The neutrophil to lymphocyte ratio can discriminate anaplastic thyroid cancer against poorly or well differentiated cancer

The World Health Organization (WHO) Classification of Tumors in 2004 defined poorly differentiated thyroid cancer (PDTC) as "follicular-cell neoplasms that show limited evidence of structural follicular cell differentiation and occupy both morphologically and behaviorally an intermediate position between well differentiated thyroid cancer (WDTC) and anaplastic thyroid cancer (ATC)" [1, 2, 3, 4]. The occurrence of a mutation in the p53 gene is considered a key event in the malignant progression of WDTCs toward the development of highly aggressive undifferentiated tumors. In both PDTCs and ATCs, prognosis is negatively affected by p53 mutations, which contribute actively to tumor maintenance, spreading, and increased resistance to conventional anticancer treatments [5, 6]. To date, most of the work has focused on the histological and immunohistochemical markers. These markers, while effective, are often expensive and time-consuming to use.

Inflammatory status can lead to enhance tumor growth, invasion, angiogenesis, and eventually metastasis [7, 8, 9]. Tumor-host interactions may have significant influences on patients' outcomes, but this effect is rarely taken into account in current diagnostic or prognostic systems. There is now accumulating evidence that the markers of the systemic inflammatory response, including cytokines, CRP, albumin, serum amyloid A, and WBC count are able to contribute as prognostic factors in cancer patients [10, 11, 12]. Neutrophil to lymphocyte ratio (NLR) is a simple index of the systemic inflammatory response, and has been shown to be a prognostic indicator in some types of cancer.

Thyroid cancer commonly shows a close association with inflammation. Several articles have shown that there is an increased incidence of differentiated thyroid cancer in patients with thyroiditis [13, 14]. High preoperative NLR has been associated with increased tumor size and high American Thyroid Association (ATA) risk of recurrence in patients with differentiated thyroid cancer [15]. Additionally, the intratumoral lymphocyte and immature dendritic cell infiltrate is reduced or absent in PDTC and ATC [16]. NLR is cheaper than serum CRP, which is now routinely measured as part of the cancer work-up, easily calculated, and universally available. We hypothesized that the NLR could be used as an effective discriminating tool in these subtypes of thyroid cancer.

Patients who underwent thyroid surgery in Gwangju and Hwasun Chonnam National University Hospitals between January 2004 and March 2009 were identified from a prospectively maintained database. The database is prospectively maintained and all patients are followed-up regularly by a team of thyroid cancer specialists. In total, 3,870 patients were enrolled, including some patients who did not undergo any surgical procedures but had a definite diagnosis on cytological examination.

The NLR was defined as the absolute neutrophil count divided by the absolute lymphocyte count. For patients who underwent surgery, the NLR was calculated from the full blood count routinely performed on the 7th-14th day before surgery. For patients who did not undergo surgery, the NLR was calculated from the full blood count performed as part of the diagnostic process. The median NLR of all 3,870 patients was 1.57, and evenly divided into 1,935 patients on each groups: the neutrophilic high NLR group and the lymphophilic low NLR group. The preoperative and postoperative profiles were analyzed.

Patients were excluded from the study for the following reasons: not enough data on follow-up, pediatric patients, patients with other cancer history or distant metastasis, and previous history of infectious conditions such as pulmonary or visceral tuberculosis.

Data were summarized with the number of subjects and the mean ± standard deviation or median (range) value. We used the t-test to compare continuous variables between each group and the chi-square test for categorical variables. Mann-Whitney U test was used to compare categorical end points and the two-sample t-test was used to compare continuous variables. A receiver operating characteristics curve was constructed to estimate the optimal cutoff value of pretreatment NLR and other variables. Subsequently, the variables with P < 0.05 entered into the multivariate analysis. Statistical significance was indicated by a P-value of <0.05. Results were analyzed using IBM SPSS Statistics ver. 21.0 (IBM Co., Armonk, NY, USA).

