Prognostic value of systemic inflammatory markers for oral cancer patients based on the 8th edition of AJCC staging system

New perspectives in the differential diagnosis of jaw lesions: Machine learning and inflammatory biomarkers

This study aimed to assess the diagnostic performance of a machine learning approach that utilized radiomic features extracted from Cone Beam Computer Tomography (CBCT) images and inflammatory biomarkers for distinguishing between Dentigerous Cysts (DCs), Odontogenic Keratocysts (OKCs), and Unicystic Ameloblastomas (UAs). This retrospective study involves 103 patients who underwent jaw lesion surgery in the Maxillofacial Surgery Unit of Federico II University Of Naples between January 2018 and January 2023. Nonparametric Wilcoxon-Mann-Whitney and Kruskal Wallis tests were used for continuous variables. Linear and non-logistic regression models (LRM and NLRM) were employed, along with machine learning techniques such as decision tree (DT), k-nearest neighbor (KNN), and support vector machine (SVM), to predict the outcomes. When individual inflammatory biomarkers were considered alone, their ability to differentiate between OKCs, UAs, and DCs was below 50 % accuracy. However, a linear regression model combining four inflammatory biomarkers achieved an accuracy of 95 % and an AUC of 0.96. The accuracy of single radiomics predictors was lower than that of inflammatory biomarkers, with an AUC of 0.83. The Fine Tree model, utilizing NLR, SII, and one radiomic feature, achieved an accuracy of 94.3 % (AUC = 0.95) on the training and testing sets, and a validation set accuracy of 100 %. The Fine Tree model demonstrated the capability to discriminate between OKCs, UAs, and DCs. However, the LRM utilizing four inflammatory biomarkers proved to be the most effective algorithm for distinguishing between OKCs, UAs, and DCs.

Salivary Neutrophil-to-Lymphocyte Ratio as a Prognostic Predictor of Oral Premalignant and Malignant Disorders: A Prospective Study

INTRODUCTION

Inflammation is a definitive characteristic of carcinogenesis. The neutrophil-to-lymphocyte ratio (NLR) is an easy and efficient indicator of inflammation and a valuable marker in individuals with malignancies. The present study was performed to ascertain NLR values in salivary samples collected from individuals with oral premalignant disorders (OPMDs) and to assess the prognostic significance of NLR in distinguishing OPMDs from oral malignancies.

MATERIALS AND METHODS

This study was conducted on 50 patients histopathologically diagnosed with OPMDs with mild dysplasia. The patients were provided with standard medicinal treatment, encouraged to quit their habits, and followed up for one year at three-month regular intervals. During the follow-up, 29 (67.4%) patients completely recovered, whereas 14 (32.6%) developed oral malignancies. Salivary samples were collected at baseline (T0) and one-year follow-up (T1). The total salivary neutrophils and lymphocytes were counted using an improved cell counting method with a Neubauer chamber. The NLR values were calculated at T0 and T1. The paired t-test was used to compare the NLR values at T0 and T1. The cutoff value of the NLR was determined using the receiver operating characteristic (ROC) curve. The Youden index was used to determine the optimal cutoff NLR values in the groups. Statistical significance was set at p ≤0.05.

RESULTS

OPMDs were predominantly observed in males, with leukoplakia being the most prevailing one. Erythroplakia exhibited the highest propensity for malignant transformation, and habitual consumption of alcohol and tobacco was identified as a risk factor for this transformation. NLR increased in both premalignant and malignant conditions. NLR value equal to or exceeding 4 was determined to be a reliable indicator for the occurrence of oral cancer in patients with OPMDs. The ROC curve analysis yielded a sensitivity and specificity of 92%, with an area under the curve (AUC) of 0.928.

CONCLUSION

The poor prognosis of oral cancers was associated with higher NLR values. NLR values in salivary samples can serve as an independent reliable predictor in oral cancer and OPMDs.

