Oral Oncology xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Oral Oncology

journal homepage: www.elsevier .com/locate /ora loncology

The benefit of pretreatment esophageal screening with image-enhancedendoscopy on the survival of patients with hypopharyngeal cancer

Wen-Lun Wang a,i, Cheng-Ping Wang b,i, Hsiu-Po Wang c, Ching-Tai Lee a, Chi-Yang Chang a,Chi-Ming Tai a, Cheng-Hao Tseng a, Tzer-Zen Hwang d, Chih-Chun Wang d, Jo-Lin Lo e, Ping-Huei Tseng c,Han-Mo Chiu c, Jang-Ming Lee f, Jenq-Yuh Ko b, Pei-Jen Lou b, Ming-Shiang Wu c, Yi-Chia Lee c,g,⇑,Jaw-Town Lin a,c,h,⇑a Department of Internal Medicine, E-Da Hospital/I-Shou University, Kaohsiung, Taiwanb Department of Otolaryngology, College of Medicine, National Taiwan University, Taipei, Taiwanc Department of Internal Medicine, College of Medicine, National Taiwan University, Taipei, Taiwand Department of Otolaryngology, E-Da Hospital/I-Shou University, Kaohsiung, Taiwane Department of Oncology, E-Da Hospital/I-Shou University, Kaohsiung, Taiwanf Department of Surgery, College of Medicine, National Taiwan University, Taipei, Taiwang Division of Biostatistics, Graduate Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwanh School of Medicine, Fu Jen Catholic University, New Taipei City, Taiwan

a r t i c l e i n f o

Article history:Received 12 February 2013Received in revised form 17 April 2013Accepted 26 April 2013Available online xxxx

Keywords:Hypopharyngeal cancerSynchronous neoplasmsEsophageal cancerSurvivalImage-enhanced endoscopy

1368-8375/$ - see front matter � 2013 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.oraloncology.2013.04.009

⇑ Corresponding authors. Addresses: Department ofMedicine, National Taiwan University Hospital, 7, ChTaiwan. Tel.: +886 2 23123456x63351; fax: +886 2 2ment of Internal Medicine, College of Medicine,Hospital, 7, Chung-Shan South Road, Taipei, TaiwanCatholic University, New Taipei City, Taiwan; DepaE-Da Hospital/I-Shou University, Kaohsiung, Taiwan. Tfax: +886 2 23947899 (J.-T. Lin).

E-mail addresses: yichialee@ntu.edu.tw (Y.-C.(J.-T. Lin).

i These authors are contributed equally to this work

Please cite this article in press as: Wang W-L etpatients with hypopharyngeal cancer. Oral Onco

s u m m a r y

Background: Synchronous esophageal cancers can suppress the survival of patients with hypopharyngealcancers. Esophageal screening with the image-enhanced endoscopy may identify more synchronous can-cers while there is no evidence to support its benefit on survival.Methods: A total of 281 and 320 patients were diagnosed with hypopharyngeal cancer before and afterthe policy of routine esophageal screening. Primary outcome measures were overall survival.Results: Among those who received screening, 49 patients (49/180, 27.2%) had synchronous esophagealcancers; treatment planning was changed in 42 (23.3%). Before and after the policy, percentages of stageI–II synchronous cancers were 20% (3/15) and 53.1% (26/49), respectively. Adjunctive therapies for syn-chronous cancers have led to a better survival after the policy than before (P = 0.002). The Cox regressionmodel quantified a survival benefit of 29% (95% CI: 11–43%) when adjusting for TNM stage of hypopha-ryngeal cancer. In post-policy period, the survival was better for those who chose screening than thosewho did not (HR: 0.57, 95% CI: 0.41–0.79). Among those without screening, there was no differencebetween the pre- and post-policy periods (HR: 0.96, 95% CI: 0.74–1.26).Conclusions: Patients with hypopharyngeal cancers may benefit from the esophageal screening withimage-enhanced endoscopy through the better detection of early-stage synchronous cancers.

� 2013 Elsevier Ltd. All rights reserved.

Introduction

Head and neck cancer represent one of the most commonmalignancies in the world. The incidence continues to increase in

ll rights reserved.

