MolDX: Biomarker Testing for Risk Stratification in DCIS

Title XVIII of the Social Security Act, §1862(a)(1)(A) allows coverage and payment for only those services that are considered to be reasonable and necessary.

42 CFR §410.32(a) Diagnostic x-ray tests, diagnostic laboratory tests, and other diagnostic tests: Conditions

CMS Internet-Only Manual, Pub. 100-02, Medicare Benefit Policy Manual, Chapter 15, §80 Requirements for Diagnostic X-Ray, Diagnostic Laboratory, and Other Diagnostic Tests, §80.1.1 Certification Changes

Biomarker tests (molecular or proteomic) that enable risk stratification of Ductal Carcinoma in situ (DCIS) patients into sufficiently low risk patients, who may safely forego the use of adjuvant radiation therapy (RT) or patients who should not forgo the use of RT or receive a higher intensity intervention are currently non-covered by this contractor. However, should evidence be further developed to meet reasonable and necessary standards, such a test may be considered for coverage when ALL the following criteria are met:

• The patient must:
• Have a diagnosis of DCIS and NOT have concurrent invasive breast carcinoma;
• Not be in consideration for mastectomy;
• Not have received RT for breast cancer in the same breast;
• Have a formulated treatment plan that includes:
• RT and consent to its use; AND
• Intent to forgo the use of RT, if determined to be placed in the “sufficiently low risk” category by the test;
• Not have been previously tested with the same or other similar test;
• Along with any sample used, be within the validated, intended use and population of the test;
• The test must:
• Have successfully completed a Technical Assessment (TA); AND
• Demonstrate significant improvement in accuracy of identification of sufficiently low-risk patients compared to existing risk stratification tools and methodologies, including non-biomarker clinical and laboratory findings or combinations of such findings (and including such findings within its internal algorithms), such as those found in nomograms and risk scoring systems, in appropriately designed and controlled studies published in peer-reviewed, published literature; OR
• Demonstrate equal or superior performance in risk stratification to any other biomarker test that has already met the criteria above;

Definitions

Sufficiently Low-Risk: For the purpose of this policy, this term is defined as a stratification group of patients wherein the absolute risk reduction of Ipsilateral Breast Tumor Recurrence (IBTR) from RT is estimated to be 5% or less when compared to Breast Conserving Surgery (BCS) alone regardless of individualized factors (extreme age, intolerance to RT, co-morbidities, etc.).

Ductal carcinoma in situ (DCIS) is a heterogeneous group of neoplastic lesions confined to the breast ducts and lobules. It is considered a pre-invasive form of ductal breast cancer with recent publications demonstrating a clonal relationship in up to 75% of DCIS lesions and subsequent invasive cancers.¹ It is one of the most commonly diagnosed breast conditions, accounting for approximately 20% of newly diagnosed breast cancers in the United States, equating to roughly 54,000 new cases yearly.² Women with a history of DCIS are at risk for local recurrence, which may be either DCIS or progression to invasive breast carcinoma. Most commonly, recurrence is in the ipsilateral breast, and there is a similar abundance of DCIS and invasive cancer when it recurs. The management of patients with DCIS is an area of controversy, and historically, treatment has included both surgical excision and radiation therapy.³ Following BCS-only therapy, local recurrences occur in approximately 25% to 30% of women by 10 years. The addition of RT (combined, herein referred to as Breast Conserving Therapy, BCT) has been reported to reduce local recurrence risk by approximately 50% regardless of most other factors, but has not been demonstrated to significantly prolong overall survival (OS).^4-10 Furthermore, RT use in DCIS is not benign and is associated with risks, including development of secondary cancers.¹¹ Despite this, the use of RT has increased over time.¹² Therefore, treating all women with radiation therapy following surgical excision may represent overtreatment for many, especially given that the majority of cases do not recur following surgery alone, and thereby incurs financial and health risks, with little impact on OS. Current guidelines recommend RT for most patients, but allow BCS-only in assumed low-risk patients in a shared decision-making process with the patient.¹³

Significant efforts have subsequently been undertaken to identify a low-risk DCIS population that could safely forgo RT. Several groups have created strategies based on identifying clinical, demographic, and laboratory factors associated with reduced risk of IBTR and assessed them in observational, retrospective, and prospective studies. Di Salvero et al demonstrated in 259 patients that a low-risk population had significantly higher Disease-Free Survival (DFS) when compared to intermediate and high-risk groups (94% vs 83%).¹⁴ Other studies have demonstrated that low-risk patients treated only with BCS have local recurrence (IBTR is used herein interchangeably with this term) rates of 4-15% at 10-15 years and this is roughly half of higher risk groups.^4,15-18 RTOG 9804 represents a pivotal study that investigated BCS-only treatment in a low-risk population defined only by histopathologic data (margin status, tumor grade, and tumor size) in a prospective study including 636 patients.⁷ At 7 years, IBTR was 0.9% with BCT and 6.7% with BCS. A follow-up study in 2021 showed a reduction (~50%) between the BCT and BCS arms at 7.1% vs. 15.1%, mirroring other studies.¹⁷ Most recently, Wright et al published 20-year results from the ECOG-ACRIN E5194 study that enrolled patients into low/intermediate or high-grade DCIS groups and treated with BCS alone and demonstrated a plateau in recurrence rates after 15 years.¹⁹ However, these studies employ different definitions and cut-offs in determining “low-risk,” and may include different combinations of histopathologic, patient demographic, and laboratory data, with different stratification groupings. Additionally, data analysis varied between isolating low-risk or combining low and intermediate risk patients.

