Tumor Board Conference from the University of Pittsburgh

The patient is a 64-year-old woman with a nonhealing oral ulcer.

Karen T. Pitman, MD: We will be discussing the case of a 64-year-oldwoman who was referred to the Department of Otolaryngology, Universityof Pittsburgh Medical Center, with the chief complaint of a nonhealingulcer in the oral cavity that had been present for 2 months. Upon furtherquestioning, the patient noted the recent onset of right otalgia and a10-pound weight loss. She denied experiencing trismus or odynophagia.

The patient's social history was remarkable for alcohol and tobaccoabuse. Her past medical and surgical history was remarkable only for periodontaldisease and the subsequent loss of all dentition 10 years ago. An otolaryngologistin the community had seen her 2 weeks prior to referral, and a biopsy ofthe oral ulcer disclosed squamous cell carcinoma.

On physical examination, the patient appeared to be adequately nourishedand to be her stated age. Inspection of the oral cavity revealed that thepatient was edentulous and had an ulcerative lesion of the right retromolartrigone, which extended into the soft palate, buccal mucosa, and base ofthe tongue. Leukoplakia was noted along the floor of the mouth and theright ventral surface of the tongue. Figure1 depicts the location of the lesion, which had its epicenter on thealveolar process of the ascending ramus of the mandible.

Flexible fiberoptic examination of the nasopharynx, hypopharynx, andlarynx was normal. There was no palpable cervical adenopathy.

Our pathologists reviewed the biopsy specimens obtained at the outsideinstitution and confirmed the diagnosis of moderately differentiated squamouscell carcinoma.

Determiningthe Extent of the Primary Lesion

Dr. Pitman: What specific concerns would you have about the extentof the primary lesion?

Carl H. Snyderman, MD: Tumors in the retromolar trigone typicallypresent at an advanced stage. Tumors in this area raise several concerns.First, the close proximity of the retromolar trigone to the mandible posesa high risk of invasion of the mandible, either by direct invasion throughthe alveolar surface or along the inferior alveolar nerve. Another concernis the deep infiltration into the pterygoid musculature, making it moredifficult to obtain clear surgical margins. With involvement of the tonguebase and soft palate, there is downgrading of oral function following surgery,which should be considered the therapy of choice. There is also the riskof a second primary tumor.

Dr. Pitman: How do you clinically assess the patient for boneinvasion?

Dr. Snyderman: One historical clue is the presence of hypesthesiaor paresthesia of the inferior alveolar nerve. In my experience, however,I have found these symptoms to be rare, and I tend to rely on radiographsto determine bony invasion. In the edentulous patient, there may be irregularitieson the mandibular surface due to prior dental extractions, which can mimicerosion by cancer. I usually start with a CT scan to evaluate the mandiblefor cortical erosion. If perineural infiltration is a concern, MRI is probablya better imaging study to look for extension of the tumor along the marrowspace.

Dr. Pitman: Both CT and MRI were obtained before the patientwas referred to our institution. Dr. Weissman, from a practical standpoint,when one encounters a patient with a tumor approximating bone who has notundergone imaging studies, what is your recommendation about the sequenceof imaging studies to obtain?

Jane L. Weissman, MD: The questions you ask will determine whichstudies you recommend. If you want to determine whether the mandible exhibitscortical erosion, a CT scan is the study of choice. Ideally, images shouldbe obtained perpendicular to the tumor mass so that you could optimallylook for enhancement along the course of the nerve within the bone. Theinferior alveolar nerve will enhance long before the CT scan shows erosionor enlargement of the osseous canal.

Radiographic Findings

Dr. Pitman: Dr. Weissman will now review the radiographic studiesobtained for this patient.

Dr. Weissman: A contrast-enhanced CT scan demonstrated the extensionsof the soft-tissue mass (Figure 2). Onthe side of the tumor, the fat is effaced, and the tumor is inseparablefrom the ascending ramus of the mandible and from the posterior aspectof the maxilla. The tumor is also inseparable from the pterygoid muscles.The tumor extends into the lateral aspect of the soft palate and uvula.At the base of the tonsillar pillar, there is a suggestion that the tumoris extending submucosally around the glossopharyngeal sulcus and into thebase of the tongue.

It is difficult to assess bone involvement from this study. Gross involvementwould be apparent; you would see frank erosion of the mandible. Gross perineuralinvolvement might be apparent with replacement of the normal marrow.

