xRead Articles - November 2022

The Laryngoscope V C 2012 The American Laryngological, Rhinological and Otological Society, Inc.

Extracapsular Dissection for Benign Parotid Tumors: A Meta-Analysis

W. Greer Albergotti, BA; Shaun A. Nguyen, MD, MA; Johannes Zenk, MD, PhD; M. Boyd Gillespie, MD, MSc

Objectives/Hypothesis: Recent studies suggest that extracapsular dissection (ECD) is an option for the resection of cer tain benign parotid tumors. This study investigates complication rates and effectiveness of ECD versus superficial parotidec

tomy (SP) for the treatment of primary benign parotid neoplasms. Study Design: Systematic literature review with meta-analysis.

Methods: Studies available for inclusion evaluated the complications and effectiveness of ECD and SP as surgical techni ques for the treatment of solitary, benign parotid tumors. An Ovid/Medline search revealed nine articles that met inclusion criteria. A critical review and meta-analysis of these articles was performed. Results: The included studies evaluated a total of 1,882 patients. There was no observed difference in tumor recurrence between the ECD and SP groups (odds ratio [OR], 0.557; 95% confidence interval [CI], 0.271-1.147). There was a significantly lower rate of transient facial nerve paresis (OR, 0.256; 95% CI, 0.174-0.377) in the ECD group (59 of 741; 8.0%) compared to the SP group (81 of 397; 20.4%); however, there was no observed difference in permanent facial paralysis between the ECD and SP groups (OR, 0.878; 95% CI, 0.282-2.730). Frey’s syndrome was less often observed (OR, 0.117; 95% CI, 0.071 0.191) after ECD (27 of 602; 4.5%) compared to SP (75 of 287; 26.1%). Conclusions: This systematic review with meta-analysis suggests that ECD has a similar recurrence rate as SP with fewer postoperative complications. ECD may be considered an alternative surgical modality for select benign parotid neoplasms. Key Words: Parotid neoplasm, parotid tumor, parotid surgery, parotidectomy, superficial parotidectomy, extracapsular dissection. Level of Evidence: 2a Laryngoscope, 122:1954–1960, 2012

INTRODUCTION Salivary gland neoplasms account for approximately 3% of all head and neck tumors. 1,2 Approximately 80% of these tumors are found in the parotid gland, most often in the superficial lobe, where 80% of the tumors are benign, with pleomorphic adenoma being the most common subtype. 2 Before the 1930s, intracapsular enu cleation, or ‘‘shelling out,’’ of the tumor from the parotid gland was considered the treatment of choice for pleo morphic adenomas because of their benign clinical course, in addition to surgeon fear of damaging the fa cial nerve. It was at this time, however, that McFarland recognized the high recurrence rate of up to 45% follow ing intracapsular enucleation. 3 Bailey later attributed this high recurrence rate to incomplete removal of the tumor and therefore advocated superficial parotidectomy From the Department of Otolaryngology–Head and Neck Surgery, Medical University of South Carolina ( W . G . A ., S . A . N ., M . B . G .), Charleston, South Carolina, U.S.A.; and the Department of Otorhinolaryngology, University of Erlangen-Nu¨ rnberg ( J . Z .), Bavaria, Germany. Editor’s Note: This Manuscript was accepted for publication April 9, 2012. The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to M. Boyd Gillespie, MD, MSc, Department of Otolaryngology, 135 Rutledge Avenue MSC 550, Charleston, SC 29425-5500. E-mail: gillesmb@musc.edu

(SP) with special attention to the removal of the capsule to cure pleomorphic adenomas. 3–5 As SP was widely adopted, recurrence rates for pleomorphic adenoma fell, settling at a current rate of approximately 2%. However, while the recurrence rate improved, the number of complications increased, includ ing a higher incidence of temporary and permanent facial nerve paralysis. Because of the frequency of com plications from a surgery for a benign condition, surgeons have questioned the need to remove the entire superficial lobe while putting the facial nerve at risk. Extracapsular dissection (ECD) has been offered as an alternative method to minimize the morbidity of parotidectomy. ECD involves careful dissection around the tumor capsule under magnified visualization with out preidentification of the facial nerve. 6 Several large series suggest that ECD has lower rates of complica tions without a higher recurrence rate compared to SP. 7,8 Nevertheless, superficial and/or partial parotidec tomy remains the gold standard treatment at most centers. The objective of this study was to compare the out comes of SP with those of ECD in the treatment of benign parotid tumors using a systematic review and meta-analysis. Specifically, we tested the null hypothesis that there is no difference between ECD and SP in the rate of tumor recurrence, facial nerve weakness, and Frey’s syndrome.

DOI: 10.1002/lary.23396

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors

1954

Fig. 1. Steps taken in the determination of papers to be included in the meta-analysis. ECD ¼ extracapsular dissection. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

MATERIALS AND METHODS An Ovid/Medline literature search was performed using the following search terms: parotid neoplasm, extracapsular dis section, capsule, ECD, and dissection. There were no limits placed on study design. We included studies that compared ECD to SP with regard to at least one outcome of interest (recurrence, facial weakness, Frey’s syndrome) in the surgical management of solitary, clinically benign parotid nodules. For the purpose of this study, ECD is defined by capsular dissection of the tumor with a thin margin of surrounding gland without planned identification of the main trunk of the facial nerve. SP is defined as both complete and partial SP, where a portion of the superficial lobe is removed with the tumor after planned identification and dissection of the facial nerve. Studies were excluded if they included recurrent or multiple tumors where the data could not be separated from primary, solitary tumors, involved nonparotid salivary tumors, or included malignant neoplasms. Article abstracts were reviewed to determine whether the inclusion and exclusion criteria were met. The initial search returned 79 results, of which 14 included original data on ECD for benign, parotid neoplasms (Fig. 1). These 14 articles were obtained and reviewed in detail. A total of eight studies were excluded because there were no data on SP (6 studies) or involved an older series later updated by a single author group (2 studies). An additional three studies were identified that met inclusion criteria through a review of references and a concurrent PubMed search. A comprehensive, retrospective data review was performed on the included articles with MedCalc 12.1 (MedCalc Software; Mariakerke, Belgium) using a fixed-effects meta-analysis model. The fixed-effects model assumes that all studies come from a common population and that the effect size (odds ratio [OR]) was not significantly different among the different trials. The Mantel-Haenszel method was used for calculating the weighted summary OR under the fixed-effects model. The heterogeneity