In an effort to optimize the low-risk designations, several groups have created scoring systems and nomograms to try to normalize and standardize the identification of low-risk patients that may safely forgo RT use. The Van Nuys Prognostic Index (VNPI) was developed in the 1990s and modified in 2010, incorporating tumor size, margin status, tumor class (grade and presence of comedonecrosis), and patient age to stratify patients.^20-24 In a study of 939 patients, wherein 604 received BCS and 345 BCT, those in the low-risk group (34%) had 5.4% IBTR with BCS and 2.5% with BCT. Despite inclusion and discussion in current guidelines, external validation has been inconsistent, as not all studies have been able to yield similar results.^13,25-28 In 2010, Rudloff et al published a nomogram of ten factors derived from a cohort of 1,868 patients at Memorial Sloan Kettering in New York (MSKCC), which was similarly based on clinical and laboratory factors, but also introduced the use of endocrine therapy.²⁹ A subsequent study evaluated these variables in 2558 patients and showed most variables to be significant in association with IBTR in BCS but not BCT.³⁰ This study also demonstrated that treatment in more recent years (after 1998) showed lower IBTR rates, even when controlling for endocrine therapy and RT use, and that there was a significant increase in RT use over time. This approach has subsequently been tested and validated in other patient populations, although performance in studies conducted at external centers has been mixed; reported concordance indices between studies have been 0.61-0.68, with the most discordance found in higher-risk patients.^31-35 The nomogram does not define risk categories but uses a point-based system to identify a 5- and 10-year risk of IBTR. Notably, the nomogram did not incorporate tumor size. Sagara et al proposed the Patient Prognostic Score, which utilized data from over 32,000 patients and was based on only 3 patho-clinical features (patient age, tumor size, and tumor histology) that were readily available in existing publicly-available datasets.³⁶ This study focused on Breast Cancer Mortality (BCM) rather than IBTR with a median follow-up of 96 months and demonstrated that BCM was 1.3% vs 0.8% in the low-risk group for BCS vs. BCT, and that this was significantly less than non-cancer causes of mortality, including from heart disease. The authors argue that such patients do not benefit from BCT over BCS. External validation is still lacking for this approach.

Biomarker tests  

To try to improve stratification performance over clinical and pathological data, several groups have attempted to define a low-risk population based on molecular or proteomic signatures. Solin et al published initial findings of a 12-gene, reverse-transcriptase polymerase chain reaction (RT-PCR) expression profile test (Oncotype DX DCIS Score, DS1, based on 7 cancer-associated genes and 5 reference genes) in 327 patients who participated in the E5294 trial and received BCS therapy, of whom 46 had IBTR.³⁷ The test was used to define a low, intermediate, and high risk group, with 10-year IBTR rates identified as 10.6%, 26.7%, and 25.9%, with ~70% of patients being placed into the low-risk group. A second study using this test was performed in 718 patients from a similar pool of patients as the original study demonstrated IBTR rates of 8%, 20.9%, and 15.5% in the low, intermediate, and high-risk groups, respectively.³⁸ In another study incorporating clinical factors as well as BCT treated patients of the same cohort, a subpopulation treated after the year 2000 had 10-year IBTR rates of 10.6% and 5% in the low-risk DS1 group treated with BCS and BCT, vs. 25.4% (BCS) and 12.6% (BCT) in the high-risk group.³⁹ These studies demonstrated the significance of the signature in multivariate analysis, collectively showed a concordance index of 0.68, and were used to create a predictive model that incorporated clinical and pathological factors to improve its ability to stratify patients.^40,41 A study also showed that the use of the test did reduce RT utilization, but patient outcomes were not assessed.⁴² An expansion to a 21-gene expression test was also devised but did not demonstrate improvement over the prior test.⁴³ However, both tests were compared and shown to better predict invasive and local recurrence when compared to a defined set of clinicopathological factors (age, tumor size, nuclear grade, multifocality, and RT use).

A second test (DCISion RT) was developed using a combination of four clinical and pathologic factors and seven molecular markers identified by immunohistochemistry into a nonlinear risk algorithmic scoring system termed Decision Score (DS2) that was validated in 526 (474 informative) BCS and BCT patients.⁴⁴ Although the score is continuous, the test defines a “low-risk” and “elevated-risk” group based on a DS2 cutoff of 3. Of the 526 patients in the study, 41% were placed in the low-risk group; the low-risk group had a 10-year IBTR rate of 8% for BCS and 7% for BCT. The cohort had 61 IBTR incidents (of the informative set) and the study’s baseline IBTR rate was 12.9% for all patients and 15% for those treated with BCS only. The elevated-risk group saw expected reductions in 10-year IBTR rates from BCT, dropping from 23% to 11%. This study concludes that the test can identify a low-risk population that does not benefit from RT given that there was only a non-significant 1% reduction in 10-year IBTR rates, whereas the high-risk population does significantly benefit. A follow-up study of 455 patients (with 53 total IBTR instances) from a separate site showed that low-risk patients (41%) had a 10% risk of IBTR with BCS and 5% with BCT, vs 30% with BCS and 10% with BCT for the elevated-risk group.⁴⁵ Multivariate analysis showed that elevated risk patients were significantly associated with higher IBTR risk when controlling for other factors (hazard ratio 1.86, P=0.048), as were the use of RT, endocrine therapy, and the presence of necrosis. This study further compared DS2 to a “RTOG 9804-like” score to demonstrate DCISion RT could identify patients who would have been considered low-risk and could instead be re-categorized to the elevated-risk category based on the DCISion RT score. This study also evaluated patients for invasive recurrence. A prospective study was subsequently performed (and subsequently re-evaluated) to show that the test impacted clinical decision making when made available; however, no patient outcomes were reviewed.^46,47 The test was also used to evaluate 504 patients (59 IBTR events, 12%) in the SweDCIS study that randomly assigned BCS vs BCT therapy in patients between 1987 and 2000.⁴⁸ These patients showed that the addition of RT reduced 10-year IBTR from 22.8% to 8.3% in the elevated-risk group, and 12.9 to 7.2% in the low-risk group. However, there is statistical significance only in the elevated-risk group (p< 0.001 vs. p=0.059), with an absolute risk reduction of 15.5% vs 5.7%. This study also introduced a lower DS2 cutoff of 2.8, as it better separated the fraction of patients that result in invasive breast cancer recurrence. Only 60% concordance was observed between RTOG 9804 criteria and DS2 scores. Vicini et al tested a new signature that could identify high risk patients who would have an increased risk for recurrence after BCT (residual risk, RRt), and evaluated data from prior studies in 3 patient tiers (low, high, and residual risk groups) using the DS2=2.8 cutoff.⁴⁹ The 10-year recurrence rate for all groups was 12% (77 events in 926 patients). The low-risk patient group (now 37% of patients) had an IBTR rate of 5% (not significantly differentiated by RT use). In the high-risk group, patients treated with RT had a 15.7% absolute IBTR rate reduction. Patients in the RRt group had a 27.4% absolute rate reduction with RT. Patients in the RRt group had a higher rate of nuclear grade 3 DCIS, larger tumors, and HER2(3+) disease. A RRt group was also evaluated specifically in HER2-positive patients from the NSABP-43 trial, which identified a subset of HER2(3+) patients with greater IBTR rates following BCS and RT; notably, the RRt group in this study also had significantly more nuclear grade 3 disease.⁵⁰ Finally, Dabbs et al published an analytical validation of the DCISion RT test.⁵¹ Concordance studies for the accuracy of risk assessment were not performed between clinical studies using DCISionRT.

