Endoscopic Carbon Dioxide Laser Surgery for Glottic Cancer Recurrence After Radiotherapy: Oncological Results

Objective  To evaluate local control, organ preservation, and complications after endoscopic laser surgery for early recurrent glottic cancer after radiotherapy.

Design  Retrospective study.

Setting  European Institute of Oncology, Milan, Italy.

Patients  The study, which was conducted between May 1999 and September 2005, included 37 consecutive patients (33 men and 4 women) with recurrent glottic cancer after radiotherapy. Selection criteria were as follows: rcTis, rcT1, or rcT2 with subglottic or supraglottic involvement of less than 5 mm and no arytenoid invasion; adequate laryngeal exposure; no previous open surgery; no contraindications to general anesthesia; and signed consent.

Interventions  Endoscopic laser surgery with curative intent using types III to V cordectomies according to the European Laryngological Association.

Main Outcome Measures  Five-year actuarial recurrence-free and overall survival, complications, and rate of laryngeal preservation.

Results  The clinical classifications of the recurrences were rcTis (n = 4), rcT1a (n = 10), rcT1b (n = 11), and rcT2 (n = 12). The pathologic classifications of the recurrences were rpT0 (n = 2), rpTis (n = 5), rpT1a (n = 9), rpT1b (n = 3), rpT2 (n = 14), and rpT3 (n = 4). The median follow-up was 44 months (range, 18-88 months). New recurrences developed in 13 patients (35%): 11 were treated by total laryngectomy, 1 by supracricoid laryngectomy, and 1 by chemotherapy. Three patients died of laryngeal cancer, 1 is alive with disease, and 1 died of a second cancer. Five-year actuarial recurrence-free and overall survival rates were 58% and 86%, respectively. The larynx was preserved in 26 patients (70%). Laryngeal stenosis was the most common major complication (in 3 of 4 women and 1 of 33 men).

Conclusions  Endoscopic laser surgery is a safe and effective salvage procedure in selected cases involving glottic recurrence after radiotherapy. Oncological results are satisfactory, and organ preservation can be achieved in a high proportion of cases; however, the risk of laryngeal stenosis is high in women.

Early glottic cancer (cTis, cT1, and selected cases of cT2) can be treated effectively by radiotherapy, open conservative surgery, or endoscopic laser surgery (ELS). Radiotherapy is considered the treatment of choice in most centers. For local recurrences after radiotherapy, which occur in 5% to 35% of cases,^1-6 the only therapeutic option of curative intent is surgery, with total laryngectomy as the standard of care in most cases.^1-7 In selected patients, partial laryngectomy may achieve local control and spare laryngeal function, with total laryngectomy considered only as a last salvage option without greatly compromising ultimate survival.⁸ Endoscopic laser surgery is a well-established procedure for the initial treatment of early glottic cancer; it is characterized by good oncological and functional results.⁹ Recently, ELS has been used to remove small recurrences after radiation failure.¹⁰ We performed a retrospective study to assess local control rates, organ preservation, and complications in patients who underwent ELS for glottic cancer recurrence after radiotherapy.

We reviewed the clinical records of patients who presented to the European Institute of Oncology, Milan, Italy, between May 1999 and September 2005 with recurrent laryngeal cancer after radiotherapy (60-72 Gy) with curative intent for early glottic cancer. Thirty-seven consecutive patients (4 women and 33 men) underwent ELS for recurrent laryngeal cancer during that period. Eligibility criteria for ELS were as follows: crTis, crT1, or crT2N0¹¹; supraglottic or subglottic extension of cancer less than 5 mm from the free edge of true vocal cords and no arytenoid invasion; good laryngeal exposure; no contraindication to general anesthesia; no previous open laryngeal surgery; and written informed consent. We also included a patient with lung metastases who refused total laryngectomy and received ELS because he met all the other eligibility criteria. Preoperative evaluation included routine blood tests, chest x-ray films, ultrasonograpy of the liver, laryngeal computed tomography, laryngostroboscopy, and biopsy with the patient under general anesthesia (or slide revision if biopsy had been performed elsewhere).

