Australian general practitioners’ attitudes and knowledge of sentinel lymph node biopsy in melanoma management

In Australia, general practitioners (GPs) are the first point of contact for the majority of patients who develop melanoma. A GP is therefore the person who will most often make a preliminary diagnosis, take the first biopsy and, following pathological confirmation of melanoma, decide on the need for referral for specialist care. Some GPs with a special interest in skin cancer also perform wide excisions and flap repairs.¹ The NSW Melanoma Patterns of Care study,² a population-based survey of in situ and invasive melanoma management in NSW in 2006–07, found the initial melanoma diagnosis was managed in general practice for 36% of patients, in skin cancer clinics (usually staffed by GPs) for 17%, in dermatology practices for 26%, by surgeons for 13% and by others for <1%. Approximately half of patients initially managed by a GP are then referred to a specialist or multidisciplinary centre.²

Melanoma prognosis is closely related to whether a melanoma has spread to regional or distant parts of the body. Sentinel lymph node biopsy (SLNB) was introduced in 1992³ and is performed after lymphatic mapping to provide information about the presence of metastatic melanoma in the regional lymph node basin. SLNB is important for staging and prognostic assessment, and to guide patient management including appropriateness for systemic adjuvant therapies and clinical trials.⁴ To optimise the accuracy of the test, SLNB should be performed in a centre with expertise in the procedure, including specialist nuclear medicine, surgery and pathology.⁵

Current Australian guidelines recommend that SLNB be considered for all patients with melanoma >1 mm in thickness and for patients with melanoma >0.8 mm in thickness with other high-risk pathological features.⁵ SLNB should be performed at the time of the primary tumour wide excision.⁵ Historically, adherence to SLNB guidelines has not been optimal, allowing for patients who may refuse or not be suitable for SLNB. Data from the NSW Melanoma Patterns of Care study indicated that SLNB was undertaken for 45% of patients diagnosed with a melanoma >0.8 mm Breslow thickness who were potentially eligible for SLNB.⁶ The Multicenter Selective Lymphadenectomy Trial (MSLT-I) showed improved disease-free survival but not overall melanoma-specific survival following SLNB.⁷ There have been conflicting interpretations of the results from MSLT-I and other studies examining the benefits of SLNB in relation to the therapeutic value^8,9 and survival benefit of SLNB.^10,11 Potential reasons for low uptake of SLNB include confusion about the evidence, lack of awareness of the guidelines, and individual preferences of GPs and patients. The aim of this study was to examine the knowledge and attitudes of GPs regarding the role of SLNB in the management of patients with invasive primary melanomas, to assist development of and adherence to guidelines.

Methods

Mixed methods study design

Quantitative and qualitative data were collected in the form of questionnaires and interviews. Informed consent was obtained.

Sampling and recruitment

Recruitment of GPs was conducted at two Australian GP meetings: the Royal Australian College of General Practitioners (RACGP) annual conference in Queensland in October 2018, and a GP skin cancer–focused continuing medical education workshop in Sydney in December 2018; participants were also recruited through other GP professional communications. GPs were eligible to participate in the study if they had practised in Australia in the previous year. Overall, 231 GPs completed a questionnaire. Of these, 23 also completed an interview and were reimbursed $100 for their time.

Questionnaire and interview guide development

A cross-sectional questionnaire and semi-structured interview guide (Supplementary file; available online only) were developed from a literature review and discussion with a multidisciplinary team of melanoma clinicians and researchers. The questionnaire was completed on paper or electronically and consisted of 24 items covering demographic characteristics, knowledge of melanoma guidelines, clinical management of melanoma, referral patterns, and attitudes to SLNB and shared care. The questionnaire data were managed using REDCap.¹²

Questionnaire data analysis

Postcodes were classified using the Accessibility Remoteness Index of Australia (ARIA)¹³ and Socio-Economic Indexes for Areas (SEIFA)¹⁴ classifications. Univariable and multivariable prevalence ratios and 95% confidence intervals (CIs) were calculated using Poisson regression models to examine factors associated with familiarity with Australian melanoma clinical practice guidelines and knowledge about and attitudes towards SLNB. Factors assessed were age, sex, type of practice, years of practice, number of invasive melanomas diagnosed in a year, ARIA, SEIFA and exposure to articles or talks about SLNB. Analyses were conducted using SAS version 9.4. The STROBE guidelines were followed to report the findings.¹⁵

Qualitative data analysis

Interviews (conducted by SR and ALS) lasted 16–39 minutes (mean length: 24 minutes) and were audio-recorded and professionally transcribed. The de-identified transcripts were read by two members of the research team (SR and ALS). Data were compared within and across interviews to identify commonalities, differences and patterns in the data and to develop a coding framework and thematic map.¹⁶ The codes and themes were discussed with the research team and refined until consensus was reached. Coding was conducted using NVivo12 (QSR International, Australia) software.

Results

Participant characteristics

Characteristics of GPs who participated in the study are shown in Table 1. Of the GPs who completed the questionnaire, 46% worked in group medical practices, 32% in independent practice, 14% in skin cancer clinics and 8% in other types of practice. The majority of GPs (83%) diagnosed fewer than six patients presenting with melanoma per year.