Comparisons between the biomarker tests and other established nomograms or scoring systems have been sparse, but systemic reviews and metanalyses were performed by Schmitz et al and Ouattara et al in 2022 and 2023.^52,53 Pooled analysis showed that the relative risk reduction with RT in high-risk and low-risk patients was significant for DCISion RT for IBTR, but not for invasive cancer in the low-risk group (Hazard Ratio = 0.58 with a confidence interval of 0.25-1.32). Oncotype DX DCIS had similar findings, with absolute risk reduction in high-risk patients at 12.7% and 6.6% in low-risk patients. Lei et al published a comparison between the radiation oncologists’ practice based on clinical factors vs. the application of the VNPI, MSKCC, and Oncotype DX DCIS tests.⁵⁴ Three oncologists’ estimates of 10-year IBTR risk were compared to those identified by the scoring systems and nomograms, and correlations were measured. No outcomes were evaluated, although the DS1 scores predicted lower 10-year IBTR risk for all patients and subgroups analyzed. Van Zee et al compared the 10-year IBTR risk results of the MSKCC nomogram and the DS1 score in 59 patients who were > 50 years of age and had a tumor size less than 2.5mm.⁵⁵ 92% of patients had concordant results when considering IBTR risk categories of <10%, 10-15%, and >15%. All discordant calls occurred in the higher MSKCC score patients and had absolute 10-year IBTR rate discordance of less than 10%, suggesting that these discrepancies alone may not account for changes in management decisions.

Contractor Advisory Committee Meeting (CAC)

To further understand provider perspectives on current practice in DCIS management and existing evidence for risk stratification strategies, a CAC was held on July 15^th, 2024. This meeting that was open to the public and based on questions posed to subject matter experts (SMEs) including breast surgeons and radiation oncologists from both academic and private practices. Both the questions asked and the transcript are available for public consumption.^56,57 Of relevance to this work, the CAC confirmed that relevant metrics for DCIS risk stratification included both IBTR and the risk of invasive cancer. The SMEs noted that decisions regarding RT utilization are based on several factors, including life expectancy, defined clinical and pathological criteria, the possible harm from radiation, financial toxicity (cost of treatment effect on patient’s well-being), and possible risk reduction by RT; these are utilized as part of a shared decision-making process with the patient. This process is highly individualized and likely not concordant between physicians or institutions; it was even stated the current process is “highly flawed.” SMEs did not agree on what 10-year IBTR rate constituted low-risk, with advocacy between 5-15%. However, the SMEs noted that low-risk patients may still choose RT because they derive some benefit, no matter how minimal, and that the knowledge that a patient would not benefit from RT would help promote RT avoidance (as determined by the use of DCISionRT). When asked at what level of risk reduction RT would they consider the use of RT to not be clinically meaningful, there was general consensus that at a 5% absolute risk reduction of IBTR (or less) confers no significant clinical impact from RT therapy.

Regarding current biomarker tests, there was general consensus that a test that can accurately predict recurrence or RT response (or lack of response) would be of value and have clinical utility and that there is a current clinical need for such a test; however, there was a spirited debate as to the current evidentiary support for such tests. Some CAC members commented on the tremendous value of relying on the DCISionRT test, claiming that it identifies patients with no relative risk reduction from RT and do not respond to RT. While there was concern about clinico-pathologic scoring system utilization and applicability, there was also concern for the lack of direct comparison for the biomarker tests to such systems. Some but not all of the committee stated they rely on current biomarker tests in stratifying patients. Given proposed theoretical tests and their performance, there was general agreement that tests that could predict RT response would be valuable, but to a limit (not overriding personal risk factors); however, several SMEs noted that they would value tests that predict invasive cancer recurrence and the relative response to RT more highly than those that predict IBTR. There was also a comment about the importance of the identification of patients in the “residual risk” group who may be undertreated.