statistic was incorporated to calculate the summary OR under the random-effects model (DerSimonian and Laird). The random effects model gave a more conservative estimate (i.e., with wider confidence interval [CI]), but the results from the two models usu ally agreed where there was no heterogeneity. The Q value, a measure of the heterogeneity of the included studies, tested the null hypothesis that the studies are homogeneous. If the test of heterogeneity was statistically significant ( P < .05), then more emphasis would have to be placed on the random-effects model. The following null hypotheses were tested comparing ECD to SP: 1) No difference exists in the rate of Frey’s syndrome; 2) No dif ference exists in the rate of transient facial nerve paresis; 3) No difference exists in the rate of permanent facial nerve paralysis; and 4) No difference exists in rate of tumor recurrence. Effect sizes were calculated using ORs with respective 95% CI. If the value 1 is not within the 95% CI, then the OR is statistically sig nificant at the 5% level ( P < .05), which means that the study did not demonstrate a significant difference between ECD and SP. RESULTS Nine studies clearly met inclusion criteria and were chosen for analysis (Table I). The nine studies evaluated included a total of 1,882 patients, of which 1,102 under went ECD and 780 underwent SP. The median follow-up time was 12 years, ranging from 2 to 32 years. The ma jority of the studies were performed with a retrospective cohort design, although one was prospective. Among the studies included, there were five countries that were represented with a disproportionate number of cases being reported from Europe. The indications used by the authors for ECD varied between studies, but the tumors were generally superficial, clinically benign tumors of the parotid gland. Several limited the size of the tumor

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors 1955

0–2 yr

Mean Follow-up Time (Range) Unknown

18 yr (11–25 yr)

54 mo (minimum 2 yr)

10.3 yr for ECD (3–21 yr); 8.3 yr for SP (3–22 yr)

Median, 18 yr; (6–26 yr)

9 yr for ECD; 8 yr for SP

194 mo (range, 117–264 mo)

Median, 12 yr (range, 5–30 yr)

No. of Subjects Surgical Selection Criteria

Gleave et al. 6 1979 United Kingdom Retrospective cohort No 369 Decision of ECD vs. SP made at time of surgery; factors that favored ECD included mobility, thin covering, size large enough to allow for digital manipulation SP performed for larger tumors or for visualization of the facial nerve

SP used for tumors in the body of the parotid or deep lobe

Natvig and Soberg 24 1994 Norway Retrospective cohort No 198 Only considered patients with pleomorphic adenoma; unclear criteria used for the determination of surgical technique

Hancock 7 1999 United Kingdom Retrospective cohort No 101 Tumors that were superficial, mobile considered for ECD; converted to SP if boundary was lost between tumor and parotid parenchyma or for deeper, less mobile tumors were smaller than 2.5 cm; unclear selection criteria for SP

pleomorphic adenoma treated with ECD vs. SP (all tumors smaller than 4 cm, superficial, mobile)

McGurk et al. 20 2003 United Kingdom Retrospective cohort No 662 ECD performed on discrete, mobile tumors less than 4 cm in diameter; ECD vs. SP decision made at time of skin flap based on mobility Uyar et al. 27 2011 Turkey Prospective cohort No 41 All patients had FNA-proven pleomorphic adenoma in the superficial lobe of the parotid; unclear criteria used for the determination of surgical technique ECD ¼ extracapsular dissection; SP ¼ superficial parotidectomy; FNA ¼ fine-needle aspiration.

Martis 12 1983 Greece Retrospective cohort No 176 ECD used for superficial tumors, up to 2.5 cm; Prichard et al. 26 1992 United Kingdom Retrospective cohort No 46 ECD used for tumors in tail of the parotid;

Marti et al. 9 2000 Greece Retrospective cohort No 249 ECD applied to benign, superficial tumors that Witt 23 2002 United States Retrospective cohort No 40 Retrospective analysis of outcomes of

TABLE I.

Characteristics of Included Studies. Authors Year Country Study Design Randomization

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors

1956

TABLE II. Data Abstraction From Included Studies.

Temporary Facial Nerve Paresis

Permanent Facial Nerve Paralysis

Frey’s Syndrome

Tumor Recurrence

Authors

ECD

SP

ECD

SP

ECD

SP

ECD

SP

Gleave et al. 6

5/257

2/112

Martis 12

0/98 0/31

0/78 1/15

Prichard et al. 26

1/31

2/15

0/31

1/15

0/31

6/15

Natvig and Soberg 24

0/5

5/193

Hancock 7 Marti et al. 9

2/28

6/73

0/28

0/73

0/28

18/73

0/28

0/73

6/139

18/110

Witt 23

0/20

4/20

0/20

0/20

0/20

2/20

0/20

0/20

McGurk et al. 20

50/503

48/159

9/503

3/159

27/503

48/159

9/503

8/159

Uyar et al. 27

0/20

3/20

0/20

0/20

0/20

1/20

0/21

0/20

ECD ¼ extracapsular dissection; SP ¼ superficial parotidectomy.

