A Posted MolDx LCD Request

 This is a publicly available example, on the CMS website, of a MolDx LCD request.  

The request is for coverage of a Castle melanoma genomic test, to be used when H&E slides of a skin lesion are not diagnostic of either a clearly benign nor malignant lesion.    MolDx had previously covered a Myriad test for this scenario.   Castle asks that the Myriad-specific LCD be broadened to be a "foundational" or I would say "umbrella" LCD, with multiple products.   Note, however, that if there hadn't been a single molecular predicate test, then the otherwise available standard(s) of care would have been the predicate that the new test exceeds.

I've listed some key features below.

#####

#####

MEDICARE RULES FOR LCD REQUESTS

Before a few years ago, Medicare did not have any rules for LCD requests.  (You just found a medical director's name and sent your request).  Here are the current more formal rules.

1.  MACs must log "LCD Requests" as a specific type of their work functions under their contract.

2.  Further instructions in an online manual, Chapter 13, 13.2.2.

https://www.cms.gov/regulations-and-guidance/guidance/manuals/downloads/pim83c13.pdf

3.  Request must be in writing and must specify

(a) A benefit category (Diagnostics are category SSA 1861(s)(3) )

(b) Language requested in LCD  ("This LCD covers service X under condition Y" etc)

(c) Justification with evidence, and PDF copies of all evidence.

(d) "Information that addresses relevance, usefulness, clinical health outcomes, or medical benefits"

(e) Fully explain the design, purpose, method, of the service

4.  Requester must be a patient, health professional, or party doing business in the MAC.

CASTLE REQUEST LETTER

MACs are now supposed to post request letters.   However, I've seen at least one that was truncated to only a table of contents.  This is a full request letter, which I think is CMS's intent.   National CMS decisions have always posted request letters in full.

The letter is 9 pages (before bibliography and appendix.)  For example, the letter is not 30 or 60 or 90 pages long.

More about the provenance of this request here:

http://www.discoveriesinhealthpolicy.com/2022/10/moldx-posts-melanoma-dlcd-at-cgs-mac.html

Letter online here:

https://www.cms.gov/medicare-coverage-database/attachments/lcd/39344_6/DecisionDxDiffDxMelanomaZ00BSCoverageRequestFINAL15JAN21.pdf

or  https://drive.google.com/file/d/1iQ1-GnJJTCLImSj0MP8Y8VymLC4WPbW9/view

TIMELINE

The Castle letter describes meetings in October and November 2020.  The PDF seems to be titled as date January 2021.   The LCD revision appeared in draft form in May 2022.   The final version is expected most likely around May 2023.

CMS requires that MACs acknowledge request letters, and log them, within 90 days.  However, CMS does not require requests to be handled in any particular priority, order, or in fact, to ever be acted upon after being logged.

THE LETTER

Page 1 is an overview and summary.   Past meetings are noted.   

Page 2 starts over with general background and a review of the diagnostic pathway (for the "difficult ambiguous biopsy")

Page 2-3 describes alternate methods.  Flow charts for the multi-pathway workflow of a case, are provided.

Page 4 introduces the specific test and how it runs and how it was developed.

Page 4-5 describe the clinical performance and compares, in detail, to laboratory alternatives including FISH.

Page 6 introduces "clinical utility."  They define this as performance in real-world ambiguous cases.   This is very important, because most tests are validated on clear-cut cases.  For example, one has 200 known benign lesions and 200 known malignant lesions, and develop a test that can sort them accurately.   However, the clinical case is to use AMBIGUOUS lesions.   So this is a separate, later study, to study the performance in AMBIGUOUS lesions which are the whole point of the test.  Castle styles this as "Clinical utility."  

Page 7 shows that for an existing 23 gene test, the test should reduce unnecessary re excisions (intrusive wide excisions) by 63%.  The new Castle 35 gene test should reduce these by 76%.

Page 7:  Describes why this reduction in extra surgery is a "health outcome improvement."  
Page 8 summarizes the key points in a two-column format (issue on the left, answer on the right).

Page 9 concludes with a final summary and leads into the bibliography.  

Page 10-12 is bibliography.  39 references.

Page 13 is a logical summary of reasons and rationale why the test is medically necessary again in two column format.

Page 14 is a gene list of the test.

I've attached my own redline copy with highlights I felt were pivotal.  A clean copy would be at the links above.

