1. IntroductionColorectal cancer (CRC) is considered as one of the most common types of malignant cancer worldwide. According to GLOBOCAN 2020, it is the third most common tumor and the second most common in terms of cancer-related deaths [1]. In particular, surgical resection remains the most effective treatment for rectal cancer. Notably, anastomotic leakage (AL) is one of the most serious complications after surgery, with an incidence rate of 0–20% [2]. In addition, it has been reported that AL is associated with higher in-hospital mortality and healthcare costs [3] as well as poor oncological prognosis in long-term follow-up [4]. In particular, with the innovation of surgical instruments and the development of minimally invasive technology, the preoperative neoadjuvant therapy paves the way for low or even ultra-low anus-preservation surgery. Meanwhile, the incidence of AL has increased correspondingly. In the era of enhanced recovery after surgery or Fast Track Surgery, early rehabilitation with reduced hospital stay has been gradually accepted by surgeons [5]. However, approximately 20% of AL cases are usually diagnosed after a mean of 6–15 days of discharge [6,7]. In addition, a previous study reported that a 2.5-day delay in insufficient AL-specific treatment could increase mortality from 24% to 39% [8]. Therefore, early AL diagnosis is of great significance in clinical practice.To date, several biomarkers of the inflammatory response, including platelet-to-lymphocyte (PLR) and neutrophil-to-lymphocyte (NLR) ratios, have been found to have the highest accuracy in predicting advanced disease among all gastrointestinal malignancies [9]. A previous high-quality meta-analysis reported that increased NLR was associated with poor overall and recurrence-free survival in patients with CRC [10]. Liu et al. reported that patients with CRC who presented with changes from a low pre-treatment NLR level to a high post-treatment NLR levels had worse overall and progression-free survival than those who presented with NLR level changes from high to low [11]. Furthermore, a previous study reported that increased preoperative NLR was associated with greater perioperative complications after colorectal surgery, with a trend toward the occurrence of AL [12]. Therefore, the predictive value of NLR in postoperative AL should be further investigated.

The present study aimed to investigate the predictive value of postoperative NLR along with PLR and lymphocyte-to-monocyte ratio (LMR) as early biomarkers of symptomatic AL in patients undergoing laparoscopic low anterior resection (LAR) for rectal cancer.

4. Discussion

In the present study, the incidence of symptomatic AL was 5.56%, and no in-hospital deaths caused by symptomatic AL were reported in our unit. To the best of our knowledge, this is the first study to use the PSM analysis to identify the predictive value of NLR for early prediction of symptomatic AL after laparoscopic LAR in rectal cancer. Notably, patients who developed symptomatic AL were mostly males and presented with a closer distance to the anal margin than those without AL. In addition, NLR was found to have a more accurate predictive value for symptomatic AL on postoperative day 5 with and without PSM analysis. Furthermore, NLR cut-off scores before and after PSM analysis were 6.97 and 6.54, respectively.

