Medical Policy

This document addresses molecular signature testing (for example, PrismRA, Scipher Medicine, Waltham, MA) to predict response to Tumor Necrosis Factor inhibitor (TNFi) therapy. The molecular signature test includes RNA sequencing and gene expression data to determine if an individual is unlikely to respond to TNFi therapy.

Note: For a high-level overview of this document, please see “Summary for Members and Families” below.

Investigational and Not Medically Necessary:

Molecular signature testing to predict response to Tumor Necrosis Factor inhibitor (TNFi) therapy is considered investigational and not medically necessary for all uses, including but not limited to guiding treatment for rheumatoid arthritis.

This document describes clinical studies and expert recommendations, and explains why molecular signature testing to predict response to Tumor Necrosis Factor inhibitor (TNFi) therapy is not considered to be clinically appropriate. The following summary does not replace the medical necessity criteria or other information in this document. The summary may not contain all of the relevant criteria or information. This summary is not medical advice. Please check with your healthcare provider for any advice about your health.

Key Information

TNFis are medicines that reduce inflammation by blocking a protein called tumor necrosis factor (TNF). They are used to treat certain conditions called autoimmune diseases, where the body’s defense system attacks itself. Examples of autoimmune diseases include rheumatoid arthritis (RA), psoriatic arthritis, ankylosing spondylitis, Crohn disease, ulcerative colitis, and psoriasis. TNFi medicines include adalimumab (Humira^®), certolizumab pegol (Cimzia^®), etanercept (Enbrel^®), golimumab (Simponi^®), and infliximab (Remicade^®).

Molecular signature testing is a type of medical test that looks for patterns in genes, proteins, or other markers in a blood or tissue sample. Researchers are studying whether molecular signature testing can help predict how well a person with an autoimmune disease will respond to treatment with a TNFi, help choose the most effective treatment, and avoid medicines that may not work. However, their routine use for this purpose has not been proven to improve health outcomes such as decreased disease severity.

One available molecular signature test is called the PrismRA test.

What the Studies Show

Researchers have studied the PrismRA test and similar molecular signature tests in people with RA. Several studies found that the test was often able to identify people who did not respond to TNFi medicines. Some studies also suggested that treatment decisions guided by test results may improve disease activity scores. However, important questions about the use of molecular signature testing remain. Many studies included low numbers of people, and used highly selected populations that may not reflect the broader group of people who would receive the test in routine clinical practice. In addition, many studies did not use comparison groups, which is important to be able to know how a medical treatment works in relation to another. Such studies also are more likely to be affected by factors that may change the study outcome but are unrelated to the treatment in question, making it harder to determine whether the test itself improves clinical outcomes. Some studies showed that the test was better at identifying people who do not respond to specific drugs than identifying those that do. Most studies did not prove that using the test leads to better health outcomes such as less disease activity, fewer complications, improved quality of life, or less joint damage. Additional studies are needed in larger and more diverse groups of people.

The American College of Rheumatology (ACR) does not currently recommend the use of molecular signature testing as part of routine RA treatment decisions. The test itself may not cause significant physical harm beyond blood collection, but inaccurate results could lead to treatment choices that do not provide benefit.

Is this Clinically Appropriate?

Molecular signature testing to predict response to TNFi therapy is not clinically appropriate because it has not been proven to improve health and because the benefits and risks of changing treatment based on such test results remain uncertain.

(Return to Description/Scope)

Summary

Molecular signature testing offers the potential to improve treatment results by avoiding Tumor Necrosis Factor inhibitor (TNFi) therapy when an individual is predicted to be a non-responder. Early studies of molecular signature tests to predict an individual’s response to TNFi therapy for rheumatoid arthritis (RA) have indicated some predictive capabilities, such as establishing a classifier model with promising specificity. However, there remain significant limitations. The current model shows a high positive predictive value (PPV) for non-response but falls short in negative predictive accuracy. More peer-reviewed studies are needed to confirm test accuracy in identifying responders and non-responders, especially in diverse populations. These studies should also show whether using the test actually improves health outcomes for people with RA. Current standardized endorsed treatment guidelines do not recommend use of molecular signature testing in the treatment of RA.

