Response to Comments: MolDX: Melanoma Risk Stratification Molecular Testing

The Decision-DX-Melanoma test is a clinically-useful gene expression profile test for patients with cutaneous melanoma that stratifies the s-year risk of developing metastasis as low or high. It has been validated in retrospective and prospective studies which consistently find shortened recurrence-free, distant metastasis-free, and melanoma-specific survival in patients with a high-risk Class 2 result. Multiple studies have reported the prognostic accuracy of the test independent of other pathological features, indicating that the test adds information about the genetic composition of the tumor that contributes to a patient's individual risk that is otherwise inaccessible by the physician. This added information provides a more comprehensive assessment of a patient's risk of disease progression to improve clinical decision-making. The test has a current LCD providing coverage to guide sentinel lymph node biopsy (SLNB) decisions and a current draft LCD to extend this coverage to guide intensity and modality of patient surveillance for early detection of metastasis. This letter provides my full support for this extension of coverage.

Melanoma metastases present in sentinel lymph nodes as detected by sentinel lymph node biopsy is an accepted practice to detect patients with higher risk of metastasis and melanoma-related death and qualifies patients for referral for adjuvant therapy. However, not every patient qualifies for consideration of sentinel lymph node biopsy, as the false negative rate and complication rate for the procedure only warrants consideration with the risk of positivity is beyond 5%. Thus, there is great utility in a non-invasive test, such as Decision-Dx-Melanoma, that can provide accurate prognostic information without sentinel lymph node biopsy information. Recent studies comparing the predictive value of sentinel lymph node biopsy to the Decision-Dx-Melanoma test shows that the positive predictive value and specificity of the tests are similar, however, the Decision-Dx-Melanoma test has higher sensitivity and detects more patients with eventual recurrence, distant metastasis, and death from melanoma than a positive sentinel lymph node. ln addition, the Decision-Dx-Melanoma test has a very high negative predictive value, higher than that of a positive sentinel lymph node, indicating that a patient with a low-risk result can rest assured they have a truly low-risk of disease progression. When sentinel lymph node biopsy is indicated, the combination of nodal status and Decision-Dx-Melanoma enhances the sensitivity in detecting recurrence, metastasis, and death beyond either method alone. Thus, the use of Decision-Dx-Melanoma to guide sentinel lymph node biopsy decisions (as is covered under the current LCD) and to provide information on the risk of recurrence in patients with or without sentinel lymph node biopsy information creates a powerful tool to improve the management of melanoma patients.

Thank you for your comments.

As a physician who treats many melanoma patients, I would like to add my support for the proposed Local Coverage Determination for the DecisionDx-Melanoma gene expression profile test. In my practice I use DecisionDx-Melanoma to help find patients at higher risk for metastasis (a Class 2 test result) so I can move them to follow-up protocols that are appropriate for their risk. The DecisionDx-Melanoma test provides an independent prediction of risk for metastasis that is clinically important and adds to American Joint Committee on Cancer (AJCC) staging. The existing AJCC staging system misses many high-risk patients, so those who are not identified will lose the opportunity to undergo frequent surveillance that may enable me to detect any metastasis at an early stage. Many studies show that melanoma treatments (both surgical and medical therapies) are more effective and demonstrate better patient outcomes when tumor burden is low. As I see in my practice every day, if identification of metastatic disease can be done at an early stage, patients have a higher chance of improved outcomes. DecisionDx-Melanoma is an important part of my patient care and I encourage you to fully support the proposed LCD.

Thank you for your comments.

Thank you for allowing public comments. I wanted to make a statement on the application of DecisionDx-Melanoma. I have been utilizing this test for the last 4-5 years in my clinical practice (over 200 tests ordered) to assist my patients in understanding their personal risk for melanoma recurrence. Armed with these data, we have been able to follow high risk patients more closely and catch recurrences early, allowing treatment and in some instances cure. Patients at low risk are provided with some piece of mind, which makes a huge difference in their quality of life.

The data has only gotten stronger over the years and supports this being a medically necessary and covered test for patients under both criterion you have laid out in the proposed LCD. Based on the most recent data out of UT Southwestern the proposed coverage policy will open this test up to a broader population where the impact is even greater and a specified population where providers like myself can appropriately stratify risk. I strongly encourage you to hold the proposed policy into final form.