to 2.5 cm, and others used a cutoff of 4 cm. One study suggested that the tumor should be large enough for digital manipulation to be considered for ECD. With the exception of one study, fine-needle aspiration cytology (FNAC) and preoperative imaging were not routinely used unless there was suspicion of a malignant process. The number of patients in each study included in this analysis varied from 40 to 662, with the percentage of patients who underwent ECD in each study also varying widely, from 2.5% to 75% (Table II). The recurrence rates for ECD and SP were 1.5% (14 of 963 cases) and 2.4% (16 of 670 cases), respectively; these findings were not statistically significant (OR, 0.557; 95% CI, 0.271-1.147) (Fig. 2). There was no differ ence in follow-up time specified in most of the studies; however, two of the articles mentioned a slightly longer follow-up for ECD cases compared to SP (9 and 10.3 years vs. 8 and 8.3 years, respectively).

The rates of transient facial nerve paresis for ECD and SP were 8.0% (59 of 741) and 20.4% (81 of 397), respectively, demonstrating a mean reduction of 75% in transient facial nerve paresis for ECD when compared to SP (OR, 0.256; 95% CI, 0.174-0.377) (Fig. 3). There was no difference in the rate of permanent facial nerve paral ysis for the two techniques; however, 1.4% (8 of 590) of ECD and 1.1% (3 of 268) of SP cases experienced this complication (OR, 0.878; 95% CI, 0.282-2.730) (Fig. 4). Symptomatic Frey’s syndrome was reported by 4.5% (27 of 602) of ECD patients compared to 26.1% (75 of 287) of SP patients, which is an 88% reduction in symp tomatic Frey’s syndrome in the ECD group compared to the SP group (OR, 0.117; 95% CI, 0.071-0.191) (Fig. 5). None of the reports elucidated the methodology for eval uating Frey’s syndrome. The studies likewise provided insufficient detail to determine the timing of the evalua tion of Frey’s syndrome, which may be relevant because

Fig. 2. Forest plot of recurrence rates comparing extracapsular dissection (ECD) to superficial parotidectomy (SP). Dots to the left of the line favor ECD. Any finding that crosses the midline is con sidered to be a nonsignificant finding. Meta-analysis fails to reject the null hypothesis that there is no difference in the recurrence rates of benign parotid tumors between ECD and SP. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Fig. 3. Forest plot of transient facial nerve rates comparing extracapsular dissection (ECD) to superficial parotidectomy (SP). Dots to the left of the line favor ECD. Any finding that crosses the midline is considered to be a nonsignificant finding. Meta analysis reveals significantly decreased rates of transient facial nerve damage following ECD as compared with SP. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors 1957

study found a 4% risk of facial paresis after ECD of tumors 4 cm or less in size compared to a 21% paresis rate in tumors larger than 4 cm. 13 Although most groups only consider ECD for superficial lobe tumors, several have used it for deep lobe parotid tumors as well. 14–16 This practice should not be widely employed because of an observed recurrence rate of 10% in deep lobe lesions compared to a rate of 3% for superficial tumors. 15 There fore, if considering ECD, the tumor should be solitary, clinically benign, mobile in two planes, smaller than 4 cm, and located in the superficial lobe. Preoperative imaging is likely to assist in the determination of whether these criteria are met. Most of the studies included in this analysis did not use preoperative imaging or FNAC for the evaluation of tumor histology; however, it should be noted that the majority of these studies were published before the rou tine use of these technologies. More recent articles on ECD not included in this analysis have used preopera tive imaging and FNAC in their eligibility criteria of tumors for the application of ECD. 11,14,17,18 We would recommend the use of FNAC and appropriate imaging in the evaluation of tumors being considered for ECD. Pre operative imaging will reveal whether a given tumor is solitary, has a size of 4 cm, is isolated to the superficial lobe, has no suspicious adenopathy, and is therefore a potential candidate for ECD. FNAC will demonstrate features worrisome for malignancy in up to 80% of ma lignant tumors; however, the surgeon should have a low threshold for conversion to SP if there are any intraoper ative concerns given the relatively high 20% false negative rate for malignancy on FNAC. ECD was developed as a surgical modality after it was demonstrated that SP frequently involves an ele ment of ECD in areas where the tumor capsule makes contact with the facial nerve. Although contact between the tumor and the nerve has been reported to be as

Fig. 4. Forest plot of permanent facial nerve paresis rates compar ing extracapsular dissection (ECD) to superficial parotidectomy (SP). Dots to the left of the line favor ECD. Any finding that crosses the midline is considered to be a nonsignificant finding. Meta-analy sis reveals no difference in the rates of permanent facial nerve damage following ECD versus SP. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

symptomatic Frey’s syndrome in the immediate postop erative period may lessen over time.

DISCUSSION Surgery remains the mainstay of treatment for both benign and malignant neoplasms of the parotid gland. Surgery for parotid gland tumors has two primary goals: complete removal of the tumor and functional preserva tion of the facial nerve. In particular, the surgical method should ensure complete tumor removal to pre vent recurrent tumor, which is more difficult to surgically cure and involves greater risk of permanent facial nerve injury. 6 Although a variety of surgical tech niques exist, SP remains the most widely practiced despite increasing consideration of minimally invasive techniques. ECD is a minimally invasive approach that differs from classic enucleation, which involves the incision and shelling out of the contents of the tumor capsule. In ECD, the dense parotid fascia overlying the tumor is sharply incised followed by a blunt dissection to the level of the tumor. 9,10 Under magnified visualization, a loose areolar plane may be seen 2 to 3 mm adjacent to the tu mor and is the preferred plane of dissection. Careful dissection continues along the tumor capsule to prevent rupture of the small outpouchings of the tumor that may be encountered. In this method, unlike other forms of parotidectomy, planned identification of the facial nerve is not performed, although use of a facial nerve integrity monitor is advocated. Branches of the facial nerve may be encountered deep to the tumor and must be carefully dissected away from intervening parotid tissue. In general, ECD has been applied to small, benign, superficial parotid tumors. Most authors apply the method to smaller tumors with reported cutoffs of between 2.5 and 4 cm. 11,12 The risk of facial nerve injury increases during ECD with increasing tumor size. One