Prognostic Significance of Pre-Treatment Neutrophil-Lymphocyte Ratio and Platelet-Lymphocyte Ratio in Head and Neck Malignancies

Introduction

Carcinoma is the second most common cause of death worldwide. The neutrophil-lymphocyte ratio (NLR) and the platelet-lymphocyte ratio (PLR) are essential markers of inflammation and tumorigenesis in various cancers including head and neck cancers. Pretreatment platelet- lymphocytic ratio can be used as an independent predictor of mortality whereas neutrophil- lymphocytic ratio is an independent predictor of recurrence. The main aim of this study is to compare the pre-treatment neutrophil lymphocyte ratio and platelet-lymphocyte ratio in the patients of head and neck malignancies with those of the control group.

Material and Method

100 patients with histologically diagnosed cases of head and neck malignancies. Age and sex matched healthy subjects attending Otorhinolaryngology out-patient department for any other complaints (100 control subjects). Complete blood count had been done to calculate absolute neutrophil count and absolute lymphocyte count.

Results

The mean age of the subjects in the study group was 55.73 ± 11.56 years. In control group, the mean age group was 54.11 ± 10.46 years. NLR and PLR significantly increased in cases than controls. NLR associated with T stage, histological type and histological grade but not with site and nodal involvement. PLR associated with T stage, metastasis but not with the histological grade, histological type, site and nodal involvement.

Conclusion

From this study, we conclude that pre-treatment NLR and PLR were closely associated both with the size of primary tumor and also with the stage of malignant disease in patients of head and neck malignancies.

[Oral cavity cancer: A distinct entity]

Oral Squamous cell carcinoma represent the 17th most frequent cancer in the world. The main risk factors are alcohol and tobacco consumption but dietary, familial, genetic, or oral diseases may be involved in oral carcinogenesis. Diagnosis is made on biopsy, but detection remains late, leading to a poor prognosis. New technologies could reduce these delays, notably Artificial Intelligence and the quantitative evaluation of salivary biological markers. Currently, management of oral cancer consists in surgery, which can be mutilating despite possible reconstructions. In the future, immunotherapies could become a therapeutic alternative and the immune microenvironment could constitute a source of prognostic markers.

The Role of IL-17 in the Pathogenesis of Oral Squamous Cell Carcinoma

Elucidating the inflammatory mechanisms underlying formation and progression of oral squamous cell carcinoma (OSCC) is crucial for discovering new targeted therapeutics. The proinflammatory cytokine IL-17 has proven roles in tumor formation, growth, and metastasis. The presence of IL-17 is demonstrated in both in vitro and in vivo models, and in OSCC patients, is mostly accompanied by enhanced proliferation and invasiveness of cancer cells. Here we review the known facts regarding the role of IL-17 in OSCC pathogenesis, namely the IL-17 mediated production of proinflammatory mediators that mobilize and activate myeloid cells with suppressive and proangiogenic activities and proliferative signals that directly induce proliferation of cancer cells and stem cells. The possibility of a potential IL-17 blockade in OSCC therapy is also discussed.

A Nomogram Incorporating Neutrophil-to-Lymphocyte Ratio and Squamous Cell Carcinoma Antigen Predicts the Prognosis of Oral Cancers

We introduced a novel squamous cell carcinoma inflammatory index (SCI) and explored its prognostic utility for individuals with operable oral cavity squamous cell carcinomas (OSCCs). We retrospectively analyzed data from 288 patients who were given a diagnosis of primary OSCC from January 2008 to December 2017. The SCI value was derived by multiplying the serum squamous cell carcinoma antigen and neutrophil-to-lymphocyte ratio values. We appraised the associations of the SCI with survival outcomes by performing Cox proportional hazards and Kaplan-Meier analyses. We constructed a nomogram for survival predictions by incorporating independent prognostic factors in a multivariable analysis. By executing a receiver operating characteristic curve analysis, we identified the SCI cutoff to be 3.45, and 188 and 100 patients had SCI values of <3.45 and ≥3.45, respectively. The patients with a high SCI (≥3.45) were associated with worse disease-free survival and overall survival than those with a low SCI (<3.45). An elevated preoperative SCI (≥3.45) predicted adverse overall survival (hazard ratio [HR] = 2.378; p < 0.002) and disease-free survival (HR = 2.219; p < 0.001). The SCI-based nomogram accurately predicted overall survival (concordance index: 0.779). Our findings indicate that SCI is a valuable biomarker that is highly associated with patient survival outcomes in OSCC.