The patients underwent general anesthesia and were intubated with laser-safe tubing (Laser Flex Tracheal Tube; Mallinckrodt Inc, St Louis, Missouri). Intraoperative staging included evaluation with 30° and 70° rigid endoscopes to determine vocal rigidity and depth of infiltration and to look for any involvement of the anterior commissure, subglottic region, or ventricle. Cordectomies, types III to V, were performed¹² with a 25-W carbon dioxide laser (Gebrüder Martin, Tuttlingen, Germany) (output power, 0.8-3.4 W in superpulse mode; beam width, 150 μm). Specimens were always removed en bloc.

Specimens were orientated, fixed in 10% buffered formalin for 24 hours, and paraffin embedded. Five-micrometer sections were cut and stained with hematoxylin-eosin. Resection margins were stained with Indian ink, and the distance between tumor and margins was measured. Margins were defined as negative if they were more than 1 mm from the tumor, close if they were 1 mm or less from the tumor, and positive if cancer was present on 1 or more margins.

Patients with negative margins underwent a clinical, fiberscopic, or laryngostroboscopic checkup every 3 months during the first year and every 4 to 6 months thereafter. Those with close margins or positive margins with laryngeal intraepithelial neoplasia I or II were checked monthly for 6 months, every 2 months for another 6 months, and every 3 months during the second year. Patients with positive margins underwent a second ELS 30 to 40 days later. If there was massive positivity, however, subtotal laryngectomy (where possible) or total laryngectomy was scheduled.

Patient characteristics were summarized as frequencies and percentages for categorical variables and as medians and ranges for continuous variables. Overall survival was defined as the time from surgery to death or latest follow-up visit. Relapse-free survival (RFS) was defined as the time from surgery to recurrence or latest follow-up visit. Survival probabilities over time were estimated by the Kaplan-Meier method,¹³ and survival differences between subsets of patients were assessed with the log-rank test.¹⁴ Because of the small number of patients and events, multivariate analysis was not performed. The analyses were performed using SAS version 8.2 (SAS Institute Inc, Cary, NC).

The median hospital stay was 3 days (range, 1-6 days); 29 patients (78%) were discharged on the third postoperative day. No immediate complications occurred; no patient required tracheostomy or a feeding tube. Laryngeal stenosis occurred in 3 women (75%) and 1 man (3%). One of the women received a permanent tracheostomy. Table 1 shows smoking status, clinical stage before ELS, and tumor site. After ELS, most patients continued smoking; 13 patients (35%) stopped smoking; and 2 nonsmokers (5%) started smoking. Table 2 shows definitive tumor characteristics. The anterior commissure was involved in 29 (78%) of cases. Recurrences were locally advanced (rpT2 and rpT3) in 18 patients (49%). Margins were free or close in 25 patients (68%). Oncological results are shown in Table 3: 32 patients (86%) were alive without disease, and the larynx was preserved in 26 patients (70.2%). A second primary tumor developed in 5 patients.

The results of the analysis of the influence of various factors on events are shown in Table 4. A total of 14 events were reported: 10 locoregional recurrences, 3 deaths due to laryngeal cancer, and 1 death due to pancreatic cancer 20 months after ELS. The latter patient was censored for local recurrence because of short follow-up. Positive and close margins significantly predicted RFS, while anterior commissure involvement had no influence on RFS.

The 5-year RFS rate was 58% (Figure 1), and the 5-year overall survival rate was 86% (Figure 2). The 5-year RFS rate was significantly lower (P = .002) in cases with positive or close margins than in those with free margins (Figure 3). The 5-year RFS rate was also significantly worse (P = .03) among patients with locally advanced cancer (Figure 4), and the groups also differed significantly (P = .01) at 12 months.

