Artículo Clínico

---

Preliminary study of the sentinel node in oral cancer:
in conjunction with 12 cases
Estudio preliminar del ganglio centinela en el cáncer oral:
a propósito de 12 casos

 

C.I. Salazar Fernandez¹, S. Gallana Álvarez², A. Pérez Sánchez³, J. Torres⁴, A. Rollan¹,
F. Mayorga¹, R. Del Rosario², J.M. Pérez Sánchez⁵

---

Abstract: IManagement of the N0 neck in patients with head and neck squamous cell carcinoma remains controversial. Objective. To evaluate the feasibility and predictive ability of the sentinel node (SN) localization-biopsy technique for patients with squamous cell carcinoma of the oral cavity and clinically negative necks. Methodology. We realize a prospective study of 12 consecutive patients at present, with squamous cell carcinoma oral and clinically negative necks. These patients had not recived treatment against tumor and they were assited by Maxillofacial Surgery Service of HUVM from Seville. All patients received a cervical Tc99mlymphoscintigraphy to localize the sentinel node and a new dose before surgery. Intraoperatively, the sentinel node is localized and it is removed separately before tumor resection and elective functional neck dissection. The different anatomic specimens are sent for independent histological examination. Results. lympholocalization 91%, radiolocalization 100%, false negative 0%, sensitivity and NPV 100%, likely hood ratios > 10 and likely negative ratios < 0.1. Conclusions. The technique allows identification of SN-metastases and shows promise in guiding functional neck dissection (FIS-2002 subvention). Keywords: Sentinel node; Oral squamous cell carcinoma; Cervical lymphoscintigraphy; Elective cervical dissection.

Resumen: El manejo de los cuellos N0 en pacientes con carcinoma epidermoide de cabeza y cuello es controvertido. Objetivo. Demostrar la eficacia diagnóstica de la biopsia del ganglio centinela (GC) en los pacientes con carcinoma epidermoide oral con cuello clinicamente negativo. Metodología. Se realiza un estudio prospectivo de 12 pacientes consecutivos, hasta el momento, con carcinoma epidermoide oral de cualquier tamaño y cuello clinicamente negativo que no habian recibido tratamiento antitumoral, asistidos en el S. de Cirugía Maxilofacial del área del H.U.V.M de Sevilla. A estos pacientes se les realiza una linfografía cervical con nanocoloides -Tc 99 para localizar el GC, y una dosis de recuerdo antes de iniciar el ttº quirúrgico. Durante la cirugía se localiza el GC con la sonda y se extirpa, se completa la disección cervical funcional y la extirpación de la lesión con posteriores estudios histológicos independientes. Resultados. índice de linfolocalización: 91%, índice de radiolocalización 100%, falsos negativos 0%, la sensibilidad y VPN del 100%, cocientes de probabilidades positivo > 10 y negativo < 0,1. Conclusiones. La técnica es eficaz para detectar GC-metástasico y es una guia prometedora para indicar la disección cervical (Subvención FIS 2002). Palabras clave: Ganglio centinela; Carcinoma de células escamosas oral; Linfografía cervical; Disección cervical electiva.

---

1 Médico Adjunto del Servicio de Cirugía Oral y Maxilofacial.
2 Médico Residente del Servicio de Cirugía Oral y Maxilofacial.
3 Médico Adjunto del Servicio de ORL.
4 Médico Residente del Servicio de Anatomía Patológica.
5 Jefe de Servicio de Cirugía Oral y Maxilofacial.
Hospital Universitario Virgen Macarena, Sevilla. España

Correspondencia:
Clara Salazar Fernandez
Plza. Vicente Aleixandre 2-5 B, 41013 Sevilla, España.
E-mail: amperezs@wol.es

 

Introduction

The risk of suffering metastasis of the neck in oral cancer varies between 10-45%. This risk together with the belief that if metastases of the neck are removed in sub-clinical (non-palpable) stages, survival rates improve by 50%,¹ has resulted in elective or prophylactic neck dissection (END) becoming generalized. However END is effective in only 26% of patients with occult metastasis, 74% are over treated with the degree of morbidity and cost that this entails.² On the other hand, if observation of the neck is decided upon, metastasis will appear in 30-40% of patients who will need therapeutic dissection of the neck and survival rates will be worse.³ Forecasting would therefore be of great use in order to determine what patients were going to have sub-clinical metastasis of the neck and who would benefit from END.

