New Techniques to Detect Unknown Primaries in Cervical Lymph Node Metastasis

The LaryngoscopeVC 2010 The American Laryngological,Rhinological and Otological Society, Inc.

New Techniques to Detect UnknownPrimaries in Cervical Lymph NodeMetastasis

Akihiro Sakai, MD; Kenji Okami, MD, PhD; Koji Ebisumoto, MD; Ryosuke Sugimoto, MD;

Daisuke Maki, MD; Masahiro Iida, MD, PhD

Objective: Various methods have been reportedfor the detection of unknown primaries in cervicallymph node metastasis. Recently, we applied new op-tical devices and modifications of endoscopic techni-ques for the detection of primary lesions, andobtained excellent results. The detection rate of thenew method was compared with that of previousmethods.

Study Design: Retrospective.Methods: A total of 51 patients with cervical

lymph node metastasis from an unknown primarysite (CUP) were referred to our hospital between Jan-uary 2000 and May 2009, and were retrospectivelyanalyzed. Between 2000 and 2005, the observation bynormal video-endoscope in straight head position wasperformed to detect the primary lesions in the outpa-tient setting. Since 2006, a new method for detectionof primary lesions has been employed. The methodincludes the use of new optical devices (hooded video-endoscope and narrow-band imaging endoscope) anddifferent head positions (head torsion technique, Val-salva maneuver, and the Killian position).

Results: The detection rate of primary lesionsusing the new method was 71% (15 of 21), which wasbetter than the 30% (10 of 30) obtained with the con-ventional method. All primary lesions identified usingthe new method were located in the hypopharynx.

Conclusions: The new method was able todetect primary lesions in 71% of cases with CUP. Thehigher rate achieved with the new method was prob-ably due to the clear visualization of the hypophar-ynx. Thus, the new method was shown to be usefulfor the detection of primary lesions in cases of CUP.

Key Words: Unknown primary, hooded videoendoscope, NBI endoscope, head positioning.

Level of Evidence: 3aLaryngoscope, 120:1779–1783, 2010

INTRODUCTIONVarious methods have been reported for the detection

of unknown primaries in cervical metastasis.1–3 However,primary lesions remain undetected in some cases. Otherapproaches are desired to obtain a higher detection rate. Asa result of recent advances in optical devices, tiny cancerouslesions arising from digestive tract mucosa have become de-tectable.4–6 Recently, we employed a new method fordetection of primary lesions in cases with CUP. This methodincluded the use of new optical devices and modifications ofendoscopic techniques such as different head positions. Inthis study, we investigated the usefulness of the newmethodby comparing it with the conventional method.

MATERIALS AND METHODSA total of 51 patients with cervical lymph node metastasis

from an unknown primary site (CUP), who were referred to ourhospital between January 2000 and May 2009, were retrospec-tively analyzed. Between 2000 and 2005, the observation bynormal video-endoscope in straight head position was performedto detect the primary lesions in the outpatient setting. Since2006, a new method using new optical devices (hooded videoendoscope7 and narrow-band imaging [NBI] endoscope) and dif-ferent head positions (head torsion technique, Valsalvamaneuver, and the Killian position9) has been employed for thedetection of primary lesions in the outpatient setting asdescribed below. After pharyngeal anesthesia with 4% lidocaine,the pharynx and larynx were examined endoscopically. The newmethod was employed in 21 patients with CUP between 2006and 2009, and the detection rate of primary lesions wasassessed. This was compared with the detection rate obtainedin 30 patients examined using the conventional video-endo-scopic method between 2000 and 2005.

Optical DevicesWe used an EH-1530T2 (Pentax, Tokyo, Japan) hooded

video endoscope (Fig. 1) and an ENF TYPE VQ (Olympus,Tokyo, Japan) NBI endoscope.

From the Department of Otolaryngology, Tokai University, School ofMedicine, Isehara, Japan.

Editor’s Note: This Manuscript was accepted for publication April20, 2010.

The authors have no financial disclosures or conflicts of intereststo disclose.

Send correspondence to Akihiro Sakai, MD, Department of Otolar-yngology Tokai University, School of Medicine, Isehara, 259-1193 Japan.E-mail: [email protected]

DOI: 10.1002/lary.21030

Laryngoscope 120: September 2010 Sakai et al.: Detection of Unknown Primaries

1779

Different Head Positions (Fig. 2)

Head torsion technique. The patient’s head was rotatedlaterally to open the contralateral pyriform sinus (PS).

Valsalva maneuver. Pressure was applied to the hypo-pharyngeal area by holding breath to open the PS bilaterally.