Total laryngectomy is considered the treatment of choice for glottic cancer recurrence after radiotherapy for several reasons^1-7: Locoregional recurrences are often larger than the original cancer. Viani et al¹⁵ noted that only 6.5% to 10% of laryngeal recurrences are classified as rcT1 or rcT2, and in their series nearly 50% of the recurrences were classified as transglottic rcT3 or rcT4. Furthermore, cardiovascular comorbidities, old age, and refusal to accept a potentially long rehabilitation program often contraindicate open partial laryngectomy.^7,8 In some cases, patients who are eligible for a conservative approach undergo total laryngectomy because the surgeon does not have experience in endoscopic or open conservative surgical procedures.^7,8 In fact, approximately only one-third of patients with local glottic recurrence after radiotherapy undergo conservative surgery.^15-26 There are relatively few reports of oncological results in large series of cases treated by ELS. Several articles emphasize good oncological and functional results with open conservative surgery; we refer only to those in which more than 20 patients were analyzed.^16-22 Open surgical techniques used for salvage surgery are frontolateral and vertical hemilaryngectomy and supracricoid laryngectomy.

There are several advantages to using ELS for salvage surgery in recurrent laryngeal cancer compared with open conservative surgery,^10,22-26 including shortened hospital stay with consequently reduced costs, low complication rate, possibility of avoiding tracheostomy, good functional results (especially in swallowing), and fewer contraindications (advanced age and presence of comorbidities that contraindicate open surgery are not usually a problem with ELS).^10,22,25 Open conservative surgery achieves good oncological results: local recurrence rate, 5% to 25%; 5-year RFS rate, 74% to 100%; and rate of laryngeal preservation, 74% to 90%.^16-22 The reported rates for salvage ELS are as follows: local relapses, 6% to 59%; 3-year RFS, 91% to 100%; and preservation of larynx, 50% to 71%.^10,23-26 Our experience of about 35% recurrence after salvage ELS is consistent with the rates published in previous studies.

High recurrence rates after ELS are probably the result of understaging of the cancer both before radiotherapy and before ELS, which can lead to undertreatment and positive or close resection margins. The margins were positive or close in 20 (54%) of our cases (Table 2). Clinical and fiberscopic preoperative staging is often not reliable²⁷ in laryngeal cancer. Multislice laryngeal computed tomography and careful intraoperative endoscopic evaluation may improve diagnostic accuracy.

Late complications of ELS can be serious. Laryngeal stenosis occurred in 4 of our 37 patients, including 3 women, 1 of whom required permanent tracheotomy. Women have a smaller glottis and subglottis than men, and when the resection extends to both cords, stenosis is highly likely to develop. In cases involving a small larynx, the high power of the laser may also contribute to complications, because excessive heat can be dispersed to surrounding tissues, damaging them and causing stenosis.

The main limitation of this study is its retrospective nature: it includes cases initially considered suitable for ELS, some of which would now be excluded. The experience gained with these cases and their complications allowed us to refine indications and to improve the technique. We now perform multislice computed tomography of the larynx before surgery in order to identify or assess involvement of the cartilage and paraglottic space, and cases with involvement in these areas are not indicated for ELS. Similarly, our experience suggests that open surgery or a lower-power laser may be more appropriate for women with recurrence of laryngeal cancer, particularly if both cords are involved.

Overall survival in our series was relatively unaffected by the high relapse rate. Preservation of the cartilage, without opening the neck, may have contributed to the good overall survival because any subsequent recurrence is generally contained within the larynx. Furthermore, careful postoperative clinical and fiberscopic follow-up evaluations can identify recurrences early and permit effective salvage surgery.

Correspondence: Mohssen Ansarin, MD, Division of Head and Neck Surgery, European Institute of Oncology, Via Ripamonti 435, 20141 Milano, Italy (mohssen.ansarin@ieo.it).

Submitted for Publication: May 27, 2007; final revision received June 3, 2007; accepted June 25, 2007.

Author Contributions: Drs Ansarin and Chiesa 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: Ansarin, Zurlo, and Chiesa. Acquisition of data: Planicka, Cattaneo, Rotundo, and Maffini. Analysis and interpretation of data: Santoro and Alterio. Drafting of the manuscript: Ansarin, Planicka, Cattaneo, Santoro, Zurlo, and Maffini. Critical revision of the manuscript for important intellectual content: Rotundo, Alterio, and Chiesa. Statistical analysis: Rotundo and Santoro. Administrative, technical, and material support: Ansarin, Zurlo, Maffini, and Alterio. Study supervision: Chiesa.

Financial Disclosure: None reported.

Previous Presentation: This article was presented at The American Head & Neck Society 2007 Annual Meeting; April 28, 2007; San Diego, California.