Impact of Reflex Testing for BRAF Mutational Status in Advanced Melanoma

Approval of this study by the institutional ethics review board was waived as this was a quality improvement study. The laboratory information system of the Eastern Ontario Regional Laboratory Association (EORLA; Ottawa, Ontario, Canada) was queried to retrospectively review specimen accessions that had both pathology report text containing the term melanoma and an order for BRAF molecular testing, received in the laboratory from January 1, 2019, to January 28, 2020. Cases received from January 1, 2019, to July 25, 2019, and from July 26, 2019, to January 28, 2020, were designated as before or after implementation of reflex BRAF testing guidelines, respectively. All EORLA hospital sites were included, representative of an academic tertiary care center (The Ottawa Hospital, Ottawa, Canada) and 3 affiliated community hospitals (Queensway Carleton Hospital and Montfort Hospital in Ottawa, Canada, and Cornwall Community Hospital in Cornwall, Canada). External consultation cases were excluded as these cases were referred to EORLA specifically for molecular testing.

The pathology reports corresponding to each accession were reviewed to include only bona fide cases of melanoma with nodal, intralymphatic (in transit, microsatellite or satellite), or distant metastases for further analysis, inclusive of all stage III and IV disease as defined by the 8th edition AJCC (American Joint Committee on Cancer) Cancer Staging Manual,¹⁵  as these patients were eligible for funded testing through the Ontario provincial health plan. Patient data including age, sex, disease stage, and BRAF mutational status were recorded from the laboratory information system.

Cases with reflex testing were defined as those with BRAF testing ordered on or before the date that the pathology report was finalized; all other cases were considered as having routine testing. Cases of reflex testing were further categorized as being clinician requested (noted on the pathology requisition) or pathologist initiated at the time of diagnosis.

The electronic medical record for each patient was reviewed to determine subsequent clinical treatment decisions. Factors including adjuvant or primary use of systemic therapy, choice of therapeutic agent, date of first dosage, and whether there was intentional treatment delay via interval CLND, were all noted.

To assess for trends in BRAF test ordering patterns over the analyzed time interval, numbers of BRAF orders for both total melanoma cases and advanced (stage III/IV) melanoma cases were plotted by month and analyzed by linear regression.

Various laboratory TAT metrics were calculated. Total TAT was determined as number of days from specimen receipt to reporting of BRAF mutational status, and time intervals for each major step in the laboratory process contributing to total TAT were also assessed. TAT metrics were compared between cases grouped on the basis of whether routine or reflex BRAF testing was performed.

For patients who received systemic therapy, time from the date of pathologic diagnosis of stage III/IV melanoma to date of first dose of systemic therapy was calculated, and analyzed on the basis of usage of routine or reflex BRAF testing. Subgroup analysis was also performed on cases with or without interval CLND.

Statistical analysis was performed with Graphpad Prism 7 software (San Diego, California). Fisher exact test and Student t test were used for comparative analyses, with a 2-tailed P < .05 considered statistically significant.

In total, 85 cases of melanoma were identified for which BRAF testing was ordered, of which 41 and 44 were received in the periods before and after reflex testing guideline implementation, respectively. Of these cases, 65 represented advanced (stage III or IV) melanoma and were included for further analysis, with 28 and 37 of these cases received in the preguideline and postguideline periods, respectively. Excluded were cases where BRAF testing was ordered before known metastases, such as cases with high-risk stage II disease with patients enrolled in a clinical trial, cases of primary mucosal melanoma, and cases of early-stage cutaneous disease for which BRAF testing was requested for unspecified reasons. Linear regression analysis performed on both total melanoma cases for which BRAF testing was ordered (Figure, A) and included cases of advanced melanoma (Figure, B) showed no significant change in the rate of BRAF testing; however, there was a slight upward trend toward increased BRAF testing over time for cases of advanced melanoma.

