morbidity and mortality rates of CRC are currently ranked third and second globally, respectively (1). CRC is the fifth most common reason for cancer-related death in China (2). Despite advancements in diagnostic methods, the majority of CRC patients still have intermediate and late-stage diagnosis. The 5-year survival for CRC has increased from 50% to 64% in the last 40 years due to early screening, advances in imaging and treatment (3). Studies have found that about one-third of patients who have received therapeutic surgery for CRC experience postoperative recurrence (4). More than half of the cases can reduce the incidence by adjusting risk factors, and the mortality can be reduced through correct screening and follow-up. Early diagnosis may lead to a higher cure rate, and effective monitoring after surgery can help identify the progression of the tumor and prompt treatment. Therefore, it is critical to find more economical and effective prognostic indicators to decide appropriate treatment regimens.
Systemic inflammation is considered as a marker of cancer and is associated with tumor development, metastasis and prognosis (5). Recent studies suggested that systemic inflammatory markers and biomarker combinations could be used as prognostic biomarkers for cancers. Preoperative examination results can help predict complications, and some preventive measures may be taken based on the examination results. IIBs are a group of markers that represent the host's inflammatory and immunological condition. IIBs, such as neutrophil/lymphocyte ratio (NLR), pan-immune-inflammation value (PIV), platelet/lymphocyte ratio (PLR), and systemic immune inflammation index (SII), are closely related to CRC (6-9). LDH to ALB ratio (LAR), a novel biomarker, has been demonstrated to be related to a bad prognosis in pancreatic and esophageal malignancies (10,11). However, the significance of LAR in CRC patients in the Chinese population has been less reported. The latest results from Hu et al. (12) suggest that LAR could be a prognostic factor for OS and DFS. However, the relationship between LAR and clinical features was not explored. In this study, we investigated the correlation between IIBs and TNM stage and tumor differentiation in CRC patients, further verifying the role of LAR in CRC.

Characteristics of patients
382 patients diagnosed as CRC were recruited from the First Affiliated Hospital of Sun Yat-sen University between October 2015 and November 2019. 382 CRC patients were randomly specified to the training cohort (n = 249) and the validation group (n = 133) ( Figure 1). From the date of operation until the date of death or the final follow-up, overall survival (OS) was calculated. We followed up the CRC patients and the endpoints of this study were OS. According to the NCCN guideline (13) for CRC, all CRC patients were diagnosed as CRC by postoperative pathology or tissue biopsy. Meanwhile, the patients did not have other tumor diseases at the same time. The exclusion criteria are as follows: (1) pathology did not support the diagnosis of CRC; (2) recurrence of CRC; (3) received an anti-tumor therapy before resection; (4)suffering from infectious diseases and autoimmune diseases before resection; (5)incomplete clinical data.
This retrospective study was approved by the Human Research Ethics Committees of the First Affiliated Hospital of Sun Yat-sen University (approval number: [2021] No.299). Informed consent was obtained from all participants.

Statistical analysis
Categorical data are shown as percentages, whereas continuous data are shown as median (range). To examine the independent prognostic risk variables for CRC, a Cox regression model was utilized. The multivariate cox regression analysis included significant components found in the univariate study. These analyses were performed with SPSS 23.0 (SPSS, United States). The nomogram for prognostic factors associated with OS was established with the rms package in R version 4.2.1 (http://www.r-project.org/). The performance of nomogram was evaluated by ROC curves and calibration curves. The improvement in survival prediction accuracy is measured by time-dependent receiver operating characteristic curves with areas under the curves (AUCs). The prognostic prognosis was more accurate the higher the AUC. The net benefit of the nomogram was assessed with the aid of decision curve analysis (DCA). Statistical significance was defined as p < 0.05.

Characteristics of CRC patients
The clinicopathological characteristics of the training and validation cohorts were showed in Table 1. Of the 382 patients including 153 males and 96 females in the training cohort, 154 patients were diagnosed with TNM stage I + II, while 95 patients with TNM stage III + IV.
Among the 133 patients including 78 males and 55 females in the validation cohort, 67 patients were diagnosed with TNM stage I + II, while 66 patients with TNM stage III + IV. Overall, 86.3%, and 13.7% of the patients in the training cohort had well and moderate differentiation and undifferentiated differentiation, respectively, and 82.7% and 17.3% had well and moderate differentiation and undifferentiated differentiation, respectively in the validation cohort.

Relationship between IIBs and clinicopathological features in all CRC patients
LAR and IIBs levels at different TNM stage and tumor differentiation degree were observed in all CRC patients. The levels of LAR, SII, NLR and PLR in patients with stage IV were higher than those in patients with stage I, stage II and stage III (p < 0.05). The undifferentiated group had significantly higher SII and PLR levels (p < 0.05). There was no significant difference in LAR and NLR levels among all CRC patients in different tumor differentiation degree groups (p > 0.05) (Figures 2 and 3, Tables 2 and 3).