Fig. 5. Forest plot of Frey’s syndrome rates comparing extracap sular dissection (ECD) to superficial parotidectomy (SP). Dots to the left of the line favor ECD. Any finding that crosses the midline is considered to be a nonsignificant finding. Meta-analysis reveals decreased incidence of Frey’s syndrome in ECD versus SP. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors

1958

increased risk of recurrence between ECD and SP. 23 Other authors have found similar results. 24 Rupture of the capsule with macroscopic tumor spill during surgery cannot be considered a curative opera tion. 24,25 Rupture has been reported to occur in 5% to 10% of SPs. 23,25 Published rates of capsular rupture with ECD range from 0% to 9.2% with a calculated mean of 5.8%. 7,13,15,16,18,23 Rupture of the capsule is cer tainly a surgical complication to be avoided; however, it does not appear that there is any increased risk in this adverse outcome with the use of ECD. 23 ECD should be avoided if there is any concern for malignancy in the preoperative work-up, including a risk of malignancy identified by FNAC or imaging. Intraopera tive findings suspicious for malignancy such as tethering, induration, nerve invasion, or suspicious lymphadenopa thy should likewise lead to conversion to a superficial or total parotidectomy. Every precaution should be taken; however, it is inevitable that a few cases of malignancy will be subject to resection by ECD. McGurk et al. reported that in their series of clinically benign parotid tumors, 32 of 662 patients were ultimately found to have carcinoma, of whom 12 were subjected to ECD, with 10 year survival rates being broadly similar whether the patient received SP or ECD. 20 Although it is not clear that there is any difference in outcomes between surgical tech niques in these clinically benign but ultimately malignant tumors, ECD should be avoided. In the case that malig nancy is discovered postoperatively, completion SP should be considered and remains a viable option because the main landmarks of the facial nerve are still largely intact. There are several weaknesses of this analysis, including selection bias to which procedure was per formed and lack of randomization. It is thus very likely that patients who underwent the two methods had dif ferent types of tumors and that ECD was selectively applied to more manageable nodules. An additional weakness of several studies is the reporting of only the results for pleomorphic adenoma. Although this is the entity most targeted by ECD, it is important to note the recurrence rates and complications for other benign tumors and tumors that were clinically benign but fol lowing resection were found to not be benign. CONCLUSION No difference in the rate of tumor recurrence or permanent facial paralysis was noted between SP and ECD when applied to solitary, clinically benign nodules of the parotid gland. ECD appears to have a decreased incidence of transient facial nerve palsy and Frey’s syndrome, although there is the possibility that the technique was applied to less challenging tumors on average. Although ECD should not be seen as a replacement for SP, it appears to be safe and a poten tially advantageous alternative when applied to specific parotid tumors by experienced surgeons. SP remains the standard of care for treatment of benign parotid tumors owing to its proven track record, but ECD may be considered by surgeons trained in its application.

frequent as 98% of cases, most reports indicate it is in contact 39% to 51% of the time. 11,18–20 Thus, in theory the risk to the facial nerve should be the same or less with ECD compared to SP, which involves planned dis section of a longer segment of facial nerve. This theory is validated by the finding of a significantly lower rate of temporary facial paresis following ECD compared to SP in this systematic review. It should be noted, how ever, that the lower rate of facial paresis observed in these studies came from high-volume tertiary referral centers. Therefore, there is the potential that ECD could result in higher rates of facial nerve paralysis in less-experienced hands if the surgeon is less familiar with the common pathways and patterns of facial nerve branches. Frey’s syndrome is a relatively common occurrence following surgery of the parotid gland, and its incidence increases with progressively invasive surgery owing to further disruption of parotid tissue. An incidence as high as 50% has been observed after SP depending on the method used to assess gustatory sweating. Sympto matic complaints of Frey’s syndrome, however, are less common, with 5% of patients undergoing SP bothered enough to seek treatment. 21,22 This review found a significantly lower rate of Frey’s syndrome after ECD (4.5%) compared to SP (25%). All the included studies assessed gustatory sweating clinically without starch io dine testing, which could have increased the sensitivity but has questionable clinical relevance. The authors did not clarify whether patients were evaluated for Frey’s syndrome in the immediate postoperative period or more distant from the resection. Certainly, the prevalence of Frey’s syndrome in this population would be expected to decrease with distance from surgery. It is also not clear from the articles how many of the patients reporting these symptoms actually sought treatment for it. It is worth noting that this review could not control for the potential that ECD and SP may have been applied to different subsets of tumor. For instance, the largest se ries by McGurk et al. evaluated a total of 662 patients with clinically benign parotid tumors and determined the type of surgery based upon the tumor’s exam once the skin flap had been raised. The likelihood that more chal lenging tumors underwent SP in that study is apparent when comparing the rate of malignancy in these clinically benign tumors (12.5% in SP vs. 2.4% in ECD). 20 Because of this methodology, the meta-analysis was run without this study without a statistical difference in the recurrence rate (1.6% in SP vs. 1.1% in ECD). However, it cannot be overlooked that in all of the studies, SP was likely applied to more challenging tumors. It should also be noted that an ideal follow-up period to truly assess recurrence would be 10 years due to the late recurrence of pleomorphic ade noma. Although most of the studies we evaluated had a mean follow-up of more than 10 years, several did not, and it can be expected that the absolute value for recur rence is somewhat higher than we have reported. Unlike capsular rupture, capsular exposure does not appear to affect recurrence. Witt studied 60 cases of pleomorphic adenoma and found that microscopic capsu lar exposure occurs universally and results in no