Thank you for your comments.

My name is (removed) and I am a medical oncologist at the VCU Massey Cancer Center and melanoma specialist. I am writing in support of the proposed LCD DecisionDx-Melanoma (DL37725). This expanded coverage provided access to more of my melanoma patients who need the most information to understand their risk of recurrence. We recognize that thin melanomas account for the highest number of deaths from melanoma, compared to thick melanomas (Whiteman 2015), despite the lower risk of the group as a whole. The complexity of thin melanomas is the identification of those at risk of recurrence and death. The DecisionDx-Melanoma assay can provide valuable information for risk stratification. The results of this test are utilized to guide patient management and surveillance plans. For example, a patient with a Class 1A test result identifies low-risk patients that can be spared unnecessary surgical intervention (e.g. sentinel lymph node biopsy [SLNB]) and providers are able to reduce intensive surveillance and treatment. In contrast, a patient with a Class 2B test result is at high risk for metastatic disease and can benefit from intensive management, including SLNB and higher frequency surveillance, along with referrals to clinical trials for high risk early stage disease. In my practice, I offer CT surveillance for early stage class 2B melanomas, and access to trials. For class 1A early stage patients, I may avoid CT scans all together, and generally do not recommend treatment or participation in therapeutic clinical trials. This assay has a direct and meaningful impact on patient care decisions. Thank you for allowing me to comment on the proposed LCD DecisionDx-Melanoma (DL37725).

Thank you for your comments.

I am writing to encourage support for the DecisionDx-Melanoma LCD. The DecisionDx-Melanoma test is well-supported by published literature showing that it can accurately stratify patients diagnosed with cutaneous melanoma into high- and low-risk groups based on tumor biology. This information cannot be provided by anatomic or pathologic staging. The DecisionDx-Melanoma test result can help clinicians detect metastasis at an earlier stage by identifying high-risk patients who were otherwise missed by traditional staging. More intensive surveillance for these patients can detect metastases at a time when tumor burden is low, offering a higher chance at improved surgical and therapeutic outcomes based on multiple published studies. For patients with melanoma 0.3 mm or greater thickness, the DecisionDx-Melanoma test will guide clinical patient management decisions regarding frequency and type of imaging, frequency of office visits, referrals to specialties, and potential for adjuvant therapies. This will allow healthcare resources to be applied to the highest risk patients and enable those at lower risk to receive risk-appropriate care. Because of this I urge you to support the expanded policy as drafted.

Thank you for your comments.

The purpose of my email is to encourage Palmetto to finalize the proposed LCD for DecisionDx-Melanoma as written. At present, the eighth edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual offers valuable prognostic information for patients. In particular, for those with localized disease, Breslow depth offers important staging and prognostic information, and for some, sentinel lymph node biopsy (SLNB) can provide population-based prognostic information.

Nevertheless, studies demonstrate that intermediate thickness SLNB-negative cutaneous melanomas cause twice as many deaths as intermediate thickness SLNB-positive cutaneous melanomas. Moreover, deaths from AJCC T1 cutaneous melanomas (≤1 mm) outnumber deaths from both T2 (1.01–2 mm) and T4 cutaneous melanomas (>4 mm), with similar mortality figures for T1 and T3 cutaneous melanomas (2.01–4 mm).

The DecisionDx-Melanoma test has had a significant positive impact of the management of melanoma. For all of these reasons, I encourage Palmetto to finalize the proposed LCD for DecisionDx-Melanoma as written.

Thank you for your comments.

There are severe limitations of the current available staging criteria in identifying patients with “low-risk” cutaneous melanomas who will go on to develop metastatic disease. Contrary to a positive SLNB, a negative SLNB, which most patients have, does not necessarily equate to a “low-risk” disease.

Furthermore, up to 50% of SLNB eligible patients do not undergo SLNB. These findings are troubling because SLNB is the lone prognostic test listed in current National Comprehensive Cancer Network (NCCN) guidelines. For patients with cutaneous melanoma, who do not meet criteria for SLNB, are unwilling or unable to undergo an additional invasive procedure, or have a negative SLNB, there are no prognostic tests available under current NCCN guidelines to evaluate the metastatic risk.

I support the expanded coverage criteria and urge you to finalize as proposed.

Thank you for your comments.