Familiarity with Australian clinical practice guidelines for SLNB

Questionnaire results

Self-reported familiarity with the guidelines for melanoma management varied: 30% were unfamiliar or somewhat unfamiliar; 38% were a little familiar; and 32% were quite familiar or very familiar. Overall, 31% of GPs reported they had checked the recent update of the melanoma guidelines. In univariable analyses, GPs were more likely to claim familiarity with the guidelines if they saw ≥6 patients/year with melanoma when compared with ≤5 patients/year (prevalence ratio 2.84; 95% CI: 1.77, 4.55), worked in a skin cancer clinic when compared with medical centre (prevalence ratio 2.90; 95% CI: 1.60, 5.24) or had been exposed to articles or talks about SLNB (prevalence ratio 4.19; 95% CI: 2.15, 8.18). In multivariable analysis, exposure to articles or talks about SLNB was the only factor that remained statistically significantly associated with familiarity of the guidelines (prevalence ratio 3.55; 95% CI: 1.71, 7.39).

Of the 164 GPs (72%) who reported that they would discuss SLNB and would refer eligible patients for discussion of the risks and benefits of SLNB with a surgeon (Table 2), 39% indicated they would tell a patient that SLNB would be appropriate if a melanoma’s Breslow thickness was >1.0–2.0 mm; responses for different Breslow thickness criteria are shown in Table 3. Three-quarters of GPs indicated they would arrange other tests if their patient was eligible for SLNB. The most commonly ordered tests were computed tomography (CT) of the chest, abdomen and pelvis (36%); whole-body positron emission tomography (PET)/CT (28%); and regional node ultrasonography (18%; Table 3), despite the Australian guidelines recommending no investigations prior to SLNB.

Attitudes and views about SLNB for melanoma management

Questionnaire results

Overall, 68% of GPs believed that SLNB had an important role in the management of patients with melanoma, 30% were unsure and 2% disagreed (Table 2). Most GPs (60%) had been exposed to articles or talks about SLNB in the past three years (Table 2); in a multivariable analysis, this was associated with less uncertainty about the importance of SLNB (prevalence ratio 0.58; 95% CI: 0.35, 0.96) and less hesitation in discussing SLNB with patients (prevalence ratio OR 0.58; 95% CI: 0.34, 0.99).

Of the 164 GPs (72%) who said that they would discuss SLNB with their patients, 71% indicated a role of SLNB for staging and prognostic information, 67% thought SLNB was useful to assess suitability for adjuvant systemic therapies for patients who were node-positive, 62% considered that the results would influence patient management, 40% considered there was a likely survival benefit for performing SLNB and 30% indicated that SLNB assisted selection of patients for completion of lymphadenectomy (Table 3). The factors most influential in their decision to discuss or recommend SLNB were Breslow thickness (92%), lymphovascular invasion (66%), presence of palpable lymph nodes (67%) and the likelihood that the result would influence patient management (54%; Table 3).

Table 3 also presents GPs’ preferences for referral of patients for whom SLNB would be applicable for definitive management, and preferences for the degree of involvement in patient follow-up care following a positive or negative SLNB result.

Of the 64 GPs (28%) who indicated they would not usually discuss SLNB with or recommend SLNB for a patient with newly diagnosed melanoma eligible for SLNB, 47% said they did not know enough about SLNB. Under the option ‘Other, please specify’, 29% reported that recommendations were the role of whoever was caring for the patient’s melanoma. Other reasons (not shown) included unconfirmed survival benefit (14%), access to facilities (9%) and lack of additional prognostic information (8%).

Interview results

The questionnaire data indicated a substantial proportion of participants were unable to correctly identify the Breslow thickness at which patients should be referred for discussion of SLNB. The interview data (Table 4) provided a possible explanation for this knowledge gap by revealing that many of the GPs did not perceive the SLNB guidelines to be relevant to their practice and many had a limited working knowledge of these guidelines. They frequently expressed the belief that decisions related to SLNB were the responsibility of the specialist and not the GP. The interview data also indicated considerable variation in GPs’ knowledge of SLNB guidelines and perceived utility of SLNB (Table 4). There was also considerable variation in the time point at which a GP would refer a patient, with some referring for biopsy, some performing biopsies but referring once a melanoma diagnosis was confirmed, and others performing wide local excisions.

Discussion

The data indicate that while GPs have generally favourable attitudes to SLNB, and they understand that Breslow thickness is important for guiding selection of patients for SLNB, they lack knowledge about the Breslow thickness cut-off points specified in the guidelines for referral of patients eligible for SLNB. GPs who saw higher volumes of patients with invasive melanoma were more likely to be familiar with the melanoma management guidelines; however, they were in the minority, and many of the participants did not feel sufficiently informed about the role of SLNB or which investigations should be performed at the time of SLNB. GPs reporting exposure to articles or talks about SLNB in the past three years were more likely to believe SLNB had an important role and to discuss SLNB with their patients.