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors 1959

BIBLIOGRAPHY 1. Leegaard T, Lindeman H. Salivary-gland tumours. Clinical picture and treatment. Acta Otolaryngol Suppl 1969;263:155–159. 2. Eneroth CM. Histological and clinical aspects of parotid tumours. Acta Otolaryngol Suppl 1964;188:S191–S199. 3. McFarland J. 300 mixed tumors of the salivary glands of which 69 recurred. Surg Gynecol Obstet 1936;63:457–468. 4. Bailey H. The surgical anatomy of the parotid gland. BMJ 1948;2: 245–248. 5. Janes RM. The treatment of tumours of the salivary glands by radical excision. Can Med Assoc J 1940;43:554–559. 6. Gleave EN, Whittaker JS, Nicholson A. Salivary tumours–experience over thirty years. Clin Otolaryngol Allied Sci 1979;4:247–257. 7. Hancock BD. Clinically benign parotid tumours: local dissection as an al ternative to superficial parotidectomy in selected cases. Ann R Coll Surg Engl 1999;81:299–301. 8. McGurk M, Renehan A, Gleave EN, Hancock BD. Clinical significance of the tumour capsule in the treatment of parotid pleomorphic adenomas. Br J Surg 1996;83:1747–1749. 9. Marti K, Zografos GC, Martis C. Extracapsular excision of small benign tumors of the parotid gland. J Surg Oncol 2000;75:208–209. 10. Nahlieli O, ed. Modern Management Preserving the Salivary Glands . Israel: Isradon Publishing House; 2007:272. 11. Smith SL, Komisar A. Limited parotidectomy: the role of extracapsular dissection in parotid gland neoplasms. Laryngoscope 2007;117: 1163–1167. 12. Martis C. Parotid benign tumors: comments on surgical treatment of 263 cases. Int J Oral Surg 1983;12:211–220. 13. Piekarski J, Nejc D, Szymczak W, Wronski K, Jeziorski A. Results of extracapsular dissection of pleomorphic adenoma of parotid gland. J Oral Maxillofac Surg 2004;62:1198–1202. 14. Fukushima M, Miyaguchi M, Kitahara T. Extracapsular dissection: mini mally invasive surgery applied to patients with parotid pleomorphic ad enoma. Acta Otolaryngol 2011;131:653–659.

15. Dallera P, Marchetti C, Campobassi A. Local capsular dissection of parotid pleomorphic adenomas. Int J Oral Maxillofac Surg 1993;22:154–157. 16. Anderson R. Benign Mixed Tumors of the Parotid Gland. International Congress Series 1975 . Amsterdam: Excerpta Medica; New York: Ameri can Elsevier; 1975:245–249. 17. George KS, McGurk M. Extracapsular dissection–minimal resection for be nign parotid tumours. Br J Oral Maxillofac Surg 2011;49:451–454. 18. Shehata EA. Extra-capsular dissection for benign parotid tumours. Int J Oral Maxillofac Surg 2010;39:140–144. 19. Iizuka K, Ishikawa K. Surgical techniques for benign parotid tumors: seg mental resection vs extracapsular lumpectomy. Acta Otolaryngol Suppl 1998;537:75–81. 20. McGurk M, Thomas BL, Renehan AG. Extracapsular dissection for clini cally benign parotid lumps: reduced morbidity without oncological com promise. Br J Surg 2003;89:1610–1613. 21. Bradley PJ. Pleomorphic salivary adenoma of the parotid gland: which operation to perform? Curr Opin Otolaryngol Head Neck Surg 2004;12: 69–70. 22. Guntinas-Lichius O, Klussmann JP, Wittekindt C, Stennert E. Parotidec tomy for benign parotid disease at a university teaching hospital: out come of 963 operations. Laryngoscope 2006;116:534–540. 23. Witt RL. The significance of the margin in parotid surgery for pleomorphic adenoma. Laryngoscope 2002;112:2141–2154. 24. Natvig K, Soberg R. Relationship of intraoperative rupture of pleomorphic adenomas to recurrence: an 11-25 year follow-up study. Head Neck 1994;16:213–217. 25. Henriksson G, Westrin KM, Carlsoo B, Silfversward C. Recurrent primary pleomorphic adenomas of salivary gland origin: intrasurgical rupture, histopathologic features, and pseudopodia. Cancer 1998;82:617–620. 26. Prichard AJ, Barton RP, Narula AA. Complications of superficial paroti dectomy versus extracapsular lumpectomy in the treatment of benign parotid lesions. J R Coll Surg Edinb 1992;37:155–158. 27. Uyar Y, Caglak F, Keles B, Yildirim G, Salturk Z. Extracapsular dissection versus superficial parotidectomy in pleomorphic adenomas of the pa rotid gland. Kulak Burun Bogaz Ihtis Derg 2011;21:76–79.

Laryngoscope 122: September 2012

Albergotti et al.: ECD for Benign Parotid Tumors

1960

The Laryngoscope © 2021 The American Laryngological, Rhinological and Otological Society, Inc.