I appreciate the comment period to provide my perspective on the proposed LCD on DecisionDx-Melanoma. For years I have seen node negative patients go on to develop metastases and question why. The DecisionDx-Melanoma test is the best independent predictor of risk of metastasis. Without knowing who is at high risk for metastasis, we cannot direct surveillance efforts to those who need it. It has been recently demonstrated that intensive surveillance in high-risk patients identifies 80% of metastases before they become symptomatic, allowing identification of patients with low burden of metastatic disease. Multiple studies have shown that surgical management and systemic therapies in melanoma are more effective when disease burden is low, with the majority of complete remissions in patients with lowest tumor burden. Thus, early identification of metastatic disease translates to better outcomes for these patients.

Thank you for your comments.

One of the most challenging aspects of any melanoma case is determining the proper management plan. This is critical from both my perspective as the healthcare provider as well as that of the patient; we both want the most risk-appropriate course of action. DecisionDx-Melanoma allows for exactly that. By including the 0.3mm/greater risk stratification section in a new coverage guidance, CMS is helping support the best clinical decision-making opportunity for patients with a melanoma diagnosis. Thank you for doing your part in the fight against skin cancer.

DecisionDx-Melanoma is a multigene expression test that analyzes 28 genes demonstrated to have differential expression between metastatic and nonmetastatic melanoma tumors. It is performed on formalin-fixed paraffin-embedded tumor tissue specimens of a patient’s primary CM tumor. The test uses a radial basis machine algorithm to classify patients into four risk groups with increasing probability of recurrence and/or metastasis within 5 years of diagnosis: low (Class 1A), intermediate (Class 1B or 2A), or high (Class 2B) risk.

Accurately predict risk of recurrence through four prospective, independent studies totaling 788 patients, and three multicenter, prospectively designed archival tissue studies including 690 patients.^1-5

Have a strong negative predictive value (NPV) for recurrence-free (RFS), distant metastasis-free (DMFS), and melanoma-specific (MSS) survival (99% MSS NPV for Stage I-III patients), and 70% sensitivity to identify patients who developed distant metastasis or died from melanoma, including SLN-negative patients.^6-9

Impact clinical management decisions for 47-53% of patients tested in several retrospective and prospective studies.^10-13

Have the potential for up to 70% reduction in SLNB procedures in the T1-T2 ≥55 year old population, with decrease in cost of care and ability to focus healthcare resources to those patients that need them most.¹⁴

Thank you for your comments.

We recommend that the DecisionDX-Melanoma (DL37725) Local Coverage Determination be approved because the utility and validity of the test has been demonstrated in large prospective and retrospective, single and multicenter studies. Additionally, the test has been independently validated increasing our confidence of the results. These studies show that the DecisionDX-Melanoma test can identify patients at low risk of SLN positivity and that the test accurately predicts risk of metastasis. There has been significant uncertainty in the field about the adequacy and consistency of microstaging melanoma patients. Therefore, there has been a need for a more reliable and unbiased technique to influence patient management decisions, and the DecisionDX-melanoma test meets this need.

Thank you for your comments.

I am writing to express my support for the DecisionDX-Melanoma (DL37725) Local Coverage Determination. I am a dermatologist in private practice and have reviewed the literature on the test and have implemented it into my private practice. The literature supports that a DecisionDX-melanoma Class 1 test result is associated with high survival rates and less than 5% likelihood of sentinel lymph node positivity. Based on the evidence I have incorporated the test into a risk assessment algorithm in order to prevent unnecessary sentinel lymph node biopsies and their associated morbidities from surgical complications. It is also useful to alleviate patient anxiety about metastases risk. Use of this test to guide SLNB determination could reduce unnecessary biopsies by more than 50%. I can attest that physicians including myself are using the test to guide their management decisions.

Thank you for your comments.

The purpose of my email is to provide my perspective on the utility of the genetic expression profile test, DecisionDx-Melanoma. More specifically, I am requesting that the expanded LCD in your current proposal be finalized as it is currently written.

The DecisionDx-Melanoma test is a gene expression profile assay for patients with cutaneous melanoma that differentiates between patients who are at low risk or high risk of developing metastatic disease. This test is medically necessary since there is great heterogeneity that exists in clinical outcomes for melanoma patients despite the currently available staging criteria.