The axial T1-weighted MR images before contrast demonstrate that thetumor does abut the signal void of the cortex of the ascending ramus; thisis seen better after gadolinium enhancement and with fat suppression (Figure3). The MRI shows that the tumor is inseparable from the mandible,but the subtle irregularity of the cortex is imperceptible. You can seethe tumor infiltrating into the confluent pillar, again probably submucosally,and extending downward.

The images of the neck show some suspicious lymph nodes. By size criteria,these nodes are normal, but the inhomogeneous enhancement suggests thepossibility of tumor involvement.

In summary, the radiographic studies reveal a tumor in the retromolartrigone, extending to the lateral aspect of the soft palate and along theglossopharyngeal sulcus to the tongue base. There is no other tumor involvementof the mandible, but subtle cortical erosion cannot be assessed by thesestudies. Examination of the neck shows a suspicious mass in the right neck(level II), and the left neck looks normal.

What Further Studies Are Needed?

Dr. Pitman: Are there any other studies that could be obtainedto help assess for mandibular invasion?

Dr. Weissman: A software package for the CT scanner called theDentaScan is used in our institution to evaluate the thickness of the alveolusin edentulous patients in whom the dentist is planning to do dental implants.What makes it useful for the dentist and for our purposes is that thisprogram allows you to obtain extremely thin slices through the alveolarprocess. Typically, 1.0-mm intervals are utilized, and they are scannedwithout an interslice gap. This is usually done without contrast for thepurpose of looking for obvious bony involvement, rather than the extentof the soft-tissue component of the tumor.

The DentaScan software reformats direct coronal and axial images. Ithink that perhaps the axial images and the 1.0-mm thin slices are themost helpful part of this study for evaluating both the maxillary buttressand the ascending ramus of the mandible.

Jonas T. Johnson, MD: I would argue that further imaging of thispatient is unnecessary, because it is inconceivable that you would surgicallymanage a T3 lesion of the retromolar trigone without a segmental mandibulectomy.The only exception would be if you were considering the possibility ofirradiating the patient, in which case, further imaging might be neededto convince yourself that irradiation would be a bad choice. The retromolartrigone comprises the soft tissue overlying the mandible, and less thana segmented resection would be an inadequate oncologic operation. Conceptually,what you need to ask yourself is how extensive a resection is needed. Inthis case, I think a segmental mandibulectomy is a foregone conclusion.

Dr. Pitman: Before proceeding with the formulation of the treatmentplan for this lesion, the presence of a second primary carcinoma or distantmetastasis must be ruled out. What further studies should be obtained forthis purpose?

Jennifer R. Grandis, MD: The lung and esophagus are among themost common sites of second primary tumors in patients with upper aerodigestivetract squamous cell carcinoma. These tumors would be discovered by radiographicstudies, such as a barium swallow and chest radiograph. Examination underanesthesia and direct laryngoscopy are also performed. I do not routinelydo bronchoscopy to rule out endobronchial primaries. Distant metastasesare relatively uncommon, and the most likely location is the chest.

Dr. Pitman: Do you think that flexible fiberoptic laryngoscopyperformed in the office is satisfactory for an endoscopic examination,or do you need to repeat the examination in the operating room?

David E. Eibling, MD: The answer to that question, I think, dependson the patient. I occasionally perform flexible fiberoptic laryngoscopyin the office. However, I routinely repeat it in the operating room, althoughthe question of whether this is necessary is valid. In some cases, youreally are less able to examine certain areas, such as the post-cricoidarea and pyriform sinus, in the office; these are best seen in the operatingroom.

Dr. Johnson: I would still repeat laryngoscopy, in part, becauseI agree that you probably see some areas, such as the pyriform sinus, betterdirectly than indirectly. I also do esophagoscopy in the operating room.Since I am doing endoscopy, it seems appropriate to look both places, butI have wondered if that is necessary because we have already done flexibleendoscopy in the office and billed for it.

Eugene N. Myers, MD: I think that the two examinations are complementary,and I do not exclude either one. The only way to evaluate vocal cord movementis in the office when the patient is awake. Often, because of a tumor thatis either bulky or exophytic and hangs over the glottis or into the pyriformsinus where there is a lot of edema, you really do not obtain an accurateview of the extent of the tumor on an office examination.

