Abstract
Objective:
The high rate of positive margins after transoral laser microsurgery (TLM) remains a matter of debate. This study investigates the effect of intraoperative narrow band imaging (NBI) examination on the incidence of positive superficial surgical margins in early glottic cancer treated by TLM.
Methods:
Between January 2012 and October 2013, 82 patients affected by Tis-T1a glottic cancer were treated with TLM by type I or II cordectomies. Intraoperative NBI evaluation was performed using 0-degree and 70-degree rigid telescopes. Surgical specimens were oriented by marking the superior edge with black ink and sent to a dedicated pathologist. Comparison between the rate of positive superficial margins in the present cohort and in a matched historical control group treated in the same way without intraoperative NBI was calculated by chi-square test.
Results:
At histopathological examination, all surgical margins were negative in 70 patients, whereas 7 had positive deep margins, 2 close, and 3 positive superficial margins. The rate of positive superficial margins was thus 3.6% in the present group and 23.7% in the control cohort (P < .001).
Conclusion:
Routine use of intraoperative NBI increases the accuracy of neoplastic superficial spreading evaluation during TLM for early glottic cancer.
Introduction
One of the most important factors decreasing local as well as loco-regional surgical control in head and neck cancer is involvement of resection margins.1,2 Therefore, the first aim of any oncologic surgical treatment should be complete tumor resection within adequate free margins. Among the wide gamma of different therapeutic options for laryngeal tumors, transoral laser microsurgery (TLM) is nowadays universally recognized as a minimally invasive approach and a safe standard management option for early glottic cancer. In fact, it represents a valid alternative to radiotherapy (RT) and/or open-neck conservative surgery in terms of oncologic and functional results, allowing low morbidity, good vocal outcomes, and excellent cost-effectiveness ratio.2-5 However, to achieve these advantages, every effort should be made during TLM to optimize the balance between oncologic radicality and vocal outcomes, minimizing the rim of healthy tissue resected with the tumor itself.5-6 TLM glottic resections are usually considered oncologically acceptable within “ultra-narrow” margins of 1 to 2 mm of healthy tissue. However, these margins are usually chosen based on the surgeon’s intraoperative clinical impression, and the definitive histopathological report may reveal positive and/or close margins in a high proportion of cases with subsequent need for “second-look” microlaryngoscopies or close follow-up.5,7,8 The impact of positive margins after TLM for glottic cancer on local control is still a matter of debate: the large number of margins judged as positive is not always clinically reflected in a corresponding high number of recurrences or treatment failures. 9 A number of different reasons may, in fact, partially explain this discrepancy (postresection and postfixation coarctation of the surgical specimen, laser artifacts at the level of the cutting edges, and manipulation of the small piece of excised tissue).
The introduction of new endoscopic devices and techniques, as well as the development of the “biologic endoscopy” concept, 10 have already been proven to be of enormous value in defining tumor superficial extension and delineating its resection margins. Among these diagnostic tools, narrow band imaging (NBI) seems to be 1 of the more promising: it consists of use of filtered wavelengths that enhance the microvascular abnormalities associated with the preneoplastic and neoplastic changes of the mucosal lining of the upper aerodigestive tract.10-12 Several reports in the literature have already addressed the diagnostic value of NBI in identifying epithelial abnormalities and detecting persistent/recurrent disease after (chemo)-RT and surgery or synchronous and metachronous tumors within the nasopharynx, oral cavity, oropharynx, hypopharynx, esophagus, and larynx. To the best of our knowledge, there have been no studies with the specific aim of investigating the impact of intraoperative NBI examination on the incidence of positive superficial surgical margins in the TLM treatment of selected patients affected by early glottic cancer.
Materials and Methods
Between January 2012 and October 2013, 82 consecutive patients (group A) affected by glottic lesions staged as cT1aN0 according to the TNM system 13 underwent TLM as a definitive treatment at the Department of Otorhinolaryngology–Head and Neck Surgery of the University of Genoa, Italy. There were 70 men and 12 women. Their ages ranged from 30 to 83 years (median = 67 years). Inclusion criteria were (1) untreated early glottic cancer (cT1aN0) without involvement of the anterior commissure; (2) no previous biopsy; (3) adequate laryngoscopic exposure of the glottic plane; and (4) no contraindication to general anesthesia. Informed consent was obtained from all patients.