WSJ x 2: Fungal Infections Surge in US and Globally

 https://www.wsj.com/articles/deadly-fungi-are-infecting-more-americans-11666568576

https://www.wsj.com/articles/fatal-fungi-threaten-global-health-who-says-11666708805

An Example of Early ADLT Price Changes: 0340U

 These are the original MOLDX prices for SIGNATERA as set by MOLDX:

• For an exome and four plasma tests package, MolDx paid $5897.   For an exome and one plasma test, Moldx paid $3878.   For four plasma tests (no exome) MolDx paid $3177.

If you compare the $5897 price and the $3177 price, the latter dropping off the exome and lowering the price, MolDx "seemed" to value an exome at around $2800 (just using 2nd grade arithmetic).  

MolDx seemed to value each single PCR bespoke test at somewhere about $800 (using $3177/4).   MolDx late dropped specified test multiples from its DEX descriptors for policy and audit reasons.

For more on the above analysis, see earlier blog:

http://www.discoveriesinhealthpolicy.com/2021/06/medicares-somewhat-confusing-path-to.html

###

0340U

This is a new code for SIGNATERA which appears for the first time on the October 2022 fee schedule (being active for use on 10/1/2022).   

Signatera got ADLT pricing from July 2021 to March 2022 even though it did not have a PLA code yet (price per CMS ADLT table).   Code 0340U is now published and explicitly excludes prior exome testing from the code name (code work), and, pays $3500 per MRD test.   

For the CMS first-year ADLT rate, see here:

https://www.cms.gov/files/document/advanced-diagnostic-laboratory-tests-under-medicare-clfs.pdf

Presumably, under ADLT regulations, Natera reported first year pricing around March 2022 to CMS, which CMS would use to set its next price.  In fact, in the October 2022 CLFS, the 0340U price rises from $3500 to $3920.

See October 2022 fee schedule here:

(Note the code is "pan cancer" but Medicare coverage, as of 10/2022, is only for CRC MRD, immuno-therapy checkpoint drug response, and, bladder cancer MRD.  Also, the code specifies that prior tumor/germline testing was conducted.)

Note that Natera reported this $3920 price data to CMS in April 2022, and was aware that it would set the new price by May 2022.  (That is, the $3920 was noted in public Natera statements by May 2022 - here.  It's quirky there price went up long before they got a specific code to reflect this fact on a fee schedule on the CMS website.  The original plan for ADLT codes was that either AMA PLA would issue a code or failing that CMS itself would issue one quarterly, which would have been back in mid 2021.)

When will the price for 0340U change again?  I believe the $3920 will be  stable for a couple years.  

As I understand the ADLT pricing rules, Natera will gather its commercial pricing in 1H2023, report in 1Q2024, and that will set its new price for CY2025.    (And this ADLT pricing repeats annually, but also slowly as just shown).   

SInce most payments will be from Medicare Advantage plans in the first couple years, this will probably tend to replicate the $3920 price at least for a while.   At least til 2025, maybe til 2026, 2027.   

A Medicare Advantage regulation seems to state that Medicare Advantage plans shall pay out of network providers (when necessary to use them) at the prevailing Part B rates.  

See:  https://brucedocumentblog.blogspot.com/2022/04/medicare-advantage-and-price-based.html 

My Comment: Using Gapfill Codes as Crosswalk Targets

Sent: Monday, October 24, 2022 at 11:20:04 AM PDT

Subject: CLFS: Request comment on "Gapfill codes as Crosswalk codes"

I have attended many past CLFS pricing meetings, and nearly every year, CMS has remarked that a code currently in gapfill may not be used as a crosswalk target.

This topic arose again during the July 2022 expert advisory panel meetings, and CMS announced provisionally that gapfill codes could, now, be used as crosswalk codes.   However, verbally, CMS also requested public comment.

The index case seemed like a rational example of using a gapfill code as a new crosswalk code.   The Kiatek company had its code 0248U in current gapfill, and had very similar tests 0324U and 0325U up for pricing.  Here, the crosswalk-to-gapfill seemed like a "hand in glove" fit.   The expert panel also recommended that at least two other gapfill codes (0264U and 0267U) be used as crosswalk targets for other tests.

However, there are several reasons why a gapfill code should NOT be used as a crosswalk target.

1.  NOT PRICED ON FEE SCHEDULE

In the past, CMS stated that a gapfill code not priced on the fee schedule should not be a target for crosswalk.   This fits the most reasonable reading of statute and regulation, that prices be assigned to a new code by assigning the price of a code "on the fee schedule."  

An unpriced code is not really "on the fee schedule" - it may appear in a table of all code numbers, but it doesn't have a "fee" on a "fee schedule."  To argue it is "on the fee schedule" of fees, because it appears there with no price, is a stretch.