Discussion

The use of signature molecular testing has been investigated for predicting response to TNFi therapy using RNA sequencing and gene expression data. The available literature addresses a single test, the PrismRA, developed to assist providers in therapy selection by identifying individuals that are predicted to be unlikely to respond to TNFi therapy.

In 2020, Mellors and colleagues reported the results of a cross-platform, cross-cohort study to develop a classification algorithm to predict whether individuals diagnosed with RA will respond to anti-TNF drug treatment. The authors of the study analyzed gene expression biomarkers for 376 participants who had participated in another trial of TNFi effectiveness (the CERTAIN trial by the Consortium of Rheumatology Researchers of North America [CORRONA]; Pappas, 2014). Serum samples had been collected prior to longitudinal evaluations over at least 6 months of follow-up while being treated with a TNFi (adalimumab, certolizumab pegol, etanercept, golimumab, or infliximab). Response to treatment was measured as the percentage of participants who experienced at least a 50% reduction in symptom severity at 6 months as measured by the American College of Rheumatology (ACR) core set of RA markers (ACR50 score). The authors of this study noted that only 30.2% of individuals saw this level of benefit from TNFi treatment. In other words, 30.2% were responders and 69.8% were nonresponders. The authors also note that this response rate is similar to that seen in other studies (Alonso-Ruiz, 2008).

Mellors and colleagues described a study involving the analysis of serum biomarkers from 58 participants as a discovery cohort, 143 participants as a training cohort and 175 participants used as a validation cohort. For the discovery cohort, they evaluated the biomarker profiles of 58 anti-TNFi treatment-naïve individuals. Mann-Whitney U tests were used to eliminate biomarkers that were not statistically different between responders and non-responders. Further statistical analysis identified 70 biomarker features that were most different between responders and non-responders. These features included gene expression products, single-nucleotide polymorphisms (SNPs), and clinical factors. All members of the discovery cohort were female, but the relationship of this 70-feature set to response/non-response was validated for the entire training set including males and females.

As a second step, the authors iteratively analyzed their candidate biomarker set in samples from the 143-participant training set. They created a Random Forest model using the 25 features most associated with non-response. They next evaluated the ability of this model to predict response to therapy for the 175 individuals in the validation cohort. None of the validation participants had been included in the discovery or training sets.

Of those who responded to treatment, 86.8% (46/53) had a test result indicating that they were likely to respond. Half of the participants in the validation set who did not respond to treatment had a test that indicated they were likely to respond. Among participants with a test indicating nonresponse, the ratio of nonresponders to responders was 3.79. For those with tests indicating response, the ratio of nonresponders to responders was 0.58. This indicates that individuals with a positive test are 6.3 (3.79/0.58) times as likely to be nonresponders as they are to be responders.

This study has several limitations. Although the authors showed correlation with non-response for this small set of participants, this study does not show that use of this model can prospectively improve response rates among individuals with RA proposed for TNFi treatment. SNP alleles may vary among ethnic groups and the training and validation cohorts were predominately White (88.1% and 84.6%, respectively). The authors report no significant difference in the panel’s ability to predict response among individuals of other ethnicities. Additional work is needed to prospectively evaluate prediction performance in diverse populations.

