Distribution of days of stay in hospital.
Grant DG, Hinni ML, Salassa JR, Perry WC, Hayden RE, Casler JD. Oropharyngeal CancerA Case for Single Modality Treatment With Transoral Laser Microsurgery. Arch Otolaryngol Head Neck Surg. 2009;135(12):1225-1230. doi:10.1001/archoto.2009.185
To demonstrate the role of transoral laser microsurgery (TLM) in the treatment of oropharyngeal cancer.
A 2-center retrospective case series analysis.
Two tertiary care medical centers.
The study population comprised 69 patients with previously untreated select T1 to T3, N0 to N2 squamous cell carcinoma of the oropharynx, of whom 44 (74%) had no indication for adjuvant RT and 25 (36%) had an indication for adjuvant RT to the neck alone but declined radiotherapy. The primary tumor sites were the tonsil (n = 28 [41%]), tongue base (n = 28 [41%]), pharyngeal wall (n = 8 [12%]), soft palate (n = 4 [6%]), and vallecula (n = 1 [1%]).
Transoral laser microsurgery in 69 patients, with neck dissection in 59 patients (83%).
Main Outcome Measures
Complications, local and regional control, overall and disease-specific survival, swallow function, and feeding tube dependence.
Over the mean follow-up period of 44 months, 66 of 69 patients had no disease recurrence at the primary site. The 5-year local control estimate was 94%. The mean duration of hospitalization was 3 days. There were no major complications relating to TLM. No patient required a permanent feeding or tracheostomy tube. For stage I, II, and III disease, the 5-year Kaplan-Meier estimates of locoregional control were 90%, 73%, and 70%, respectively. The 5-year overall survival estimate was 86%.
Transoral laser microsurgery alone with or without neck dissection is an effective approach for select T1 to T3, N0, or N1 oropharyngeal cancer. Low levels of morbidity, short treatment duration, and excellent disease control make it an attractive therapeutic strategy. The treatment option of endoscopic-assisted laser microsurgery should be discussed by the multidisciplinary team for patients presenting with tumors suitable for this approach.
The role of primary surgery in the treatment of oropharyngeal cancer remains somewhat contentious.1- 5 Critics point to the effectiveness of primary radiotherapy (RT) or concurrent chemoradiotherapy (CRT) and question the role of primary surgery particularly when many patients will receive adjuvant RT anyway. As experience with transoral laser microsurgery (TLM) increases, its value in managing carcinoma of the oropharynx is becoming more apparent. Proponents of TLM point to accurate staging of the primary tumor, short duration of treatment and low morbidity, and excellent functional outcomes compared with RT and CRT. To clarify the effectiveness of TLM as a primary surgery modality, we examine the results of TLM with or without neck dissection and without RT or chemotherapy in the treatment of cancer of the oropharynx.
Since 1996, prospective data has been collected on patients undergoing TLM at Mayo Clinic Jacksonville (Jacksonville, Florida) and Mayo Clinic Arizona (Phoenix, Arizona). Information relating to diagnoses, primary tumor site, and staging and histopathologic findings has all been carefully recorded. We have also documented complications, tracheostomy tube and feeding tube use, and duration of hospital stay. Swallowing function before and after TLM was measured with a clinically relevant functional outcome swallow scale. A search of the database identifying all patients with a diagnosis of previously untreated squamous cell carcinoma involving the oropharynx (including the tonsil, tongue base, soft palate, and pharyngeal wall) was made and a data set exported to SPSS version 16.0 (SPSS Inc, Chicago, Illinois) for further analysis.
All patients underwent TLM with curative intent. Other than biopsy procedures, no patient received treatment prior to definitive surgery. In all cases the treatment choices were discussed at the regular institutional head and neck multidisciplinary meeting. The decision to proceed with a surgical rather than a nonsurgical treatment option was a joint one between both patient and clinician. In TLM no strict exclusion criteria apply; however, relative contraindications include patient or tumor factors that might result in inadequate endoscopic access or tumor exposure.
In all cases the primary tumor was removed with orotracheal intubation using a microscope-mounted carbon dioxide laser or, more recently, a hollow fiber carbon dioxide laser system (OmniGuide Inc, Cambridge, Massachusetts) with the patient under general anesthesia. Exposure of the tumor was achieved with a variety of specialized endoscopes, and excision was performed either circumferentially or in a piecemeal fashion to ensure adequate surgical margins. In our institutions, tumor mapping and margin analysis is performed using intraoperative frozen sections. In addition, after surgery, margin specimens are sent for permanent paraffin sectioning and further reporting.