Before reflex testing guidelines, BRAF testing for most cases (24 of 28; 86%) was ordered by the clinician after the pathology report was issued (Table 1). This was significantly reduced to 13 of 37 cases (35%; P < .001) after guideline implementation, corresponding to a significant increase in pathologist-initiated reflex testing from 3 of 28 cases (11%) to 23 of 37 cases (62%; P < .001). It was rare for clinicians to explicitly request reflex BRAF testing on the initial pathology requisition (“clinician-requested reflex testing”), occurring in just 2 cases overall. In both cases, the patients had a known history of melanoma, presented with extensive metastatic disease, and received immunotherapy empirically before biopsy. Aside from the differences in BRAF test ordering methods, there were no significant differences in patient age bracket, sex, disease stage, BRAF mutational status, or treatment decisions between the preguideline and postguideline periods.

BRAF mutations were detected in 32 of 65 cases (49%), consistent with the frequency cited in the literature.¹⁶  Most patients (43 of 65; 66%) received systemic therapy after pathologic diagnosis either in the form of adjuvant treatment in resected stage III disease or primary treatment in unresectable stage III or stage IV disease. Of the patients with BRAF mutations, 8 of 32 (25%) received targeted therapy, with the remainder receiving immunotherapy. Three patients, all with stage IV disease, received immunotherapy empirically before a pathologic diagnosis of advanced melanoma and were not included in subsequent TTT analysis.

Cases that ultimately underwent routine or reflex testing, regardless of time period, were compared with respect to laboratory TAT metrics (Table 2). Cases considered as routine testing included those where pathologists did not adhere to the implemented reflex testing guidelines. For cases with reflex testing, mean total TAT from the date of specimen receipt to the date of final BRAF result was reduced by 65% from 52.5 ± 5.6 to 18.6 ± 1.0 days (P < .001). As expected, this corresponded to the elimination of the postanalytic time interval from the date of the finalized pathology report to the date that BRAF testing was ordered, from 31.6 ± 5.5 to −1.1 ± 0.4 days (P < .001). Otherwise, no significant differences in TAT were observed in the preanalytic phase of specimen processing (time from specimen receipt to pathologist assignment), or the analytic phases of pathologist diagnosis (pathologist assignment to finalized report) and BRAF assay completion (BRAF ordered to BRAF resulted).

TTT from pathologic diagnosis of advanced disease to first dose of systemic therapy was calculated for patients who received systemic therapy, inclusive of 28 cases with routine testing and 15 cases with reflex testing (Table 3). For patients who received reflex testing, overall mean TTT was reduced by 45% from 86.9 ± 9.4 to 47.9 ± 7.8 days (P = .009). To control for intentional treatment delays, subgroup analysis was performed as based on the presence or absence of interval CLND, which was done in a subset of patients with stage III disease. For patients who did not undergo interval CLND, TTT was significantly reduced from 71.7 ± 11.4 to 37.7 ± 4.6 days (P = .02). In contrast, no significant difference in TTT was observed in patients for whom interval CLND was performed (118.9 ± 10.9 versus 110.5 ± 22.5 days; P = .75).

Pathologist adherence to reflex testing guidelines varied depending on practice type (Supplemental Table 1, see supplemental digital content at https://meridian.allenpress.com/aplm in the December 2022 table of contents). Post guideline implementation, the rate of pathologist-initiated reflex testing in eligible cases of advanced melanoma was highest for subspecialized academic dermatopathologists (14 of 18; 78%), followed by academic non-dermatopathologists (9 of 14; 64%), and lowest for pathologists at affiliated community hospitals practicing in a general sign-out model (1 of 5; 20%). Adherence to reflex testing guidelines was also noted to trend toward improved consistency toward the end of the analyzed period.

To our knowledge, the present study is the first to assess the potential impact of reflex BRAF testing in melanoma. We demonstrate that reflex BRAF testing in advanced melanoma significantly improves laboratory TAT with clinical translation to a reduction in TTT. These results are consistent with recent studies showing the impact of reflex biomarker testing in lung cancer^12,14  and serve to highlight the importance of collaboration between oncologists and pathologists for optimizing patient care in an era of genomically selected therapies.¹³ 

Resource management and appropriateness of testing is an important potential barrier for implementation of laboratory quality improvement initiatives. This is especially relevant in our practice within Canada, and in other jurisdictions with publicly funded health care systems constrained by finite resources. A potential drawback to reflexive testing is the possibility that testing could be ordered unnecessarily, such as in cases where targeted therapy is contraindicated or unfeasible. In our study, we did not observe a statistically significant change in the overall rate of BRAF molecular testing in total or advanced melanomas over our analyzed period before and after implementation of pathologist-initiated reflex testing. Although there was a slight upward trend in the rate of BRAF testing in advanced melanomas, review of the electronic medical record for the patients included in the study did not reveal any clear instances where testing was unwarranted, nor did we receive feedback from clinicians regarding cases where BRAF molecular testing was performed but not needed.