At present, the eighth edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual offers valuable prognostic information for patients. In particular, for those with localized disease, Breslow depth offers important staging and prognostic information, and for some, sentinel lymph node biopsy (SLNB) can provide population-based prognostic information. Nevertheless, studies demonstrate that intermediate thickness SLNB-negative cutaneous melanomas cause twice as many deaths as intermediate thickness SLNB-positive cutaneous melanomas. Moreover, deaths from AJCC T1 cutaneous melanomas (≤1 mm) outnumber deaths from both T2 (1.01–2 mm) and T4 cutaneous melanomas (>4 mm), with similar mortality figures for T1 and T3 cutaneous melanomas (2.01–4 mm).

The above findings highlight the limitations of the currently available staging criteria in identifying patients with “low-risk” cutaneous melanomas who will go on to develop metastatic disease. Contrary to a positive SLNB, a negative SLNB, which most patients have, does not necessarily equate to a “low-risk” disease. Furthermore, up to 50% of SLNB eligible patients do not undergo SLNB. These findings are troubling because SLNB is the lone prognostic test listed in current National Comprehensive Cancer Network (NCCN) guidelines. For patients with cutaneous melanoma, who do not meet criteria for SLNB, are unwilling or unable to undergo an additional invasive procedure, or have a negative SLNB, there are no prognostic tests available under current NCCN guidelines to evaluate the metastatic risk.

The results of the DecisionDx-Melanoma test are utilized to guide patient management and surveillance plans. For example, a patient with a Class 1A test result identifies low-risk patients that can be spared unnecessary surgical intervention (e.g. sentinel lymph node biopsy [SLNB]) and intensive surveillance and treatment. In contrast, a patient with a Class 2B test result is at high risk for metastatic disease and can benefit from intensive management, including SLNB and high intensity surveillance.

Without knowing who is at high risk for metastasis, we cannot direct surveillance efforts to those who need it. It has been recently demonstrated that intensive surveillance in high-risk patients identifies 80% of metastases before they become symptomatic, allowing identification of patients with low burden of metastatic disease. Multiple studies have shown that surgical management and systemic therapies in melanoma are more effective when disease burden is low, with the majority of complete remissions in patients with lowest tumor burden. Thus, early identification of metastatic disease translates to better outcomes for these patients.

I have used the test for my melanoma patients over the past several years and am familiar with the published clinical studies. My use is in line with the two current indications 1) Sentinel lymph node biopsy (SLNB) guidance for patients with T1-T2 melanoma and 2) Determination of management plans based on risk of recurrence. I have seen the full spectrum of utility with the test in my practice including patients with low risk factors according to anatomopathologic staging who had a Class 2 (high risk) result and patients with high risk factors according to anatomopathologic staging who had a Class 1 (low risk) result.

The patients I take care and myself would appreciate your organization will finalize the expanded LCD in your proposal as it is currently written.

Thank you for your comments.

This letter provides complete support for the extension of the LCD for Decision-Dx-Melanoma from guidance of sentinel lymph node biopsy decisions to encompass the intended use of the test – detection of melanoma patients with high and low-risk for development of metastasis to appropriately guide intensity of follow-up. The Decision-Dx-Melanoma test predicts the 5-year risk of recurrence and metastasis resulting in significant stratification in survival outcomes between low-risk and high-risk patients and is independent from staging features.

While stratification of risk by staging has improved with implementation of AJCC 8th edition staging guidelines, the fact remains that the small percentage of Stage I and II melanoma patients who experience disease progression continue to account for a substantial portion of melanoma-related deaths because of the prevalence of early stage disease. The Decision-Dx-Melanoma test provides additional prognostic information that stratifies the risk of patients within AJCC stages to identify patients with high-risk tumor biology in need of close follow-up, even within patients traditionally considered to be low-risk. In a recent study by Gastman, et.al, Stage I and II patients with a negative sentinel lymph node biopsy result were significantly stratified by Decision-Dx Melanoma results such that 92.9% of patients with a low-risk Class 1A were free of distant metastasis 5 years after diagnosis, compared to only 84.2% of patients with a high-risk Class 2B result. Similar results are reported in a prospective study of Stage IB-II patients in which all recurrence events occurred in patients with high-risk Class 2 results (Podlipnik, et. al.). This separation of risk based on the objective tumor biology provides information beyond traditional clinical and pathological features to further clarify risk a provide more comprehensive prognostication to guide patient care. Harnessing the power of gene expression profiling to further personalize patient management improves clinical decision making to improve patient outcomes.