Based on the evaluated data herein, it is apparent that DCIS patients, and in particular older Medicare beneficiaries who have a lesser risk of recurrence than younger patients, are often overtreated and should be considered for RT with some reservation, even if there is some benefit regarding IBTR risk. However, current practice as per guideline is to treat with RT as a default unless patients have contraindications to RT or are considered low risk and physicians and patients agree to waive RT use. Furthermore, per expert comment in the CAC and reported elsewhere, even in the low-risk population many physicians and patients will opt for RT to not “leave anything on the table,” leading to even higher rates of RT utilization. It is apparent that if there is any perceived risk reduction for RT, its use is preferred by practitioners for most patients.^12,43,56 Based on the evidence reviewed and feedback from the CAC, it is thus relevant to identify a population of patients for whom there is no reasonable benefit from RT and are sufficiently low risk to safely preclude RT use regardless of other factors.

DCIS risk stratification has been controversial and evidence difficult to evaluate given several major challenges. First, DCIS is a disease that recurs both in low rates and often after substantial periods of time, and medical management has changed considerably over time. For example, the implementation of hormonal therapy has improved IBTR rates and is commonly used in DCIS; studies consistently show that a major significant factor to improved IBTR rates is when the patient accrual occurred- those that include patients before hormonal therapy and revised surgical protocols show poorer outcomes than more modern cohorts. As such, estimates of average DCIS IBTR rates being as high as 30% in 10 years is likely an overestimate; more recent studies show that the stratified “high-risk” BCS-only patients have lower rates. This implies that comparison of patient cohorts that include old and new patient data may not be of value as studies would have to control for such factors. Second, DCIS recurrence is not associated with meaningful increase in patient mortality; as such, treatment is aimed at reducing morbidity. Decisions rendered (by the physician and patient) must include a careful balance of the risk of morbidity from recurrence vs. that from RT given numerous external factors. Even when these factors are held constant, physicians will make different recommendations.⁵⁴ In this context, it is apparent to this contractor that having a defined and reproducible high/low risk classification will likely improve RT overutilization, possibly regardless of source. Third, there is no clear definition or standardization of the term “low-risk.” It is possible that the ambiguous nature of the term may have contributed to RT over-utilization, particularly when highly individualized factors are included in its formulation. Studies that aim to measure patient performance in risk-stratified groups rarely agree on defining a low-risk population or how to analyze risk groups, making any analyses between such studies and drawing larger conclusions about expected performance of such patients impossible. Most studies pre-select a group of defined clinical or pathological values (or biomarker status). We propose here that it would be more valuable to providers to define risk by expected IBTR rates (such as expected 10-year IBTR rates) regardless of methodology rather than specific and sporadic set of clinicopathologic variables (assuming quality evidence development and confidence in the results). Based on CAC feedback, that cutoff could be ~10% IBTR risk at 10 years (longer time periods are both less relevant to the Medicare population and will suffer from less data). This definition would bring a definitive value to the term and would identify patients who will not find meaningful IBTR reduction from RT, even if expected relative risk reduction is maintained near 50% regardless of other factors (the absolute risk reduction is 5%). Such patients would be sufficiently low-risk and would not derive meaningful benefit from RT; such a designation may be sufficient to arm providers with messaging to preclude RT over-utilization even if derived solely (if accurately) from clinicopathologic factors. While feedback and limited data also support the use of invasive cancer risk in managing decisions, it is not clear that providers would make different decisions with invasive cancer risk than IBTR risk. Consistent with current practice, patients should also be dissuaded from RT use even if they are not in this “low-risk” group based on highly personalized risk criteria (extreme age, intolerance to RT, other comorbidities, etc.). Based on these data and SME feedback, any test that can accurately and reliably identify a patient population that achieves 5% or less absolute risk reduction for IBTR is reasonable and necessary if those values cannot be attained by clinicopathological factors alone.

Clinicopathologic variables can be used to identify risk, but different strategies and nomograms used for this purpose do not perform equally. It may be reasonable to use RTOG 9804 criteria because of its ease of use, but this is outperformed by other strategies (scores and nomograms) cited above. Both the VNPI and MSKCC strategies are well-studied and externally validated, with mixed results, often with inaccuracies in the higher-risk patients and thus not relevant to the assessment of RT avoidance. This contractor is confident that biomarkers are likely to improve the process of risk stratification and identify, with greater confidence, those who can safely forgo RT when compared to clinicopathological factors. Yet, clinicopathologic risk assessments have demonstrated the ability to identify sufficiently low-risk patients and can be accessed by anyone. The advancement of newer technologies must come with direct comparison to the current best options available and be founded on the best-available evidence. To date, the only published direct comparisons of a biomarker test to one of these clinicopathologic risk assessments (in a Medicare-relevant population) has not demonstrated a significant difference in risk stratification⁵⁵ or improved performance.⁵⁴ It is expected that such comparisons be used to demonstrate incremental value over existing methodologies. For coverage, a biomarker must demonstrate that it is more accurate than such existing systems. Further, the data suggest that biomarker tests may place a significantly greater percentage of patients in a sufficiently low risk group; this contractor would consider such information as relevant to the assessment of accuracy. The observations of the residual risk group identified in Vicini et al^50,51 are noted but should be considered exploratory until it becomes apparent what specific action a physician should take with that information. Further, the observations made by reviewing RRt in HER2 positive DCIS is interesting, but HER2 status is not routinely assessed in DCIS, making comparisons between cohorts problematic and the utility of the biosignatures questionable in routine clinical practice.