Previous studies have shown that a high preoperative NLR often indicates a poor prognosis [10,11], and only a few studies have reported on complications after rectal cancer surgery. First, Josse et al. reported that a preoperative NLR of ≥2.3 was associated with postoperative surgical complications and found a trend toward AL incidence without statistical significance [12]. Subsequently, Caputo et al. found that patients with a preoperative NLR above the cut-off score had significantly higher rates of postoperative complications after rectal surgery [18]. They also suggested that the cut-off values of NLR before and after neoadjuvant therapy were 2.8 and 3.8, with a reported AUC value of 0.476 and 0.692, respectively [18]. However, the present study found no statistical significance in the preoperative NLR scores between the two groups. In addition, the abovementioned two cut-off scores were relatively low, which seems unconvincing for clinical diagnosis and treatment. In addition, lower AUC values (0.476 and 0.692) reported in the two previous studies were based on lower-level statistical evidence [12,18]. Therefore, the validity of the preoperative NLR remains controversial.There are only a few studies on the postoperative NLR level, which have not come to a consistent conclusion. In the early stages of inflammation, rapidly activated neutrophils release chemokines and cytokines while migrating to the site of inflammation; moreover, the neutrophils increase as inflammation progresses. Lymphocytes reflect the immune status of the system, and their immune activity is inhibited by bacteria and other various factors, which decrease as the inflammation progresses. However, the abovementioned phenomenon existed in a relatively late period and could not reflect the progress of the inflammation with time. Therefore, the elevated NLR levels can be used as an important marker of inflammation and are more reliable than calculating neutrophils or lymphocytes alone [19]. A retrospective study by Diana et al. with a small sample size (116 patients) reported that no statistical difference was found in the NLR levels between patients with and without AL in the first 5 postoperative days [20]. However, the largest multicenter study by Paliogiannis et al. included 1,432 patients and demonstrated that the postoperative NLR cut-off score of 7.1 showed the best efficacy in AL prediction (AUC: 0.744, 95% CI: 0.719–0.768), with the sensitivity and specificity of 72.73% and 73.44%, respectively [21]. Another study by Milk et al. reported that an NLR cut-off score of 6.5 on postoperative day 4 had a sensitivity, specificity, positive predictive value, and negative predictive value of 69%, 78%, 49%, and 88% for AL diagnosis, respectively [22]. Our study revealed that the AUC of 0.802 at postoperative day 5 below an NLR cut-off score of 6.97 was assessed as good, with a sensitivity and specificity of 76.5% and 80.5%, respectively. This value is close to the NLR cut-off score of the two previous studies by Paliogiannis et al. and Mik et al. who reported an NLR cut-off of 7.1 and 6.5 at postoperative day 4, respectively. Notably, based on the difference in test time, this closer value has a certain reference significance [21,22]. It is believed that the greater substantial growth of the postoperative NLR compared with the preoperative NLR usually indicates excessive inflammation. In particular, these inflammatory reactions can be caused by the body’s self-repair process or post-surgical complications. Therefore, postoperative NLR is believed to have some importance in the early prediction of symptomatic AL in rectal cancer.Several other indices of inflammation, including PLR, C-reactive protein (CRP), and procalcitonin (PCT), have been reported in previous studies. Jones et al. reported that both postoperative NLR and PLR levels were associated with serious postoperative complications [23]. However, the specific cut-off scores by the ROC analysis were not reported in detail. Subsequently, Diana et al. further reported that CRP was significantly better than NLR in AL diagnosis and the best AUC was calculated using ROC curves on postoperative day 5, with a CRP value of >54 mg/dL (AUC: 0.81, sensitivity: 89%, specificity: 61%) (20). Similarly, in the study by Walker et al. involving 136 patients, they reported that CRP and NLR instead of PCT were effective predictors of anastomotic dehiscence [24]. They also found that a cut-off of CRP (105 mg/L) on postoperative day 5 demonstrated the best AUC of 0.81, with the highest sensitivity (100%) and specificity (56.5%) [24]. Notably, platelets play an important role in the body’s blood clotting function. Moreover, the tumor itself leads to an increase in the platelet count and causes the blood to become hypercoagulable, thereby leading to various postoperative complications, impairing the blood supply to tissues, and interfering with tissue healing. In the present study, the postoperative PLR value was found to have no significance in the early prediction of symptomatic AL; however, this warrants further exploration. In addition, Jabłońska B. et al. reported that lower total lymphocyte counts and serum albumin levels were found in patients who developed complications after distal pancreatectomy [25]. Moreover, this study found that patients who developed symptomatic AL had lower total lymphocyte counts and LMR levels compared to patients without AL. However, the AUC, sensitivity, and specificity values that were analyzed using the ROC curve did not yield meaningful results.

To the best of our knowledge, this is the first study to use the PSM analysis to further demonstrate the role of the NLR in early prediction of postoperative symptomatic AL after rectal surgery. However, three main limitations of the study should also be considered. The first limitation is the single-center retrospective design and the heterogeneity of the patients in this study. The second limitation is that several drugs, including nonsteroidal anti-inflammatory drugs and antibiotics, may affect blood counts and should therefore be controlled for in future studies. The third limitation is that although we used the PSM analysis, the total number of matched samples (61 patients) does not meet the ideal standard (1:3) owing to the small sample size of this study. Therefore, further prospective multicenter studies of higher quality are needed to verify the study results.