We are currently analyzing our own database and the preliminary numbers seems to validate the above mentioned studies.

Thank you for your comments.

Dear MolDX In response to the Local Coverage Determination (LCD) DecisionDX-Melanoma (DL37725), I am fully supportive of the proposed LCD in its entirety. The DecisionDX-Melanoma test is an independent factor that has an impact on the management decisions I make with my cutaneous melanoma patients particularly for guiding frequency of follow-up visits and imaging used at those visits.

Thank you for your comments.

I am a board certified dermatologist, Mohs surgeon and skin cancer specialist who treats many patients diagnosed with melanoma.

The purpose of my email is to support the proposed LCD (DL37725) for DecisionDx-Melanoma. The DecisionDx-Melanoma test is a gene expression profile assay for patients with cutaneous melanoma that differentiates between patients who are at low risk or high risk of developing metastatic disease. I support your expanded coverage criteria:

Patients diagnosed with cutaneous melanoma ≥ 0.3 mm without distant metastases in Breslow thickness where additional risk stratification information beyond anatomic and pathologic staging will influence management decisions regarding the following:

• Sentinel Lymph Node Biopsy decision (T1-T2 only)
• Frequency and modality of imaging in follow-up
• Frequency of follow up visits in follow-up
• Referral to Medical Oncology
• Appropriateness of adjuvant therapy

Thank you for your comments.

I strongly urge you to approve the DecisionDX-Melanoma (DL38018) Local Coverage Determination to extend coverage of DecisionDX-Melanoma to influence management decisions for patients with melanoma, particularly: follow up frequency, imaging decisions and frequency, specialty referral, and potential use of adjuvant therapy. I personally use the test to guide my management decisions for my cutaneous melanoma patients. Additionally, the test is used by a number of my colleagues due to the wide body of evidence supporting use of the test to better assess prognosis and guide management of patients with melanoma. We are confident in the test results and have first-hand experience using the test in the clinic. Once again, my colleagues and I use this test to guide our management decisions in patients with invasive melanoma, and I strongly urge you to approve the extended coverage proposed in this LCD.

Thank you for your comments.

The Decision-DX Melanoma gene expression profile predicts 5-year risk of metastasis in patients with melanoma as low risk with a Class 1 result or high risk with a Class 2 result. A recent draft Limited Coverage Determination to expand clinical indications for coverage to include the clinical indication for prediction of risk of recurrence to appropriately guide frequency of follow-up, use and modality of imaging, referral to medical oncology and consideration of adjuvant therapy. This letter provides my full and enthusiastic support for this coverage expansion.

The Decision-DX Melanoma test has been previously published in prospective and retrospective studies to be a significant predictor of recurrence and distant metastasis and is independent of other clinical and pathological factors. Across studies, Decision-DX Melanoma is consistently among the strongest predictors for disease progression, with high hazard ratios for prediction of recurrence and distant metastasis and highly significant p-values. These studies indicate Decision-DX Melanoma is a powerful clinical tool that adds clinically actionable prognostic information over and beyond information obtained from clinical and pathological features. Recent utility studies show that the test informs patient treatment in 50% of patients tested, both decreasing intensity of follow-up in patients with low-risk Class 1 results and increasing intensity in patients with high-risk Class 2 results. Appropriate follow-up of patients based on their individual risk assessment ensures the best use of healthcare resources and focuses them on patients most likely to benefit from them.

Thank you for your comments.