Pre- and Intraoperative Assessment
Impairment of the mucosal wave due to tumor involvement of the lamina propria was evaluated during preoperative diagnostic work-up by videolaryngostroboscopy (Kay Digital Strobe 9106; Kay Elemetrics Corporation, Pine Brook, New Jersey, USA) coupled with 70-degree rigid telescopes. The atypical vascular patterns within and surrounding the lesion were evaluated under local anesthesia by a transnasal flexible videoendoscope (Olympus Medical System Corporation, Tokyo, Japan) with high-definition television (HDTV) white light (WL) and NBI. 14
Intraoperative rigid endoscopy with 0-degree and 70-degree angled telescopes (Karl Storz, Tuttlingen, Germany) was accomplished during microlaryngoscopy under general anesthesia to reassess and better delineate the superficial resection margins. The superficial extension of the lesion was thoroughly examined by switching from HDTV-WL to HDTV-NBI light to identify the neo-angiogenic patterns within the lesions and the surrounding tissues (Figure 1). Moreover, saline infusion (SI) into the Reinke’s space using an angled needle was routinely performed.
(A) Intraoperative examination by high-definition television (HDTV) white light (WL) with 0-degree rigid telescope of cT1a of the anterior third of the right vocal fold and its presumptive resection margins (dotted line) based on WL alone. (B) By switching to HDTV Narrow Band Imaging mode, thick dark spots typically associated with carcinoma in situ became clearly evident at the level of the right paracommissural area and the anterior third of the right floor of the ventricle. The superficial resection margins of type II cordectomy were therefore extended as depicted by the dotted line. The histopathological report confirmed invasive cancer (pT1a) with negative deep and superficial margins.
A type I cordectomy was applied in patients with intraepithelial lesions apparently limited to the basement membrane (normal mucosal wave at videolaryngostroboscopy and complete hydrodissection of the mucoligamentous plane at SI). Type II cordectomy was reserved for lesions with impaired mucosal wave and/or incomplete ballooning of the Reinke’s space at SI.14-17
Surgical Technique
The surgical procedure was carried out under general anesthesia after orotracheal intubation using a laser-safe endotracheal tube (Xomed; Medtronic, Paris, France) with internal diameters ranging from 4.5 to 5.5 mm. Laryngeal exposure was obtained using laryngoscopes of different sizes and shapes: the Microfrance laryngoscope (Microfrance, Paris, France) with wide distal diameter was used most often.
Transoral laser microsurgery was always performed using a CO2 laser (Lumenis Encore Ultrapulse, Tel Aviv, Israel) combined with an Acublade micromanipulator with ultrapulse delivery in continuous mode (1-3 W). “En bloc” resection was performed by the same surgeon (G.P.) and included type I and II cordectomies on the basis of the aforementioned pre- and intraoperative diagnostic work-up. Surgical specimens were oriented at the end of resection by marking the superior edge with black ink, fixed in 10% buffered formalin, and sent to a dedicated pathologist. All the surgical margins (superficial and deep) were assessed, and the distance between tumor and margins measured. Negative margins were defined as more than 1 mm of healthy tissue from the tumor, close was defined as 1 mm or less, and positive was defined as presence of tumor at the level of 1 or more superficial and/or deep resection margins. 18
Follow-up
All patients with negative margins were followed with periodic videolaryngostroboscopy and transnasal flexible videoendoscopy with HDTV-WL and NBI, scheduled every 2 months in the first year after surgery and less frequently thereafter. In the event of close and/or positive superficial surgical margins, a strict “wait and see” policy was adopted, with monthly endoscopic controls by videolaryngostroboscopy and HDTV-WL and NBI. In all cases with positive deep margins, a second TLM was scheduled and complementary RT planned in case of remaining positive deep margins after endoscopic re-excision.
Statistical Analysis
The rate of positive superficial margins in the present cohort (group A) was compared with that of positive margins observed in a matched historical control group of 152 patients (group B) treated between 1999 and 2006 in the same way without using intraoperative HDTV-NBI evaluation. To reduce the possible bias related to the surgical learning curve elapsed from 1999 to 2013, we further subdivided group B into 2 subsets: patients treated between 1999 and 2002 (group B1) and those operated on from 2003 to 2006 (group B2). Comparison of the rates of positive margins between groups A and B and groups B1 and B2 were assessed using chi-square test with a threshold of statistical significance set at a P value of .001.
Results
Patient demographics, pT category, types of cordectomy, and histological details of groups A, B, B1, and B2 are summarized in Table 1.
Patient Characteristics, pT Category, Types of Cordectomy, and Margin Status of Groups A, B, B1, and B2.
In group A (treated after pre- and intraoperative HDTV-NBI examination), 70 (85.5%) patients had all (superficial and deep) negative margins. In the remaining patients, 3 specimens (3.6%) showed superficial positive margins (2 pTis treated by type I cordectomy and 1 pT1a treated by type II cordectomy) in terms of squamous epithelial neoplasia (SIN) grade III. Two (2.4%) patients had close superficial margins (all pT1a treated by type II cordectomy) in terms of SIN grade III. In group B (treated without pre- and intraoperative HDTV-NBI evaluation), 36 (23.7%) patients presented positive superficial margins and none had close superficial margins. Deep positive margins were herein not considered for group A or group B since the role of NBI in their detection is null. A statistically significant difference in the rates of positive superficial margins between groups A and B was found by chi-square test (P < .001). Comparing groups B1 and B2, no statistically significant difference in terms of superficial positive margins (20.6% vs 28.3%) was found.
