Pretreatment C‐reactive protein/albumin ratio is associated with poor survival in patients with stage IB‐IIA cervical cancer

Abstract Previous studies have shown that the C‐reactive protein/albumin ratio (CAR) is a prognostic indicator in multiple types of carcinomas. This study is the first to evaluate the prognostic significance of CAR in stage IB‐IIA cervical cancer patients treated with radical surgery, as well as that of several other inflammation‐based factors, including the neutrophil‐to‐lymphocyte ratio (NLR), platelet‐to‐lymphocyte ratio (PLR), and prognostic nutritional index (PNI). A total of 235 patients were enrolled in this study. The optimal cut‐off values of CAR and other inflammation‐based factors were determined by receiver operating characteristic curves. The Kaplan–Meier method and Cox regression model analysis were performed to determine the independent predictors of progression‐free survival (PFS) and overall survival (OS). At a cut‐off value of 0.15, patients with a high CAR had significantly shorter PFS and OS than those with a lower CAR (P < 0.001). A higher CAR was significantly associated with elevated scores of NLR and PLR and a decreased PNI (P < 0.001). Univariate analyses showed that elevated CAR preoperatively was significantly associated with poor survival; a similar trend was also noted for the NLR, PLR, and PNI. Multivariate analyses demonstrated that only CAR was an independent indicator for PFS (hazard ratio [HR]: 5.164; 95% confidence interval [CI]: 2.495–10.687; P < 0.001) and OS (HR: 4.729; 95% CI: 2.263–9.882; P < 0.001). In conclusion, preoperative CAR is a novel and superior predictor of poor survival in patients with stage IB‐IIA cervical cancer.

Introduction after induction by inflammatory or tumor cells [6,7]. Human papillomavirus (HPV) can promote the development of cervical cancer by triggering a series inflammation and inflammatory molecules in the microenvironment [8]. It was previously shown that one way to improve the nutritional status is to attenuate the systemic inflammatory response [9]. Basic and clinical research results have suggested that hypoalbuminemia, malnutrition, and cancer cachexia are all consequences of the body's systemic inflammatory response to malignancy [10]. Zheng et al. also showed that albumin is an independent prognostic indicator in early operable cervical cancer [7]. Recently, several inflammationand nutrition-based scores such as the platelet-to-lymphocyte ratio (PLR), neutrophil-to-lymphocyte ratio (NLR), and prognostic nutritional index (PNI) have been demonstrated as significant predictors in cervical cancer as well as other cancers [11][12][13][14]. The PNI is calculated according to the following formula: serum albumin value (g/L) + 0.005 × lymphocyte counts (per mm³) in peripheral blood [14]. However, the most common measure of the systemic inflammatory response in cancer patients is the serum concentration of C-reactive protein (CRP), which can identify patients who will likely develop cachexia and have a poor survival and response to treatment after combining with albumin to a scale [9,15]. As a novel inflammation-and nutritionbased score, the CRP/albumin ratio (CAR) is a combination of CRP and albumin. It has been reported as an independent indicator of poor prognosis in different malignant carcinomas including hepatic cancer [16], lung cancer [17], colon cancer [18], esophageal cancer [19], and pancreatic cancer [20]. Currently, there are no reports concerning CAR as a predictor in patients with cervical cancer. Hence, we retrospectively studied the intrinsic tumor factors along with inflammationbased scores to identify superior prognostic factors in cervical cancer patients with stage IB-IIA disease.

Methods Patients
A total of 235 IB-IIA cervical cancer patients who had radical hysterectomy and pelvic lymphadenectomy at Nanfang Hospital of Southern Medical University from January 2005 to December 2009 were included in this study. Patients who had an active or chronic infection received chemotherapy or radiotherapy before surgery, hematological and autoimmune disorders, or did not have available follow-up data were excluded. Patients whose clinical information could not be completely obtained were also excluded from our analysis. The tumor stage of all cases was defined according to the FIGO 2009 criteria for cervical cancer. Clinical data including the treatment programs (surgery, chemoradiotherapy, or radiotherapy), maximal tumor size, and histopathological parameters were retrospectively achieved from the medical data of cervical cancer patients. Laboratory parameters including the neutrophil, lymphocyte, and platelet counts, and CRP and albumin levels were obtained within 2 days before radical surgery. This study was approved by the medical ethics committee of Nanfang Hospital of Southern Medical University. All the methods were performed in accordance with the relevant guidelines and regulations. Written informed consent was obtained for each patient.

Clinical management
All the patients with FIGO stage IB-IIA cervical cancer were managed by radical hysterectomy and pelvic lymphadenectomy. Para-aortic lymph node dissection was performed in patients who had suspicious para-aortic lymph node metastasis. Postoperative radiotherapy with or without concurrent platinum-based chemotherapy was treated depending on the postoperative pathology. PFS was defined as the time from the start of surgery to recurrence or death. OS was defined as the time from the start of the surgery to the last contact or death. All of the cervical cancer patients were followed up every 3 months for the first 2 years, every 6 months for the next 5 years, and annually thereafter until January 2014. Gynecological examinations, laboratory checks, and imaging methods (e.g., head and abdominal ultrasound scanning, chest X-ray, electrocardiography, and computed tomography) were conducted during follow-up evaluations.