However, current diagnostic imaging techniques do not allow confirming with total certainty the presence of cervical metastasis in clinically negative necks which would allow those patients who are to undergo END of oral cancer to be selected. Ultrasound of the neck seems to be the diagnostic method with greater sensitivity and PET scanning seems to have greater specificity for neck staging in oral cancer.^4,5 For most of the authors revised sensitivity, specificity and accuracy of ultrasound of the neck varied between 84-92%, 68-83% and 76-86% respectively^4-7, for PET this was 70%, 82% and 75% respectively⁴, for CT it was 66%, 74% and 70% respectively^4,6 and for MRI this varied between 64-73%, 69-95% and 66% respectively^4,8. An attempt is being made to improve on these values using techniques such as MR imaging with superparamagnetic iron oxide. Mack et al⁹ report sensitivity at 86%, and a specificity at 100% with a PPV of 100% and a NPV of 99% in a study carried out in 27 patients with this pathology.

The sentinel node is the first lymph node into which a localized tumor will drain. Various authors have shown that, with breast cancer and melanoma, if the first stage of lymphatic drainage (sentinel node) is not affected in the histological study, the possibility of micrometastasis in less than 5%^10-14. On the basis of this, patients with these tumors will receive lymphatic drainage only when the biopsy of the sentinel node shows micrometastasis of the disease. This technique, applied to squamous cell carcinoma of the head and neck, is still in an initial stage and the comparatively little literature published to date shows methodical defects (limited samples of heterogeneous localization, do not demonstrate the diagnostic feasibility of the technique).^15,16

Objectives

The objectives of the study are: 1. To evaluate the diagnostic feasibility of the sentinel node biopsy technique in order to detect occult metastasis in TN0 squamous cell carcinoma of the oral cavity; 2. To evaluate this technique as a staging method; 3. To establish the type of treatment for the neck, which would result in a spectacular change in the treatment of the disease and in costs (Table 1);.⁴ To understand the individual lymphatic drainage pattern of oral squamous cell carcinoma.

Figure 1. DC: Neck dissection; RT: Radiotherapy.

Material and method

A prospective following of 96 consecutive patients with clinical TN0 oral squamous cell carcinoma were attended by the Maxillofacial Service of the Hospital Virgen Macarena of Seville from March 2002.

The criteria for being included in the study were: patients with primary squamous cell carcinoma confirmed by means of an incisional biopsy, located in the oral cavity (carcinomas localized in the posterior tonsillar pillar, the tonsils or pharynx were not included) with a size of T1- T4 N0 according to the IUAC and the AJCC (evaluated by means of clinical exploration and a cervical TAC), with no previous anti-tumor treatment nor with secondary oral tumors. Patients with allergies to the components of the radiotracer were excluded from the study, together with pregnant or nursing mothers, and patients who refused to participate having been given the relevant information.

Methodology of the study was carried out in three stages.

1. Lymphoscintigraphy. 18 hours before surgical intervention (tumor removal and END), each patient was given a submucosal injection of 84 MBq 99m Tc-nanocolloid (in a minimum of four points). The same person always carried this out. After this a cervical lymphoscintigraphy was done, also carried out by the same person. A dynamic study was started in an anterior sense two minutes after the injection, followed by a static study in an anterior-posterior and lateral sense, marking on the skin of the patient the approximate position of the sentinel node. The study was considered finalized when the sentinel node is detected, or 24 hours after the nanocolloid injection, and a blind evaluation was carried out by 2 doctors of the Service of Nuclear Medicine (always the same ones).

2. Surgery. Two hours before surgery, the peritumoral submucous injection was repeated with the same dose of 99m Tc-nanocolloids in the same points as at the start, and by the same person. Once the patient was anesthetized, the localization of the sentinel node/s were checked by means of cutaneous sweeping using a gamma detection probe (GPS navigator. Tyco). Once located, a biopsy of the sentinel nodes was carried out, followed by dissection of the neck which included lymphatic levels I to V. The radioactivity of the sentinel node was measured in situ, from the bed, from the sentinel node ex situ as well as from the cervical bed and from the dissection specimen, in order to capture residual activity. The sentinel nodes and the rest of the operative sample were placed in separate jars so that an anatomopathological study could be made in order to give the reference of the anatomic level of their localization.