I am writing to express support of the draft Local Coverage Determination (LCD) for the Decision-Dx-Melanoma gene expression profile test for cutaneous melanoma. This draft LCD extends coverage of the Decision-Dx Melanoma to predict risk for disease progression to identify the appropriate level of follow-up and surveillance for melanoma patients. DecisionDx-Melanoma stratifies patients into a low-risk (Class 1) and high-risk (Class 2) probability of recurrence and distant metastasis within 5 years. It has been demonstrated to be an accurate, independent predictor of recurrence and distant metastasis for patients with cutaneous melanoma in numerous retrospective and prospective studies supporting its clinical validity and utility. The majority of melanoma patients present with early stage disease, traditionally regarded to have favorable overall prognosis. While only a small fraction of early stage patients experiences distant metastasis or death from melanoma, this fraction constitutes the majority of deaths from melanoma due to the large number of patients in this population. Gastman, et. al. reported significant stratification of risk of recurrence, distant metastasis, and death in an archival cohort of Stage I-IIA patients, such that only a GEP Class 2 result in the context of other clinical and pathological features was a significant predictor of outcome. Similarly, in patients with thin tumors ≤ 1 mm, 5.3% of patients had a high-risk Class 2B result and had significant poorer recurrence-free survival (64.6%) than low-risk Class 1A patients (96.8%). The addition of this valuable prognostic tool to the dermatologist’s toolbox enables detection of the “needle in a haystack” patient with low-risk melanoma by staging criteria, but high-risk tumor biology to guide appropriate patient management.

Thank you for your comments.

Today, the DecisionDx-Melanoma test is used by healthcare providers in the immediate post-diagnostic period to adjust treatment plans for patients with invasive cutaneous melanoma (CM) based on their risk of recurrence/metastasis. The clinical validity and utility of the test has been the subject of 20 peer-reviewed articles.^1-20

In 2018, the DecisionDx-Melanoma test was requested for over 12,450 patients (representing use in approximately 1 in 7 CM patients), demonstrating that this test has been accepted in the medical community as safe and effective and is in line with generally accepted professional medical standards. In 2018 more than 217 insurance companies across the United States approved and paid for the DecisionDx-Melanoma test. Thus, the DecisionDx-Melanoma test meets medical necessity and reasonableness criteria.

According to clinical guidelines, intensive surveillance (frequent follow-up and imaging) is recommended for patients with Stage IIB-IIIC “high-risk” melanoma¹⁶, however only a fraction of these patients develop metastatic disease and die from melanoma.²¹ Conversely, earlier stage patients who have biologically aggressive disease are missed by clinicopathologic staging factors, as nearly two-thirds of patients who recur and die from melanoma belong to Stage I and IIA “low-risk” for whom intensive surveillance is not recommended in current guidelines.^17-20 Thus, improvement in the risk stratification for CM patients to guide management decisions within the current guideline framework is an unmet and clinically important need.

DecisionDx-Melanoma provides a validated, accurate assessment of risk of metastasis that is independent of traditional clinical and pathologic factors and has been validated in multiple prospective and archival studies as a robust predictor of recurrence and which is used by physicians to adjust care in accordance with these guidelines.^1-6,8-18,21

Molecular classification of melanoma with the DecisionDx-Melanoma test can facilitate the determination of individual metastatic risk, which, along with clinicopathologic factors can help inform improved risk-appropriate management decisions, both in T1-T2 and the T3-T4 populations. Patients who are anatomically classified by AJCC stage as higher risk but have biologically indolent disease can benefit from the high negative predictive value (NPV) of a Class 1A result (99% for melanoma-specific survival). For these Class 1A patients a physician can then safely de-escalate or use lower intensity management (lower frequency of follow up, no imaging and no referral) and, if T1-T2, avoid an unnecessary surgical procedure (SLNB) which directly reduces healthcare resource utilization. In addition, the stage I and II patients who have intrinsic high-risk tumor biology (Class 2) can be promptly identified as such at diagnosis and can be offered the benefit of high intensity management with more frequent visits, surveillance imaging to effectively identify metastatic disease early and referral to surgical/medical oncology consistent with their risk of metastasis.

Accuracy of the test to determine metastatic risk in CM patients has been evaluated in three prospectively designed studies of archived formalin-fixed, paraffin-embedded (FFPE) tissue from patients with known outcomes totaling 690 patients.^1,2,9,14 And results from a combined prospective registry and four prospective studies have confirmed the prognostic accuracy of DecisionDx-Melanoma in >700 patients.^{6,11,17,18,22} Together these studies demonstrate the accuracy of DecisionDx-Melanoma to identify CM patients at low or high risk of recurrence and/or metastatic disease. Importantly, the risk associated with a Class 1 or 2 result is consistent across multiple different multi- and single-center cohorts and in both archival tissue and prospective studies. All of these studies have also demonstrated that the test results are independent from traditional clinicopathologic factors and that a DecisionDx-Melanoma result provides individualized risk information with improved sensitivity, specificity and NPV over SLNB that can inform patient management. In addition, two recent independent manuscripts reviewed the published evidence for the DecisionDx-Melanoma test and concluded the test is supported by sufficient evidence (level IIA in the SORT system) to support clinical use in cutaneous melanoma patients.^20,23