Statistical analysis
Statistical analyses were performed using Statistical Product and Service Solutions 20.0 software (IBM, Corporation, Armonk, NY). The best cutoff points of NLR, PLR, PNI, and CAR were determined using receiver operating characteristic (ROC) curve analysis. The relationships between CAR and other variables were performed using Pearson's chi-square test. Significant prognostic variables in univariate analyses were selected for multivariate Cox regression model analyses to determine independent prognostic factors using the forward stepwise method. Estimates of PFS and OS among the classification groups were calculated using the Kaplan-Meier method, and two-sided 95% confidence intervals were obtained. A two-sided P < 0.05 was considered statistically significant.

Discussion
Chronic inflammation is involved in cellular transformation, survival, proliferation, invasion, and metastasis of tumor cells [8]. As an inflammation-sensitive marker,  serum CRP has been demonstrated to be an independent prognostic indicator for clinical outcomes and posttherapy monitoring in cervical cancer [21,22]. Interleukin-6 (IL-6) can be released from the tumor microenvironment by leukocytes and promote the generation of CRP [23,24]. High CRP is most likely to be associated with tumor necrosis, local tissue damage, and inflammation response in cancer patients [25]. A study by Polterauer et al. demonstrated that polymorphisms of the CRP gene can weaken the immune defense of tumor patients and increase the serum CRP levels [22]. CRP-lowering drugs, such as COX inhibitors and lipid-lowering agents, have already been confirmed as effective therapeutic targets for patients undergoing cardiovascular therapy, providing promising clues for theoretical efficacy in cancer prevention and therapy [26]. Many epidemiological studies have shown that the long-term use of COX inhibitors and lipid-lowering agents is related to decreased risk of various human cancers including cervical cancer [26,27]. Previous studies have shown that statins, CRP-lowering drugs, have antitumor proliferative, antitumor angiogenic, and antitumor metastatic properties, and therefore, represent a novel treatment method for cancer prevention and treatment [28]. Furthermore, anti-IL-6 therapy not only decreases CAR, C-reactive protein/albumin ratio; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; PNI, prognostic nutritional index.
W. Zhang et al. CAR and cervical cancer CRP levels but also improves the nutritional status of cancer patients by increasing their appetite [29].
Malnutrition accounts for 20% of all cancer-related deaths [30]. Approximately 40-80% of all cancer patients will experience malnutrition at some stage during the clinical course of their disease [30]. In fact, studies have demonstrated that approximately 62-88% of gynecological cancer patients will experience malnutrition [31,32]. Various factors including hemoglobin, total protein, albumin, and transferrin are used to evaluate the nutritional status in patients with gynecological cancer [33]. Malnutrition and inflammation suppresses the synthesis of serum albumin, which can reflect the nutritional status of patients, as well as the severity, progression, and prognosis of disease [34]. IL-6 can regulate albumin production and decrease the serum albumin concentration [35]. Similarly, the serum albumin level, which correlates well to body immunity and nutrition status, may increase the CRP concentration and induce inflammation [36,37]. Serum albumin has been shown as an independent predictor of clinical outcomes in various cancers, such as lung cancer, breast cancer, colorectal cancer, ovarian, and cervical cancer [33,38]. In addition, many studies have combined CRP and albumin to create new inflammationbased scores such as the Glasgow Prognostic Score (GPS) and CAR. Both the GPS and CAR have been confirmed as independent indicators of poor survival in various cancer types [39,40]. To date, the relationship between CAR and cervical cancer prognosis is obscure. Hence, we decided to combine CRP and serum albumin to create a new inflammatory prognostic factor in stage IB-IIA cervical cancer.
This study was the first to report the correlation between CAR and clinical outcomes in cervical cancer patients. We assessed and compared the prognostic value of the preoperative PNI, PLR, NLR, and CAR in 235 patients with stage IB-IIA cervical cancer by retrospectively analyzing the clinical characteristics. A 0.15 cut-off value for CAR was used to divide patients into two groups. According to the chi-square test, the CAR had a significant relationship with other inflammation-based scores, but no significant relationship with all of the intrinsic factors. When patients were stratified into low-CAR and high-CAR subgroups, patients with a higher CAR had a significantly shorter PFS and OS than those with a lower CAR. Univariate analyses indicated that tumor stage, maximum tumor size, adjuvant therapy, and all of the inflammation-based scores (PLR, NLR, PNI, and CAR) were significantly related to a poor PFS and OS in cervical cancer patients of stage IB-IIA. Multivariate analyses demonstrated that only tumor stage, adjuvant therapy, and CAR were independent prognostic factors for predicting poor PFS and OS in cervical cancer patients.
However, this study had several limitations. First, it was inevitably affected by residual confounding factors. Second, we only included stage IB-IIA cervical cancer patients. Third, the number of patients was relatively small. Finally, it was single retrospective study.
In conclusion, the preoperative CAR is a novel and superior prognostic factor in cervical cancer patients with stage IB-IIA and radical surgery. More clinical trials are needed to further elaborate upon the relationship between CAR and the prognosis of stage IB-IIA cervical cancer.