We consider a node as being radiopositive when a reading of at least 15 emissions in 10 seconds is made, and when there is three times the count of the surgical bed. The sentinel node is the one located in the area marked by the lymphography which gives the highest reading.

3. Pathological Anatomy. The anatomopathological study of the sentinel nodes supported by most authors includes multisection and immunohistochemical techniques (anticytokeratin antibodies AE1/AE3)^17-20. We follow the steps set out by Turner et al²¹, who carry out the histopathological validation of the sentinel node in squamous cell carcinoma of the breast. They showed that if the sentinel node is tumor free using hematoxylin-eosin dye and immunohistochemical techniques, the probability that a node, which is not a sentinel node and which is hematoxylin-eosin negative, could be affected is less than 0.1%.

The PA study of the neck dissection (of non-sentinel nodes that are greater than 0.5 cm) that is considered the quality control test or gold standard consists in:

1. Hilus bisection or through the long axis.

2. Dyeing using hematoxylin-eosin and studying a section of 250 micras of each half.

3. The non-sentinel nodes of the neck larger than 2.5cm are processed in the same way as sentinel nodes.

The PA study of the sentinel nodes consists in:

1. The sentinel nodes are sectioned across the hilus or along their long axis.

2. Blocks with a thickness of 250 micras are then obtained from each half of the bi-sectioned node which are dyed with hematoxylin-eosin. If there is tumor presence the sentinel node is considered positive.

3. If on the first examination the sentinel node is considered tumor free, from each half layers of 150-200 micras are obtained in order to examine them by means of immunohistochemical techniques. If there are epithelial tumor deposits the result is considered positive.

4. If a tumor is not detected in the previous step, the sentinel node analyzed is considered negative.

The anatomopathological studies of the sentinel node and the dissection of the neck are evaluated independently and blindly by two anatomopathologists (always the same ones).

Patient following is carried out every month for the first two years, every three months in the third year and after that every six months.

Statistical method

A sample size of 96 patients giving a confidence interval of 95% and a precision of +/- 2, estimating a priori technical sensitivity at 99% with regard to the gold standard and increasing it by 20% because of possible losses.

The diagnostic validity is given by: The pre-test probability (prevalence); the post-test probability; sensitivity and specificity (internal value of the technique); the positive predictive value and the negative predictive value (security of the technique); the likelihood ratios for positive results and for negative results, in order to give technique precision. The ROC curves are plotted (receiver operating characteristic curve) after establishing certain cut off values so that high sensitivity of the test is ensured (avoiding possible false negatives) as well as the diagnostic Concordance between observers using the kappa analysis/kappa balance.

Results

The SN technique was used on 12 patients (7 men and 6 women) with an age range between 45-74 years. The primary tumor was situated on the tongue in five cases, on the floor of the mouth in one case, in the gums in four cases and in the retromolar trigone in two cases. Seven primary tumors were staged as T1 and five as T4. Lymphographies of the neck showed single primary SN in ten cases, 2 primary SNs in one case and only in one patient was it silente (Localization rate of the SN by means of lymphography: 91%). The remaining 11 had SNs in homolateral neck with a radiolocalization SN rate of 100%. Nine patients (82%) presented negative anatomopathological (AP) SNs with negative AP neck dissection, presenting a TN0 stage. Two patients (18%) presented positive AP SNs and positive AP neck dissection presenting a TN2b (IV) stage according to the AJCC in 1997. Of these 2 patients with positive AP SNs, one (50%) was negative with the hematoxylin-eosin technique and positive with the immunohistochemical technique.

We can therefore state that the pre-test Probability of presenting metastasis of the neck was 18% and that the post-test Probability of presenting metastasis with the positive SN technique was 100%. The percentage of false negatives was 0% and sensitivity was 100%. The specificity and positive predictive value was by definition 100% however, we obtained a negative predictive value of 100% with a positive likelihood ratio of (2/0) >10 and a negative likelihood rate of (0/9) < 0.1

There was inter-observer diagnostic concordance in the localization of the SN in the lymphography in all cases except for in 2 patients. There was inter-observer diagnostic concordance in the AP study of the SNs in all cases.

27% of our patients (3 cases) presented lymphatic drainage with «stage jumps» (SN in level II in one case, level III in another case and level IV in another). A summary of results in reflected in table Nº 2.