Killian position. The patient’s neck was bent forwardand the chin was depressed far enough to look down at the um-bilicus so that the postcricoid (PC) region was visible.

RESULTS (TABLE I)Primary lesions were detected in 12 (40%) (hypo-

pharynx, 9; larynx, 1; thyroid gland, 1; oropharynx, 1)

out of 30 patients with CUP in whom the conventional

method was employed. During follow-up, primary lesions

were found in 5 patients among the remaining 18 patients

in whom primary lesions had not been initially detected

(esophagus, 2; gingiva, 1; buccal mucosa, 1; bladder, 1).

Fig. 1. Hooded video-endoscope. An EH-1530T2 (Pentax, Tokyo, Japan) hooded video-endoscope was used. The video endoscope isequipped with a transparent hood at its tip. The hood kept the mucosa and its secretion away from the objective lens. The endoscope alsohad a continuous air insufflation channel that enabled opening of the narrow viewing fields around the apex of pyriform sinus (PS) and thecervical esophagus, resulting in a clearer and wider view of these inaccessible areas.

Fig. 2. Different head positionsused. During head torsion, thepatient’s head was rotated laterallyto open the contralateral PS. Duringthe Valsalva maneuver, pressurewas applied to the hypopharyngealarea by holding breath to open thePS bilaterally. In the Killian position,the patient’s neck was bent forwardand the chin was depressed farenough to look down at the umbili-cus so that the postcricoid (PC)region was visible.


1780

Primary lesions were detected in 15 (71%) (hypo-pharynx, 12; oropharynx, 2; esophagus, 1) out of 21patients with CUP in whom the new method wasemployed. There was significant difference for detectionof primary lesions (P ¼ .03, chi-square test).

Among the remaining six patients, primary lesionswere subsequently found in three (one each in esopha-

gus, oropharynx, and nasopharynx detected withpositron emission tomography [PET], tonsillectomy, andrandom biopsy, respectively).

Images of representative primary lesions detectedwith new methods are shown in Figures 3, 4, and 5.

DISCUSSIONA large number of patients are referred to our hos-

pital after being diagnosed with CUP, and the detectionof primary lesion before treatment is necessary in thesecases. If the primary lesion cannot be detected by

TABLE I.Results of This Study.

The detection rate for primary lesions with the new method was 71% (15/21) compared to 40% (10/30) with the conventional method. Using the newmethod, hypopharyngeal cancer was identified in 12 out of 15 cases.

Fig. 4. Detection of a lesion in the Killian position. The tumor atthe PC region can be detected in the Killian position.

Fig. 3. Detection of a lesion using the hooded video-endoscope.A small mucosal cancer was demonstrated at the right PS usingthe hooded video endscope.


1781

routing examination including inspection, palpation, andvideo-endoscopy, additional efforts are required. Methodscurrently used include random biopsy of the epipharynx,oropharynx, hypopharynx, tonsillectomy, and imagingstudies such as computed tomography (CT), magneticresonance imaging (MRI), and PET/CT. Recently, someauthors reported that primary lesions could be detectedby tonsillectomy and PET/CT,1–3 resulting in the currentacceptance of such methods. It is also believed that pal-pation of the region from the base of the tongue to thetonsils under general anesthesia is important for detect-ing submucosal lesions.10

Because an improved ability to detect primary lesionscan contribute significantly in therapy, we applied ournew method for preliminary examinations in the outpa-tient setting. As a result, primary lesions were detected inmore than 70% of cases with CUP. This higher rate isattributed to the detection of very small cancers usingnew optical devices and modifications of endoscopic techni-ques, due to which the hypopharynx was visualized moreclearly. We describe our new technique in detail below.

Techniques to Expand the Hypopharyngeal AreaCircumferential observation of hypopharyngeal mu-

cosa is difficult during conventional endoscopy becauseof its anatomically closed nature, pharyngeal reflex, andsalivary accumulation. Therefore, we devised methods toexpand the hypopharyngeal area to improve visualiza-tion of the mucosa.

The new procedure involves the following techni-ques: pharyngeal anesthesia and different head positions(head torsion technique, Valsalva maneuver, and theKillian position).

Pharyngeal anesthesia. The pharyngeal reflexcould be inhibited by mucosal anesthesia with 4% lido-caine, enabling a closer examination of the mucosa.

Different head positions. A large section of themucosa of the hypopharyngeal area, which is usuallyclosed, became visible when examined under differenthead positions. Head torsion technique and Valsalva ma-neuver were helpful for opening the PS. The Killian

position is effective for observation of the PC region (Fig.4). We were able to detect tumors in different areasusing different head positions.