Internal Medicine, College ofung-Shan South Road, Taipei,3412775 (Y.-C. Lee). Depart-National Taiwan University; School of Medicine, Fu Jenrtment of Internal Medicine,el.: +886 2 23123456x65695;

Lee), jawtown@ntu.edu.tw

.

al. The benefit of pretreatmentl (2013), http://dx.doi.org/10.1

geographic areas where the use of tobacco, alcohol, and betel nutis prevalent. Despite multidisciplinary treatment, including sur-gery, radiotherapy, and chemotherapy, the prognoses remainpoor.1,2 Patients with head and neck cancers, especially those withhypopharyngeal cancer, often demonstrate the concept of fieldcancerization,3,4 and develop simultaneous esophageal tumors.5–8

This phenomenon may partly explain the poor survival, despite ad-vances in treatment for primary tumor control. Without pretreat-ment endoscopic screening, these concomitant esophagealtumors can become symptomatic and even metastasize, whichsuppresses the overall survival.9

Currently, the image-enhanced endoscopy, such as narrow-band imaging and Lugol chromoendoscopy, has substantially im-proved the ability to detect synchronous esophageal tumors(�30%).8,10–12 However, despites the potential to alter treatment

esophageal screening with image-enhanced endoscopy on the survival of016/j.oraloncology.2013.04.009

2 W.-L. Wang et al. / Oral Oncology xxx (2013) xxx–xxx

planning, identification of synchronous lesions does not necessar-ily correlate with improved survival (i.e., lead time bias). Therefore,it is crucial to provide solid evidence to support the routine appli-cation of this procedure.

In 2007, routine administration of image-enhanced endoscopicscreening for patients with hypopharyngeal tumors (prior to theinitiation of treatment) was begun in two medical centers inTaiwan. The primary aim of this study was to compare the survivalrates of patients between the pre-policy (2003–2006) and post-policy (2007–2010) periods and to quantify the survival benefit re-lated to the screening intervention and adjunctive therapies. Thefindings of our longitudinal cohort study were further validatedby adjusting for the improvement in the quality of cancer treat-ment and care over time.

Materials and methods

Patients

Patients with hypopharyngeal cancer were enrolled from theNational Taiwan University Hospital and E-DA Hospital between2003 and 2010. All subjects were diagnosed histologically withsquamous cell carcinoma. Their demographic characteristics, socialhabits, and medical histories were obtained from the medical re-cords or by face-to-face interviews. Survival was determined fromthe medical records or by telephone contact when appropriate. Forpatients with no death event, the follow-up time was calculated asthe time from the first diagnosed date to the December 31, 2011.The hospitals’ ethics committees approved the study protocol(Nos.: 201104007RC and EMRP-100-039).

Implementation of the screening strategy

In 2007, a strategy was initiated to routinely screen the esoph-agus with image-enhanced endoscopy at both hospitals. All

Figure 1 The study flow of the esophageal screening with image-enhanced endoscopy. Uwas seen over right hypopharynx (A); white-light imaging could not identify the early esunstained area was found after being sprayed with Lugol’s solution (D). Histology confirmtransnasal endoscopy and the magnifying endoscopy are shown in Supplemental Fig. 3A

Please cite this article in press as: Wang W-L et al. The benefit of pretreatmentpatients with hypopharyngeal cancer. Oral Oncol (2013), http://dx.doi.org/10.1

patients received endoscopic screening with traditional white-light, narrow-band image (Evis Lucera Spectrum video imagingsystem, Olympus Optical Co.) and Lugol staining (spray of 3% Lu-gol’s solution) and the choice of endoscopic instrument dependedupon the severity of the hypopharyngeal tumor-related esophagealobstruction and/or airway compromise (Fig. 1 and SupplementalFig. 3). As previously reported, almost all patients could tolerateendoscopic screening without significant complications.10–12 Aflowchart (Supplemental Fig. 1) was designed to aid further treat-ment planning based on the presence of a synchronous esophagealtumor. Clinical staging for hypopharyngeal and esophageal cancerwas determined by computed tomography, magnetic resonanceimaging, and/or positron emission tomography following theTNM classification system recommended by the American JointCommittee on Cancer.13 Positron emission tomography was con-sidered for patients with stage III–IV hypopharyngeal cancers andfor those with esophageal cancers if no evidence of M1 disease, fol-lowing the National Comprehensive Cancer Network guidelines.14