Patient following to date is of 2-17 months. 2 have died, one of them for post-operative reasons not related with the tumor (UDTH) and the other patient developed a second tumor in the lung. The other patients are alive and they do not present metastasis of the neck nor regional relapse.

Discussion

The diagnosis of sub-clinical metastases of the neck in squamous cell carcinoma of the head and neck still constitutes an important and decisive problem in the treatment of these tumors. Even though the sensitivity of the imaging techniques for metastasic node detection in the neck is gradually improving, specificity remains low. Laborious techniques have, as a result, been described using puncture-aspiration guided by MRI and CT. Puncture using ultrasound-guided aspiration is more practical with a sensitivity ratio of (98 versus 88%) specificity (95% versus 82%) and diagnostic accuracy (97% versus 87%) superior to conventional puncture- aspiration.²² New systems are also appearing for tumor clinico-histological evaluation in relation with the risk of metastasis in the neck, such as the one established by Martinez- Gimeno et al²³ which establishes a scoring system according to various clinico-histological parameters for the tumor (microvascular invasion, differentiation grade, tumor size, tumor thickness, inflammatory infiltration, tumor interphase and perineural invasion) and their correlation with metastases of the neck, establishing a high risk of metastasis of the neck for a score of over 17.

The objective of our study was to evaluate the feasibility and applicability of the SN in clinical N0 oral cancer in order to decide on treatment for the neck, using the results of the SN technique in 12 patients with clinical TN0 oral squamous carcinoma. In our opinion it is very important to have well-defined criteria in order to carry out the technique, and we consider the dynamic phase of the lymphography to be equally useful for defining the location of the SN in order to avoid considering this same node as the next lymphatic station. In our study a 91% rate of lymph localization was achieved, and a rate of radiolocalization of 100% coinciding with the literature,^24-28 and always within the first 15 minutes of injecting the tracer. In contrast with other previous studies^15,29 we were able to locate and remove a single SN in the majority of our patients. We suggest that the removal of the SN should be done just after lifting the skin flaps, as it would not be realistic to look for it in the surgical specimen, and the technique for a node biopsy should be followed. Also, the interferences in the activity of the primary tumor in the localization of the CN can be reduced by removing the lesion before localizing the SN or by changing the angle of the probe. We have had a very poor experience with regard to dying the SN in blue as extracting the primary tumor is made more difficult, and we are in agreement with the difficulties that have been published.^15,24,25

In out study we present a hidden rate of metastasis which is slightly lower (2 patients: 18%) than that of other authors^{24- 30,32,33} probably due to the small sample, both cases presenting an AP N2b stage (stage IV). Of these one case was negative with hematoxylin-eosin and positive with the immunohistochemical technique. Villareal et al.³⁰ in a study of 22 patients with TN0 oral cancer, noted that of six positive SNs one was recaptured with immunohistochemical techniques. Ross et al³² similarly, in their series with 48 patients with TN0 oropharynx, presented two cases (13%) of negative SNs with hematoxylin-eosin that were positive with immunohistochemical techniques. In their series with 17 patients with TN0 oropharynx and larynx, Barzan et al³³, reported that of the 3 positive SNs, 2 had micrometastasis. This finding is equally reflected by other authors in similar studies^24,26-29. We therefore consider immunohistochemical and multisection analysis of great importance for technical precision.

Our rate of false negatives was 0% although values in the literature are recorded between 0-10%^24-31,34,35. The contrast between our results and other broader articles that have been published, and more recently other similar studies, is reflected in table 3. Our most important finding was that all occult metastases of the neck were detected by means of a positive SN, and we were able to guarantee that all patients presenting a negative SN had an equally negative neck dissection. Therefore, neck dissection is only indicated in 18% of cases and in 82% of cases it should have been avoided.

Oral cancer does not follow lymphatic draining which is scaled from level I to level IV as has been thought conventionally, rather individual lymphatic draining takes place as shown in the different levels of the SNs: 27% of our patients presented individual lymphatic draining with «level jumps». Ionna et al.³⁵ in their series with 41 patients with T1-2 tongue and floor of the mouth cancer found a «jump in scale» in 18% of cases, coinciding with similar levels reflected in the literature.^{24,26-27,29,31- 34} Therefore systematic neck dissection is not justified for only level I and level II in neck treatment of oral cancer.