Preoperative serum interleukin-6 level in head and neck cancer reflects systemic inflammatory response and is a predictor of postoperative prognosis

BACKGROUND

The outcome of head and neck cancer has improved in recent years but survival is not yet satisfactory. Interleukin (IL)-6 is a representative inflammatory cytokine and inducer of systemic inflammatory response. It is not known whether preoperative serum level of IL-6 is a prognostic factor in head and neck cancer surgery.

METHODS

We studied 181 consecutive patients who underwent head and neck surgery with free tissue transfer reconstruction (HNS-FTTR) between September 2016 and December 2020. Whether preoperative serum IL-6 level was a prognostic risk factor was retrospectively investigated by univariate and multivariate analyses. We also investigated the association between preoperative IL-6 level and representative systemic inflammatory response markers.

RESULTS

The preoperative IL-6 ≥ 8 pg/mL group had a significantly worse prognosis than the preoperative IL-6 < 8 pg/mL group (overall survival [OS]: hazard ratio [HR] 3.098, P = 0.0006; disease-specific survival [DSS]: HR 3.335, P = 0.0008). In multivariate analysis, IL-6 ≥ 8 pg/mL and age ≥ 70 years were independent poor prognostic factors for OS (HR 1.860, P = 0.0435 and HR 1.883, P = 0.0233, respectively). The only independent poor prognostic factor for DSS was IL-6 ≥ 8 pg/mL (HR 2.052, P = 0.0329). Serum albumin was significantly lower and serum C-reactive protein and neutrophil-to-lymphocyte ratio were significantly higher in the IL-6 ≥ 8 pg/mL group than in the IL-6 < 8 pg/mL group (all P < 0.0001).

CONCLUSIONS

Preoperative serum IL-6 level is an independent poor prognostic factor for both OS and DSS after HNS-FTTR, reflecting the degree of preoperative systemic inflammatory response.

Pattern and localization of perineural invasion predict poor survival in oral tongue carcinoma

BACKGROUND

Survival rate for oral tongue squamous cell carcinoma (OTSCC) is still poor and, despite Tumor-Node-Metastasis staging system has been recently updated, patients included under the same stage still show difference in prognosis. Perineural invasion (PNI) emerged to be an indicator of tumor aggressive behavior and unfortunate events. In this study, we investigate the clinic and prognostic value of PNI in a cohort of OTSCC patients.

METHODS

About 200 patients with OTSCC were retrospectively evaluated the presence of PNI. PNI was furtherly descripted as uni-/multifocal and as intra-/peritumoral. Disease-Specific and Relapse-Free Survival (DSS; RFS) were estimated; moreover, we included PNI in the current AJCC 8th Staging System, improving the prognostication model.

RESULTS

Perineural invasion was found in 40.5% of patients. Intratumoral PNI predicted patients at high risk of being diagnosed with lymph-node metastasis. Tumors with positive PNI reported a worse DSS (Hazard Ratio=1.878, p-value = 0.008). Moreover, patients exhibiting both multifocal intra- and peritumoral PNI reported poorest DSS (Hazard Ratio = 2.409, p-value = 0.010). Patients were reclassified in a new staging system in case of multifocal PNI, providing better stratification capacity.

CONCLUSIONS

Perineural invasion might serve as an additional prognostic factor in OTSCC, and by integrating PNI in the staging system, further improvements in prognostication might be reached.

Association of Neutrophil-to-Lymphocyte Ratio and Bloodstream Infections with Survival after Curative-Intent Treatment in Elderly Patients with Oral Cavity Squamous Cell Carcinoma