Designing the TreatmentPlan

Dr. Pitman: When designing the treatment plan for a particularpatient, we take into consideration various characteristics of the patient,the tumor, and the treatment team. For example, this patient was referredto the Department of Otolaryngology because of the multidisciplinary expertiserequired for the treatment and rehabilitation of her cancer. The membersof our treatment team include a medical oncologist, radiation oncologist,speech therapist, dietitians, physical therapist, and skilled nurses, aswell as head and neck oncologic and reconstructive surgeons.

With this in mind, I would like to discuss some of the characteristicsof this patient and the tumor that would influence the recommended treatment.I will start by asking, how does the patient's overall health and functionalstatus influence your decision to offer this patient either primary surgicalor nonsurgical therapy?

Dr. Eibling: I think that when you talk to patients with a headand neck cancer, you very quickly get some idea as to whether or not theyare going to be candidates for extirpation of the tumor. If, for example,the patient is wheelchair-bound and is on oxygen, and the chart is quitethick with previous cardiac disease and so forth, one has to very carefullyentertain whether or not to perform surgery. We see patients like thisat the Veteran's Administration Medical Center almost weekly who are notsurgical candidates based on the fact that they are not likely to survivethe surgery or to have a long enough period of remaining life, based oncomorbidities or performance status, to reap the benefits of extirpation.

On the other hand, if the patient is relatively healthy, I proceed witha discussion about what treatment the patient wants to pursue. Some patientsare very much opposed to any form of surgery, and these patients becomenonsurgical candidates based on their personal preference. In contrast,many patients say, "I want you to fix the problem." With thesepatients, we discuss proceeding with surgery, as that is going to givethem the best chance of a cure.

Dr. Pitman: How would the patient's nutritional status influenceyour treatment plan?

Dr. Grandis: The patient should be in the best possible conditionfor surgery, and maximizing nutritional status in malnourished patientspreoperatively via supplemental or tube feedings should be considered.Several studies over the years have suggested that if the patient's nutritionalstatus is good, perioperative morbidity is minimized.

Dr. Pitman: How do the characteristics of this particular tumorinfluence your treatment recommendation?

Dr. Myers: In our department, we have taken the attitude that,with most tumors, particularly more advanced-stage bulky lesions, patientsdo better with primary surgical removal followed by adjunctive therapy.This patient is a case in point. According to both the physical examinationand imaging studies, she has a high-volume tumor; there also some suspicionof bone involvement.

Dr. Pitman: What is the probability of occult cervical metastasisin this patient?

Dr. Snyderman: Using both physical examination and radiographs,the patient would be staged clinically N0. Therefore, her risk of occultmetastases is 20% to 30%. With a tumor this large, I would treat the neck.

Dr. Pitman: What guides your decision regarding the extent ofmandibular resection, and what are your thoughts concerning the airwayand alternative feeding in the perioperative period?

Dr. Johnson: I think that this patient has a bulky T3 lesion,and thus, she would really be best served by doing a segmental mandibulectomy.A marginal mandibulectomy of the ascending ramus of the mandible can beused perhaps for superficial lesions of the retromolar or trigone. However,I have been dissatisfied with the functional results of that procedure,and frankly I would recommend a segmental mandibulectomy even for patientswith those lesions.

In this particular patient, I would remove the entire neural canal,and thus, I would make my anterior osteotomy at the mental foramen. I thinkthat it would be hard to argue against removing the coronoid and condylein this case because it would facilitate obtaining a good margin on thepterygoid. The real issue is to attain an adequate deep soft-tissue marginon the pterygoids. If you compromise the mandibular resection, you willcompromise the deep margin.

In this patient, healing probably will be complete in 7 to 10 days,at which point, she could be discharged. The problem is that in 7 to 10days, she will not be rehabilitated, and by that I mean she may not yetbe able to protect her airway and she probably will not be able to swallowsatisfactorily. If we want to discharge people 7 to 10 days after havinga major oropharyngeal resection, we need to make preparations to managethem with tube feedings and tracheotomy care as an outpatient. If, on theother hand, we plan to rehabilitate them in the hospital as we did 5 yearsago, this will take about a month, but I think that the payors are no longerinterested in such a plan. Therefore, at present, I think that it is prudentto put a percutaneous endoscopic gastrostomy or gastrostomy tube in everypatient who is undergoing an operation for a T3 tumor of the oral cavityor oropharynx.

