The Value of Preoperative Blood Neutrophil-Lymphocyte, White Blood Cell-Lymphocyte, Monocyte-Lymphocyte, and Platelet-Lymphocyte Ratios in Predicting the Malignant Potential of Gastrointestinal Stromal Tumor: A Retrospective Study

Background: To investigate the predictive value of the neutrophil-lymphocyte ratio (NLR), white blood cell-lymphocyte ratio (WLR), monocyte-lymphocyte ratio (MLR), and platelet-lymphocyte ratio (PLR) in grading the malignant potential of the gastrointestinal stromal tumour (GIST). Methods: We retrospectively reviewed 843 GIST patients who underwent surgery during hospitalisation from June 2016 to July 2020 and 374 unrelated healthy controls. Demographic data were collected for each participant, including age, gender, body mass index, laboratory test results, and pathological results. We compared the NLR, WLR, MLR, and PLR between GIST and healthy individuals, as well as among patients with different risk grades of GIST. Results: NLR, WLR, MLR, and PLR values of patients with GIST were signicantly increased compared with those of the control group (P<0.05). These values also showed statistically signicant differences among the four risk degree groups. The NLR, WLR, MLR, and PLR values of the high-risk group were signicantly higher than those of the very low-risk group, low-risk group, and intermediate-risk group (p<0.001). The NLR, WLR, MLR, and PLR showed a high predictive value in the high-risk group, with an area under the receiver operating characteristic curve of 0.802, 0.824, 0.866, and 0.814, and a cut-off value of 2.059, 3.347, 0.221, and 141.4, respectively. Conclusion: Preoperative NLR, WLR, MLR, and PLR have a high predictive value for grading the malignant potential of GIST, especially in the high-risk group. These values provide a convenient and inexpensive way to determine preoperative chemotherapy and surgical strategies.


Introduction
Gastrointestinal stromal tumour (GIST) is the most common mesenchymal tumour of the gastrointestinal tract, with an unadjusted annual incidence of approximately 1.5 per 100,000 [1]. The most common primary tumour site is the stomach (40-70%), followed by the small intestine (20-40%) and the colorectum (5-15%), and it can also occasionally occur in the oesophagus, mesentery, omentum, and retroperitoneum [2]. GISTs originate from Cajal cells in the gastrointestinal tract wall, which are the precursors of gastrointestinal stromal cells and are located in the myenteric plexus of the lamina propria [3].
Complete surgical resection remains a standard treatment for primary resectable GIST [4]. The latest American National Comprehensive Cancer Network guidelines indicate that surgery aims to completely excise the GIST without rupturing the tumour capsule and damaging the nearby organs; otherwise, preoperative imatinib mesylate treatment should be considered after the tumour grade and histological subtype have been established with a biopsy [5]. Imatinib, a small molecular inhibitor of receptor tyrosine kinases, is one of the few tumour-targeting drugs successfully applied in clinical practice, and has dramatically changed the paradigm of GIST treatment [6]. Preoperative imatinib can shrink the tumour without targeting the surrounding organ, increasing the chance of complete tumour resection without extensive organ resection [7][8][9].
The National Institutes of Health (NIH) criteria have been widely used by clinicians, which stratify GISTs into four risk degrees classi cation: "very low-risk," low-risk," "intermediate-risk," and "high-risk" [8]. The NIH criteria not only provide useful information to predict patient prognosis, but can also determine GIST patients who may bene t from adjuvant therapy [10]. Its predictive value has been evaluated in several large retrospective cohort studies [11]. These criteria depend on pathological ndings; however, due to the soft texture of the GIST tumour, an improper preoperative biopsy can lead to implantation metastasis and haemorrhage.
Above all, preoperative grading is of great signi cance for predicting the prognosis and determining the preoperative chemotherapy strategy. However, due to the risk and cumbersome task of performing a biopsy, it is necessary to nd an alternative accurate and convenient method of preoperative grading.
In recent years, the value of cancer-related in ammatory factors in determining patient survival and disease progression has emerged for various types of cancer [12]. Systemic in ammation can promote tumour progression and metastasis by inhibiting apoptosis and promoting angiogenesis. The neutrophillymphocyte ratio (NLR) is an economical, convenient, and repeatable measure that can be used to detect the systemic in ammatory response, and a high NLR is closely related to the prognosis of a variety of tumours, including GIST [13][14][15][16]. The peripheral blood cell-lymphocyte ratio (WLR), monocyte-lymphocyte ratio (MLR), and platelet-lymphocyte ratio (PLR) have also been shown to be prognostic factors for various malignant solid tumours [17,18].
This study aimed to assess the predictive value of preoperative peripheral blood NLR, WLR, MLR, and PLR for classifying the risk of GIST, and compare differences in their predictive value between different risk grades of GIST.