Cox regression analyses of prognostic factors for OS
Univariate analysis identified that TNM III + IV stage, age, lymphocytes, RBC,TP, HGB, CREA, ALB, NLR and LAR were

Nomogram for OS establishment and validation
A comprehensive nomogram was developed to predict the survival rate of CRC patients, integrating the age, TNM stage, lymphocytes, and LAR ( Figure 4). The LAR-nomogram's C-index was 0.69 (95% CI: 0.63-0.74, p < 0.001). Using timedependent ROC curves, the nomogram's prediction accuracy was assessed ( Figure 5)

Relationship between LAR and clinicopathological features in all CRC patients
Patients were divided into the high LAR (> median LAR) group and the low LAR (< median LAR) group (Table 5). There were statistically significant differences in SII, NLR, PLR, T stage, M stage and TNM stage between patients with high LAR group and Frontiers in Surgery low LAR group (p < 0.05, Table 3). However, there were no differences in tumor differentiation and N-stage. CRC patients with high LAR were more likely to have high SII, NLR and PLR levels compared with those in the low LAR group. Spearman's rank correlation coefficient showed that LAR was positively correlated with SII (r = 0.25, p < 0.001), NLR (r = 0.25, p < 0.001) and PLR (r = 0.24, p < 0.001). In addition, the scatter diagram and thermogram show the correlation between LAR and the variables (Figure 8).

Discussion
The recurrence and prognosis of CRC are closely related to the TNM stage. However, the prognosis of patients with the same stage and similar treatment regimens is not completely consistent (14). Numerous studies have demonstrated inflammatory cytokines can be used as prognostic markers in human malignancies (15,16) and have shown that LAR and IIBs are associated with prognosis of tumors. Based on these considerations, this study was designed to explore the relationship between LAR and IIBs and CRC patients in a Chinese population. Our results indicated that LAR was a significant independent risk factor for OS. IIBs, including SII, NLR, and PLR are independent of OS. The TNM IV stage group had significantly higher LAR, SII, NLR and PLR levels. LAR and NLR showed no significant difference in different tumor differentiation groups, while SII and PLR in undifferentiated groups were significantly higher than those in well and moderate groups. T4 stage, M0 stage and TNM IV stage groups had higher LAR levels, suggesting that high levels of LAR are associated with poor outcomes in CRC patients.
Metabolic changes are one of the important features of tumors (17). The glucose metabolic reprogramming plays an important role in tumor genesis and development. In the tumor microenvironment, LDH is one of the key enzymes in the reprogramming of glucose metabolism and can catalyze the conversion of pyruvate to lactic acid. Studies have shown that even when oxygen levels are normal, tumor cells still metabolize glucose mainly through the glycolysis pathway, producing large amounts of lactic acid. The lactic acid effusion into the stroma of tumor cells makes the tumor microenvironment acidic (18), promotes tumor invasion, migration (19, 20) and immune escape (21,22). Meanwhile, LDH can regulate tumor angiogenesis (23). LDH is expressed in many tissues. When cells are damaged, LDH is released into the bloodstream. Increased serum LDH levels due to tissue destruction by tumor growth and metastasis suggest that LDH may be a potential diagnostic marker for cancer. Studies have revealed that LDH and the prognosis of

Tumor differentiation
Well and moderate Reference      (11). Meanwhile, Ulas et al. reported that the high LAR is an independent prognostic factor for patients with OS in CRC, but did not provide a prognostic model for CRC (29). Therefore, this study developed a CRC prediction model and examined the connection between LAR and prognosis in CRC patients. The findings demonstrated that LAR is a significant    (31). The inconsistencies in the findings may be related to population characteristics. The connection between IIBs and CRC has to be confirmed by larger investigations. Some restrictions applied to our investigation. First, the results need to be further validated in multi-center studies with more patients because this was a retrospective, single-center investigation. Second, our study did not compare preoperative and postoperative changes in each index. Third, due to the limitations of the study, there may be some inevitable bias in our study, which may affect the results. Meanwhile, our results did not provide a cut-off threshold for clinical evaluation of patient outcomes. To confirm the predictive usefulness of LAR in CRC patients, further prospective studies are required.
In conclusion, our study identified LAR as a potential combination of biomarkers for the prognosis of patients with CRC. IIBs (SII, NLR and PLR) are not associated with the prognosis of CRC tumors, but with clinical features. High levels of LAR are associated with a poor prognosis for CRC. Preoperatively LAR can help physicians make more effective perioperative management and adjuvant treatment decisions after surgery. Correlation between LAR and SII, NLR and PLR (A). (B) Correlation between LAR and SII; (C) Correlation between LAR and NLR; (D) Correlation between LAR and PLR.