Different Surgical Strategies in the Prevention of Frey Syndrome: A Systematic Review and Meta-analysis Armando De Virgilio, MD, PhD ; Andrea Costantino, MD ; Elena Russo, MD; Fabio Ferreli, MD ; Raul Pellini, MD; Gerardo Petruzzi, MD ; Jacopo Zocchi, MD ; Giuseppe Spriano, MD; Giuseppe Mercante, MD Objective/Hypothesis: To de fi ne the best surgical technique able to reduce Frey syndrome (FS) incidence after parotidectomy. Study Design: Systematic review and network meta-analysis. Methods: An arm-based network analysis was conducted using a Bayesian hierarchical model. The primary and second ary outcomes were the incidence of subjective (clinical) and objective (positive starch-iodine test result) FS, respectively. Results: A total of 3830 patients with a median age of 50.35 years (n = 2323; IQR 44.25 – 54.18) were included for six interventions [temporoparietal fascia (TPFF), free fat graft (FFG), acellular dermal matrix (ADM), sternocleidomastoid muscle (SCM) fl ap, and super fi cial musculoaponeurotic system (SMAS) fl ap]. If compared to no treatment, the greatest reduction of subjective (clinical) FS incidence was measured for the TPFF (OR: 0.07, CI: 0.004 – 0.57), the ADM (OR: 0.09, CI: 0.02 – 0.35), and the FFG (OR: 0.11, CI: 0.03 – 0.42) techniques. However, a signi fi cant difference was measured also for the SCM fl ap (OR: 0.38, CI: 0.18 – 0.73) and for the SMAS fl ap (OR: 0.42, CI: 0.19 – 0.97). All treatments showed a signi fi cant reduction of the objective FS incidence if compared to no treatment (FFG, OR: 0.06, CI: 0.002 – 0.62; TPFF, OR: 0.07, CI: 0.01 – 0.33; ADM, OR: 0.11, CI: 0.03 – 0.44; SMAS, OR: 0.36, CI: 0.17 – 0.71; SCM, OR: 0.40, CI: 0.19 – 0.74). Conclusions: TPFF, ADM, and FFG seem to be the best treatment strategies to prevent FS after parotidectomy. Further randomized controlled trials comparing these techniques should be conducted to de fi ne speci fi c indications. Key Words: Parotidectomy, free fat graft, temporoparietal fascia, acellular dermal matrix. Laryngoscope , 131:1761 – 1768, 2021

cells of the parotid acini. 3,4 The reported incidence of FS after parotidectomy vary between 12.5% and 62% based on the presence of symptoms, and up to 98% when a Minor ’ s starch iodine test is performed. 5 – 7 The timing of its development ranged from 20 days to 22 months with a median of 11 months. 8 FS often leads to a signi fi cant degree of patient dis satisfaction, and several papers showed that it represents the most serious self-perceived sequelae, along with cos metic dis fi gurement, with signi fi cant potential negative social and psychological implications. 9 – 12 Although the primary aim of a parotidectomy is to completely remove the tumor preserving the facial nerve branches, the possi bility to prevent the occurrence of FS should not be underestimated. Many surgical techniques aiming to prevent this complication have been described. 7,13 – 17 A previous meta analysis performed by Curry et al. 18 demonstrated that the use of these techniques decreases the likelihood of FS symptoms and/or a positive starch-iodine test result. However, the analysis performed by the authors does not allow to de fi ne the best surgical treatment. Some random ized clinical trials 6,19 – 24 and meta-analysis 25 – 28 compared speci fi c techniques with no treatment con fi rming that the prevention of the FS is recommended. Nevertheless, these data are not able to suggest, which technique should be used, and the choice of a speci fi c treatment is nowadays based on the surgeons ’ preference and experience.

INTRODUCTION Frey syndrome (FS), also known as gustatory sweat ing or auriculotemporal nerve syndrome, is characterized by abnormal fl ushing, sweating, and burning sensation of the skin in the preauricular region, induced by mastica tion and salivatory stimulation. 1 This phenomenon was fi rst described by Duphenix in 1757, 2 while the role of the auriculotemporal nerve in the pathogenesis of this condi tion was fi rst reported by Frey in 1923. 1 FS is considered caused by an aberrant nervous regeneration of parasym pathetic fi bers after the disruption of the parotid bed. The fi bers reinnervate sweat glands and the subcutaneous vascular plexus, instead of connecting the secretomotor From the Department of Biomedical Sciences ( A . D . V ., A . C ., E . R ., F . F ., G . S ., G . M .), Humanitas University, Via Rita Levi Montalcini 4, Pieve Emanuele – Milan, Italy; Otorhinolaryngology Unit ( A . D . V ., A . C ., E . R ., G . S ., G . M .), IRCCS Humanitas Research Hospital, Via Manzoni 56, Rozzano – Milan, Italy; and the Department of Otolaryngology Head and Neck Surgery ( R . P ., G . P ., J . Z .), Regina Elena National Cancer Institute, IRCCS, Rome, Italy. Additional supporting information may be found in the online version of this article. Editor ’ s Note: This Manuscript was accepted for publication on January 12, 2021. G . S . and G . M . contributed equally to this work. The authors have no funding, fi nancial relationships, or con fl icts of interest to disclose. Send correspondence to Andrea Costantino, MD, Otorhinolaryngol ogy Unit, IRCCS Humanitas Clinical and Research Center, Via Manzoni 56, 20089 Rozzano, MI, Italy. E-mail: andrea.costantino94@gmail.com