Study design and patient selected
Patients who underwent resection of a primary GIST without preoperative chemotherapy between 2016 and 2020 were identi ed from Wuhan Union hospitals. The clinical, surgical, and pathological data of the patients were retrospectively collected from clinical les. The inclusion criteria were as follows: (1) the patient was initially diagnosed as having a primary completely resectable GIST; (2) the patient had complete peripheral blood test results within 7 days before surgery; (3) the patient had complete followup data; and (4) the patient provided informed consent. Patients with the following conditions were excluded from the study: (1) patients who underwent preoperative chemoradiotherapy and targeted therapy with imatinib or its analogues; (2) patients with other cancers; (3) patients with various types of in ammation within 7 days before surgery; and (4) patients with severe cardiopulmonary diseases. The study was approved by the ethics committee of Wuhan Union Medical College Hospital (No. 2018-S377), and all patients provided written informed consent to participate.
De nition and classi cation criteria NLR was de ned as the ratio of neutrophils to lymphocytes, WLR as the ratio of white blood cells to lymphocytes, MLR as the ratio of monocytes to lymphocytes, and PLR as the ratio of platelets to lymphocytes. According to the NIH criteria for GIST, all patients were classi ed into very low-risk, low-risk, intermediate-risk, and high-risk based on tumour size, mitotic count, tumour sites, and rupture status.

Statistical analysis
All statistical analyses were performed using the R language environment version 3.5.2. Continuous variables were presented as mean ± standard deviation or median, and categorical variables were expressed as a percentage. Students t-test was performed to compare the difference in continuous variables. After the normal test for each group of data, the data did not follow the normal distribution. Kruskal-Wallis H test was used for the comparison between multiple groups of data, and Bonfferoni corrected A value that was used for the comparison between the two groups. The ROC curve was drawn with speci city as the abscisic coordinate and sensitivity as the ordinate coordinate. P values less than 0.05 were considered statistically signi cant.

Patient characteristics
A total of 843 GIST patients who underwent surgery during hospitalisation from June 2016 to July 2020 and 374 unrelated healthy controls were included. In the GIST patient, males accounted for 55.16% of participants, and the average age was 55.17 years. There were 141 cases in the very low-risk group, 265 cases in the low-risk group, 94 cases in the intermediate-risk group, and 343 cases in the high-risk group.
Three hundred and seventy-six healthy subjects from Wuhan Union Hospital were included as healthy controls during the same period. There were no statistically signi cant differences in age, sex, or body mass index between the two groups (P > 0.05). Compared with the control group, the NLR, PLR, WLR, and MLR values of GIST patients were signi cantly increased (P < 0.001) ( Table 1). Furthermore, the NLR, WLR, MLR, and PLR values of GIST patients of all risk grades were signi cantly higher than patients in the control group (P < 0.05, respectively) ( Table 2). Upon comparing the different risk groups, the NLR, WLR, MLR, and PLR values in the high-risk group were signi cantly higher than that in the very low-risk group, low-risk group, and intermediate-risk group (P < 0.05). In contrast, there was no statistically signi cant difference among the very low-risk group, low-risk group, and intermediate-risk group (P > 0.05).  P1: between control group and extra-low level group; P2: between control group and low level group; P3: between control group and medium level group; P4: between control group and high level group.
Receiver operating characteristic curve analysis in the very low-risk group According to the receiver operating characteristic curve (ROC), the MLR was found to have a relatively good predictive power in the very low-risk, low-risk, and intermediate-risk groups. However, the NLR, WLR, and PLR were found to have a relatively poorer predictive power in the very low-risk, low-risk, and intermediate-risk groups (Table 3).

