High-risk cutaneous squamous cell carcinoma (SCC) survey questions with number of surgeons responding to each question. CT indicates computed tomography, MRI, magnetic resonance imaging; PET, positron emission tomography; XRT, radiation therapy.
Perineural invasion survey questions with number of surgeons responding to each question. CT indicates computed tomography, MRI, magnetic resonance imaging; PET, positron emission tomography; and SCC, squamous cell carcinoma.
Jambusaria-Pahlajani A, Hess SD, Katz KA, Berg D, Schmults CD. Uncertainty in the Perioperative Management of High-Risk Cutaneous Squamous Cell Carcinoma Among Mohs Surgeons. Arch Dermatol. 2010;146(11):1225-1231. doi:10.1001/archdermatol.2010.323" cmelink="1
Copyright 2010 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2010
To evaluate whether Mohs surgeons' management of high-risk cutaneous squamous cell carcinoma (HRCSCC) is uniform regarding radiologic nodal staging (RNS) and adjuvant radiation therapy (ART).
A survey study of randomly selected, fellowship-trained Mohs surgeons.
An academic medical center.
American College of Mohs Surgery members who responded to an e-mail invitation completed either a survey regarding management of HRCSCC (n = 117) or SCC with perineural invasion (PNI) (n = 118). Participants totaled approximately 25% of the American College of Mohs Surgery membership.
Main Outcome Measures
(1) Percentage of patients with HRCSCC referred for RNS, sentinel lymph node biopsy (SLNB), or ART over the preceding 12 months; (2) top factors leading surgeons to consider RNS, SLNB, or ART; and (3) acceptance of ART for clinical scenarios of various degrees of PNI.
Most respondents cited PNI and in-transit metastasis as top factors leading to consideration of RNS, SLNB, or ART. Otherwise, there was no consensus regarding use of, or indications for, RNS, SLNB, or ART.
The lack of consistency between experts indicates that there is equipoise regarding indications for RNS and ART in HRCSCC. There is also wide variation in RNS and ART practices among Mohs surgeons who are specifically trained to manage HRCSCC. Clinical trials should therefore be conducted in these areas as there is no clear standard of care.
Cutaneous squamous cell carcinoma (CSCC) is the second most common type of skin cancer, and its incidence is rising.1 Although overall cure rates are high, a subset of tumors recur or metastasize,2- 4 resulting in a reported 4% risk of nodal metastasis and a 1.5% risk of disease-specific death overall.5While characteristics of these high-risk tumors have not been consistently defined, such factors as poorly differentiated histologic features, tumor diameter of 2 cm or more, invasion deeper than fat, and perineural invasion (PNI) have been associated with poor outcomes.2- 4,6,7 Host factors, including immunosuppression, also appear to affect prognosis.8- 10
Several studies have shown that PNI, particularly in cases of CSCC in which patients experience neurologic symptoms, carries a particularly poor prognosis.2,4,11,12 More commonly, patients with CSCC with PNI are asymptomatic, and PNI is discovered incidentally on biopsy or during surgical resection. The prognosis and management of such asymptomatic PNI (termed subclinical PNI) has not been well defined.
Though not precisely defined, CSCC with an elevated risk of recurrence or metastasis has been termed high-risk CSCC (HRCSCC). There are several published guidelines designed to help physicians manage HRCSCC (Table 1).13- 17 These guidelines have been developed through consensus by panels of persons who are experienced in CSCC treatment and who attempt to make recommendations based on the existing medical literature. Many of these consensus panels are multidisciplinary13,14,16 and have been composed of dermatologists,13- 17 dermatologic surgeons,13,15,17 pathologists,13,17 surgical oncologists,13 otorhinolaryngologists,13,16 plastic surgeons,14 medical oncologists,13,14 and radiation oncologists.13,14,16 Current recommendations vary. Surgical excision with clear margins is emphasized, and most consensus guidelines recommend meticulous margin assessment to confirm that the tumor has been removed. However, indications for radiologic nodal staging (RNS) or adjuvant radiation therapy (ART) are not consistent across guidelines and are often not mentioned, likely because of insufficient data on which to base clear recommendations. Decisions regarding RNS and ART are key elements in treating patients with HRCSCC, yet clinicians have little data and no clear guidelines to aid decision making.
Clinical equipoise refers to situations in which “there is genuine uncertainty within the expert medical community—not necessarily on the part of individual investigators—about the preferred treatment.”18 When equipoise exists, a clinical trial should be conducted because there is no clear standard of care.18 The lack of consistency in current expert recommendations regarding RNS and ART in HRCSCC indicates that clinical equipoise is present with regard to these aspects of management. We undertook this study to provide further evidence that there is uncertainty about these elements of management among physicians who treat HRCSCC. We conducted a questionnaire-based study to determine whether there is a lack of uniformity among Mohs surgeons regarding RNS and ART in the management of HRCSCC. Although many different subspecialists are involved in the treatment of HRCSCC, we chose to survey management practices among Mohs surgeons because their clinical practice is largely focused on the management of skin cancer, including HRCSCC. Therefore, as a group, Mohs surgeons are likely to have accumulated a large clinical experience in managing HRCSCC, particularly in operable cases without clinical evidence of nodal metastases, in which decisions regarding RNS and ART commonly arise.