DOI: 10.1002/lary.29414

Laryngoscope 131: August 2021

De Virgilio et al.: Surgical prevention of Frey syndrome 1761

Risk of Bias and Study Quality Assessment Methodological quality of included studies was assessed inde pendently by two separate authors (E.R. and A.D.V.). The quality of each study was assessed according to the Methodological Index for Non-randomized Studies (MINORS) quality assessment tool 30 or the Risk of Bias Assessment Tool for randomized controlled trials (RCTs) recommended by the Cochrane Collaboration, 31 as appropri ate. A funnel plot was created using the effect size for each of the fi ve intervention in order to examine a potential publication bias. Data Synthesis and Statistical Analysis The data from each study were transcribed in tabular form. Categorical variables were summarized by counts and percentage, while continuous variables were reported as median ± interquartile range (IQR: 25th and 75th). Given that only a few RCTs have been published on this topic, an arm-based network analysis was conducted to compare different surgical interventions. In particular, a meta-analysis of proportions was performed for both subjective and objective FS incidence. Net work meta-analysis is a statistical approach able to synthetize direct and indirect evidences about multiple treatment comparisons from independent studies. In particular, an arm-based approach describes population-averaged absolute effect sizes for each treatment arm, and it differs from the contrast-based method that focuses on relative effect size within each study. As a consequence, an arm-based analy sis allows for the inclusion of single-arm studies other than compara tive prospective or retrospective studies, enhancing the robustness of the meta-analysis in the setting of a lack of studies that directly com pare different treatment strategies. 32 A Bayesian hierarchical model based in Markov Chain Monte Carlo (MCMC) convergence simula tion was used. 33 In particular, the number of iterations for the adap tation process was set at 5.000, while the number of iterations in each chain was 100.000. A Bayesian framework incorporates a prior probability distribution based on external evidences, and a likelihood estimate based on the data reported by the included studies. The results of Bayesian inference are indeed probabilistic. The MCMC algorithm generate posterior samples that are used to estimate the parameters of interest (e.g., relative treatment effects and the corresponding credible intervals). Moreover, posterior samples can empirically rank treatments in terms of their probabilities to be the best treatment on the basis of the estimated effect size. Network meta-analysis introduce several criteria to deter mine if one intervention is better than another. The absolute effect size is calculated for each treatment arm in the arm-based approach, but a measure of relative effect size is also derived. In particular, we chose the odds ratio (OR) given that also retro spective studies have been included in the analysis. Moreover, the Bayesian approach allows for treatment ranking based on these data, using the abovementioned MCMC algorithm. 34 Pairwise associations between each treatment modality were depicted by a graphical representation of the network. Network esti mates from all outcome variables were presented as OR with 95% credible intervals (CIs). Rank probability plots (rankogram) were constructed to display the calculated treatment ranks. Posterior density plots were also used to determine the absolute risks of the various treatment alternatives and to identify potential overlapping. Cochrane ’ s Q method was used to assess between studies heterogeneity in each treatment group. I 2 was calculated as a measure of heterogeneity. An I 2 value represents the percentage of total variation across studies caused by heterogeneity rather than by chance. According to the Cochrane criteria, values from 0% to 40% may signify low heterogeneity, 30% to 60% may repre sent moderate heterogeneity, 50% to 90% may represent sub stantial heterogeneity, and 75% to 100% represent considerable heterogeneity. Scatter plots (Baujat plot) were created for each

The aim of this systematic review and meta-analysis was indeed to compare different treatments in the pre vention of FS after parotidectomy. In particular, we tried to de fi ne the best surgical technique able to reduce the subjective and objective FS incidence. MATERIALS AND METHODS The study was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 29 Institutional review board approval and informed consent were not required for this review of previously published studies. No review protocol was registered for this study. Eligibility Criteria This systematic review was conducted according to PICOS acronym: Patients (P), adults underwent partial or total parotidectomy; Intervention (I), temporoparietal fascia (TPFF), free fat graft (FFG), acellular dermal matrix (ADM), sternocleidomastoid muscle (SCM) fl ap, and super fi cial musculoaponeurotic system (SMAS) fl ap; Comparator (C), observation; Outcomes (O), FS inci dence (subjective and objective); Study design (S), both retrospective and prospective studies. Exclusion criteria were as follows: non-English language; insuf fi cient reported data; not extractable data; multiple tech niques; subgroup analysis of patients from a larger study; article type was either review, case report, conference abstract, letter to the editor, or book chapter. In addition, treatments, including less than 50 patients in both outcomes (subjective and objective FS incidence) were excluded to reduce the distortion of the pooled estimates. Data Source and Study Searching A comprehensive electronic search was performed on PubMed/MEDLINE, Cochrane Library, and Google Scholar data bases. Relevant keywords, phrases, and MeSH terms were adjusted to fi t the speci fi c requirements for each of the individual databases. The Medical Subject Heading keywords and phrases searched included: “ parotidectomy, ” “ parotid, ” “ Frey syndrome, ” “ reconstruction, ” and “ prevention. ” To minimize the risk of miss ing relevant data, a cross-reference search of the selected articles was performed, and the “ cited by ” function on Google Scholar was also used to obtain other relevant articles for the study. The last search was conducted on August 15, 2020. Data Collection Process Two independent reviewers (E.R. and A.D.V.) separately conducted the electronic search. The title and the abstract of all studies were reviewed to identify potentially relevant studies. The full-text of studies that appeared to be relevant to our search were downloaded and analyzed to determine eligibility. A third author (G.S.) resolved the con fl ict between reviewers. Data extraction from the included studies was systematically made using a structured form. In particular, one author (E.R.) indepen dently compiled a standardized form to extract the following characteristics of included studies: study design, number of patients, patient demographics, outcome de fi nitions, treatment protocols, and duration of follow-up. Then, another author (A.C.) checked for accuracy. Studies ’ general characteristics are shown in Supporting Table 1. The FS subjective and objective incidence rate was calculated for each study (Supporting Table 1) to be included in the network meta-analysis.