SurgicalTreatment and Reconstruction Performed in This Patient

Dr. Pitman: The patient was taken to the operating room, wheredirect laryngoscopy and esophagoscopy were performed. These examinationsconfirmed the office examination and radiographic findings. Subsequently,a segmental mandibulectomy, partial maxillectomy, and excision of involvedbuccal mucosa, soft palate, and posterior pharyngeal wall were performedto remove the primary tumor. A right modified radical type II neck dissectionand left selective neck dissection, including levels I to IV, were performed.Tracheotomy and gastrostomy tubes were also placed.

The surgical defect included the right hemimandible from the condylarneck to the parasymphysis, and included the maxillary tuberosity and portionsof the soft palate, buccal mucosa, tongue base, and posterolateral pharyngealwall.

Dr. Pittman: Dr. Russavage, from the Department of Plastic andReconstructive Surgery, will now discuss the reconstruction that was performedfollowing tumor ablation.

James M. Russavage, MD: Before discussing the reconstructiveprocedure that was performed in this patient, I will review the basic factorsthat we consider when choosing a flap appropriate for a specific defect.The primary goal is to replace tissues that have been removed with similartissues, such as soft tissue or bone. Also, we want to minimize the effectto the donor site.

Moreover, want to try to do the reconstruction in one stage if possible.We can usually work in two teams; for example, one team can work on theextremity while the other team exposes the vessels in the neck. The scapula,for example, is not a very good flap to use with a two-team approach becausethe patient has to be repositioned.

Obviously, we want to optimize the pedicle position so that it is relatedto the length of the pedicle with respect to the skin flap. We want torestore form and function, and we want to coordinate the reconstructionwith our oncologic colleagues.

To reconstruct a defect of the retromolar trigone, several differentflaps can be used. A skin graft is an option for small defects. Pedicleflaps, such as the pectoralis or latissimus dorsi, are good flaps for soft-tissuerepair. When bone and soft-tissue reconstructions are chosen, the iliaccrest, fibula, radial forearms, and scapular and lateral arm have all beenused and all have their champions.

We like to use the fibula and the radial forearm as our primary flaps.Some studies have suggested that restoring sensory innervation in the flapsaids function, and therefore, we innervate these flaps. Whenever possible,we incorporate innervated muscle to provide some function.

We want to reproduce the thin, pliable mucosa of the oral cavity thathas been removed, and using a fasciocutaneous flap, such as the radialforearm or fibular or lateral arm, is optimal. External skin color matchis another consideration; in this case, cervical pectoral flaps will optimizethe color match. We want to choose a fixator that basically minimizes morbiditywith respect to plate dehiscence and provides the best contour.

We do not immediately place dental implants in patients who are likelyto receive postoperative irradiation. Since the incidence of infectionof implants in radiated vascularized bone is high, we delay putting implantsin these vascularized bone grafts for 1 year after radiation.

In addition, there have been a number of developments with reconstructionplates. For example, there are now plates that use screws that are morelike implants. The screws and plate work more as an internal/external fixatorthan as a compression plate and provide less compressive vasculopathy tothe vascularized bone.

This particular patient had a large defect that crossed the mid-lineon the soft palate, tongue base, and posterior pharyngeal wall. Bone andsoft tissue were required for the reconstruction.

Since repair of the soft-tissue defect required a lot of folding, Ichose the radial forearm for reconstruction. In this case, the inset extendedto the contralateral tonsillar pillar and the posterior pharyngeal wallon the opposite side and was folded underneath to get to the base of thetongue and retromolar trigone area.

Other free flaps commonly used to provide bone and soft tissue, suchas the iliac crest, rectus, and scapula, all have much thicker soft-tissuepaddles. Thus, the degree of folding that had to be achieved in this casewithout obstructing the airway was just impossible to accomplish with anyflap other than the radial forearm.

This patient's postoperative course was complicated by a fractured radiuseven though we took the appropriate precautions of doing a keel-shapedosteotomy of the radius. The fractured radius was treated with an openreduction and a plate fixation.

Dr. Eibling: Are implants feasible following radiation?

Dr. Russavage: There are some European studies showing that after1 year the incidence of infection in radiated vascularized bone drops from30% to 5%.