The process of the development of malignancy has been firmly associated with impaired function of the immune system [18]. Recently, many research centers have begun examining the host's inflammatory response to tumors and the systemic effects exerted by tumors in causing up-regulation of the inflammatory process, and thus increasing the propensity of metastasis through the inhibition of apoptosis, promotion of angiogenesis, and DNA damage [7, 19, 20, 21]. Prior studies have demonstrated an association between simple inflammatory markers (blood neutrophil, lymphocyte, and platelet counts) and adverse effects in certain types of cancer (stomach, colon, bladder, esophageal, pulmonary, ovarian, pancreas, renal and others). Representatively, Bruckner et al. [22] found that pretreatment absolute neutrophil and lymphocyte count were independent prognostic indicators for patients with metastatic gastric cancer. Later, the NLR was observed to be significantly elevated in the advanced stages of colorectal cancer [23].

In our study, advanced stage of the disease, such as T3 or T4, cancer-specific death, and the proportion of PDTC to ATC was significantly more distributed in the high NLR group. Some studies have shown that high preoperative NLR is associated with an increased tumor size and high ATA risk of recurrence in patients with differentiated thyroid cancer [15]. Unfortunately, in our study, tumor size and cancer-specific deaths in the subanalysis of WDTC with NLR were not different, and we did not discriminate between malignant cancers and benign tumors. However, the proportion of PDTC and ATC was significantly higher in the high NLR group. The NLR was the only diagnostic tool for discriminating between WDTC, PDTC and ATC.

PDTC lies both morphologically and behaviorally between WDTC and undifferentiated ATC [4, 17, 24]. PDTC and ATC have poor prognoses and rare incidence rates compared to WDTC. Since the original description of PDTC in 1983 [25], PDTC has been introduced as a separate entity in the 2004 WHO Classification of endocrine tumors [1]. From a histopathogenetic point of view, as for undifferentiated (anaplastic) carcinomas, it is generally accepted that PDTC may arise de novo or from preexisting well differentiated follicular and papillary carcinomas. However, definite molecular (and prevalence) data are difficult to obtain from the published literature, which reflect confusing terminology and the heterogeneous classification criteria debated above [17].

ATC is a rare but aggressive form of cancer that accounts for <2% of all thyroid malignancies [26]. The results of surgical resection, chemotherapy and radiotherapy, alone or in combination, are not effective and patients usually die within 6-12 months [27, 28]. This clinical behavior and outcome is dramatically different from that observed in WDTCs, which usually grow and progress slowly, rarely metastasize, and are characterized by a good or very good prognosis. Our series have shown that ATC have a distinct NLR from that of PDTC and WDTC. Therefore, we constructed a simple approach in an effort to streamline the diagnosis of ATC versus PDTC and WDTC with NLR.

The processing of discrimination between PDTC and ATC is very difficult even with microscopic examination and immunohistochemical staining and clinical outcome data. In our patients between 2004 and 2009, 14 patients had PDTC and seven had completely discriminated ATC. Two pathologists from different institutions reviewed and confirmed the diagnoses of 21 patients with PDTC or ATC. Although there is much limitations remains to be done. The incidences of PDTC and ATC are very low, so too small numbers were enrolled in this study. But our work generates important findings in the field of distribution of NLR. Importantly, the NLR was also a meaningful diagnostic tool for discriminating PDTC and ATC. The NLR can discriminate between PTC, PDTC, and ATC (P = 0.035, P = 0.002, and P = 0.025, respectively), with a cutoff value of 3.8. The use and validation of uniform internationally accepted criteria for PDTC and ATC should be encouraged to garner a better understanding of their pathogenetic origin, to search for potentially helpful diagnostic, prognostic, and predictive markers, to plan therapy, and to establish the epidemiologic distribution of these tumor entities.

In conclusion, we demonstrated that the NLR is significantly different and high in PDTC and ATC as compared to WDTC, and represents a poor prognosis for those cancers. Therefore, NLR can play a relevant role as a discriminating tool and may be considered as a new diagnostic criterion in discriminating these aggressive forms of thyroid cancer.