Laryngoscope 131: August 2021

De Virgilio et al.: Surgical prevention of Frey syndrome

1762

intervention. The horizontal axis (x-axis) shows the contribution of each study to the overall Q-test statistic for heterogeneity. The standardized difference of the overall treatment effect is plotted on the vertical axis, showing the in fl uence of each study on the pooled effect size. Studies on top right have greater in fl uence on the results and have a bigger contribution to heterogeneity. 35 Analysis of publication bias was conducted for each treat ment group individually using the effect size of the subjective FS incidence. It was performed by visual inspection of the funnel plot, that is, a graphical representation of the size of trials plot ted against the effect size they report. It assumes small studies are more likely to be susceptible to publication bias than large ones. If publication bias does exist, it is most likely to be due to small negative trials not being published. As a consequence, an asymmetry in the scatter of small studies with more studies showing a positive result than those showing a negative result is evident at visual inspection in case of publication bias. Moreover,

the Egger ’ s regression intercept was measured to statistically examine the asymmetry of the funnel plot. 36 All the analyses were performed using the R software for statistical computing (R version 4.0.1; “ pcnetmeta ” and “ meta ” packages). Statistical signi fi cance was de fi ned as P < .05.

RESULTS Search Results and Studies Description

A fl ow chart of the study identi fi cation process is shown in Figure 1. The search strategy yielded 1,375 arti cles after duplicates were removed. After title and abstract review, 1,286 articles were rejected and full-texts of the remaining 101 papers were obtained and reviewed. After applying the abovementioned eligibility criteria, a

Fig. 1. PRISMA 2009 fl ow diagram. [Color fi gure can be viewed in the online issue, which is available at www.laryngoscope.com.]

Laryngoscope 131: August 2021

De Virgilio et al.: Surgical prevention of Frey syndrome 1763

Fig. 2. Results of the network meta-analysis concerning clinical (subjective) Frey syndrome prevention. (A) network plot of included interven tions. Circles are weighted according to the number of studies, including the intervention, and lines are weighted according to the number of studies comparing the two connected techniques. (B) absolute risk of clinical (subjective) Frey syndrome incidence for each intervention. [Color fi gure can be viewed in the online issue, which is available at www.laryngoscope.com.]

TABLE I. League Table Showing Comparative Effectiveness of Analyzed Interventions in Network Meta-analysis Concerning Clinical (Subjective) Frey Syndrome Prevention. TPFF 1.33 (0.12, 26.80) ADM 1.60 (0.15, 31.89) 1.21 (0.20, 7.16) FFG 5.38 (0.60, 92.14) 4.08 (0.97, 17.74) 3.39 (0.81, 14.21) SCM 6.13 (0.68, 106.10) 4.64 (1.01, 21.17) * 3.87 (0.86, 16.81) 1.13 (0.42, 3.08) SMAS 14.37 (1.75, 233.90) * 10.95 (2.87, 43.29) * 9.11 (2.35, 35.30) * 2.67 (1.37, 5.42) * 2.38 (1.03, 5.29) * None Different treatments are reported in order of clinical (subjective) Frey syndrome prevention ranking. Comparisons should be read from left to right. The esti mate is located at the intersection of the column-de fi ning treatment and the row-de fi ning treatment. Data are reported as OR with 95% CI. An OR value greater than 1 favors the column-de fi ning treatment. * P value statistically signi fi cant.

total of 56 studies 6,8,13,19 – 24,37 – 83 were included in the quali tative and quantitative synthesis. The reasons behind the exclusions of 45 studies are shown in Figure 1. Study ’ s general characteristics are shown in Supporting Table 1. A total of 3830 patients with a median age of 50.35 years (n = 2323; IQR 44.25 – 54.18) were included for six different interventions and 96 treatment arms. A median follow-up of 24 months (n = 2401; IQR 16 – 36.9) was reported by 40 studies. The median follow-up for each intervention is shown in Supporting Table 2. The extent of parotid surgery was reported by 44 studies (n = 2913). The majority of patients (n = 2161; 74.2%) underwent a super fi cial/partial parotidectomy, while only 752 (25.8%) patients were treated with a total/ subtotal parotidectomy. Histopathologic nature of the parotid tumor was reported by 44 studies (n = 2527), with a predominance of benign lesions (n = 2304; 91.2%) com pared to malignant tumors (n = 223; 8.8%).

or retrospective 13,38,39,41,42,44,48,49,51,53 – 55,57,58,61 – 67,70,72,74,75,78,82,83 (n = 2301, 60.1%) non-randomized studies. In particular, 12 prospective (n = 349, 9.1%) and seven retro spective (n = 378, 9.9%) single-arm studies were included, while 22 retrospective case – control (n = 1923, 50.2%) and eight prospective non-randomized studies (n = 890, 23.2%) ful fi lled the abovementioned eligibility criteria. On the other hand, only a minority of patients (n = 290, 7.6%) were enrolled in seven RCTs. 6,19 – 24 MINORS criteria showed an overall mean score of 10.8 ± 1.4 in non-comparative studies (maximum of 16), while a mean score of 15.5 ± 1.8 was measured in compar ative studies (maximum of 24). MINORS scores of indi vidual studies are shown in Supporting Table 3. The methodological quality of included randomized trials is summarized in Supporting Figure 1. Some con cerns of reporting bias were raised for all included RCTs. Five RCTs (n = 224) showed some concerns also for selec tion bias. However, no high risk of bias was highlighted in the included RCTs. The funnel plots generated for the meta-analyses of each intervention are shown in Supporting Figure 2. Egger ’ s regression test showed a symmetric distribution of the points in all funnel plots (ADM, Intercept = 0.36,

Methodological quality and risk of bias of included studies The majority of included studies were prospec tive 37,40,43,45 – 47,50,52,56,59,60,68,69,71,73,76,77,79 – 81 (n = 1239, 32.3%)

Laryngoscope 131: August 2021

De Virgilio et al.: Surgical prevention of Frey syndrome

1764