Discussion
The impact of positive margins after TLM on local control is still a matter of debate, and many authors have described different policies to deal with this clinical problem.3,4,6,18-20 Hinni et al 8 underlined how positive surgical margins contribute to an increase in local recurrence and morbidity rates, decreasing survival. Crespo et al 19 described the role of margin status in recurrence after TLM in 40 patients with T1 and T2 glottic cancer. They observed local recurrences in 37.5% of patients with positive margins and no recurrence in cases of negative ones. Other authors have reported the relatively low effect of uncertain or superficial positive margins on local control with laser and organ preservation rates compared with residual deep positive margins.2,4,15,19-21 Michel et al 9 reported a rate of 37.5% of positive margins in primary resection by TLM for T1a glottic cancer, without any significant impact on oncologic outcomes.
Although the impact of positive margins and consequent re-resection as complementary therapy on local control and survival in early glottic cancer is not clear, Jäckel et al 2 and Remacle et al 22 attributed a questionable value to planned second-look microlaryngoscopy due to the high rate of no residual tumor in the second histopathological report. As an alternative to such a treatment policy, Remacle et al 22 and Fang et al 5 suggested using intraoperative frozen sections. Lucioni et al 1 reported an increased local control in cases of close and positive superficial margins by intraoperative laser photocoagulation of the surgical bed.
In our experience, pre- and intraoperative evaluations by HDTV-NBI represent useful diagnostic tools to optimize the bidimensional evaluation of the neoplastic boundaries and to reduce the incidence of positive superficial margins after TLM. In fact, comparison between the rates of positive superficial margins between patients treated without HDTV-NBI and those in which such a diagnostic technique was applied shows a highly statistically significant reduction (P < .001) from 23.7% to 3.6%. On the other hand, in this respect, the surgical learning curve seems to play a less important role when compared to the use of such biologic endoscopy techniques, as shown by the subanalysis conducted comparing groups B1 (years 1999-2002) and B2 (years 2003-2006). The paramount importance of HDTV-NBI in defining neoplastic superficial spreading and precisely delineating its peripheral resection margins has already been outlined by our group: Piazza et al 23 reported intraoperative up-staging due to use of NBI in 26 laryngeal tumors and the application of wider superficial resection margins by TLM than preoperatively expected (but without consequent up-staging) in 9 cases.
Narrow band imaging, however, applies filtered wavelengths that penetrate the superficial layers of mucosa, thus enhancing the visualization of mucosal and submucosal microvascular networks only. In this way, it is obvious that it does not represent an adjunctive issue in evaluation of the deep extent of neoplastic lesions. As a consequence, its use does not affect the incidence of deep positive margins10-12; this is the reason that we focused our present study on the incidence of superficial positive margins only. In case of infiltration of deep margins, no significant difference in recurrence rate was reported by Lucioni and coworkers 1 between patients managed by laser photocoagulation of the surgical bed and those who were not managed in this way, suggesting that in these cases, photocoagulation is probably unable to destroy residual tumor cells remaining in the surgical field.
According to Sigston et al, 4 accurate histopathological interpretation of surgical margins after TLM can be hampered by artifacts secondary to laser thermal damage, tissue coarctation, and mechanical manipulation of these small surgical specimens. For these reasons, accurate histopathological assessment of the specimen should be carried out in close cooperation between the surgeon and a dedicated pathologist.24,25 In this regard, HDTV-NBI may help in achieving a higher diagnostic accuracy by avoiding adjunctive tissue manipulation and staining. 23
Retreatment policies in cases of positive resection margins may vary from a second surgical procedure to postoperative RT, but both present distinctive disadvantages. A second TLM, in fact, clearly means further loss of tissue, leaving the vocal fold more scarred and usually resulting in a poorer quality of voice. 6 Complementary RT encompasses a multimodal therapeutic approach for early tumors that could have been managed by RT alone from the beginning. We therefore emphasize the importance of complete removal of the tumor at the first surgical attempt by obtaining clear margins in healthy tissue by pre- and intraoperative use of HDTV-NBI.
In conclusion, on the basis of our results, the routine use of pre- and intraoperative HDTV-NBI increases the accuracy of neoplastic superficial spreading evaluation. The significant reduction of superficial positive margins dramatically decreases the number of patients at risk and needing a compulsive follow-up policy.
Footnotes
Authors’ Note
This article was presented at the 94th annual meeting of the American Broncho-Esophagological Association; Las Vegas, Nevada; May 14-15, 2014.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.