Discussion
The occurrence of SPMs is a crucial issue in the treatment of HPSCC. Previous observations have shown that the frequency of SPMs vary from 7-36%, according to the anatomical site of original primary. In particular, second primary ESCNs are more frequently found in patients with HPSCC. This phenomenon is supported by the concept of field carcinogenesis.4-6,9,15-18The average prevalence of second primary ESCNs in patients with HPSCC in a recent systemic review was 15.2% (413 of 3386, 95% CI: 11.4-19.0).19 In terms of sublocation, the average prevalence of esophageal lesions screened both synchronously and metachronously in patients with hypopharyngeal tumors of seven studies was 28.0% (161 of 574, 95% CI: 22.5-33.5).
Our study analyzed 99 patients with HPSCC who received IEE screening of the upper gastrointestinal tract. We used the triple endoscopy combination of conventional white-light endoscopy (WLE), narrow-band imaging with magnification (NBI-M) and Lugol chromoendoscopy to detect suspicious ESCNs. The prevalence of second primary ESCNs in the ND group was 27%, and 33% in the FU group. Our results were similar to those obtained studies and represented that relatively high chance of synchronous and metachronous esophageal malignancies could be detected through IEE compared with traditional white light endoscopy.20
In the ND group, 8 of the 30 patients (27%) was found second primary ESCNs in the IEE screening with varies histology, including severe dysplasia, carcinoma in situ, and SCC. It is essential to evaluate the UADT before definitive treatments and may change the treatment planning.18
SPMs have a significant effect on the survival of patients with primary head and neck cancers, and are a major threat to the morbidity and mortality of patients with HPSCC after treatment. Second primary esophageal cancer, especially, is known to be associated with a poor prognosis.21,22 Patients with second primary ESCNs have significantly lower survival rates (HR 2.75/2.79, 95% CI 1.11~6.82/1.15~6.80, p = 0.03/0.02 in multivariate analyses) than those without second primary ESCNs.23
In the FU group, 23 of the 69 patients (33%) had second primary ESCNs in the IEE examination. According to previous studies,24,25 the risk analysis of clinical parameters revealed that alcohol exposure and N3 disease of HPSCC were the most important independent risk predictors for simultaneous esophageal lesions. Some studies showed that an age of under 50 years was also a risk predictor on univariate analysis as well.26,27Our study shows only those with previous history of SPMs-UADT had higher incidence of second primary ESCN than those without the history (35% vs. 13%, p =0.0343). Therefore, routine IEE screening of the esophagus is recommended for HPSCC patients, especially at those with previous history of SPMs in the UADT.
Most oncologists agree that early diagnosis and treatment are the best way to manage SPMs in head and neck cancer. Imaged-enhanced endoscopy has become a useful screening tool for precancerous or early cancerous lesions in the esophagus by means of dye- or optical-based techniques.28 When using the combination of chromoendoscopy with Lugol solution and NBI system with high-resolution ME, the margin and invasiveness of the neoplasia can be well delineated and predicted.28-30
More and more researchers have suggested that routine screening of ESCNs in patients with head and neck cancer is recommended. In clinical practice, however, a comprehensive UGI endoscopy may not be possible in treated HPSCC patients because of local tumor obstruction or structural changes after surgery or irradiation. It is important to find out the characteristics of high-risk second primary ESCNs occurrence groups for more detailed surveillance arrangement. Also, it is also important to know the role of endoscopy in patients without dysphagia or other GI symptoms. In our study of the FU group, there was significant difference in clinical T-classification and treatment modalities of ESCNs between patients with symptoms and those without symptoms; patients without symptoms at the time of UGI screening had less advanced T stages (44% vs. 0%, p = 0.0359), and had a higher percentage of non-minimal invasive therapy (75% vs. 0%, p = 0.0013).