The current evidence for the use of biomarker tests is lacking, with some deficiencies previously described.^52,53 Studies must demonstrate the accuracy of the test in the correct intended use population, including the Medicare population, with sufficient data to draw meaningful conclusions. If they include clinicopathologic factors, multivariate analyses performed must demonstrate the additional value of the biomarker above the clinical content alone, especially the combined value of that content. External, independent validation and a thorough analysis are expected. If multiple studies are used with different cohorts, concordance studies should be performed.

While it is desirable to have a tool that identifies a population with no relative risk reduction from RT (claimed lack of response to RT) for decision making and for supporting such decisions as a rhetorical proof, we conclude that there is insufficient data currently to convincingly support the assertion that any current biomarker test has demonstrated it can identify such a population, nor does this contractor believe this is a reasonable nor a necessary requirement for a useful test. This was a central rationale presented at the CAC by SMEs for use of such tests, yet this assertion is most easily explained by a failure to identify statistical significance of RT use in the low-risk groups by an underpowered data set from which such conclusions could not be drawn. The initial assessment that the low-risk group (identified with a DS2<3) in Bremer et al only benefited from 1% absolute risk reduction from RT was based on a very limited number of recurrences (numbers not stated, HR was 0.7 but 95% CI was 0.3-1.6) and not demonstrated in subsequent studies, where the same methodology (in another limited set of patients with 16 total recurrences in the low-risk group from a similarly-sized cohort) showed the expected 50% relative-risk reduction in these patients⁴⁵ but was not statistically significant due to numbers of recurrences (4 recurrences in the BCS only group and 12 in the BCT group). A third study with randomization for BCS and BCT showed that RT in low-risk patients (DS2<3) reduced IBTR 10-year risk by 5.7%, a relative-risk decrease of 44.2% that was not significant (p=0.059).⁴⁸ While Vicini’s reanalysis of prior data (at a lower DS2 cutoff plus RRt) seemingly greatly improves the test by removing poor-performing patients from the low-risk group (to the high and RRt groups) and shows little value from RT in the low-risk group, this inference is made from 17 or fewer recurrences total (BCS or BCT groups), where both BCT and BCS-treated patients had IBTR rates of ~5% (but the 95% confidence intervals for BCS-treated patients was 2.5%-12.1%).⁴⁹ These data cannot reasonably be interpreted as evidence that low risk patients have no relative risk reduction from RT, but raise concern that the studies are underpowered to draw such a conclusion given the expected small absolute risk reduction and small sample sizes with informative IBTR patients in the single digits for some categories (particularly in the low-risk groups).⁴⁵ Moreover, pooled analyses by Ouattara et al of these same data concluded that the low-risk group did have significant IBTR risk reduction from RT (HR=0.62, CI=0.39-0.99).⁵³ Such conclusions would be even more difficult to draw for invasive cancer recurrences as data are more limited. The demonstration that a low-risk patient attains no relative risk reduction from RT is unsupported but also unnecessary; we reason that demonstrating less than 5% absolute risk reduction benefit is sufficient to warrant preclusion of RT use regardless of personal factors given SME feedback on decision making.

Tests must be validated, and if any changes are made to the test or algorithm, they must again be externally validated in representative cohorts. Both DCISionRT and OncotypeDx DCIS have made changes to their test strategy and have not clearly and sufficiently validated the changes in independent and reliable patient cohorts. As an example, the Vicini 2023 paper re-evaluates cutoffs for DS2 at 2.8 rather than the 3.0 cutoff used in the prior published datasets wherein DCISionRT was already evaluated, and it is not clear that these data were omitted from the discovery of the new cutoff. Furthermore, the combined 926 patients only had 77 recurrences in 10 years (12%) including all risk types and BCS use, which is lower than expected for such a cohort and leaves questions as to how representative of DCIS patients this study is. While the new cutoff shows promise, it needs further external validation and concordance studies with other patient populations.

In summary, biomarker tests such as DCISionRT and OncotypeDX DCIS have some limited evidentiary support for risk stratification and may even identify sufficiently low risk patients. However, to meet reasonable and necessary thresholds, the tests must clearly demonstrate they meet analytical and clinical validity, and clinical utility requirements discussed herein. The current literature supporting their use is sufficiently flawed in demonstrating such requirements that it is the opinion of this contractor that further evidence must be developed to demonstrate with confidence whether these tests can identify a sufficiently low risk population that does not reasonably benefit from treatment with RT, or other possible valuable risk stratification categories for DCIS management. The tests’ abilities to identify sufficiently low risk patients must be evaluated directly in comparison to the best available clinicopathological tools available, including the MSKCC nomogram and the VNPI. Given that some of the variables evaluated in the nomograms and the biomarker tests overlap, it remains unclear whether it is a particular combination of some of the clinicopathologic variables (present in both the biomarker tests and the nomograms) that is driving the results of the biomarker risk assessment and it must be clearly demonstrated that the analytes measured are valid beyond the value of the combined clinicopathological factors. Changes in test methodology must be externally validated with valid and representative patient data and with concordance studies. Further, conclusions regarding risk of invasive recurrence are based on even more limited patient data, as invasive cancer rates are approximately half of IBTR rates, and it is unclear how this information would lead to different decisions beyond IBTR; any such observations should be considered exploratory. Novel risk groups such as RRt are interesting but do not clearly confer clinical utility as it is not clear how this information would alter patient management leading to improved outcomes.

This contractor believes the utility of risk stratification, based on all evidence reviewed, is in the identification of patients who may not derive meaningful benefit from RT aside from individualized risk considerations given current practices and dependence on RT use. At this time, there is no convincing evidence that any test (and their analytes measured) can identify a patient who would not reasonably benefit from RT beyond clinicopathologic factors already demonstrated to risk stratify patients; but if an absolute risk reduction is ~5% or less, most practitioners should conclude that the benefit is not meaningful and should advocate for RT avoidance. Patients should not get such a test unless it is clear they could safely forgo RT use if placed in that group. Other patient factors, such as age or co- morbidities, should also be considered before a test is ordered and assist in making a decision regarding performing such a test.