Patients with oral cavity squamous cell carcinoma (OSCC) undergoing curative-intent treatment may become immunocompromised. This study aimed to investigate the association of pretreatment sarcopenia, nutritional status, comorbidities, and blood-based inflammation prognostic biomarkers in bloodstream infection (BSI) with survival status in elderly patients with OSCC. Retrospective data were collected from 235 patients who were newly diagnosed with OSCC, were aged ≥ 65 years, had undergone curative-intent treatment, and were classified into either the BSI group or the no-BSI group within 6 months after surgery and/or adjuvant therapy initiation. Of the 235 elderly patients, 27 presented with BSI episodes. A preoperative high neutrophil-to-lymphocyte ratio (NLR) was a significant independent risk factor for BSI. BSI was not significantly associated with survival status. Ever betel nut chewing, hypoalbuminemia, and advanced tumor stage were associated with shorter overall survival. Moreover, a high NLR was an independent risk factor associated with disease-free survival. A high NLR was associated with BSI and resistance to curative-intent treatment. Pretreatment of NLR could act as an independent prognostic indicator and help inform treatment strategies for older patients with OSCC.

Defining biomarkers in oral cancer according to smoking and drinking status

Introduction

Oral Squamous Cell Carcinomas (OSCC) are mostly related to tobacco consumption eventually associated to alcohol (Smoker/Drinker patients: SD), but 25-30% of the patients have no identified risk factors (Non-Smoker/Non-Drinker patients: NSND). We hypothesized that these patients have distinguishable immune profiles that could be useful for prognosis.

Materials and Methods

Cells present in immune tumor microenvironment (TME) and blood from 87 OSCC HPV-negative patients were analyzed using a multiparameter flow cytometry assay, in a prospective case-control study. Cytokine levels in tumor supernatants and blood were determined by a cytometric bead array (CBA) assay.

Results

Normal gingiva and blood from healthy donors (HD) were used as controls. A significant increase of granulocytes (p<0.05 for blood), of monocytes-macrophages (p<0.01 for blood) and of CD4⁺ T cells expressing CD45RO and CCR6 (p<0.001 for blood; p<0.0001 for TME) as well as higher levels of IL-6 (p<0.01 for sera, p<0.05 for tumor supernatant) were observed in SD patients as compared to NSND OSCC patients and HD. High percentages of CD4⁺ T cells expressing CD45RO and CCR6 cells in tumor tissue (p=0.05) and blood (p=0.05) of SD OSCC patients were also associated with a poorer prognosis while a high percentage of regulatory T cells (Treg) in tumor tissue was associated with a more favorable prognostic factor (p=0.05). Also, a higher percentage of blood CD8⁺ T lymphocytes among CD45⁺ cells in NSND patients was associated with a better disease-free survival (p=0.004).

Conclusion

Granulocytes, monocytes-macrophages, and CD4⁺ T cells expressing CD45RO and CCR6 in blood and TME as well as serum IL-6 can therefore distinguish OSCC SD and NSND patients. Quantifying the proportion of CD4⁺ T cells expressing CD45RO and CCR6 and of Treg in SD patients and CD8⁺ T cells in NSND patients could help defining the prognostic of OSCC patients.

Prognosis of the Keratinizing Squamous Cell Carcinoma of the Tongue Based on Surveillance, Epidemiology, and End Results Database

BACKGROUND

The objective of this study is to determine the prognostic factors of keratinizing squamous cell carcinoma of the tongue (KTSCC) and to establish a prognostic nomogram of KTSCC to assist clinical diagnosis and treatment.

METHODS

This study identified 3874 patients with KTSCC from the Surveillance, Epidemiology, and End Results (SEER) database, and these patients were randomly divided into the training (70%, (n = 2711) and validation (30%, n = 1163) cohorts. Cox regression was then used to filter variables. Nomograms were then constructed based on meaningful variables. Finally, the concordance index (C-index), net reclassification index (NRI), integrated discrimination improvement (IDI), calibration charts, and decision-curve analysis (DCA), were used to evaluate the discrimination, accuracy and effectiveness of the model.

RESULTS

A nomogram model was established for predicting the 3-, 5-, and 8-year overall survival (OS) probabilities of patients with KTSCC. The model indicated that age, radiotherapy sequence, SEER stage, marital status, tumor size, American Joint Committee on Cancer (AJCC) stage, radiotherapy status, race, lymph node dissection status, and sex were factors influencing the OS of patients with KTSCC. Verified by C-index, NRI, IDI, calibration curve, and DCA curve, our model has better discrimination, calibration, accuracy and net benefit compared to the AJCC system.