In 2021, Cohen and colleagues published the results of a prospective observational study of a molecular signature response classifier (MSRC) using 23 features (PrismRA) to predict inadequate response to TNFi treatment. A total of 146 individuals completed the 24-week study with clinical assessments occurring at baseline, 3 months, and 6 months. Assessments included patient global assessment of pain, patient global assessment of disease activity, Clinical Disease Activity Index (CDAI) score, Health Assessment Questionnaire, and C-reactive protein (CRP). The 3-month visit also included molecular and clinical data such as the ACR50, ACR70, CDAI, and the Disease Activity Score-28 with CRP (DAS28-CRP) that was then used to predict response to TNFi therapy. PAXgene RNA samples were collected at all 3 visits. The MSRC detected non-response in 46/113 (40.7%). Among TNFi-exposed participants, the classifier was associated with significantly increased odds of inadequate response across most outcome measures (odds ratio [OR] range, 3.3-26.6), although the magnitude of association varied substantially across endpoints and some estimates had wide confidence intervals, reflecting limited precision. Of the 146 individuals, 26/146 (17.8%) achieved CDAI remission at the 6-month visit. A total of 22/68 (32.4%) of individuals that did not have a molecular signature for non-response achieved remission and 4/78 (5.1%) for those with the molecular signature for non-response achieved remission. A limitation of this study includes its observational design. Treatment allocation was not randomized and residual confounding cannot be excluded. The study did not assess development of antidrug antibodies which may have affected treatment response, and MSRC results were not used to inform treatment selection. Also, some of the outputs incorporated subjective assessments of disease activity, which may have introduced measurement variability into response classification. Nearly half of enrolled participants were excluded because of attrition or incomplete molecular or clinical data.

Jones (2021) reported the results of a retrospective study involving residual samples from 174 participants with RA who had participated in the NETWORK-004 study. A total of 100 participants were TNFi treatment naïve and 74 had received prior treatment with TNFi therapy. The majority of participants had received concomitant methotrexate therapy (> 80%) and treatment with all five available TNFi drugs were represented in the study population. Based on ACR50 measurements at 6 months, the authors reported a PPV of 87.7% (95% confidence interval [CI], 78-94%), sensitivity of 60.2% (95% CI: 50-69%), and specificity of 77.3% (95% CI, 65-87%).

Strand (2022a) described an interim analysis of data from the Study to Accelerate Information of Molecular Signatures (AIMS) in Rheumatoid Arthritis registry involving participants with RA who underwent testing with the 21-factor PrismRA assay and who had test-result based treatment. The report included results from 85 participants who completed 24 weeks of PrismRA-guided therapy. The primary endpoint was therapeutic responsiveness defined as ACR50 at 24 weeks. Participants were stratified into 4 groups: 1) PrismRA results indicating non-responder status and who were treated with non-TNFi therapy; 2) PrismRA results indicating non-responder status and who were treated with TNFi drugs; 3) PrismRA results indicating responder status and treated with TNFi drugs; and 4) PrismRA results indicating responder status and treated with non-TNFi therapy. In participants receiving therapy in alignment with PrismRA test results (Groups 1, 3 and 4), the 24-week mean ACR50 responses were reported to be 39.6%. In the groups where participants had a PrismRA result indicating non-response and who were treated with non-TNFi therapy, significantly improved therapeutic responses were reported (ACR50: Group 1 = 34.8% vs. Group 2 = 10.3%, p=0.05; ≥ MID [minimally important differences] in CDAI: Group 1 = 56.2% vs. Group 2 = 15.4%, p=0.009). No significant differences were observed between groups that underwent treatment with non-TNFi drugs but had differing PrismRA results (ACR50: Group 1 = 34.8% vs. Group 4 = 33.3%, p>0.05; ≥ MID in CDAI: Group 1 = 56.2% vs. Group 4 = 42.9%, p>0.05). Finally, participants with PrismRA results indicating non-response had significantly lower therapeutic responses when treated with TNFi therapy compared with participants with responder PrismRA status participants (ACR50: Group 2 = 10.3% vs. Group 3 = 45.8%, p=0.005; ≥ MID in CDAI: Group 2 = 15.4% vs. Group 3 = 45.0%, p=0.02). These interim results are promising and the final results from this trial should provide additional useful data addressing the clinical utility of the PrismRA test.