We conducted a survey study of American College of Mohs Surgery (ACMS) members. Membership in ACMS requires the completion of a 1- to 2-year fellowship in Mohs surgery (or its equivalent if training was completed before formal fellowships were available). We used a random number generator to select 500 ACMS members who were listed in the 2005 ACMS membership directory and then sent them an e-mail invitation to complete an anonymous online questionnaire. The first 250 persons on the randomly generated list were invited to complete a questionnaire regarding the perioperative management of HRCSCC (Figure 1); the remaining 250 persons were invited to complete a questionnaire focusing on the management of CSCC with subclinical PNI (Figure 2). Both surveys were administered via the Web site www.surveymonkey.com. The study was approved by the institutional review board of the University of Pennsylvania, Philadelphia.
In the questionnaires, HRCSCC was defined as any CSCC that the surgeon thought had a 10% or greater risk of metastasis, and subclinical PNI was defined as the presence of an asymptomatic CSCC in which tumor cells were abutting or wrapping around a nerve fiber that was away from the main mass of the tumor (Figures 1 and 2). Questions focused on the use of RNS, sentinel lymph node biopsy (SLNB), and ART. Questions in the PNI questionnaire also asked surgeons to imagine themselves as having tumors with varying degrees of PNI and whether they would want ART. Two and 4 weeks after sending the initial invitation, we e-mailed reminders to nonresponders. Responses were collected for 6 weeks after the initial invitations were sent.
The study was designed to have 80% power (to define a maximum 15% width of the 95% confidence interval [CI]) to determine the top 3 reasons for recommending RNS, SLNB, or ART in the management of HRCSCC, assuming that 20% or more of surgeons would rank the same high-risk factor as “most important” and that 50% of those invited would complete the survey. The percentage of patients referred for RNS, SLNB, or ART was calculated for each surgeon by dividing the number of patients with HRCSCC who were referred for each modality by the total number of patients with HRCSCC who were treated by each surgeon during the preceding 12 months. Respondents were divided into quartiles based on the number of patients with HRCSCC who were treated and the number of years in practice. Kruskal-Wallis tests were used to determine whether the percentage of patients who were referred for RNS, SLNB, or ART differed according to the number of years in practice, practice location (private practice vs academic practice vs combination of practices), or number of patients with HRCSCC who were treated during the preceding 12 months. Stata version 9.0 (Stata Corp, College Station, Texas) was used for all statistical analyses.
Among 250 surgeons who were invited to complete the HRSCC questionnaire, 117 (47%) completed at least part of the survey (Figures 1 and 2). Response to the PNI questionnaire was similar, with 118 of 250 invitees (47%) completing all or part of the survey. The characteristics of the respondents, including number of years practicing Mohs surgery, practice setting, and number of HRCSCCs managed in the previous 12 months, are listed in Table 2.
Most respondents (75%; 95% CI, 66%-82%) had not referred any patients with HRSCC for SLNB within the past 12 months. Practice patterns were otherwise divided regarding the use of RNS and ART. When RNS was ordered, management was divided regarding the imaging modality used (Table 3). The largest fraction of respondents reported using computed tomography, with 31 of 106 surgeons (29%; 95% CI, 21%-39%) routinely using this modality when looking for subclinical nodal metastasis.
Most surgeons cited PNI in patients with HRSCC as 1 of 3 top factors leading them to consider RNS (70% of respondents; 95% CI, 60%-78%), SLNB (51%; 95% CI, 41%-61%), or ART (85%; 95% CI, 77%-91%). Satellite or in-transit metastasis was also considered by most surgeons to be a top reason for RNS (66%; 95% CI, 56%-75%), SLNB (74%; 95% CI, 64%-82%), or ART (65%; 95% CI, 55%-74%). Otherwise, opinion was divided as to which risk factors warranted RNS or ART.
Most surgeons (98%; 95% CI, 94%-100%) would personally decline ordering ART for their patients if perineural inflammation (but no nerve invasion) was seen histologically. Opinion was divided for scenarios of dermal and subcutaneous nerve involvement. Most surgeons (92%; 95% CI, 86%-96%) would also accept ART for involvement of temporal nerve, particularly if it involved the parotid gland. Most surgeons (77%; 95% CI, 68%-84%) reported that they do not use cytokeratin immunostains to confirm a diagnosis of suspected PNI. Overall, there was little consensus regarding the use of additional Mohs margins or permanent sections, referral for additional non-Mohs surgery, RNS, SLNB or lymphadenectomy, ART, or close clinical follow-up alone in cases of PNI. However, in cases in which Mohs margins are unobtainable, most surgeons reported never using close follow-up alone (55%) and always referring the patient for ART (54%).