The traditional approach involved rigid laryngoscopy undergeneral anesthesia and was performed mainly for hypopharyngealtumor biopsy; esophageal examination was symptom-oriented.While the new approach began with a non-sedative esophagealscreening with hypopharyngeal tumor biopsy as indicated. If endo-scopic screening did not reveal synchronous esophageal cancer,treatment for the hypopharyngeal cancer would be the same withthe traditional approach, which would in general follow the Na-tional Comprehensive Cancer Network guidelines. If a synchronousesophageal cancer was identified, the treatment planning wouldthen include therapies for both the esophageal and hypopharyn-geal cancers. Minimally invasive treatments, such as endoscopicmucosectomy or submucosal dissection, could be considered forearly esophageal cancer.15,16 If the synchronous esophageal cancerwas in an advanced stage, concomitant esophagectomy and laryn-gectomy or definitive chemoradiotherapy with platinum-basedregimens for both hypopharyngeal and esophageal cancers would

sing a standard endoscopy (outer diameter: 9.2 mm), a tumor with uneven surfaceophageal tumor (B); a brownish area was found under narrow-band imaging (C); aned both lesions were squamous cell carcinoma. Other representative cases using the

and B.

esophageal screening with image-enhanced endoscopy on the survival of016/j.oraloncology.2013.04.009

W.-L. Wang et al. / Oral Oncology xxx (2013) xxx–xxx 3

be arranged. Patients with distant metastasis and/or poor perfor-mance status were candidates for systemic chemotherapy and/orpalliative care only.

Statistical analysis

For descriptive findings, the quantitative data were presentedas mean ± standard deviation (SD), and categorical variables pre-sented as percentages. We compared the baseline characteristicsof patients in the pre- (2003–2006) and post-policy (2007–2010)periods using the Student t test or v2 test when appropriate. Toevaluate the effect of pretreatment esophageal screening, the sur-vival rates of patients before and after the implementation of thescreening policy were compared. The cumulative survival withtime was shown with the Kaplan–Meier curves, and the statisticaldifferences between the pre- and post-policy periods were evalu-ated using the log-rank test. We also confirmed the generalizabilityof our findings by stratifying our data according to two differentinstitutions.

To quantify the effect of pretreatment esophageal screening, weused the Cox proportional hazards model to analyze the time-to-event data, right censoring at the last day of follow-up and adjust-ing for common prognostic factors, which included age, gender,difference in institutions, and the pretreatment TNM stage of can-cers. Following univariate analyses, we used the backward elimina-tion method to identify statistically significant predictors andformulate a multivariate Cox regression model. The survival bene-fit (or the absolute risk reduction) provided by pretreatmentesophageal screening could be calculated as: 1 � (adjustedHR) � 100%. Statistical significance was assigned to a two-side Pvalue < 0.05. The statistical analyses were performed using SPSSsoftware (SPSS for Windows, version 18.0; SPSS Inc, Chicago, IL).

As our study used a before-and-after design, rather than a ran-domized controlled study, the effect of pretreatment esophagealscreening had the potential to be confounded by the improvementin the quality of cancer treatment and care over time.17 To confirmour screening effect, several additional analyses were performed.First, because some of our patients did not obtain endoscopicscreening, we compared the survival rates of patients with andwithout pretreatment esophageal screening during the post-policyperiod. We assumed that, if there was a survival difference be-tween these two groups, this difference could not be explainedby the change in quality of cancer treatment and care, because theywere within the same time frame. Second, we compared the sur-vival rates of patients without pretreatment esophageal screeningbefore and after the screening policy using the Cox regression anal-ysis. We considered that, if the quality of cancer treatment and

Table 1Clinical data of patients with hypopharyngeal cancer before and after the policy of endosc

Characteristics Before the policy (n = 281)

Age (mean ± SD, range), yrs 55.5 ± 11.4 (32–89)Male gender 273 (97%)

HSCC stageStage I 11 (4%)Stage II 23 (9%)Stage III 26 (10%)Stage IVA 164 (61%)Stage IVB 27 (10%)Stage IVC 16 (6%)

Synchronous ESCC 15 (5.3%)

ESCC stageStage I–II 3 (1.1%)Stage III–IV 12 (4.2%)

Abbreviation: SD, standard deviation; ESCC, esophageal squamous cell carcinoma; HSCCThe stage of HSCC was based on the TNM system; 14 patients before the policy and 15

Please cite this article in press as: Wang W-L et al. The benefit of pretreatmentpatients with hypopharyngeal cancer. Oral Oncol (2013), http://dx.doi.org/10.1

care did improve with time, the survival should be better for thosewho were diagnosed and treated after 2007. Third, we searched thedatabase of the Taiwan Cancer Registry and the Department ofHealth to examine the time trends in the survival rate and mortal-ity rate of the upper aerodigestive tract cancer for the whole pop-ulation in Taiwan.18 We supposed such time trends might partiallyreflect the quality of cancer treatment and care over time andcould be used as an external control to confirm the effectivenessof our intervention.