N/A

N/A

1. Lips EH, Kumar T, Megalios A, et al. Genomic analysis defines clonal relationships of ductal carcinoma in situ and recurrent invasive breast cancer. Nat Genet. 2022;54(6):850-860. doi:10.1038/s41588-022-01082-3

2. Ward EM, DeSantis CE, Lin CC, et al. Cancer statistics: breast cancer in situ. CA Cancer J Clin. 2015;65(6):481-495. doi:10.3322/caac.21321

3. Zujewski JA, Harlan LC, Morrell DM, Stevens JL. Ductal carcinoma in situ: trends in treatment over time in the US. Breast Cancer Res Treat. 2011;127(1):251-257. doi:10.1007/s10549-010-1198-z

4. Early Breast Cancer Trialists' Collaborative Group (EBCTCG), Correa C, McGale P, et al. Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. J Natl Cancer Inst Monogr. 2010;2010(41):162-177. doi:10.1093/jncimonographs/lgq039

5. Cuzick J, Sestak I, Pinder SE, et al. Effect of tamoxifen and radiotherapy in women with locally excised ductal carcinoma in situ: long-term results from the UK/ANZ DCIS trial. Lancet Oncol. 2011;12(1):21-29. doi:10.1016/S1470-2045(10)70266-7

6. van Seijen M, Lips EH, Fu L, et al. Long-term risk of subsequent ipsilateral lesions after surgery with or without radiotherapy for ductal carcinoma in situ of the breast. Br J Cancer. 2021;125(10):1443-1449. doi:10.1038/s41416-021-01496-6

7. McCormick B, Winter K, Hudis C, et al. RTOG 9804: a prospective randomized trial for good-risk ductal carcinoma in situ comparing radiotherapy with observation [published correction appears in J Clin Oncol. 2015 Sep 10;33(26):2934. doi: 10.1200/JCO.2015.64.1290.]. J Clin Oncol. 2015;33(7):709-715. doi:10.1200/JCO.2014.57.9029

8. Narod SA, Iqbal J, Giannakeas V, Sopik V, Sun P. Breast cancer mortality after a diagnosis of ductal carcinoma in situ. JAMA Oncol. 2015;1(7):888-896. doi:10.1001/jamaoncol.2015.2510

9. Giannakeas V, Sopik V, Narod SA. Association of radiotherapy with survival in women treated for ductal carcinoma in situ with lumpectomy or mastectomy. JAMA Net Open.2018;1(4):e181100. doi:10.1001/jamanetworkopen.2018.1100

10. Goldberg M, Whelan TJ. Systemic Effects of radiotherapy in ductal carcinoma in situ. JAMA Netw Open. 2018;1(4):e181102. doi:10.1001/jamanetworkopen.2018.1102.

11. Withrow DR, Morton LM, Curtis RE, Schonfeld SJ, Berrington de González A. Radiotherapy for ductal carcinoma in situ and risk of second non-breast cancers. Breast Cancer Res Treat. 2017;166(1):299-306. doi:10.1007/s10549-017-4410-6

12. Worni M, Akushevich I, Greenup R, et al. Trends in treatment patterns and outcomes for ductal carcinoma in situ. J Natl Cancer Inst. 2015;107(12):djv263. doi:10.1093/jnci/djv263

13. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines). Breast Cancer. Version 4.2024. https://www.nccn.org/professionals/physician_gls/pdf/breast.pdf . Accessed 2/11/2025.

14. Di Saverio S, Catena F, Santini D, et al. 259 patients with DCIS of the breast applying USC/Van Nuys prognostic index: a retrospective review with long term follow up. Breast Cancer Res Treat. 2008;109(3):405-416. doi:10.1007/s10549-007-9668-7

15. Solin LJ, Gray R, Hughes LL, et al. Surgical excision without radiation for ductal carcinoma in situ of the breast: 12-year results from the ECOG-ACRIN E5194 study. J Clin Oncol. 2015;33(33):3938-3944. doi:10.1200/JCO.2015.60.8588

16. Smith GL. Toward minimizing overtreatment and undertreatment of ductal carcinoma in situ in the United States. J Clin Oncol. 2016;34(11):1172-1174. doi:10.1200/JCO.2015.66.2064

17. McCormick B, Winter KA, Woodward W, et al. Randomized phase III trial evaluating radiation following surgical excision for good-risk ductal carcinoma in situ: long-term report From NRG Oncology/RTOG 9804. J Clin Oncol. 2021;39(32):3574-3582. doi:10.1200/JCO.21.01083

18. Hughes LL, Wang M, Page DL, et al. Local excision alone without irradiation for ductal carcinoma in situ of the breast: a trial of the Eastern Cooperative Oncology Group. J Clin Oncol. 2009;27(32):5319-5324. doi:10.1200/JCO.2009.21.8560

19. Wright JL, Gray R, Rahbar H, et al. Lumpectomy without radiation for ductal carcinoma in situ of the breast: 20-year results from the ECOG-ACRIN E5194 study. NPJ Breast Cancer. 2024;10(1):16. doi:10.1038/s41523-024-00622-w

20. Silverstein MJ. Ductal carcinoma in situ of the breast: a surgeon's disease. Ann Surg Oncol. 1999;6(8):802-810. doi:10.1007/s10434-999-0802-0

21. Silverstein MJ, Craig PH, Lagios MD, et al. Developing a prognostic index for ductal carcinoma in situ of the breast. Are we there yet?. Cancer. 1996;78(5):1138-1140. doi:10.1002/(SICI)1097-0142(19960901)78:5<1138::AID-CNCR27>3.0.CO;2-2

