Background: Due to the occult anatomic location of the nasopharynx and frequent presence of adenoid hyperplasia, the positive rate for malignancy identification during biopsy is low, thus leading to delayed or missed diagnosis for nasopharyngeal malignancies upon initial attempt. Here, we aimed to develop an artificial intelligence tool to detect nasopharyngeal malignancies under endoscopic examination based on deep learning. Methods: An endoscopic images‐based nasopharyngeal malignancy detection model (eNPM‐DM) consisting of a fully convolutional network based on the inception architecture was developed and fine‐tuned using separate training and validation sets for both classification and segmentation. Briefly, a total of 28,966 qualified images were collected. Among these images, 27,536 biopsy‐proven images from 7951 individuals obtained from January 1st, 2008, to December 31st, 2016, were split into the training, validation and test sets at a ratio of 7:1:2 using simple randomization. Additionally, 1430 images obtained from January 1st, 2017, to March 31st, 2017, were used as a prospective test set to compare the performance of the established model against oncologist evaluation. The dice similarity coefficient (DSC) was used to evaluate the efficiency of eNPM‐DM in automatic segmentation of malignant area from the background of nasopharyngeal endoscopic images, by comparing automatic segmentation with manual segmentation performed by the experts. Results: All images were histopathologically confirmed, and included 5713 (19.7%) normal control, 19,107 (66.0%) nasopharyngeal carcinoma (NPC), 335 (1.2%) NPC and 3811 (13.2%) benign diseases. The eNPM‐DM attained an overall accuracy of 88.7% (95% confidence interval (CI) 87.8%–89.5%) in detecting malignancies in the test set. In the prospective comparison phase, eNPM‐DM outperformed the experts: the overall accuracy was 88.0% (95% CI 86.1%–89.6%) vs. 80.5% (95% CI 77.0%–84.0%). The eNPM‐DM required less time (40 s vs. 110.0 ± 5.8 min) and exhibited encouraging performance in automatic segmentation of nasopharyngeal malignant area from the background, with an average DSC of 0.78 ± 0.24 and 0.75 ± 0.26 in the test and prospective test sets, respectively. Conclusions: The eNPM‐DM outperformed oncologist evaluation in diagnostic classification of nasopharyngeal mass into benign versus malignant, and realized automatic segmentation of malignant area from the background of nasopharyngeal endoscopic images.

Nasopharyngeal carcinoma (NPC) is a malignant neoplasm originating from the nasopharynx epithelium.¹ Platinum-based concurrent chemoradiotherapy (CRT) represents the treatment backbone for locoregionally advanced NPC. Intensified chemotherapy in the induction or adjuvant setting, added to CRT, has improved survival outcomes.² In the past decade, induction chemotherapy (IC) has regained the spotlight, with a series of randomized clinical trials demonstrating its clinical benefit.^3-8 Induction chemotherapy in combination with CRT is the new standard of care recommended by the treatment guidelines for nonmetastatic stage III to IV NPC.^9,10

Cervical lymph node metastasis is common in patients with nasopharyngeal carcinoma (NPC), but occipital lymph node metastasis in NPC patients has not yet been reported. In this case report, we describe an NPC patient with occipital lymph node metastasis. The clinical presentation, diagnostic procedure, treatment, and outcome of this case were presented, with a review of the related literature.

Background: The mortality of nasopharyngeal carcinoma (NPC) is usually lower in females than in males, but the underlying mechanism remains largely unknown. We sought to describe the age-dependent patterns of gender disparities in NPC survival and explore the extent to which the confounder or mediation effects could explain these differences. Methods: A total of 11,980 patients with NPC were reviewed. The effect of gender on cancer-specific survival (CSS) was assessed using Cox regression analyses. Two propensity score methods were conducted to control the confounding bias between genders. Restricted cubic spline regression was used to model the association of gender and age with mortality flexibly. Multiple mediation analysis was applied to estimate the direct or indirect effect of gender on CSS. Results: Overall, 7026 males and 2320 females were analyzed. The crude CSS was significantly higher for females than males (10-year CSS 78.4% vs 70.3%; P < 0.001). Similar results were observed after adjusting for confounding bias. Gender disparities in NPC-specific mortality were age-dependent, where they would increase with age until peaking at age 55–60 years and decline rapidly afterward. Subgroup analyses revealed that females’ survival advantage was observed in the 18–45 age group and was more prominent in the 46–55 age group, but vanished in the > 55 age group. Either confounder or mediation effects only accounted for approximately 20% of the gender differences. Conclusions: Gender disparities in cancer-specific mortality for patients with NPC were age-dependent. The differences mostly cannot be explained by confounder or mediation effects.

Nasopharyngeal carcinoma (NPC) is a head and neck cancer with unbalanced geographical distribution; it is particularly prevalent in Southern China, East and Southeast Asia, and North Africa.¹ Worldwide, approximately 129 000 new cases of NPC and 73 000 deaths attributed to this disease are reported annually.² The incidence of synchronous metastatic disease is about 5% in newly diagnosed patients. In addition, about 20% of patients with nonmetastatic NPC develop distant failure even after definitive treatments, which remains a main cause of death in these patients.^3,4 Taken together, approximately one-fourth of patients with NPC eventually experience metastatic diseases that require systemic therapy.

Patients with metastatic nasopharyngeal carcinoma (NPC) have variable survival outcomes. We have previously shown that an elevated peripheral blood lymphocyte-to-monocyte ratio (LMR) is associated with an increased metastatic risk in patients with primary NPC. The present study aimed to investigate the prognostic value of pretreatment LMR in a large cohort of metastatic NPC patients. Clinical data of 672 patients with metastatic NPC diagnosed between January 2003 and December 2009 were analyzed. The peripheral lymphocyte and monocyte counts were retrieved, and LMR was calculated. Receiver operating characteristic (ROC) curve analysis and univariate and multivariate COX proportional hazards analyses were performed to evaluate the association of LMR with overall survival (OS). Univariate analysis revealed that an elevated absolute lymphocyte count (≥1.390 × 109/L) and LMR (≥2.475) as well as a decreased monocyte count (<0.665 × 109/L) were significantly associated with prolonged OS. Multivariate Cox proportional hazard analysis showed that LMR (hazard ratio [HR] = 0.50, 95 % confidence interval [CI] = 0.41–0.60, P < 0.001), absolute lymphocyte count (HR = 0.77, 95 % CI = 0.64–0.93, P = 0.007), and monocyte count (HR = 1.98, 95 % CI = 1.63–2.41, P < 0.001) were independent prognostic factors. By stratification analyses, only LMR remained a significant predictor of prognosis. We identified pretreatment LMR as an independent prognostic factor for patients with metastatic NPC. Independent validation of our findings is needed.