Finally, to exclude the possibility of lead-time bias related toroutine endoscopic screening, we evaluated the impact of screen-ing-detected esophageal cancer on the survival. We assumed thatthe survival rate of patients with synchronous stage I–II esopha-geal cancer might be possibly close to that of those who had nosynchronous cancer because esophageal tumor diagnosed at anearlier stage is potentially curable.15,16 By contrast, the identifica-tion of a synchronous stage III–IV esophageal cancer might haveno survival benefit.

Results

Comparing the survival rates of patients before and after the policy

A total of 601 patients with hypopharyngeal cancers wererecruited. The mean follow-up time was 38.9 months (range:1–96 months) and 30.4 months (range: 1–60 months) for the pre-and post-policy periods, respectively. Before (n = 281) and after(n = 320) the implementation of screening policy, there were 215(76.5%) and 151 (47.2%) patients, respectively, who died from hyp-opharyngeal and/or esophageal cancer, there were 10 (3.6%) and 17(5.3%) patients, respectively, who were lost to follow-up, and therewere 56 (25.9%) and 152 (47.5%) patients, respectively, who re-mained alive at the end of follow-up. The clinical characteristicsare showed in Table 1. There were no significant differences inage, gender, and pretreatment TNM stage before and after thescreening policy.

In the pre-policy period, there were 15 patients diagnosed withsynchronous esophageal tumors based on the symptom-orientedapproach; 12 of them (12/15, 80%) were advanced in stage. Whilein the post-policy period, 49 patients had synchronous esophagealtumors identified during routine endoscopic screening.

After implementation of the screening policy, patients withhypopharyngeal cancer had significantly better survival than thosepatients diagnosed before the policy (Fig. 2). This difference wassimilarly found in both institutions. The results of the Cox propor-tional hazards models are shown in Table 2. In addition to the TNM

opic screening (n = 601).

After the policy (n = 320) P value

55.3 ± 11.6 (32–92) 0.81305 (95%) 0.29

20 (7%) 0.2031 (10%) 0.5331 (10%) 0.87163 (53%) 0.0643 (14%) 0.3417 (6%) 0.8349 (15.3%) <0.001

26 (8.1%) <0.00123 (7.2%) <0.001

, hypopharyngeal squamous cell carcinoma.after the policy were undetermined.

esophageal screening with image-enhanced endoscopy on the survival of016/j.oraloncology.2013.04.009

Figure 2 The Kaplan–Meier survival curves of patients before and after theimplementation of the screening policy (v2

log-rank: 9.92, P = 0.002).

4 W.-L. Wang et al. / Oral Oncology xxx (2013) xxx–xxx

stage of hypopharyngeal cancer (adjusted HR: 1.69, 95% CI: 1.29–2.22, P < 0.001), the pretreatment esophageal screening (adjustedHR: 0.71, 95% CI: 0.57–0.89, P = 0.003) was associated with a signif-icant reduction in the mortality rate, yielding a substantial benefitof 29% in survival (95% CI: 11–43%).

Adjusting for the improvement in the quality of cancer treatment andcare over time

To verify the benefit of pretreatment endoscopic screening, westratified the patients who were diagnosed after initiation of thescreening policy into those with (n = 180) and without endoscopicscreening (n = 140). Table 3 showed that there was no significantdifference in the age, gender, and TNM stage between these two

Table 2The Cox proportional hazards models to evaluate predictors associated with the survival

Predictors Categories

Univariate analysisAge, yrs >50 vs. 650Gender Male vs. femaleInstitution EDA vs. NTUHHSCC stage IVA–C vs. I–IIIScreening policy After vs. before policy (overall)

After vs. before policy (non-screened)

Multivariate analysisHSCC stage IVA–C vs. I–IIIScreening policy After vs. before policy (overall)

Abbreviation: HR, hazard ratio; CI, confidence interval; EDA, E-DA hospital; NTUH, Nati

Table 3Clinical data of patients with and without pretreatment esophageal screening in the post-

Characteristics Without screening (n = 140)

Age (mean ± SD, range), yrs 56 ± 11.7 (34–91)Males 132 (94%)HSCC stage

Stage I 8 (6%)Stage II 13 (10%)Stage III 15 (12%)Stage IVA 63 (50%)Stage IVB 20 (15%)Stage IVC 9 (7%)

Synchronous ESCC –

ESCC stageStage I–II –Stage III–IV –

Abbreviation: SD, standard deviation; ESCC, esophageal squamous cell carcinoma; HSCCThe TNM stages of HSCC could not be determined in three patients in the screened grou