Dr. Snyderman: Exactly how did you reconstruct the mandible inthis case?

Dr. Russavage: We used the radial forearm flap with one osteotomy,and we used the 0reconstruction plate. We only got one screw in the condylarhead. Whenever possible, we do a small osteotomy along the symphysis toremove the external plate and just attach a screw to the posterior platebecause this produces a much better contour and less fullness at the levelof the chin.

Dr. Grandis: Do you use prophylactic antibiotics?

Dr. Johnson: All patients receive intravenous antibiotics. Inaddition, under our current study protocol, half of them receive topicalantibiotics and half, topical saline.

Dr. Snyderman: How long are the bone segments?

Dr. Russavage: Most of the limits that have been described forthe bone of the radial forearm are around 11 cm. This segment of radiuswas 13 cm. The question is, to what degree is the proximal aspect of thebone vascularized? However, we think that if you have some masseter andpterygoid muscle remaining at the proximal aspect of the resection bed,the reconstruction will be superior. Soft-tissue coverage is in the centerpart of the flap, and thus, the issue of the exact degree of vascularityof that segment is moot.

OtherReconstructive Options for Long Bony Defects

Dr. Eibling: Is there a better fit when one uses two flaps insome of these patients with long bony defects? Or is that better fit outweighedby the morbidity associated with two donor sites?

Dr. Russavage: Some surgeons would utilize two or three freeflaps for this defect--the jejunum to reconstruct the mucosal deformityand the fibula for the bony deformity. In my view, however, it is alwaysbest to do as much reconstruction as you can with one flap. You just multiplythe donor complications with more than one flap in a patient who is elderlyand has other comorbidities.

Dr. Snyderman: What is your experience with a Thorpe prosthesisin this situation? There continue to be reports in the literature suggestingthat this works, although our concern is the increased rate of infectionor extrusion.

Dr. Russavage: The newer-generation systems are much thinnerthan the old systems. I think the basic design of the Thorpe prosthesisis good. The problem with the plate is that it is too large and bulky.I am not an advocate of plate-only reconstructions or plate and soft-tissuereconstructions without vascularized bone. I think the bone works as abiologic barrier. The soft tissue recognizes the vascularized bone as self.Consequently, there is less likelihood of plate extrusion, either to theskin or the oral cavity, with a bony reconstruction than there is withplate and soft-tissue reconstructions. I think that any plate-alone reconstructionwill have a recognized dehiscence rate related to the soft-tissue contracturearound the foreign body.

Dr. Johnson: Do you mean that if you reconstruct the soft tissueonly, you will have a less functional reconstruction?

Dr. Russavage: There are studies that would suggest just that.My problem with these three-quarter pharyngeal defects is that once posteriorstop is lost (ie, the mandible is allowed to collapse posteriorly due toloss of posterior stability) and the mandible shifts to the mid-line withcontraction, the soft-tissue flap has a tendency to migrate toward theairway. That makes swallowing more difficult.

Dr. Johnson: So you think that the scaffolding provided by thisbony reconstruction helps?

Dr. Russavage: The bony reconstruction definitely provides bettertongue support and stabilization of the mandible and may facilitate swallowing.

PathologicFindings and the Need for Adjuvant Therapy

Dr. Pitman: I will summarize the pathologic findings in thispatient. The surgical margins were clear, and 2 of 24 lymph nodes fromthe right neck and 0 of 25 nodes from the left neck were positive for squamouscell carcinoma. Extracapsular spread was noted in one of the lymph nodes.The depth of the primary tumor was 1.5 cm, and angiolymphatic invasionwas present. There was no evidence of bone invasion. The pathologic stagewas T3, N2b, M0.

The TNM staging system communicates the size and stage of a neoplasmand is a useful guide for initial treatment planning. However, it may notbe the most accurate predictor of tumor behavior and prognosis. The pathologicfindings are perhaps more predictive of the patient's ultimate prognosis.Dr Johnson, how does the finding of extracapsular spread influence thetumor behavior and your therapeutic decision-making?

Dr. Johnson: Our observations have indicated that the presenceof cervical metastasis with extracapsular spread downgrades 2-year survivorshipby about 50%. Based on that, a patient who has these findings should beoffered adjuvant therapy.