Discussion
With the recent improvements in the risk grading criteria, and the in-depth clinical study of imatinib, the diagnosis and treatment of GISTs has dramatically improved. The risk criteria proposed by the NIH are widely used worldwide and was shown to have a reliable correlation with patient survival after GIST surgery [8,19]. Imatinib is one of the few tumour-targeting drugs with signi cant clinical e cacy that has been incorporated into the standard postoperative treatment regimen for patients with high-risk GIST.
Preoperatively, imatinib can shrink the tumour to promote the complete surgical resection and reduce the occurrence of surgical complications, especially for local high-risk GIST [20]. A precise classi cation of the risk grade can have a signi cant impact on surgery and prognosis. However, the classi cation criteria rely on pathological results, so it is important to accurately predict and classify the risk grade before surgery.
PLR, WLR, MLR, and NLR are biomarkers of the immune response, and have been shown to be prognostic factors for many types of tumours. East et al. [17] found that the NLR and WLR were independent predictors of postoperative overall survival in colorectal cancer. For GIST, Racz et al. [15] showed that a high PLR is related to a poor prognosis in GIST patients. Goh et al. [16] found that a high PLR and high NLR are independent risk factors for postoperative progression-free survival in GIST patients. Cananzi et al. [18] found that a high MLR was negatively correlated with disease-free survival after GIST surgery. Peripheral blood cell detection is a convenient and routinely performed preoperative test, but only a few studies have explored the predictive value of in ammatory biomarkers for GIST risk grading.
In the present study, we evaluated the value of in ammatory markers such as PLR, WLR, MLR, and NLR in the risk grading of GISTs, compared the differences in their predictive value between different GIST risk groups, and evaluated their diagnostic signi cance. We found that the PLR, WLR, MLR, and NLR had no obvious diagnostic signi cance in the very low-risk group, but showed good predictive power in the lowand medium-risk groups. In contrast, the diagnostic signi cance in the high-risk group was the most prominent. At present, many studies have assessed the prognostic signi cance of the NLR and PLR for GISTs. A high NLR and PLR are often associated with poor prognosis, which is consistent with our results. The cut-off values of NLR and PLR (2.059 and 131.8) showed a good predictive power in the high-risk group. Few studies have assessed the prognostic signi cance of the WLR in GISTs. Therefore, we rst evaluated the predictive signi cance of the WLR for grading the risk of GISTs, and found that this had a higher predictive value compared with the NLR and PLR. The MLR has been reported as a prognostic factor in studies involving colorectal cancer, multiple myeloma, Hodgkin's lymphoma, and other types of tumours, and has also been con rmed as an independent prognostic factor for diseasefree survival after GIST surgery in the study by Cananzi [18]. In this study, the MLR showed a better predictive value in the intermediate-risk and low-risk groups, and also exhibited the most valuable predictive value compared with other in ammatory markers in the high-risk group.
This study has some limitations. Firstly, this was a retrospective study. Although promising results were obtained, there are still few studies on the MLR and WLR for GIST, and further prospective studies are needed to verify these results. Secondly, there are numerous grading criteria for GIST, and we only evaluated the widely used NIH standards, so the practicability of this study was limited.
In summary, this study assessed a large sample of 843 patients and found that preoperative PLR, WLR, MLR, and NLR values were all good predictors for the risk of GIST; however, there were signi cant differences in the predictive value among different risk groups. In the very low-risk group, the predictive value of all four markers was de cient, but these markers showed better predictive value in the low-risk and intermediate-risk groups. However, these four markers had the best predictive value in the high-risk group, among which, the MLR was the most evident. Clinically, the peripheral blood test is a cheap and convenient method that can help determine the risk grade of GIST before surgery, and can provide a reliable basis for the selection of targeted preoperative imatinib therapy and postoperative survival evaluation.

Conclusions
Preoperative NLR, WLR, MLR, and PLR have a high predictive value for classifying the different risk grade of GIST, especially in the high-risk group, which can be used as an effective indicator in predicting malignant potential of GIST before treatment.  Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
The study ethics approval was granted from the local ethical committee of the Wuhan Union Hospital (No.2018-S377), and the study was performed in accordance with the principles of the Declaration of Helsinki.

Consent for publication
Not applicable.

Figure 1
Comparison of MLR, WLR, NLR, and PLR between different risk groups. A: Comparison of MLR between different risk groups; B: Comparison of WLR between different risk groups; C: Comparison of NLR between different risk groups; D: Comparison of PLR between different risk groups.