CONCLUSIONS

This study identified the factors that affect the survival of KTSCC patients and established a prognostic nomogram that can help clinicians predict the 3-, 5-, and 8-year survival rates of KTSCC patients.

The Combination of Inflammatory Biomarkers as Prognostic Indicator in Salivary Gland Malignancy

Background: The aim of this study was to investigate how the systemic inflammation response index (SIRI), systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR), taken individually and combined, are associated with overall survival (OS) in patients surgically treated for malignant salivary gland tumors (MSGTs). Methods: A retrospective analysis of 74 cases following surgery at our department between January 2011 and June 2018 was performed. The Receiver Operating Characteristic (ROC) curve was used to calculate the optimal cutoff values for SII, SIRI, PLR, and NLR. Survival curves of different groups at 1−3−5 years were estimated using the Kaplan−Meier method. Results: The optimal thresholds with the highest sensitivity and specificity were 3.95 for NLR, 187.6 for PLR, 917.585 for SII, and 2.045 for SIRI. The ROC curves revealed that the best combination with AUC = 0.884 was SII + SIRI. The estimated 5-year OS probability in patients with SII+ SIRI scores of 0, 1, and 2 was 96%, 87.5% and 12.5%, respectively (p < 0.001). Conclusion: SII+ SIRI can independently predict the OS of patients after MSGT surgery. The prognostic score system based on SII+ SIRI may be good clinical practice as a reference for clinical decision-making.

Prognostic Abilities of Pre- and Post-Treatment Inflammatory Markers in Oral Squamous Cell Carcinoma: Stepwise Modelling

Background and Objectives: Studies examining the importance of inflammatory markers before treatment as prognosticators of OSCC are available, but information on post-therapy inflammatory markers and their prognostic significance is limited. This study aimed to evaluate the prognostic abilities of pre- and post-treatment inflammatory markers in patients with OSCC. Materials and Methods: In this retrospective analysis, information on 151 OSCC patients’ socio-demographic, clinico-pathological, recurrence, metastasis, and survival data were gathered from clinical records. A multivariable Cox proportional hazards regression (stepwise model) was conducted to identify the prognostic predictors of OS and DFS. The multivariable models’ performances were evaluated using Harrell’s concordance statistics. Results: For OS, high pre-treatment LMR (HR 3.06, 95%CI 1.56, 5.99), and high post-treatment PLC (HR 3.35, 95%CI 1.71, 6.54) and PLR (HR 5.26, 95%CI 2.62, 10.58) were indicative of a poor prognosis. For DFS, high pre-treatment SII (HR 2.59, 95%CI 1.50, 4.48) and high post-treatment PLC (HR 1.92, 95%CI 1.11, 3.32) and PLR (HR 3.44, 95%CI 1.98, 5.07) were associated with increased mortality. The fitness of the OS and DFS stepwise Cox regression models were proven with a time-dependent AUC of 0.8787 and 0.8502, respectively. Conclusions: High pre-treatment levels of LMR and SII and high post-treatment levels of PLC and PLR are independent predictors of a poor prognosis for patients with OSCC.

Prognostic Role of High-Sensitivity Modified Glasgow Prognostic Score for Patients With Operated Oral Cavity Cancer: A Retrospective Study

AIM

We probed the prognostic value of the preoperative high-sensitivity modified Glasgow prognostic score (HS-mGPS), neutrophil/lymphocyte ratio (NLR), and platelet/lymphocyte ratio (PLR) for patients with oral cavity squamous cell carcinoma (OSCC) to identify patients with the highest risk of having poor survival outcomes.

MATERIALS AND METHODS

We executed a retrospective assessment of the records of 303 patients with OSCC who had been subjected to curative surgery between January 2008 and December 2017. The HS-mGPS was categorized using C-reactive protein and albumin thresholds of 3 mg/L and 35 g/L, respectively. Moreover, receiver operating characteristic curve analyses were executed to find out the optimal PLR and NLR cutoffs. We plotted survival curves and compared them through the use of the Kaplan-Meier method and log-rank test, respectively. Through a Cox proportional hazard model, we identified prognostic variables. We also plotted a nomogram comprising the HS-mGPS and clinicopathological factors and assessed its performance with the concordance index.