For GPs who would discuss SLNB with and recommend SLNB to an eligible patient with a newly diagnosed melanoma, Breslow thickness was the most common reason for recommending SLNB, followed by lymphovascular invasion. Interestingly, the presence of palpable lymph nodes was also a common reason for recommending SLNB, even though a palpable node implies that nodal metastases are already present, negating the need for SLNB. Familiarity with recommendations for SLNB according to Breslow thickness varied among GPs. Of GPs who indicated that they would consider discussing SLNB with or recommending SLNB for patients with melanoma, the most frequently selected Breslow thickness cut-off points were >1.0–2.0 mm (37% of GPs) and >2 mm (30% of GPs). The guidelines recommend that SLNB be considered for all patients with a Breslow thickness of >1.0 mm and for patients with Breslow thickness 0.8–1.0 mm if they have other high-risk pathological features. Patients with intermediate-thickness melanomas may have better disease-free survival and lower rates of local and regional recurrence following SLNB.¹⁷ In addition, SLNB has been shown to have prognostic value for patients with thick (>4 mm) clinically lymph node–negative melanoma.¹⁸

The results of this study indicate that many GPs do not perceive the SLNB guidelines as being relevant to them and do not know enough about SLNB to recommend the procedure. The study did not assess the details of how SLNB is discussed or communicated with patients, how that might differ with level of knowledge or how many patients may have declined SLNB; this is an area that could be addressed in future studies. The present findings are based on self-report and may be subject to social desirability bias; for example, if some GPs inaccurately reported being familiar with the guidelines for melanoma management. GPs may claim they ‘know’ the guidelines from information gained from their peers and opinion leaders but may not necessarily have a good understanding of them.¹⁹ Lack of knowledge of the guidelines may not be problematic (refer to ‘Key points’ in Table 4) if GPs are not performing biopsies of suspicious lesions; however, for GPs who do perform biopsies of suspicious lesions, it is likely some melanomas will be >0.08 mm in thickness. In this case, it is important that the GPs appreciate how the guidelines relating to SLNB may affect their practice and their referral decisions, and therefore a patient’s access to contemporary melanoma treatment including adjuvant therapy post–surgical resection. This barrier could be addressed through GP education. Another possibility for improving adherence to guidelines so that all eligible patients can be offered the procedure is the inclusion of a comment in the histopathological report of a melanoma along the lines of: ‘SLNB should be a consideration for this patient. If it is to be performed, it should be done at the same time as wide local excision’.

There is also evidence that inappropriate tests are being ordered for patients following diagnosis of melanoma. The guidelines indicate there is no evidence to support chest X-rays, PET/CT or magnetic resonance imaging for initial staging of patients with thin or intermediate Breslow thickness primary melanoma (Stage I–IIB).²⁰ PET/CT for Stage IIC patients may be considered. Nodal ultrasonography is recommended only when SLNB has been declined or is not possible,²¹ since ultrasonography has lower sensitivity and specificity when compared with SLNB.²² Access to facilities was noted as a barrier to performing SLNB by some GPs based in regional and remote centres, and not all eligible patients are referred to clinicians who perform SLNB.

A slight majority (56%) of GPs preferred to manage patient follow-up in a shared care arrangement, which relies on good communication between the GP and specialist and clear delineation of roles.^23,24 A limitation of the present study is that it was not possible to compare the results for GPs working in skin cancer clinics against those working in general practice because of relatively few GPs from skin cancer clinics participating in the study. A third of GPs were unsure about the role of SLNB, which might be related to general lack of awareness or to conflicting information in journals and presentations on the benefits and harms of SLNB.^9–14 A strength of the study is that a relatively large and representative sample of GPs was surveyed. These findings are most relevant to Australia but may be applicable to other countries with comparable health systems and melanoma management practices. Difficulties with and resistance to adhering to guideline recommendations have been acknowledged by GPs in other settings.^25–27 Reasons for non-compliance to guidelines by GPs included limited knowledge, which was affected by the time available to search for information while consulting, and the perception that guidelines may not take into account specific patient needs. Resistance may also be experienced as a result of beliefs that guidelines are ‘imposed’ or not trusting recommendations.^23–25

Some people argue that SLNB is never indicated because of the lack of survival benefit; however, all evidence-based guidelines – including national (Australian, The National Comprehensive Cancer Network, British Association of Dermatologists) and international (European Society for Medical Oncology) guidelines – recommend that SLNB be considered for a melanoma >1.0 mm in thickness or >0.8 mm Breslow thickness with additional high-risk features.

Conclusions

GPs are at the front line of melanoma management in Australia. Although GPs had a generally satisfactory level of understanding of the benefits of SLNB, many were not familiar with the guidelines regarding which patients should be considered for SLNB, and there was evidence that inappropriate investigative tests may be ordered prior to referral. These results indicate a need for further education and discussion with GPs about the role of SLNB in melanoma management, targeting those who biopsy or excise melanomas. For GPs who are not familiar with the melanoma guidelines or who are not confident in managing primary melanoma, all patients with invasive melanoma should ideally be referred to specialised units for management. These findings will inform the development, dissemination and evaluation of education strategies and other initiatives to improve GPs’ knowledge of SLNB and ultimately lead to better patient outcomes.