Indications for neck dissection in TLM are unchanged compared with those of other primary surgical techniques.6 Occasionally, some patients had a contraindication to lengthy periods of general anesthesia, and in these cases neck surgery was staged 3 to 4 weeks following TLM.
Indications for adjuvant RT at our institutions were nodal extracapsular extension, bulky or multiple positive nodes (N2 or greater), positive tumor margins, or lymphovascular invasion seen at the primary site. Precise primary tumor indications for adjuvant RT are difficult to categorize. Endoscopic surgeons realize that the invasive morphologic abnormality and overall tumor footprint can vary greatly even between tumors of identical T stage. No tumor in the present series had an indication for adjuvant RT to the primary. Those tumors with an indication would be T3 or T4 tumors, with extensive infiltrative growth patterns, angiolymphatic invasion, and/or neurotropic spread.
End points analyzed were local control, locoregional control, and overall survival. These end points and evaluation of differences between groups were calculated using the Kaplan-Meier method and log-rank test. Events were measured from the day of surgery to the date of their occurrence or the date of last follow-up. For the estimates of overall and disease-specific survival, those patients lost to follow-up less than 2 years after disease recurrence were considered to have died of disease. Patient deaths within 30 days of surgery were considered treatment related. Analysis was performed in accordance with appropriate institutional review board regulations and approval.
Between 1996 and 2008, 206 patients underwent TLM with curative intent for previously untreated cancer of the oropharynx at our institutions. Three patients (1%) had insufficient follow-up or data and were excluded. A total of 159 patients (77%) received or were advised to receive adjuvant RT (Table 1). Forty-four patients (21%) had no indication for adjuvant RT and formed the main cohort in this study. Twenty-five patients had an indication for adjuvant RT to the neck alone but declined the recommendation. These patients are included in the present study for analysis of local control and the effect of TLM alone at the primary site. Therefore, 69 patients were included in the final study group. Patient characteristics are given in Table 2.
The mean duration of hospitalization was 3 days (range, 1-17 days). The distribution of length of stay is shown in the Figure. Follow-up ranged from 3 to 146 months. The mean follow-up period was 44 months. Eight patients had surgery after March 2007 and therefore did not have 2 years of follow-up.
In the group with no indication for RT to the neck, tonsil tumors outnumbered tongue base tumors 2:1. Tongue base tumors predominated in the group of patients declining the offer of RT to the neck (Table 2). All patients had both frozen and permanent margin sections reported to be free of tumor.
Management of the neck is given in Table 3. Surgical management varied with surgeon preference; however, for node-positive necks, a minimum of a selective or modified radical neck dissection was performed whenever possible. Formal classic radical neck dissection was occasionally necessary. Bilateral neck dissections were common for midline and tongue base lesions. Six patients had their neck dissections staged between 21 and 30 days after TLM. Nine patients with clinically and radiologically N0 necks had no treatment and were observed only.
The timing and indications of tracheostomy and feeding tube use is given in Table 4.
To examine the effect of TLM at the primary site, 25 patients who declined the recommendation of adjuvant RT to the unilateral neck or bilateral neck were included in the study group for local control analysis. Eight patients were advised to receive RT for macroscopic or microscopic extracapsular extension (Table 5). Fifteen patients were advised to have RT for other adverse features as detailed in the “Methods” section. Two patients had recommendations for RT to N0 necks.
There were 4 major complications relating to TLM. Of 69 patients, 2 died within 30 days of surgery and 1 died following a myocardial infarction 2 days after surgery. The patient had a significant history of carotid vascular disease, had failed a cardiac stress test, and refused a preoperative cardiac workup. One patient committed suicide 2 weeks after surgery. Two patients had significant acute aspiration following TLM. Both had tracheostomy tubes at surgery for a T2 tonsil tumor and a T3 tongue base lesion, and both made full recoveries with normal and near-normal swallow function at the last follow-up examination. One patient had minor postoperative bleeding 14 days after surgery for a T3 tongue base tumor. The bleeding settled without further intervention. Four patients had minor complications relating to their neck surgery.
Of 69 patients, 49 (71%) were alive and without disease recurrence at the last follow-up examination. No patient required a permanent feeding or tracheostomy tube. Of these 49 patients, 48 had normal or near-normal (stage 0 or 1) swallow function. Only 1 patient had stage 2 swallow function (stable with a modified diet only).