RESULTS

The PLR and NLR cutoffs were 119.34 and 4.51, respectively. We noted an HS-mGPS of 1-2 to be associated with a shorter median overall survival (OS) and disease-fee survival (DFS) compared with an HS-mGPS of 0. Multivariate analysis revealed that an HS-mGPS of 1-2 and an NLR of ≥4.51 were independent risk factors related to poor OS and DFS. The HS-mGPS appeared to have better prognostic effect than did the PLR and NLR, and the combination of the HS-mGPS and NLR appeared to exhibit optimal discriminative ability for OS prognostication. The nomogram based on the HS-mGPS and NLR yielded accurate OS prediction (concordance index = 0.803).

CONCLUSION

Our findings suggest that preoperative HS-mGPS is a promising prognostic biomarker of OSCC, and the nomogram comprising the HS-mGPS and NLR provided accurate individualized OSCC survival predictions.

Direct Contact with Platelets Induces Podoplanin Expression and Invasion in Human Oral Squamous Cell Carcinoma Cells

Oral squamous cell carcinoma (OSCC) is mostly diagnosed at an advanced stage, with local and/or distal metastasis. Thus, locoregional and/or local control of the primary tumor is crucial for a better prognosis in patients with OSCC. Platelets have long been considered major players in cancer metastasis. Traditional antiplatelet agents, such as aspirin, are thought to be potential chemotherapeutics, but they need to be used with caution because of the increased bleeding risk. Podoplanin (PDPN)-expressing cancer cells can activate platelets and promote OSCC metastasis. However, the reciprocal effect of platelets on PDPN expression in OSCC has not been investigated. In this study, we found that direct contact with platelets upregulated PDPN and integrin β1 at the protein level and promoted invasiveness of human OSCC Ca9.22 cells that express low levels of PDPN. In another human OSCC HSC3 cell line that express PDPN at an abundant level, silencing of the PDPN gene reduced cell invasiveness. Analysis of the public database further supported the co-expression of PDPN and integrin β1 and their increased expression in metastatic tissues compared to normal and tumor tissues of the oral cavity. Taken together, these data suggest that PDPN is a potential target to regulate platelet-tumor interaction and metastasis for OSCC treatment, which can overcome the limitations of traditional antiplatelet drugs.

Systemic Inflammatory Markers as Predictors of Postoperative Complications and Survival in Patients With Advanced Head and Neck Squamous Cell Carcinoma Undergoing Free-Flap Reconstruction

PURPOSE

The aim of this study was to determine the prognostic value of systemic inflammatory indices as factors for postoperative complications and survival in patients with advanced stages of p16-negative head and neck squamous cell carcinoma undergoing free-flap reconstruction.

METHODS

This was a retrospective cohort study. The primary predictor variables were inflammatory markers such as neutrophil, lymphocyte, monocyte, and platelet count, neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio, lymphocyte-monocyte ratio, derived NLR, systemic immune-inflammation, and systemic inflammatory marker index (SIM). Multivariate regression analyses were used to measure the associations between systemic inflammatory indices and overall and disease-free survival as a primary outcome and occurrence of postoperative complications as a secondary outcome measure.

RESULTS

The sample was composed of 69 male (76.67%) and 21 female (23.33%) patients, with an average age of 61.15 ± 9.79 years. The median follow-up time was 24 months, and 73 of 91 (66.43%) patients were alive during the median follow-up. Overall disease survival correlated with systemic immune-inflammation (P = .022, cutoff >1,005.3, sensitivity 67.1%, and specificity 70.6%) and SIM (P = .0001, cutoff >4.05, sensitivity 90.4%, and specificity 41.2%), preoperative platelets (P = .036, cutoff <194, sensitivity 28.8%, and specificity 94.1%), and postoperative lymphocytes (P = .012, cutoff <0.6, sensitivity 38%, and specificity 76.5%), whereas increased SIM (P = .042, cutoff >4.05, sensitivity 91.3%, and specificity 38.1%), NLR (P = .031, cutoff >13.2, sensitivity 56.9%, and specificity 60%), and preoperative platelets (P = .006, cutoff <244, sensitivity 52.3%, and specificity 76%) were associated with adverse disease-free survival. The cumulative postoperative complication rate was 34.5%, of which 13.3% accounted for major complications, whereas derived NLR (P = .013, DF 1, χ² test 6.161, cutoff >2.3) and postoperative lymphocytes (P = .009, DF 1, χ² test 6.756, cutoff <1) correlated with occurrence of complications.