While the DecisionDx-Melanoma test’s validation studies have included unselected Stage I-III patients, a recent analysis of clinical impact was performed to determine which population derives the most benefit from testing. The results showed than in T1 tumors a separation of risk can be seen between Class 1A and 2B in tumors ≤0.3 mm and thicker, with no events in the population <0.3 mm. Furthermore, 10.8% of tumors 0.3 to 1.0mm thick had a non-Class1A result and 4.3% of tumors were Class 2. Also, one in four patients in this group had a change in management as a result of the test.¹⁹ These results suggest that the appropriate use of the 31-GEP test for management decisions was found to be in cutaneous melanoma tumors ≥0.3 mm thick as determined in the draft LCD.

Recommended Changes to the Draft LCD

After reviewing the policy, we would like to comment on a few items in the LCD that we request you consider modifying. We believe these changes may result in an LCD that will ensure the best care for Medicare beneficiaries and efficient processing of claims. Our comments are based both on our experience with performing this test plus input from the medical specialists who manage Medicare beneficiaries with cutaneous melanoma. Following are sections of the policy with our suggested modifications.

1) Coverage indications, limitations, and/or Medical Necessity (page 3)

• Coverage Indications, Limitations, and/or Medical Necessity currently states:

“DecisionDX-Melanoma is not covered in patients who are not seeking disease-modifying treatment for cutaneous melanoma.”

This language can be confusing as can be interpreted to only refer to patients in which pharmacologic therapy is being considered and the intent of the LCD is to cover surgical, surveillance and other management decisions therefore, we suggest the following language:

“DecisionDx-Melanoma is not covered in patients for which the results are not intended to guide treatment and management decisions for cutaneous melanoma.”

• DecisionDx-Melanoma is written with upper case x in the last two paragraphs of this section. Suggest correct spelling to lower case x as in DecisionDx-Melanoma.

2) Summary of Evidence (pages 3-4):

• Suggest changing “management” for “diagnosis” in the first sentence of second paragraph as the intent of the test is to influence management changes, rather than diagnosis.

LCD currently states: “In the treatment of CM, the risk that a patient has or will develop metastatic disease is central to many of the decision management choices in cutaneous melanoma, with more aggressive diagnosis or treatment strategies recommended for patients who are at a higher risk.”

We suggest: “In the treatment of CM, the risk that a patient has or will develop metastatic disease is central to many of the decision management choices in cutaneous melanoma, with more aggressive management or treatment strategies recommended for patients who are at a higher risk.”

• Suggest addition of one sentence in the second to last paragraph of page 3 regarding the published evidence for a 0.3mm cutoff for test performance.

We suggest: “Another recent study showed that the appropriate use of the DecisionDx-Melanoma test for management decisions was found to be in cutaneous melanoma tumors ≥0.3 mm thick."³⁴

• Suggest correct spelling to lower case x as in DecisionDx-Melanoma wherever the test is mentioned in this section.

3) Clinical Performance: Utility and Validity (Pages 5-6)

• Changes in references on the Summary of Evidence section are not reflected in the tables in these sections. Suggest updating the tables to reflect the reference numbers adequately

4) Bibliography (Pages 7-8)

We also recommend updating the bibliography list as follows:

• Insert new reference 25: Gastman BR, Gerami P, Kurley SJ, et al. Identification of patients at risk of metastasis using a prognostic 31-gene expression profile in subpopulations of melanoma patients with favorable outcomes by standard criteria. J Am Acad Dermatol. 2019; 80(e4):149-157.
• Insert new reference 35: Marks E, Caruso HG, Kurley SJ, et al. Establishing an evidence-based decision point for clinical use of the 31-gene expression profile test in cutaneous melanoma. SKIN J Cutaneous Med 2019; 3:239-249.

We commend this LCD, which guarantees that Medicare beneficiaries who are diagnosed with Cutaneous Melanoma will have access to this important test that informs treatment and management decisions for melanoma patients, and appreciate your consideration of these suggested modifications.

Thank you for your comments and helpful edits. Many comments received have been incorporated into the policy.