Strand (2022b) published a second interim study using AIMS registry data. The report included data from 470 participants with moderate to severe RA who had treatment decisions based on PrismRA results and had data at 12-week follow-up. Data from 274 participants with 24-week data were also reported. The primary and secondary endpoints of the clinical outcomes analysis were changes from baseline in absolute CDAI scores at 24 weeks and 12 weeks. The authors reported that participants with PrismRA non-responder status who received non-TNFi therapy experienced up to 1.8-fold greater improvement in CDAI scores when compared to participants with non-responder status who received TNFi therapy (12 weeks: 12.2 vs. 8.0, p=0.083; 24 weeks: 14.2 vs. 7.8, p=0.009). Participants with high baseline disease activity and PrismRA non-responder status had a 1.7-fold greater improvement in CDAI scores when treated with a non-TNFi therapy compared to PrismRA non-responder status participants treated with TNFs (8.9 vs. 15.1, no p-value provided). This corresponded to 38.9% of participants achieving a lower disease activity level in response to TNFi therapy compared to 55.7% with non-TNFi therapy, a 43.2% relative improvement. Participants with moderate baseline disease activity had a 3.1-fold greater improvement (3.1 vs. 9.7, respectively, no p-value provided). A greater proportion of PrismRA non-responder status participants experienced worsening CDAI scores when treated with TNFi therapy compared with non-TNFi therapy (high baseline disease activity: 27.8% vs. 17.1%, no p-value provided). Finally, compared to participants with PrismRA responder status who received non-TNFi therapy, those receiving TNFi therapy had 1.6-fold greater improvement in CDAI scores at 12 weeks (8.0 vs. 10.2; p=0.087) and a 1.9-fold improvement at 24 weeks (7.8 vs. 12.7; p=0.012). As with the previously discussed Strand study, these interim results are promising and indicate potential clinical utility for the use of PrismRA test results in informing treatment decisions in individuals with RA. Full results from the completed study will provide a better picture of this issue.

In 2022, Curtis published a retrospective comparative cohort study using data from 627 participants from the AIMS registry compared to 2721 propensity-match control participants from electronic health records. A statistical matching method (PSM) was applied to allow comparison of the two groups, leading to a total of 489 PrismRA-guided treatment/experimental group participants and 761 control group participants being included in the final analysis. Thirty-eight experimental group participants and 494 control group participants switched biologic or targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) during the 6-month study course. The authors noted that the PrismRA group had higher baseline disease activity vs. the control group. In the PrismRA group 368 participants (59%) were identified as having non-response signatures to TNFi and 441 (70%) were prescribed a b/tsDMARD that aligned with PrismRA result. After PSM, the odds of responding to b/tsDMARD therapy at 6 months were significantly greater in the PrismRA group vs. controls (p<0.0001). Changes in the response rates in PrismRA group were between 1.6 and 2.9 times greater than those in the control group, depending upon the measure used. In the PrismRA group the ability to detect a non-response signature was evaluated in TNFi treated participants (n=369), with 54% who did not respond also having a TNFi non-response signature. Conversely, 88% of PrismRA group participants with a non-response signature did not respond to TNFi therapy (ACR50: PPV=88%, sensitivity=54%, specificity=70%, area under the curve [AUC]=0.65). Finally, PrismRA group participants predicted to be likely non-responders to TNFi therapy were approximately 3 times less likely to satisfy ACR50 criteria (OR, 2.73) after 24 weeks of TNFi therapy. As with the prior studies, these results indicate that use of the PrismRA test can help identify the likelihood of drug response in individuals for whom TNFi therapy is recommended. However, these results do not demonstrate changes in health outcomes such as avoided complications, disease flareups, and morbidity.

In 2024, Curtis and colleagues reported results of a study to examine the impact of MSRC test results on treatment decisions by rheumatologists. Using data from the AIMS registry, this study assessed selection of b/tsDMARDs after MSRC testing by treating physicians, the rate of b/tsDMARD prescriptions aligning with MSRC results, and the percentage of physicians utilizing MSRC results for decision making. Of 1018 participants surveyed, 70.7% had treatment selected after receiving MSRC results and 75.6% of these individuals received a b/tsDMARD aligned with MSRC results. Among physicians treating these individuals, 84.6% reported using MSRC results to guide treatment selection. Although this study reports a high rate of adoption of the MSRC test among physicians and practices participating in the AIMS registry, it did not correlate test-aligned treatment decisions with clinical outcomes such as remission or quality of life. This weakens claims of utility from a patient-centric view. The study’s reliance on retrospective reporting of how treatment decisions were made may be subject to recall and social desirability biases. Additionally, the study did not examine how previous treatments and active comorbidities may have influenced prescribing decisions.