CONCLUSIONS

Inflammatory indices as measures of inflammation-related systemic dysfunction may be associated with adverse survival in patients with head and neck squamous cell carcinoma and occurrence of postoperative complications and with specific cutoff values.

Prediction models applying machine learning to oral cavity cancer outcomes: A systematic review

OBJECTIVES

Machine learning platforms are now being introduced into modern oncological practice for classification and prediction of patient outcomes. To determine the current status of the application of these learning models as adjunctive decision-making tools in oral cavity cancer management, this systematic review aims to summarize the accuracy of machine-learning based models for disease outcomes.

METHODS

Electronic databases including PubMed, Scopus, EMBASE, Cochrane Library, LILACS, SciELO, PsychINFO, and Web of Science were searched up until December 21, 2020. Pertinent articles detailing the development and accuracy of machine learning prediction models for oral cavity cancer outcomes were selected in a two-stage process. Quality assessment was conducted using the Quality in Prognosis Studies (QUIPS) tool and results of base studies were qualitatively synthesized by all authors. Outcomes of interest were malignant transformation of precancer lesions, cervical lymph node metastasis, as well as treatment response, and prognosis of oral cavity cancer.

RESULTS

Twenty-seven articles out of 950 citations identified from electronic and manual searching were included in this study. Five studies had low bias concerns on the QUIPS tool. Prediction of malignant transformation, cervical lymph node metastasis, treatment response, and prognosis were reported in three, six, eight, and eleven articles respectively. Accuracy of these learning models on the internal or external validation sets ranged from 0.85 to 0.97 for malignant transformation prediction, 0.78-0.91 for cervical lymph node metastasis prediction, 0.64-1.00 for treatment response prediction, and 0.71-0.99 for prognosis prediction. In general, most trained algorithms predicting these outcomes performed better than alternate methods of prediction. We also found that models including molecular markers in training data had better accuracy estimates for malignant transformation, treatment response, and prognosis prediction.

CONCLUSION

Machine learning algorithms have a satisfactory to excellent accuracy for predicting three of four oral cavity cancer outcomes i.e., malignant transformation, nodal metastasis, and prognosis. However, considering the training approach of many available classifiers, these models may not be streamlined enough for clinical application currently.

Perineural invasion, lactate dehydrogenase, globulin, and serum sodium predicting occult metastasis in oral cancer

OBJECTIVE

This study aimed to develop a nomogram to predict the neck occult metastasis in early (T1-T2 cN0) oral squamous cell carcinoma (OSCC).

MATERIALS AND METHODS

The nomogram was developed in a training cohort of 336 early OSCC patients and was validated in a validation cohort including 88 patients. Independent predictors were calculated by univariate and multivariate logistic regression analyses.

RESULTS

In univariate logistical regression analysis, gender, perineural invasion (PNI), blood vessel invasion, mean corpuscular hemoglobin, aspartate aminotransferase, prealbumin, globulin (GLO), lactate dehydrogenase (LDH), serum sodium (NA), and serum chloride were significant associated with neck occult metastasis. Multivariate logistical regression analysis identified PNI (p < .001), LDH (p = .003), GLO (p = .019), and NA (p = .020) as independent predictors of neck occult metastasis. Cut-off values for LDH, GLO, and NA obtained from AUC were 142.5, 26.35, and 139.5, respectively. The nomogram based on PNI and categorical GLO, LDH, and NA exhibited a strong discrimination, with a C-indexes of 0.748 (95%CI = 0.688 to 0.810) in the training cohort and 0.751 (95%CI = 0.639 to 0.863) in the validation cohort.

CONCLUSIONS

A nomogram based on PNI, LDH, GLO, and NA for predicting the risk of neck lymph nodes occult metastasis in OSCC could help surgeons with therapy decision-making.