Please cite this article in press as: Wang W-L et al. The benefit of pretreatmentpatients with hypopharyngeal cancer. Oral Oncol (2013), http://dx.doi.org/10.1

groups. Among the 49 patients (49/180, 27.2%) who had synchro-nous esophageal cancers, 26 (26/49, 53.1%) were diagnosed withstage I–II esophageal cancer and 42 (42/49, 85.7%) received con-comitant therapies for both hypopharyngeal and esophageal can-cers. The remaining 7 patients (7/49, 14.3%) could only be offeredpalliative care due to poor performance status and/or distantmetastases. In summary, the original treatment planning waschanged in 42 (42/180, 23.3%) patients due to pretreatment endo-scopic screening.

In the post-policy period, among those without esophagealscreening, 14 patients (14/140, 10%) died of symptomatic esopha-geal cancer during follow-up, and 50% of them died within 1 yearof the diagnosis of esophageal cancer (median: 4 months; range:1–35 months). The Cox proportion hazards model consistentlyshowed that the pretreatment esophageal screening was associ-ated with a significant reduction in the mortality rate (adjustedHR: 0.57, 95% CI: 0.41–0.79, P = 0.001), yielding a substantial ben-efit of 43% (95% CI: 21–59%) in survival (Table 4).

We also compared the survival of those patients who did not un-dergo esophageal screening before (n = 281) and after (n = 140)implementation of the screening policy. As shown in Table 2, therewas no significant difference between these two groups of patients(HR: 0.96, 95% CI: 0.74–1.26, P = 0.78). The finding was consistentwith the time trend of survival rate for hypopharyngeal cancers ofthe whole population in Taiwan. Namely, if no routine esophagealscreening policy, the survival did not show significant improve-ment during the period of 2003–2010 (Supplemental Table 1).

The magnitude and independence of the contribution of screening-detected esophageal cancer on the survival

To evaluate the impact of synchronous esophageal cancers onthe survival, the Kaplan–Meier survival curves showed patients

rates of patients with hypopharyngeal cancer (n = 601).

Case no. HR (95% CI) P value

374/227 1.12 (0.90–1.38) 0.31578/23 2.17 (1.08–4.38) 0.03226/375 0.88 (0.71–1.09) 0.23430/142 1.74 (1.33–2.28) <0.001320/281 0.71 (0.58–0.88) 0.002140/281 0.96 (0.74–1.26) 0.78

430/142 1.69 (1.29–2.22) <0.001320/281 0.71 (0.57–0.89) 0.003

onal Taiwan University Hospital; HSCC, hypopharyngeal squamous cell carcinoma.

policy period (n = 320).

With screening (n = 180) P value

55 ± 11.5 (32–92) 0.34173 (96%) 0.60

12 (7%) 0.8518 (10%) 1.0016 (9%) 0.45100 (56%) 0.2123 (13%) 0.658 (5%) 0.3549 (27%)

26 (53%)23 (47%)

, hypopharyngeal squamous cell carcinoma.p and 12 patients in the non-screened group.

esophageal screening with image-enhanced endoscopy on the survival of016/j.oraloncology.2013.04.009

Table 4The Cox proportional hazards models to evaluate predictors associated with the survival rates of patients with hypopharyngeal cancer in the post-policy period (n = 320).

Predictors Categories Case no. HR (95% CI) P value

Univariate analysisAge, yrs >50 vs. 650 197/123 0.98 (0.71–1.36) 0.92Gender Male vs. female 305/15 3.43 (1.09–10.76) 0.04Institution EDA vs. NTUH 154/166 0.98 (0.72–1.36) 0.92HSCC stage IVA–C vs. I–III 223/82 1.52 (1.03–2.26) 0.036Esophageal screening Yes vs. no 180/140 0.61 (0.45–0.84) 0.003

Multivariate analysisHSCC stage IVA–C vs. I–III 223/82 1.51 (1.01–2.24) 0.04Esophageal screening Yes vs. no 180/140 0.57 (0.41–0.79) 0.001

Abbreviation: HR, hazard ratio; CI, confidence interval; EDA, E-DA hospital; NTUH, National Taiwan University Hospital; HSCC, hypopharyngeal squamous cell carcinoma.

Figure 3 The Kaplan–Meier survival curves of patients with synchronous stage I–IIand stage III–IV esophageal cancer (v2

log-rank: 6.88, P = 0.009).