22. Silverstein MJ, Lagios MD, Craig PH, et al. A prognostic index for ductal carcinoma in situ of the breast. Cancer. 1996;77(11):2267-2274. doi:10.1002/(SICI)1097-0142(19960601)77:11<2267::AID-CNCR13>3.0.CO;2-V

23. Silverstein MJ, Lagios MD, Groshen S, et al. The influence of margin width on local control of ductal carcinoma in situ of the breast. N Engl J Med. 1999;340(19):1455-1461. doi:10.1056/NEJM199905133401902

24. Silverstein MJ, Lagios MD. Choosing treatment for patients with ductal carcinoma in situ: fine tuning the University of Southern California/Van Nuys prognostic index. J Natl Cancer Inst Monogr. 2010;2010(41):193-196. doi:10.1093/jncimonographs/lgq040

25. Kunkiel M, Niwinska A. Assessment of the usefulness of prognostic Van Nuys prognostic index in the treatment in ductal carcinoma in situ in 15-year observation. Sci Rep. 2021;11(1):22645. doi:10.1038/s41598-021-02126-0

26. Whitfield R, Kollias J, de Silva P, Turner J, Maddern G. Management of ductal carcinoma in situ according to Van Nuys prognostic index in Australia and New Zealand. ANZ J Surg. 2012;82(7-8):518-523. doi:10.1111/j.1445-2197.2012.06133.x

27. Gilleard O, Goodman A, Cooper M, Davies M, Dunn J. The significance of the Van Nuys prognostic index in the management of ductal carcinoma in situ. World J Surg Oncol. 2008;6:61. doi:10.1186/1477-7819-6-61

28. MacAusland SG, Hepel JT, Chong FK, et al. An attempt to independently verify the utility of the Van Nuys prognostic index for ductal carcinoma in situ. Cancer. 2007;110(12):2648-2653. doi:10.1002/cncr.23089

29. Rudloff U, Jacks LM, Goldberg JI, et al. Nomogram for predicting the risk of local recurrence after breast-conserving surgery for ductal carcinoma in situ. J Clin Oncol. 2010;28(23):3762-3769. doi:10.1200/JCO.2009.26.8847

30. Subhedar P, Olcese C, Patil S, Morrow M, Van Zee KJ. Decreasing recurrence rates for ductal carcinoma in situ: analysis of 2996 women treated with breast-conserving surgery over 30 years [published correction appears in Ann Surg Oncol. 2015 Dec;22 Suppl 3:S1618. doi: 10.1245/s10434-015-4884-6.]. Ann Surg Oncol. 2015;22(10):3273-3281. doi:10.1245/s10434-015-4740-8

31. Collins LC, Achacoso N, Haque R, et al. Risk prediction for local breast cancer recurrence among women with DCIS treated in a community practice: a nested, case-control study. Ann Surg Oncol. 2015;22 Suppl 3:S502-S508. doi:10.1245/s10434-015-4641-x

32. Yi M, Meric-Bernstam F, Kuerer HM, et al. Evaluation of a breast cancer nomogram for predicting risk of ipsilateral breast tumor recurrences in patients with ductal carcinoma in situ after local excision [published correction appears in J Clin Oncol. 2012 Jul 1;30(19):2424]. J Clin Oncol. 2012;30(6):600-607. doi:10.1200/JCO.2011.36.4976

33. Sweldens C, Peeters S, van Limbergen E, et al. Local relapse after breast-conserving therapy for ductal carcinoma in situ: a European single-center experience and external validation of the Memorial Sloan-Kettering Cancer Center DCIS nomogram. Cancer J. 2014;20(1):1-7. doi:10.1097/PPO.0000000000000025

34. Wang F, Li H, Tan PH, et al. Validation of a nomogram in the prediction of local recurrence risks after conserving surgery for Asian women with ductal carcinoma in situ of the breast. Clin Oncol (R Coll Radiol). 2014;26(11):684-691. doi:10.1016/j.clon.2014.08.004

35. Oses G, Mension E, Pumarola C, et al. Analysis of local recurrence risk in ductal carcinoma in situ and external validation of the Memorial Sloan Kettering Cancer Center nomogram. Cancers (Basel). 2023;15(8):2392. doi:10.3390/cancers15082392

36. Sagara Y, Freedman RA, Vaz-Luis I, et al. Patient prognostic score and associations with survival improvement offered by radiotherapy after breast-conserving surgery for ductal carcinoma in situ: a population-based longitudinal cohort study. J Clin Oncol. 2016;34(11):1190-1196. doi:10.1200/JCO.2015.65.1869

37. Solin LJ, Gray R, Baehner FL, et al. A multigene expression assay to predict local recurrence risk for ductal carcinoma in situ of the breast. J Natl Cancer Inst. 2013;105(10):701-710. doi:10.1093/jnci/djt067

38. Rakovitch E, Nofech-Mozes S, Hanna W, et al. A population-based validation study of the DCIS score predicting recurrence risk in individuals treated by breast-conserving surgery alone. Breast Cancer Res Treat. 2015;152(2):389-398. doi:10.1007/s10549-015-3464-6

39. Rakovitch E, Nofech-Mozes S, Hanna W, et al. Multigene expression assay and benefit of radiotherapy after breast conservation in ductal carcinoma in situ. J Natl Cancer Inst. 2017;109(4):djw256. doi:10.1093/jnci/djw256

40. Paszat L, Sutradhar R, Zhou L, Nofech-Mozes S, Rakovitch E. Including the ductal carcinoma-in-situ (DCIS) score in the development of a multivariable prediction model for recurrence after excision of DCIS. Clin Breast Cancer. 2019;19(1):35-46. doi:10.1016/j.clbc.2018.07.018