Three patients experienced tumor recurrence at the primary site between 9 and 21 months after surgery (Table 6). In all patients, local tumor recurrences were successfully salvaged with further TLM with or without adjuvant RT. One patient had multiple recurrences of carcinoma in situ at several sites within the oropharynx. The patient was alive 6 years after primary surgery. The 5-year Kaplan-Meier local control estimate was 94%. For T1 tumors, the 5-year estimate of local control was 90%, and for T2 tumors, 94%. There was no statistically significant difference between recurrence rates in these groups. No patient with T3 or T4 disease had locally recurrent disease.
Of 44 patients, 4 experienced recurrence in the neck. In all patients, the tumor sites were successfully salvaged, and they were disease free 2, 3, 4, and 7 years after treatment for recurrence (Table 6). In this group, the overall 5-year Kaplan-Meier estimates of locoregional control were 82%. For stage I, II, and III disease, the 5-year Kaplan-Meier estimates of locoregional control were 90%, 73%, and 70%, respectively.
Of the 25 patients who declined RT to the neck, 4 experienced regional failure. In this group, the 5-year estimates of locoregional control was 74%. There was no statistically significant difference in locoregional control between groups despite patients declining, when having an indication, the offer of adjuvant RT. One patient who experienced regional recurrence would only allow a node excision and not a complete neck dissection at primary treatment. For this reason the patient was advised to receive RT. The patient declined and was salvaged with neck dissection again declining adjuvant RT (extracapsular extension). The patient survived and died of other causes 5 years later. One patient refused salvage treatment and died of disease. One patient had palliative surgery and died of disease. One patient, a 92-year-old man with T2N0 tongue base cancer, was advised to undergo TLM to the primary site and RT to the bilateral neck. The patient declined and developed regional recurrence, which was salvaged with RT. The patient was lost to follow-up.
For all 44 patients, the 5-year overall survival estimate was 86%. The 5-year overall survival estimates for stage I and II disease and stage III and IV disease were 79% and 86%, respectively. Disease-specific survival for stage I and II disease and stage III and IVA disease were 88% and 86%, respectively.
In this group, 21 patients had stage IVA disease. Eleven patients were alive at the last follow-up examination. Three patients had disease-related deaths. One patient was lost to follow-up within 2 years of recurrence and therefore considered dead of disease. Five patients died of other causes. The 5-year overall survival estimate for stage IV disease in this group was 49%. Five-year disease-specific survival was 72%.
Transoral laser microsurgery is an effective oncologic surgical treatment for T1 to T3, N0 to N1, and select N2 oropharyngeal cancer. In the present series, TLM controlled the disease at the primary site in 66 of 69 patients. The tumor sites were successfully salvaged with further TLM for all 3 patients who developed recurrence at the primary site. Our overall local control rate was 94%. For T1 and T2 tumors, we obtained local control rates of 90% and 94%, respectively. Transoral laser microsurgery compares favorably with other forms of surgery and RT to the primary site. Historically, local control rates for conventional surgery with or without adjuvant RT and for primary RT with or without neck dissection are reported as broadly comparable.1 For similarly staged disease, contemporary studies report local control rates of 87% to 98% with primary surgical approaches.2- 4 These analyses, however, include patients who have received adjuvant RT for neck indications. As the oropharynx is included in RT planning fields, the oncologic effectiveness of surgery at the primary site is therefore sometimes unclear. We have shown that our technique of careful tumor assessment and excision with clear surgical margins, facilitated with intraoperative frozen sections, can provide favorable local control rates without any dependence on RT.
Primary neck surgery without adjuvant RT is an effective approach for N0 and N1 and select N2 disease. In our main cohort, 35 of 44 patients had neck dissections. Nine patients had clinically and radiologically N0 necks that were observed only. Among these 44 patients, we observed 4 disease recurrences in the neck. Two failures occurred in cN0 necks (ie, observed only), both in T2 lesions and both at 3 months. Two further recurrences occurred in tongue base tumors with pN0 and pN1 necks. In all patients, the tumor sites were successfully salvaged, and all were disease free at last the follow-up examination. We had no regional failures in the 4 patients with select pN2 disease. The Kaplan-Meier estimate of 5-year locoregional control was 82%.