W.-L. Wang et al. / Oral Oncology xxx (2013) xxx–xxx 5

with stage III–IV synchronous esophageal cancers had significantlyworse survival than those with stage I–II cancers (Fig. 3). We per-formed additional Cox proportion hazards models focusing on pa-tients (n = 180) who had undergone endoscopic screening toevaluate the magnitude of screening-detected esophageal canceron the survival. As shown in Table 5, there was no significant dif-ference between those who were detected with the stage I–IIesophageal cancer and those who were free from esophageal can-cer (HR: 1.48, 95% CI: 0.80–2.75, P = 0.21). Multivariate analysesshowed that both the stage IVA–C hypopharyngeal cancer (adjustedHR: 1.93, 95% CI: 1.09–3.45, P = 0.02) and the stage III–IV esopha-geal cancer (adjusted HR: 4.59, 95% CI: 2.58–8.15, P < 0.001) weresignificantly associated with the poor survival.

Table 5The Cox proportional hazards models to evaluate the magnitude and independence of the chad undergone pretreatment endoscopic screening (n = 180).

Predictors Categories

Univariate analysisAge, yrs >50 vs. 650Gender Male vs. femaleInstitution EDA vs. NTUHHSCC stage IVA–C vs. I–IIIEarlier ESCC stage I–II vs. no ESCCAdvanced ESCC stage III–IV vs. I–II and no

Multivariate analysisHSCC stage IVA–C vs. I–IIIAdvanced ESCC stage III–IV vs. I–II and no

Abbreviation: HR, hazard ratio; CI, confidence interval; EDA, E-DA hospital; NTUH, NatiESCC, esophageal squamous cell carcinoma.The TNM stages of HSCC could not be determined in three patients in the screening groumodel.

Please cite this article in press as: Wang W-L et al. The benefit of pretreatmentpatients with hypopharyngeal cancer. Oral Oncol (2013), http://dx.doi.org/10.1

Discussion

The present study not only confirmed a high prevalence of syn-chronous esophageal cancer in patients with hypopharyngeal can-cer when image-enhanced endoscopy was applied, but alsodemonstrated that the strategy of routine pretreatment esophagealscreening could significantly improve survival.

The occurrence of a second primary esophageal cancer, whensymptomatic, was associated with a dismal prognosis, and only�20% of patients survived for 2 years.9,19 Therefore, we highlightedthe importance of ‘‘routine’’ esophageal screening, rather than‘‘symptom-oriented’’. By routine screening, our detection rate ofsynchronous esophageal neoplasm (27.2%) was much higher thanprevious reports.20–23 Notably, a majority of the synchronousesophageal cancers (53.1%; 26 out of 49) were earlier in stage,compared to 20% (3 out of 15) of those with symptom-orientedstrategy in the pre-policy period. Although one previous studyshowed contradictory results in regard to the benefit of routineendoscopy due to a low yield of second primary tumors,24 weshould emphasize the recent advance of image-enhanced endos-copy, which could substantially increase the detection rate ofearly-stage esophageal cancer than before. The 2-year and 3-yearsurvival rates for patients with synchronous esophageal cancer inour cohort were 50% and 40%, respectively, which were higher thanprevious reports,9,19 and further emphasizes the importance ofearly detection. On the other hand, despite being detected withsynchronous cancer, some patients (7/180, 3.9%) could receive pal-liative therapy only and may not benefit from the routine screen-ing policy. The ratio was close to the detection rate (15/281,5.3%) of synchronous cancers during the pre-policy period, sug-gesting that lead-time bias has indeed occurred when the endo-scopic screening was only based on the presence of overtdysphagic symptom.

ontribution of synchronous esophageal cancer staging on the survival of patients who

Case no. HR (95% CI) P value

109/71 1.08 (0.69–1.71) 0.73173/7 1.78 (0.44–7.27) 0.4297/83 0.81 (0.52–1.27) 0.35131/46 1.56 (0.90–2.70) 0.1226/131 1.48 (0.80–2.75) 0.2123/157 3.71 (2.16–6.35) <0.001

131/46 1.93 (1.09–3.45) 0.0223/157 4.59 (2.58–8.15) <0.001

onal Taiwan University Hospital; HSCC, hypopharyngeal squamous cell carcinoma;

p. No significant interaction was found in the multivariate Cox proportional hazards

esophageal screening with image-enhanced endoscopy on the survival of016/j.oraloncology.2013.04.009

6 W.-L. Wang et al. / Oral Oncology xxx (2013) xxx–xxx

Only a few studies have focused on the strategies for patientswith advanced cancers at both sites (see Supplemental Table 2).Although concomitant surgery has ever been recommended, highmorbidity and mortality remain as serious concerns. In our pa-tients (16/23, �70%) who underwent treatment for synchronousadvanced cancers, 10 received concomitant surgery and 6 receivedconcomitant chemoradiotherapy. Their 3-year survival rates wereonly 25% and 33%, respectively, and there were no significant dif-ferences (Supplemental Fig. 2). However, the sample size wassmall, and the study design was not randomized.