Lewis and colleagues (2025) performed a post-hoc biomarker analysis of pretreatment synovial RNA sequencing (RNA-Seq) data from participants enrolled in the biopsy-based, precision-medicine Stratification of Biologic Therapies for RA by Pathobiology (STRAP) trial (n=208) to classify gene response signatures to these randomized therapies:

• etanercept (TNF-inhibitor),
• tocilizumab (interleukin-6 receptor inhibitor) and
• rituximab (anti-CD20 B-cell depleting antibody).

Machine learning models incorporating RNA-Seq data predicted clinical response to the three different therapies at the 16-week primary endpoint with AUC values of 0.763, 0.748 and 0.754, respectively. Predictive signatures were incorporated into a synovium-specific 524-gene nCounter panel (507 target genes, 17 housekeeping genes) and evaluated using residual synovial biopsy RNA from STRAP participants, demonstrating similar discrimination (AUC 0.82-0.87). Independent external validation was performed only for the tocilizumab and rituximab models using the R4RA cohort; no comparable external validation cohort was available for the etanercept model. Although these findings suggest that synovial transcriptomic signatures may help stratify response to biologic therapies, the clinical utility of biomarker-guided treatment selection has not yet been demonstrated. The ongoing 3TR PRECIS-THE-RA trial will directly test whether assigning individuals with RA to either a TNF-inhibitor or an IL6-inhibitor based on the probability of response determined through a synovial biopsy nCounter assay will enrich for clinical response compared to the control arm, in which participants will receive one of the two drugs randomly.

The current version of the American College of Rheumatology (ACR) Guideline for the Treatment of Rheumatoid Arthritis does not address the use of molecular signature testing (Fraenkel, 2021).

Other Relevant Information

According to the Centers for Medicare & Medicaid Services (CMS) Local Coverage Determination (LCD) L39424, titled MolDX: Molecular Biomarker Testing to Guide Targeted Therapy Selection in Rheumatoid Arthritis, limited coverage is supported for molecular biomarker tests to guide targeted therapy selection in RA. Although the MSRC test will only identify about 55% of the TNFi non-responders and the utility is limited to only a subset of the RA population, testing is still recommended for those individuals who meet criteria. The LCD acknowledges that international, national, and society guidelines have not yet endorsed predictive (for response to therapy) biomarker testing in RA, however:

Despite the many limitations of predictive biomarker tests, a review of the evidence, as imperfect as it may be, supports the limited use of these services given their demonstrated validity and utility. When a NR [non-response] signature is obtained by the MSRC, nearly 90% of those patients will prove to not clinically respond to TNFi therapies using multiple validated disease response criteria including the ACR50 and CDAI. In these patients, a non-TNFi is likely the best therapeutic option.

Rheumatoid arthritis (RA) is an autoimmune disease with no current cure or known underlying cause. Although arthritis is almost always present and is often the most troubling manifestation of the disorder, the disease is systemic and can have severe effect on other organs such as the eyes, lungs, heart, and integument.

“Classic” RA is characterized by pain, swelling, and stiffness in many joints. Small joints in the wrists, hands, and feet are typically involved early in the course of illness. Shoulders, elbows, knees, and ankles are often affected. Involvement of the jaw, spine, or hips is less common.

The symptoms of RA reflect an intense autoimmune inflammation. Symptoms commonly fluctuate over time with periods of little activity and “flare” periods with severe overt inflammation. Inflammation of the joints will lead to progressive disabling joint destruction in most untreated individuals. While symptoms may wax and wane, joint destruction can progress even in the absence of severe symptoms. Current guidelines recommend aggressive control of inflammation as soon as possible after the disease is recognized.