We observed that overall survival was poorer in patients with stage I disease compared with stage II. We believe that this can be explained by the nature of the deaths recorded and their impact on the relatively small numbers in this group. Of the 4 patients with stage I disease who died, 2 were patients lost to follow-up after recurrence and therefore were considered disease deaths. The other 2 deaths were unrelated to their primary cancer. Our overall survival rate was 86%, which is comparable to similar stage series treated with a primary surgical approach with or without RT.2,3
Transoral laser microsurgery with or without neck dissection for oropharyngeal squamous cell cancer has considerable advantages over RT or concurrent CRT including lower morbidity, short duration of treatment, and patient acceptability. Transoral laser microsurgery reestablishes the principle of surgery first (as in the open surgery first era) to render local and regional disease control and more accurately identify the requirement for and extent of any adjuvant therapy. The technique of TLM allows for the precise assessment of tumor depth and infiltration, and the careful margin analysis helps preserve both normal tissue and organ function.7,8 The systematic review by Parsons et al1 concluded that surgery with or without adjuvant RT was significantly more morbid than primary RT with or without neck dissection. The authors reported that for all stages of disease, major complication rates of 3.5% to 8% were observed for RT and 22% to 34% were observed when surgery was used as a primary treatment.1 Transoral laser microsurgery is acceptable to patients because it has few permanent adverse effects and has low morbidity and mortality. Adverse function in patients with oropharyngeal cancer has been shown to be correlated with the use of RT.9 We experienced no major complications relating to primary surgery, and 48 of 49 patients had stage 0 or 1 posttreatment swallow function after TLM (only 1 patient had stage 2). Furthermore, TLM can be completed during a hospital stay of 2 to 3 days compared with several weeks of daily treatment required to complete a full course of RT. This can be inferred as a significant cost reduction in disease treatment.
The primary management of oropharyngeal cancer generates much discussion at the multidisciplinary team meeting. The debate over the optimum primary treatment modality is often as much a philosophical one as it is clinical. In this study the patients treated with TLM with or without neck dissection alone comprised T1, T2 and T3, N0, N1, and select N2 tumors. National Comprehensive Cancer Network guidelines (2008) recommend either primary surgery or primary RT as appropriate first line treatment choices for T1 to T2, N0 to N1 disease.6 These guidelines, however, also recommend adjuvant RT and chemotherapy after surgery for N1 disease with suspicious features such as extracapsular spread, advanced T category, and multiple or bulky nodal disease. Furthermore, the guidelines also recommend consideration of adjuvant RT even in single positive node neck disease.6 Many institutions and multidisciplinary teams therefore favor up-front primary RT (with or without chemotherapy), given the comparable outcomes for surgery and RT. Furthermore, critics of surgical techniques such as TLM point out that as many cases are followed by adjuvant RT to the neck, with overlap of the primary site, the effectiveness of the technique as a primary treatment choice is unclear. We believe that our results challenge this assumption. It is extremely unlikely that a randomized controlled trial comparing RT with surgery for oropharyngeal cancer will ever be carried out. Retrospective cohort analyses such as these will remain important points of reference for the effectiveness of primary treatment modalities. The present study is not without weaknesses, however. Despite the relatively small numbers involved, our results are encouraging. It may be that with further development of intensity-modulated RT protocols, a new paradigm of transoral primary surgery with neck dissection first followed by “selective” adjuvant neck radiation for high-risk features may evolve and offer patients the benefit of both surgery and select RT for the optimization of disease control.
In conclusion, TLM alone with or without neck dissection is an effective approach for select T1 to T3, node-positive oropharyngeal cancer. Low morbidity, short treatment duration, positive functional outcomes, and excellent disease control make it an attractive therapeutic strategy. The treatment option of TLM should be discussed by the multidisciplinary team for patients presenting with tumors suitable for this approach.
Correspondence: David G. Grant, MD, Department of Otolaryngology–Head & Neck Surgery, Mayo Clinic, 4500 San Pablo Rd, Jacksonville, FL 32224 (firstname.lastname@example.org).
Submitted for Publication: May 1, 2009; final revision received June 14, 2009; accepted August 3, 2009.
Author Contributions: Drs Grant, Hinni, and Salassa had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Grant, Hinni, and Salassa. Acquisition of data: Grant, Hinni, Salassa, Perry, Hayden, and Casler. Analysis and interpretation of data: Grant. Drafting of the manuscript: Grant and Perry. Critical revision of the manuscript for important intellectual content: Grant, Hinni, Salassa, Hayden, and Casler. Statistical analysis: Grant. Administrative, technical, and material support: Hinni and Perry. Study supervision: Hinni, Salassa, and Hayden.
Financial Disclosure: None reported.
Previous Presentation: This study was presented at the American Head and Neck Society 2009 Annual Meeting; May 31, 2009; Phoenix, Arizona.