Our study has some limitations. First, our study was not a ran-domized controlled trial, but ethical considerations precluded usfrom adopting such a study design because esophageal tumorswere indeed prevalent among them. Instead, we carefully inter-preted our findings by adjusting for the effect of prognostic factors,considering the improvement in the quality of cancer treatmentand care over time, and excluding the possibility of lead time bias.However, we cannot exclude the possibility that the potential ef-fects of co-morbidities as a variable were not considered in ourstudy and the advances in treatment over the later study periodthat may have influenced the results. Second, we may underesti-mate the benefit by comparing the pre-policy with the post-policyperiod because some did not receive screening. In addition, with-out randomized allocation, we may overestimate the benefit bycomparing the screened with the non-screened group followingpolicy. With a retrospective study design, we did not collect thereasons why patients did not comply to the endoscopic examina-tion, such as patients too frail, physicians not request, and patientshaving already performed other radiological examinations that notindicated the need for endoscopic screening. However, using dif-ferent sizes of endoscopy tailored to the individual patient (Supple-mental Fig. 3), the screening was indeed tolerable for almost allpatients.10–12,25,26 Finally, using our study design, a shorter fol-low-up was inevitable for cases recruited during the post-policyperiod. However, these data were mostly non-informative due tothe reach of our predetermined observation time. In addition, suchcancers showed the steepest decline of survival within 2 years offollow-up (Supplemental Fig. 4), thus our follow-up time shouldbe sufficient. However, the long-term benefit should be verifiedby continuous follow-up. Further clinical trials are also warrantedto verify our findings.

In conclusion, synchronous esophageal cancer is common in pa-tients with hypopharyngeal cancer when an endoscopic screeningstrategy is implemented. The effort to find synchronous cancers isworthwhile because it is potentially associated with survival ben-efit. For patients with hypopharyngeal cancer, such a screeningmay be considered as a part of staging workup.

Conflict of interest statement

None declared.

Acknowledgements

This study was supported by research grants from the NationalScience Council (NSC96-2314-B-002-092-MY3), E-Da Hospital(EDAHP98022), and the Taipei Institute of Pathology.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.oraloncology.2013.04.009.

Please cite this article in press as: Wang W-L et al. The benefit of pretreatmentpatients with hypopharyngeal cancer. Oral Oncol (2013), http://dx.doi.org/10.1

References

1. Taiwan Cancer Registry (1972–2008). Department of Health, Executive Yuan.Cancer Registry Annual Report. <http://crs.cph.ntu.edu.tw>.

2. National Cancer Institute: Surveillance, Epidemiology, and End Results, SEERCancer Statistics Review, 1975–2009. <http://seer.cancer.gov/about/overview.html> [accessed 16.04.13].

3. Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratifiedsquamous epithelium; clinical implications of multicentric origin. Cancer1953;6(5):963–8.

4. Lee YC, Wang HP, Wang CP, Ko JY, Lee JM, Chiu HM, et al. Revisit of fieldcancerization in squamous cell carcinoma of upper aerodigestive tract: betterrisk assessment with epigenetic markers. Cancer Prev Res (Phila)2011;4(12):1982–92.

5. Haughey BH, Gates GA, Arfken CL, Harvey J. Meta-analysis of second malignanttumors in head and neck cancer: the case for an endoscopic screening protocol.Ann Otol Rhinol Laryngol 1992;101(2 Pt 1):105–12.

6. Morris LG, Sikora AG, Patel SG, Hayes RB, Ganly I. Second primary cancers afteran index head and neck cancer: subsite-specific trends in the era of humanpapillomavirus-associated oropharyngeal cancer. J Clin Oncol2011;29(6):739–46.

7. Muto M, Hironaka S, Nakane M, Boku N, Ohtsu A, Yoshida S. Association ofmultiple Lugol-voiding lesions with synchronous and metachronousesophageal squamous cell carcinoma in patients with head and neck cancer.Gastrointest Endosc 2002;56(4):517–21.