Conclusions

We believe that our pilot study promises results which will validate the SN technique in oral cancer and which will allow the establishment of a new concept, that of superelective neck dissection in clinical TN0 patients on predicting the presence of occult neck metastasis and establishing the type of neck treatment to follow, and on making better staging with immunohistochemical techniques.

References

1. Silverman SJ. Early diagnosis of oral cancer. Cancer 1988;62:1796-99.         [ Links ]

2. Salazar Fernández CI. Tratamiento quirúrgico del cáncer de lengua y suelo de boca en estadios precoces: análisis de efectividad, morbilidad y consumo de recursos. Tesis doctoral. Universidad de Sevilla, España. Facultad de Medicina, 2000.         [ Links ]

3. Weiss MH, Harrison LB, Isaacs RS. Use of decision analysis in planning a management strategy for the stage N0 neck. Arch Otolaryngol Head Neck Surg 1994; 120:699-702.         [ Links ]

4. Stuckensen T, Kovacs AF, Adams S, y cols. Staging of the neck in patients with oral cavity squamous cell carcinomas. A prospective comparision of PET, ultrasound, CT and MRI. J Craniomaxillofac Surg 2000;28:319-24.         [ Links ]

5. McGuirt WF, Greven K, Willians D, Keyes JW, Watson N, et al. PET scanning in head and neck oncology: a review. Head Neck 1998;208-15.         [ Links ]

6. Yuasa K, Kawasu T, Nagata T, et al. Computed tomography and ultrasonography of metastatic cervical lymph nodes in oral squamous cell carcinoma. Dentomaxillofac Radiol 2000;29:238-44.         [ Links ]

7. Hodder SC, Evans RM, Patton DW, y cols. Ultrasound and fine needle aspiration cytology in the staging of neck lymph nodes in oral squamous cell carcinoma. Br J Oral Maxillofac Surg 2000;38:430-6.         [ Links ]

8. Hao SP, Ng SH. Magnetic resonance imaginig versus clinical palpation in evaluating cervical metastasis from head and neck cancer. Otolaryngol Head Neck Surg 2000;123:324-7.         [ Links ]

9. Mack MG, Balzer JO, Straub R, Eichler K, Volg TJ. Superparamagnetic iron oxide enhanced MR imaging of head and neck lymph nodes. Radiology 2002;222:239-44.         [ Links ]

10. Morton DL, Duan-Ren W, Wong JH, Economous JS, Cagle LA, Storm FK, y cols. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392-9.         [ Links ]

11. Thompson JF, McCarthy WH, Bosch CMJ. Sentinel lymph node status as an indicator of the presence of metastatic melanoma in regional lymph nodes. Melanoma Res 1995;5:255-60.         [ Links ]

12. Albertini JJ, Cruse CW, Rapaport D. Intraoperative radiolymphoscintigraphy improves sentinel lymph node identification for patients with melanoma. Ann Surg 1996;223:217-24.         [ Links ]

13. Alex JC, Krag DN. Gamma-probe-guided localization of lymph nodes. Surg Oncol 1993;2:137-44.         [ Links ]

14. Alex JC, Weaver DL, Fairbank JT. Gamma-probe-guided lymph nodes localization in malignant melanoma. Surg Oncol 1993;2:303-8.         [ Links ]

15. Alex JC, Sasaki CT, Krag DN, Wenig B, Pyle PB. Sentinel lymph node radiolocalization in head and neck squamous cell carcinoma. Laryngoscope 2000;110:198-203.         [ Links ]

16. Taylor RJ, Wahl RL, Sharma PK. Sentinel node localization in oral cavity ond oropharynx squamous cell cancer. Arch Otolaryngol Head Neck Surg 2001;127: 970-4.         [ Links ]

17. Waddington WA, Keshtgar MRS, Taylor I, Lakhani SR, Short MD, Ell PJ. Radiation safety of the sentinel lymph node technique in breast cancer. Eur J Nucl Med 2000;27:377-91.         [ Links ]

18. Van Diest PJ, Peterse HL, Borgstein PJ. Pathological examination of sentinel lymph nodes. Eur J Nucl Med 1999;26:43-9.         [ Links ]