Surgeons who treated more patients with HRCSCC referred significantly fewer patients for ART (Kruskal-Wallis test, P < .001 [Table 4]). Similarly, those who treated more PNI SCCs recommended ART significantly less often when presented with clinical scenarios of intermediate nerve involvement (Figure 2, questions 7 and 8). The difference was most apparent in question 8, in which 65 of 90 respondents (72%) who had treated fewer than 8 cases of PNI in the past year would accept ART vs only 10 of 28 respondents (36%) who had treated 8 or more cases of PNI (χ2 = 12.3; P < .001). Otherwise, years in practice, practice location, and number of HRCSCCs or PNI cases treated had no significant effect on approach to managing HRCSCC or SCC with PNI (P > .05).
Equipoise regarding RNS and ART in HRCSCC is evidenced by inconsistent and nonspecific clinical recommendations developed by multidisciplinary expert panel members who have reviewed the available data (Table 1).13- 17 We hypothesized that there is subsequent variation among fellowship-trained Mohs surgeons, a subset of physicians who are specifically trained to treat HRCSCC, in these aspects of HRCSCC management. The results of our study demonstrate that there is little consistency in current practice.
Most surgeons cited satellite or in-transit metastasis or PNI as top factors leading them to consider recommending RNS or ART. Therefore, these factors appear to be considered particularly high risk. However, our findings indicate that surgeons make management decisions differently based on the extent of PNI. There was a consensus against the use of ART in cases of perineural inflammation without evidence of nerve invasion. Yet, opinion was divided regarding the acceptance of ART when dermal or subcutaneous nerves are involved. Consensus reemerges in favor of ART in scenarios of named nerve involvement, with an 80% acceptance of ART for temporal nerve involvement and a 92% acceptance when parotidectomy is required. These findings that more extensive nerve involvement may warrant more aggressive treatment than less extensive nerve involvement likely reflect available data indicating that named nerve involvement is associated with a poor prognosis12 and that small nerve involvement may not carry a poor prognosis.19 Although this latter study was not yet published when data were collected, clinicians might have a sense that small nerve invasion is not as worrisome as more pronounced nerve involvement based on their experience. This clinical acumen might explain why surgeons with greater experience in treating PNI and HRCSCC referred significantly fewer patients with HRCSCC for ART and accepted ART themselves only in scenarios of more advanced nerve involvement.
Most surgeons (75%) had not recommended SLNB for patients with HRCSCC. However, a significant minority (25%) had used this diagnostic tool, which may represent an emerging concept in HRCSCC management for certain patient subsets. Our study found no other areas of consensus regarding RNS or ART decision making. The results of this study are not surprising given that there are no clinical trials regarding RNS or ART in HRCSCC. Two reviews of published data regarding SLNB and ART have highlighted the lack of information and areas for future work.20,21 The lack of data has likely made it difficult for expert panels to make firm recommendations in these areas and has led to the current state of equipoise.
Several limitations were present in our study. First, only fellowship-trained Mohs surgeons in the United States were sampled. Other physicians who treat patients with HRCSCC, including non–Mohs-trained dermatologists and craniofacial, head and neck, and plastic surgeons, may use RNS and ART differently. Countries outside the United States with different health care systems may also have different thresholds for using RNS and ART options. Whether treatment of HRCSCC varies among physicians from different subspecialties and in different countries merits further study. However, it is unlikely that care is uniform given the equipoise that is present in current recommendations, as discussed herein. A second limitation is that surgeons who responded to the questionnaire might treat patients differently from nonresponders. This limitation is present in any survey study. However, the 47% response rate was expected at the outset as per our power calculations, and the acceptable 97% CIs indicate that the sample size was adequate. The randomly selected 235 study participants represented approximately 25% of US fellowship-trained Mohs surgeons (ACMS) at the time of data collection. This is a high proportion of a physician subgroup to be included in a survey study, which increases the likelihood that the study results are reflective of the fellowship-trained Mohs surgeon population at large. Third, HRCSCC was defined as a 10% or higher estimated risk of metastasis. Results might differ for other definitions of HRCSCC.
In conclusion, the lack of clarity and consistency between multidisciplinary expert consensus guidelines13,14,17 indicates that equipoise exists regarding indications for RNS and ART in HRCSCC. Clinical trials should therefore be conducted in these areas as there is currently no set standard of care. The results of this study demonstrate that there is wide variability among fellowship-trained Mohs surgeons in the use of, and perceived indications for, RNS and ART in HRCSCC. These results underscore the need for high-quality data on which guidelines of care can be based. Clear guidelines providing indications for RNS and ART may improve patient outcomes.
Correspondence: Chrysalyne D. Schmults, MD, MSCE, Mohs Micrographic Surgery Center, 1153 Centre St, Ste 434, Jamaica Plain, MA 02130 (email@example.com).
Accepted for Publication: May 18, 2010.
Author Contributions: All authors 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: Hess, Katz, Berg, and Schmults. Acquisition of data: Jambusari-Pahlajani and Hess. Analysis and interpretation of data: Jambusaria-Pahlajani, Hess, Katz, and Schmults. Drafting of the manuscript: Jambusaria-Pahlajani, Hess, and Schmults. Statistical analysis: Jambusaria-Pahlajani, Katz, and Schmults. Administrative, technical, and material support: Hess. Study supervision: Schmults.
Financial Disclosure: None reported.