8. Wang WL, Lee CT, Lee YC, Hwang TZ, Wang CC, Hwang JC, et al. Risk factors fordeveloping synchronous esophageal neoplasia in patients with head and neckcancer. Head Neck 2011;33(1):77–81.

9. Leon X, Ferlito A, Myer 3rd CM, Saffiotti U, Shaha AR, Bradley PJ, et al. Secondprimary tumors in head and neck cancer patients. Acta Otolaryngol2002;122(7):765–78.

10. Lee YC, Wang CP, Chen CC, Chiu HM, Ko JY, Lou PJ, et al. Transnasal endoscopywith narrow-band imaging and Lugol staining to screen patients with head andneck cancer whose condition limits oral intubation with standard endoscope(with video). Gastrointest Endosc 2009;69(3 Pt 1):408–17.

11. Lee CT, Chang CY, Lee YC, Tai CM, Wang WL, Tseng PH, et al. Narrow-bandimaging with magnifying endoscopy for the screening of esophageal cancer inpatients with primary head and neck cancers. Endoscopy 2010;42(8):613–9.

12. Wang CP, Lee YC, Yang TL, Lou PJ, Ko JY. Application of unsedated transnasalesophagogastroduodenoscopy in the diagnosis of hypopharyngeal cancer. HeadNeck 2009;31(2):153–7.

13. Greene FL, Page DL, Fleming ID. AJCC cancer staging manual. 6th ed. New York,NY: Springer-Verlag; 2002. p. 31–46 and 91–98.

14. National Comprehensive Cancer Network Clinical Practice Guidelines inOncology. <http://www.nccn.org> [accessed 16.04.13].

15. Lee CT, Chang CY, Tai CM, Wang WL, Tseng CH, Hwang JC, et al. Endoscopicsubmucosal dissection for early esophageal neoplasia: a single centerexperience in South Taiwan. J Formos Med Assoc 2012;111(3):132–9.

16. Ono S, Fujishiro M, Niimi K, Goto O, Kodashima S, Yamamichi N, et al. Long-term outcomes of endoscopic submucosal dissection for superficial esophagealsquamous cell neoplasms. Gastrointest Endosc 2009;70(5):860–6.

17. Cook TD, Campbell DT. Quasi-Experiments: interrupted time-series designs. In:quasi-experimentation: design & analysis issues for field settings. HoughtonMifflin Co, Boston; 1979. p. 207–232.

18. Department of Health EY, Taiwan. Cause of Death Statistics. The CollaborationCenter of Health Information Application (CCHIA); 2010.

19. Chen MC, Huang WC, Chan CH, Chen PT, Lee KD. Impact of second primaryesophageal or lung cancer on survival of patients with head and neck cancer.Oral Oncol 2010;46(4):249–54.

20. Benninger MS, Shariff A, Blazoff K. Symptom-directed selective endoscopy:long-term efficacy. Arch Otolaryngol Head Neck Surg 2001;127(7):770–3.

21. Rennemo E, Zatterstrom U, Boysen M. Synchronous second primary tumors in2,016 head and neck cancer patients: role of symptom-directed panendoscopy.Laryngoscope 2011;121(2):304–9.

22. Lee KD, Lu CH, Chen PT, Chan CH, Lin JT, Huang CE, et al. The incidence and riskof developing a second primary esophageal cancer in patients with oral andpharyngeal carcinoma: a population-based study in Taiwan over a 25 yearperiod. BMC Cancer 2009;9:373.

23. Warnakulasuriya KA, Robinson D, Evans H. Multiple primary tumours followinghead and neck cancer in southern England during 1961–98. J Oral Pathol Med2003;32(8):443–9.

24. Davidson J, Gilbert R, Irish J, Witterick I, Brown D, Birt D, et al. The role ofpanendoscopy in the management of mucosal head and neck malignancy-aprospective evaluation. Head Neck 2000;22(5):449–54.

25. Takenaka R, Kawahara Y, Okada H, Hori K, Inoue M, Kawano S, et al. Narrow-band imaging provides reliable screening for esophageal malignancy inpatients with head and neck cancers. Am J Gastroenterol 2009;104(12):2942–8.

26. Muto M, Minashi K, Yano T, Saito Y, Oda I, Nonaka S, et al. Early detection ofsuperficial squamous cell carcinoma in the head and neck region andesophagus by narrow band imaging: a multicenter randomized controlledtrial. J Clin Oncol 2010;28(9):1566–72.

esophageal screening with image-enhanced endoscopy on the survival of016/j.oraloncology.2013.04.009