41. Rakovitch E, Gray R, Baehner FL, et al. Refined estimates of local recurrence risks by DCIS score adjusting for clinicopathological features: a combined analysis of ECOG-ACRIN E5194 and Ontario DCIS cohort studies. Breast Cancer Res Treat. 2018;169(2):359-369. doi:10.1007/s10549-018-4693-2

42. Rakovitch E, Parpia S, Koch A, et al. DUCHESS: an evaluation of the ductal carcinoma in situ score for decisions on radiotherapy in patients with low/intermediate-risk DCIS. Breast Cancer Res Treat. 2021;188(1):133-139. doi:10.1007/s10549-021-06187-7

43. Hahn E, Sutradhar R, Paszat L, et al. Molecular expression assays improve the prediction of local and invasive local recurrence after breast-conserving surgery for ductal carcinoma in situ. J Clin Oncol. 2024;42(27):3196-3206. doi:10.1200/JCO.23.02276

44. Bremer T, Whitworth PW, Patel R, et al. A biological signature for breast ductal carcinoma in situto predict radiotherapy benefit and assess recurrence risk. Clin Cancer Res. 2018;24(23):5895-5901. doi:10.1158/1078-0432.CCR-18-0842

45. Weinmann S, Leo MC, Francisco M, et al. Validation of a ductal carcinoma in situbiomarker profile for risk of recurrence after breast-conserving surgery with and without radiotherapy. Clin Cancer Res. 2020;26(15):4054-4063. doi:10.1158/1078-0432.CCR-19-1152

46. Shah C, Bremer T, Cox C, et al. The clinical utility of DCISionRT^®on radiation therapy decision making in patients with ductal carcinoma in situ following breast-conserving surgery [published correction appears in Ann Surg Oncol. 2021 Dec;28(Suppl 3):878. doi: 10.1245/s10434-021-10138-3.]. Ann Surg Oncol. 2021;28(11):5974-5984. doi:10.1245/s10434-021-09903-1

47. Shah C, Whitworth P, Vicini FA, et al. The clinical utility of a 7-gene biosignature on radiation therapy decision making in patients with ductal carcinoma in situ following breast-conserving surgery: an updated analysis of the DCISionRT^®PREDICT study. Ann Surg Oncol. 2024;31(9):5919-5928. doi:10.1245/s10434-024-15566-5

48. Wärnberg F, Karlsson P, Holmberg E, et al. Prognostic risk assessment and prediction of radiotherapy benefit for women with ductal carcinoma in situ (DCIS) of the breast, in a randomized clinical trial (SweDCIS). Cancers (Basel). 2021;13(23):6103. doi:10.3390/cancers13236103

49. Vicini FA, Mann GB, Shah C, et al. A novel biosignature identifies patients with DCIS with high risk of local recurrence after breast conserving surgery and radiation therapy. Int J Radiat Oncol Biol Phys. 2023;115(1):93-102. doi:10.1016/j.ijrobp.2022.06.072

50. Vicini F, Shah C, Mittal K, et al. A 7-gene biosignature for ductal carcinoma in situ of the breast identifies subpopulations of HER2-positive patients with distinct recurrence rates after breast-conserving surgery and radiation therapy. Clin Breast Cancer. 2025;25(2):e152-e158.e1. doi:10.1016/j.clbc.2024.08.016

51. Dabbs D, Mittal K, Heineman S, et al. Analytical validation of the 7-gene biosignature for prediction of recurrence risk and radiation therapy benefit for breast ductal carcinoma in situ. Front Oncol. 2023;13:1069059. doi:10.3389/fonc.2023.1069059

52. Schmitz RSJM, Wilthagen EA, van Duijnhoven F, et al. Prediction models and decision aids for women with ductal carcinoma in situ: a systematic literature review. Cancers (Basel). 2022;14(13):3259. doi:10.3390/cancers14133259

53. Ouattara D, Mathelin C, Özmen T, Lodi M. Molecular signatures in ductal carcinoma in situ (DCIS): a systematic review and meta-analysis. J Clin Med. 2023;12(5):2036. doi:10.3390/jcm12052036

54. Lei RY, Carter DL, Antell AG, et al. A comparison of predicted ipsilateral tumor recurrence risks in patients with ductal Carcinoma in situ of the breast after breast-conserving surgery by breast radiation oncologists, the Van Nuys prognostic index, the Memorial Sloan Kettering Cancer Center DCIS nomogram, and the 12-gene DCIS score assay. Adv Radiat Oncol. 2020;6(2):100607. doi:10.1016/j.adro.2020.10.020

55. Van Zee KJ, Zabor EC, Di Donato R, et al. Comparison of local recurrence risk estimates after breast-conserving surgery for DCIS: DCIS nomogram versus refined Oncotype DX Breast DCIS Score. Ann Surg Oncol. 2019;26(10):3282-3288. doi:10.1245/s10434-019-07537-y

56. CAC- Biomarker Testing for Risk Stratification in DCIS Questions. July 15, 2024. Accessed 2/11/2025. CAC Biomarker Testing for Risk Stratification in DCIS Questions.

57. MolDX: BIOMARKER RISK STRATIFICATION TESTING IN DCIS-CONTRACTOR ADVISORY COMMITTEE (CAC) MEETING. July 15, 2024. Accessed 2/11/2025. July 15, 2024, Multi-Jurisdictional Contractor Advisory Contractor (CAC) Meeting MoIDX: Biomarker Risk Stratification Testing in DCIS Transcript

• Other (LCD request from laboratory. )

• Biomarker Testing
• DCIS