Generally effective disease-modifying antirheumatic drugs (DMARDs) include non-steroidal anti-inflammatory drugs, methotrexate, corticosteroids, leflunamide, sulfasalazine, and tumor necrosis factors (TNFs). Advanced immune suppression may also be achieved with janus kinase inhibitors, abatacept, or anti-interleukin 6 receptor antagonists. Each of these treatments can have severe adverse effects. Finding the best combination of treatments for each individual is often challenging. Care should be guided by an experienced practitioner with expert knowledge of RA and its treatment.

Therapy is typically guided by composite metrics incorporating objective findings (such as the number of inflamed or tender joints) and subjective measures (such as pain, fatigue, or stiffness). Reductions of 20%, 50%, or 70% in core metrics specified by the American College of Rheumatology (ACR) form the basis for the ACR20, ACR50, and ACR70 scores, respectively. The European League Against Rheumatism (EULAR) has a different metric called the Disease Activity Score Using 28 Joints (DAS28). ACR20, ACR50, ACR70, and DAS28 are commonly used as outcomes measures in clinical research. The Simplified Disease Activity Index (SDAI) and the Clinical Disease Activity Index (CDAI) tools can be used to guide therapy response in clinical practice.

The ACR 2021 RA treatment guideline strongly recommends initiating Tumor Necrosis Factor inhibitor (TNFi) therapy when disease activity is moderate or high despite disease-modifying anti-rheumatic drugs (DMARDs), with the recommendation to alter therapy if needed due to comorbidities (Fraenkel, 2021). The TNFi therapies include adalimumab, certolizumab pegol, etanercept, golimumab, or infliximab.

The molecular signature test analyzes 23 biological features including single-nucleotide polymorphisms (SNPs) derived from RNA sequencing, gene expression levels, serum peptides, and other disease-associated clinical features to identify a molecular signature of an individual that may or may not respond to TNFi therapy. The results of the test are sorted into categories (very high, high, moderate, no signal) that are purported to predict the likelihood of nonresponse to TNFi therapy.

The following codes for treatments and procedures applicable to this document are included below for informational purposes. Inclusion or exclusion of a procedure, diagnosis or device code(s) does not constitute or imply member coverage or provider reimbursement policy. Please refer to the member's contract benefits in effect at the time of service to determine coverage or non-coverage of these services as it applies to an individual member.

When services are Investigational and Not Medically Necessary:
When the code describes a procedure indicated in the Position Statement section as investigational and not medically necessary.

Peer Reviewed Publications:

1. Alonso-Ruiz A, Pijoan JI, Ansuategui E, et al. Tumor necrosis factor alpha drugs in rheumatoid arthritis: systematic review and meta analysis of efficacy and safety. BMC Musculoskelet Disord. 2008; 9:52.
2. Cohen S, Wells AF, Curtis JR, et al. A molecular signature response classifier to predict inadequate response to tumor necrosis factor-α inhibitors: the NETWORK-004 prospective observational study. Rheumatol Ther. 2021; 8(3):1159-1176.
3. Curtis JR, Strand V, Golombek S, et al. Patient outcomes improve when a molecular signature test guides treatment decision-making in rheumatoid arthritis. Expert Rev Mol Diagn. 2022 Nov 3:1-10. Online ahead of print.
4. Curtis JR, Strand V, Golombek SJ, et al. Decision impact analysis to measure the influence of molecular signature response classifier testing on treatment selection in rheumatoid arthritis. Rheumatol Ther. 2024; 11(1):61-77.
5. Jones A, Rapisardo S, Zhang L, et al. Analytical and clinical validation of an RNA sequencing-based assay for quantitative, accurate evaluation of a molecular signature response classifier in rheumatoid arthritis. Expert Rev Mol Diagn. 2021; 21(11):1235-1243.
6. Lewis MJ, Çubuk C, Surace AEA, et al. Deep molecular profiling of synovial biopsies in the STRAP trial identifies signatures predictive of treatment response to biologic therapies in rheumatoid arthritis. Nat Commun. 2025; 16(1):5374.
7. Mellors T, Withers JB, Ameli A, et al. Clinical validation of a blood-based predictive test for stratification of response to tumor necrosis factor inhibitor therapies in rheumatoid arthritis patients. Network and Systems Medicine. 2020; 3(1):1-14.
8. Pappas DA, Kremer JM, Reed G, et al. Design characteristics of the CORRONA CERTAIN study: a comparative effectiveness study of biologic agents for rheumatoid arthritis patients. BMC Musculoskelet Disord. 2014; 15:113.
9. Strand V, Cohen SB, Curtis JR, et al. Clinical utility of therapy selection informed by predicted nonresponse to tumor necrosis factor-ɑ inhibitors: an analysis from the Study to Accelerate Information of Molecular Signatures (AIMS) in rheumatoid arthritis. Expert Rev Mol Diagn. 2022a; 22(1):101-109.
10. Strand V, Zhang L, Arnaud A, et al. Improvement in clinical disease activity index when treatment selection is informed by the tumor necrosis factor-ɑ inhibitor molecular signature response classifier: analysis from the study to accelerate information of molecular signatures in rheumatoid arthritis. Expert Opin Biol Ther. 2022b; 22(6):801-807.