We found that reflex testing expedited TTT by approximately 1 month, particularly in patients who did not undergo intentional treatment delay via interval CLND. It was unsurprising that TTT was unchanged in patients who underwent interval CLND, since BRAF testing could be ordered by the clinician and resulted during that interval. It is yet unclear whether this reduction in TTT bears clinical significance with regard to patient outcome. Regardless, in an era of patient-centered care, reducing the number of follow-up visits required and eliminating potentially anxiety-provoking waiting times should alone be viewed as a considerable benefit.

Ongoing advances in melanoma treatment suggest a broadening role for systemic therapy in the future, and increasing utility for reflex BRAF testing. Clinical trials are currently underway to assess effectiveness of adjuvant immunotherapy and TKIs in high-risk stage II disease.¹⁷  After completion of this study, eligibility guidelines for BRAF testing have changed in Ontario and all patients with melanoma of Breslow depth greater than 2 mm may receive funded BRAF mutational testing, increasing the number of eligible patients. Furthermore, there is an expected decline in the usage of CLND owing to a lack of survival benefit and new indications for adjuvant treatment options,^2,18  which would increase the proportion of patients who could see a TTT benefit of reflex testing.

As to be expected for the early stages of a new initiative, our study revealed imperfect adherence amongst pathologists after implementing reflex testing guidelines. Even amongst academic dermatopathologists who were aware of the reflex testing initiative, there was a notable nonadherence rate that could be attributed to human error and forgetfulness. Community pathologists in a general sign-out practice were most likely to be nonadherent, probably reflecting a lower frequency of melanoma cases seen by this cohort. In line with this finding, a previous study has shown that the practice of reflex microsatellite instability testing was slower to be adopted in community hospitals than in national cancer centers.¹⁹  In our affiliated community hospitals, there were no resource-specific barriers to implementation, as they use the same central laboratory as the academic hospitals. To improve awareness, the pathologists received email feedback whenever a clinician requested BRAF testing on a missed case that met criteria for reflex testing. We observed that overall guideline adherence improved over time; however, it would be prudent for any laboratory implementing reflex BRAF testing to consider additional methods to ensure a widespread and timely adoption by all members of the department. This may include email reminders, or targeted reminders through the laboratory information system.

Throughout the study period, positive feedback was received from clinicians regarding the reflex testing initiative, who found it useful to have the pathology report and BRAF testing results by the time a patient presented for either postoperative follow-up, or for their first medical oncology appointment. It has also resulted in a net decrease in overall workload for pathologists, as fewer steps are required to initiate testing. One area for further improvement in overall TAT to BRAF molecular testing results is on biopsy samples originally processed at external private laboratories that do not have access to molecular testing. Implementation of reflex BRAF testing in these private laboratories would likely result in further reductions in TAT and TTT, as it would eliminate the delay associated with retrieval and send out of materials.

The primary limitation to this study was a small sample size resulting from a relatively short-term preinterventional and postinterventional analysis, which was necessitated by the recency of our reflex testing guideline implementation. Furthermore, variation in clinical course between individual patients, including possible treatment delays due to insurance funding, illness, or complications were not accounted for in the inclusion. Longer-term follow-up studies would be necessary to study impacts on patient outcomes, and for additional subgroup analysis. Finally, for the purposes of time and cost savings, immunohistochemistry with a VE1 monoclonal antibody detecting BRAF V600E mutations has also emerged as a potential first-line screening test²⁰ ; however, at the time of this study, BRAF mutational analysis by immunohistochemistry was not funded by the Ontario provincial health plan and was thus beyond the scope of this study.

Overall, our results support the routine use of pathologist-initiated reflexive BRAF testing in advanced melanomas. Reflexive testing eliminates a substantial postanalytic TAT for clinician-submitted requests, and translates to a significant reduction in TTT.