19. Gennari R, Stoldt HS, Bartolomei M, Zurrida S, Testori A, Mazzarol G, y cols. Sentinel node localization: a new perspective in the treatment of nodal metastases. Int J Oncol 1999;15:25-32.         [ Links ]

20. Meyers JS. Sentinel lymph node biopsy: strategies for pathologic examination of the specimen. J Surg Oncol 1998;69:212-8.         [ Links ]

21. Turner RR, Ollila DW, Krasner DL, Giulliano AE. Histopathologic validation of the sentinel lymph node hipótesis for breast carcinoma. Ann Surg 1997;226:271-8.         [ Links ]

22. Baatenburg de Jong RJ, Rongen RJ, Verwoerd CD, Overhagen H, Lameris JS y cols. Ultrasound-Guided fine needle aspiration biopsy of neck nodes. Arch Otolaryngol Head Neck Surg 1991;117:402-4.         [ Links ]

23. Martínez Gimeno C, Moro Rodríguez E, Navarro Vila C, López Varela C. Squamous cell carcinoma of the oral cavity: a clinicopathologic scoring system for evaluating risk of cervical lymph node metastasis. Laryngoscope 1995;105:728-33.         [ Links ]

24. Stoeckli SJ, Steinert H, Pfaltz M, Schmid S. Sentinel lymph node evaluation in squamous cell carcinoma of the head and neck. Otolaryngol Head Neck Surg 2001;125:221-6.         [ Links ]

25. Pitman KT, Johnson JT, Edington H. Lymphatic mapping with isosulfan blue dye in squamous cell carcinoma of the head and neck. Arch Otolaryngol Head Neck Surg 1998;124:790-3.         [ Links ]

26. Shoaib T, Soutar DS, Prosser JE, McDonald DG, Dunaway DJ, Gray HW, y cols. The accuracy of head and neck cercinoma sentinel lymph node biopsy in the clinically N0 neck. Cancer 2001;91:2077- 83.         [ Links ]

27. Civantos FJ, Gomez C, Duque C, Pedroso F,Goodwin WJ, y cols. Sentinel node biopsy in oral cavity cancer: correlation with PET scan and inmunohistochemistry. Head Neck 2003;25:1-9.         [ Links ]

28. Pastore A, Turetta GD, Tarabini A, Turetta D, Feggi L, Pelucchi S. Sentinel lymph node analysis in squamous carcinoma of the oral cavity and oropharynx. Tumori 2002;88:S58-60.         [ Links ]

29. Shoaib T, Soutar DS, Prosser JE, Dunaway DJ, Gray HW, y cols. A suggested method for sentinel node biopsy in squamous cell carcinoma of the head and neck. Head Neck 1999;21:728-33.         [ Links ]

30. Villareal Renedo PM, Rayo Madrid I, Pimentel Leo JJ, Serrano Vicente J, y cols. Estadiaje cervical mediante biopsia de los ganglios centinelas cervicales en el carcinoma de células escamosas de cabeza y cuello. Rev Esp Cirug Oral y Maxilofac 2002;24:223-35.         [ Links ]

31. Mozillo N, Chiesa F, Botti G, Caraco C, Lastoria S, y cols. Sentinel node biopsy in head and neck cancer. Ann Surg Oncol 2001;8(suppl 9):103 S-105S.         [ Links ]

32. Ross G, Shoaib T, Soutar DS, Camilleri IG, Gray HW, Bessent RG, Robertson AG y cols. The use of sentinel node biopsy to upstage the clinically N0 neck in head and neck cancer. Arch Otolaryngol Head Neck Surg 2002;128:1287-91.         [ Links ]

33. Barzan L, Sulfaro S, Alberti F, Politi D, Marus W, Pin M, y cols. Gamma probe accuracy in detecting the sentinel lymph node in clinically N0 squamous cell carcinoma of the head and neck. Ann Oto Rhinol Laryngol 2002;111:794-8.         [ Links ]

34. Dunne AA, Kulkens C, Ramaswamy A, Folz BJ, y cols. Value of sentinel lymphonodectomy in head and neck cancer patients without evidence of lymphogenic metastatic disease. Auris Nasu Larynx 2001;28: 339-44.         [ Links ]

35. Ionna F, Chiesa F, Longo F, Manola M, Villano S, y cols. Prognostic value of sentinel node in oral cancer. Tumori 2002;88:S18-19.         [ Links ]