New Optical Device (Hooded Video-endoscopeand NBI System)

Hooded video endoscope. Even after following themethods described above, there were some cases whosehypopharyngeal area was not sufficiently visualized. Insuch cases, the hooded video endoscope is useful. Awider area of the hypopharyngeal mucosa can beobserved by mechanically expanding the hypopharyngealarea with a transparent hood (Fig. 3). In addition, thehood prevents the tip of the endoscope from either touch-ing or getting too close to the mucosa. Furthermore, thehooded endoscope has an air insufflation channel thatprevents saliva and secretory fluids from interferingwith the visual field.

NBI system. The NBI system is comprised of avideo endoscope and a light source with new narrow-band filters that enhanced the imaging of microvesselson mucosal surfaces. The system visualizes the neoplas-tic vascularity of superficial lesions clearly and is usefulfor early detection of esophageal, oropharyngeal, andhypopharyngeal cancer.4–6,11 As the system has becomemore popular, more cases of superficial cancer, whichcan be treated with minimally invasive surgery, havebeen reported. This could result in a revolutionary con-cept for treatment of mucosal cancers. Remarkableinnovation has been achieved in optical devices used forgastroesophageal endoscopy. Early detection of lesions inpharyngolaryngeal region will be possible with the appli-cation of such devices. These new optical devices willincrease the detection rate of early primary cancers incases that would have been otherwise diagnosed withCUP (Fig. 5). Although only hypopharyngeal cancerswere detected in this study, NBI endoscope can alsodetect lesions of other head and neck cancers.6

Fig. 5. Detection of a lesion using the NBI endoscope. On the NBI image, the lesion can be easily recognized as a brownish area withabnormal microvessel proliferation (intraepithelial papillary capillary loop; IPCL).12


1782

CONCLUSIONIn this study, the new method using new optical

devices and modifications of an endoscopic techniquewas demonstrated to increase the detection rate of pri-mary lesions in patients with CUP. We successfullyidentified very small lesions, especially in the hypophar-ynx, that are usually not detected with the conventionalmethod. This new method is useful for the detection ofprimary lesions in patients with CUP.

BIBLIOGRAPHY

1. McQuone SJ, Eisele DW, Lee DJ, et al. Occult tonsillar car-cinoma in the unknown primary. Laryngoscope 1998;108:1605–1610.

2. Jesse RH, Perez CA, Fletcher GH. Cervical lymph node me-tastasis: unknown primary cancer. Cancer 1973;31:854–859.

3. Roh JL, Kim JS, Lee JH, et al. Utility of combined (18)F-fluorodeoxyglucose-positron emission tomography andcomputed tomography in patients with cervical metasta-sis from unknown primary tumors. Oral Oncol 2009;45:218–224.

4. Watanabe A. The value of narrow band imaging endoscopefor early head and neck cancers. Otolaryngol Head NeckSurg 2008;138:446–451.

5. Muto M, Nakane M, Katada C, et al. Squamous cell carci-noma in situ at oropharyngeal and hypopharyngeal mu-cosal sites. Cancer 2004;101:375–381.

6. Ugumori T, Muto M, Hayashi R, et al. Prospective study ofearly detection of pharyngeal superficial carcinoma withthe narrowband imaging laryngoscope. Head Neck 2009;31:189–194.

7. Sato K, Nakashima T. Office-based videoendoscopy for thehypopharynx and cervical esophagus. Am J Otolaryngol2002;23:341–344.

8. Tsunoda A, Ishihara A, Kishimoto S, et al. Head torsiontechnique for detailed observation of larynx and hypo-pharynx. J Laryngol Otol 2007;121:489–490.

9. Denker A, Kahler O, eds. Handbuch der Hals Nasen Ohren-heilkunde, Band 1. Berlin: Julius-Springer; 1925:762–843.

10. Chepeha D, Koch W, Pitman K. Management of unknownprimary tumor. Head Neck 2003;25:499–504.

11. Gono T, Yamazaki K, Douguchi N, et al. Endoscopic obser-vation of tissue by narrowband illumination. Opt Rev2003;10:211–215.

12. Inoue H, Sasajima K, Sugaya S, et al. Endoscopic in vivoevaluation of tissue atypia in the esophagus using anewly designed integrated endocytoscope: a pilot trial.Endoscopy 2006;38:891–895.


1783

Documents

New Techniques to Detect Unknown Primaries in Cervical Lymph Node Metastasis