Government Agency, Medical Society, and Other Authoritative Publications:

1. Centers for Medicare and Medicaid Services (CMS). Local Coverage Determination (LCD): MolDX: Molecular Biomarker Testing to Guide Targeted Therapy Selection in Rheumatoid Arthritis.
   • LCD L39424. Available at: https://www.cms.gov/medicare-coverage-database/view/lcd.aspx?lcdid=39424&ver=6&keyword=Molecular+Biomarker+Testing+to+Guide+Targeted+Therapy+Selection+in+Rheumato&keywordType=starts&areaId=all&docType=F&contractOption=all&sortBy=relevance&bc=1. Accessed on July 16, 2026.
   • LCD L39427. Available at:https://www.cms.gov/medicare-coverage-database/view/lcd.aspx?lcdid=39427&ver=7&keyword=Molecular+Biomarker+Testing+to+Guide+Targeted+Therapy+Selection+in+Rheumato&keywordType=starts&areaId=all&docType=F&contractOption=all&sortBy=relevance&bc=1. Accessed on July 16, 2026.
   • LCD L39467. Available at: https://www.cms.gov/medicare-coverage-database/view/lcd.aspx?lcdid=39467&ver=9&keyword=L39467&keywordType=starts&areaId=all&docType=NCA,CAL,NCD,MEDCAC,TA,MCD,6,3,5,1,F,P&contractOption=all&sortBy=relevance&bc=1. Accessed on July 16, 2026.
   • LCD L39481. Available at: https://www.cms.gov/medicare-coverage-database/view/lcd.aspx?lcdid=39481&ver=8&keyword=Molecular+Biomarker+Testing+to+Guide+Targeted+Therapy+Selection+in+Rheumato&keywordType=starts&areaId=all&docType=F&contractOption=all&sortBy=relevance&bc=1. Accessed on June 9, 2026.
2. Fraenkel L, Bathon JM, England BR, et al. 2021 American College of Rheumatology guideline for the treatment of rheumatoid arthritis. Arthritis Rheumatol. 2021; 73(7):1108-1123.

Molecular signature response classifier (MSRC)
PrismRA

The use of specific product names is illustrative only. It is not intended to be a recommendation of one product over another, and is not intended to represent a complete listing of all products available.

Federal and State law, as well as contract language, including definitions and specific contract provisions/exclusions, take precedence over Medical Policy and must be considered first in determining eligibility for coverage. The member’s contract benefits in effect on the date that services are rendered must be used. Medical Policy, which addresses medical efficacy, should be considered before utilizing medical opinion in adjudication. Medical technology is constantly evolving, and we reserve the right to review and update Medical Policy periodically.

No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, or otherwise, without permission from the health plan.

© CPT Only – American Medical Association