What constitutes adequate adjuvant therapy is debatable. Most expertswould suggest that irradiation is the minimum. As you know, at the Universityof Pittsburgh Medical Center, we offer adjuvant chemoradiation to patientswith extracapsular spread because radiation alone results in what we considerto be unacceptable 2-year survival rates.

Dr. Pitman: Do the findings of angiolymphatic invasion or perineuralinvasion, in and of themselves, influence your recommendation for adjuvanttherapy?

Dr. Myers: Several pathologic findings are known to identifyan aggressive tumor and to predict a poor prognosis. These include extracapsularspread, tumor depth, and perineural invasion. If this patient had not hadinvolvement of any lymph nodes, I would still have recommended radiotherapybased on the presence of perineural invasion, because of the risk of localrecurrence and distant metastasis. Angiolymphatic invasion is also a negativefinding, but it has not been studied specifically, and we do not know itsimpact on recurrence or survival.

Dr. Johnson: Dr. Fagan, is perineural spread an independent prognosticator?

Dr. Johannes J. Fagan: I recently did a retrospective analysisof our data. This analysis indicated that perineural spread is a significantprognosticator, although this is a preliminary finding.

Recommended AdjuvantTherapy Protocol

Dr. Pitman: Dr. Agarwala, would you describe the chemotherapeuticprotocol that would be recommended for this patient?

Sanjiv S. Agarwala, MD: This patient is at very high risk forrecurrence. The question of whether to add concurrent chemotherapy to thepostoperative radiotherapy is as yet unanswered in randomized trials. WhatI encourage patients to do, if they're willing, is to enroll in such aprotocol.

One such protocol is looking at the addition of a radiation sensitizer,usually one of the platinum compounds, to standard radiation therapy tosee whether it improves survival. The entry criteria for this study areextracapsular and perineural invasion. That is what I would recommend forthis patient.

Dr. Pitman: Dr Cano will address the minimum radiation dose.

Elmer R. Cano, MD: The radiation field must extend from the baseof the skull to the infraclavicular fossa. We use a higher dose at theprimary site. The minimum dose at the primary site is 6,000 cGy.

Recommended Follow-up

Dr. Pitman: Dr. Myers, what follow-up would you recommend forthis patient?

Dr. Myers: I usually see the patient in the office frequentlyover the first several months to monitor wound healing and swallowing rehabilitation--about every 2 weeks over a 6-week period. By 6 weeks, the wound is usuallyhealed enough to begin radiotherapy. Once the radiotherapy has been completed,I usually see the patient every 3 months over the first year and every6 months for the next several years.

My major concerns during follow-up are the development of recurrenceand the appearance of second primaries. Another consideration is prosthodonticrehabilitation, which adds a great deal to the patient's quality of lifeby providing prosthetic devices which enhance speaking and swallowing.

I do not use routine radiographic scanning to follow patients becauseI don't think that it is cost-effective. It is very difficult to detectthe presence of recurrence on CT or MRI in patients who have had majorhead and neck surgery, reconstructive procedures, and radiotherapy. Rather,I usually rely on the history and physical examination. If the historyor physical findings suggest a possible recurrence, then further imagingstudies can be helpful.

Participants

Karen T. Pitman, MD
Moderator
Fellow, Head and Neck Oncologic Surgery, Department of Otolaryngology
(Currently at Naval Medical Center, Portsmouth, Virginia)

Sanjiv S. Agarwala, MD
Assistant Professor, Department of Medicine

Elmer R. Cano, MD
Associate Professor, Department of Radiation Oncology

David E. Eibling, MD
Professor, Department of Otolaryngology

Johannes J. Fagan, MB, ChB
Fellow, Head and Neck, Oncologic Surgery, Department of Otolaryngology

Jennifer R. Grandis, MD
Assistant Professor, Department of Otolaryngology

Jonas T. Johnson, MD
Professor, Departments of Otolaryngology and Radiation Oncology, ViceChairman, Department of Otolaryngology

Eugene N. Myers, MD
Professor and Chairman, Department of Otolaryngology

James M. Russavage, MD
Assistant Professor, Division of Plastic and Reconstructive Surgery,Department of Surgery

Carl H. Snyderman, MD
Associate Professor, Departments of Otolaryngology and NeurologicalSurgery

Jane L. Weismann, MD
Associate Professor, Departments of Radiology and Otolaryngology