2.  LACK OF INFORMATION THAT IS PUBLICLY REQUIRED BY CMS IN FED REG

In the Federal Register, CMS lists the information requested when making proposals for pricing, such as test resources, method, purpose, etc., and the price of crosswalk targets.   

For gapfill codes, the price of the crosswalk target cannot be submitted by the requester.   In addition, while some new codes are well-known tests (with publications, etc), other new PLA codes may have essentially no publications or information about them.   Therefore, they can't be discussed and supported as rational crosswalk targets.

3.  SOME PLA CODES IN GAPFILL ARE THERE DUE TO "NO INFORMATION"

Let's say a new PLA code appears, 0999U.  The company makes no public presentation.  There are not any publications and no helpful website.   Therefore, the panel and CMS assign the code to gapfill.   

The next spring, this repeats.  The company either no longer exists, or fails to submit information to the MACs, or submits "goofy" unhelpful information to the MACs.   The pricing of this code in gapfill could be very erratic.   

Therefore, it is unfair to assign a useful new code, which may be the lifeblood of a new company that is working hard on clinical utility and pricing, to crosswalk it to this mystery and perhaps very capriciously priced gapfill code 0999U.

Bruce Quinn MD
Pathologist

Policy Consultant

My 3 Blogs on Medicare Transgender Surgery

 In October 2022, the Annals of Internal Medicine had a long historical article (Magrath) and op ed (Keuroghlian) on transgender surgery, focused on Johns Hopkins clinical services.

I had written a long blog on Medicare's history when there was a judicial ruling "against non coverage" in 2014 and when there was a follow-up NCD in 2016.

October 2022 Update (re Magrath)

http://www.discoveriesinhealthpolicy.com/2022/10/side-note-new-2022-review-history-of.html

###

May 2014 (Judicial actions)

http://www.discoveriesinhealthpolicy.com/2014/05/judges-reject-decades-old-medicare-ban.html

June 2016 (NCD)

http://www.discoveriesinhealthpolicy.com/2016/06/cms-posts-draft-decision-on-gender.html

CMS National Decision Home Page

https://www.cms.gov/medicare-coverage-database/view/ncacal-tracking-sheet.aspx?ncaid=282&bc=AAAAAAAAACAA&=

Guidelines That Underplay Genomic Testing

Recent guidance documents for managing major depression, Barrett's esophagus, and CRC screening, give either no mention or neutral mention to the addition of molecular diagnostics.

____

In the October 2022 Annals of Internal Medicine, discussion of a new DOD/VA guidance document for major depression.   Evidence for pharmacogenetic testing in depression is rating as too equivocal to evaluate to a conclusion.

The VA/DOD guideline in full, is here:

https://www.healthquality.va.gov/guidelines/MH/mdd/index.asp

The summary by McQuaid et al. is here:

https://www.acpjournals.org/doi/10.7326/M22-1603

###

A 2022 review of Barrett's esophagus, diagnosis and management, in JAMA does not mention molecular diagnostics.  Sharma 2022:

https://jamanetwork.com/journals/jama/fullarticle/2795263

Similarly, a 2022 guidance document in Amer J Gastroenterol by Shaheen et al. does not recommend molecular testing:

https://journals.lww.com/ajg/fulltext/2022/04000/diagnosis_and_management_of_barrett_s_esophagus_.17.aspx

###

A new 2022 review article on "How would you screen for CRC?" in Annals of Internal Medicine notes that one of the available methods is stool-based DNA-FIT, but does not provide any directional recommendation for using this much-more-expensive technology over other basic FIT technology.  Burns et al.:

https://www.acpjournals.org/doi/10.7326/M22-1961

Commercial Coverage DecisionDx

 https://www.anthem.com/dam/medpolicies/abc/active/policies/mp_pw_c148391.html

https://www.bcbsla.com/-/media/Medical%20Policies/2020/08/03/17/34/Gene%20Expression%20Profiling%20for%20Cutaneous%20Melanoma%20CC%2000622%2020180815_accessible%20pdf.pdf

https://www.bcbsri.com/providers/sites/providers/files/policies/2021/01/2020%20Gene%20Expression%20Profiling%20for%20Cutaneous%20Melanoma_New%20Code%20Update.pdf

http://www.aetna.com/cpb/medical/data/300_399/0352.html

SCOTUS ALINA 2019: Can't do retroactive changes, APA vs CMS-specific clause for APA

 https://www.supremecourt.gov/opinions/18pdf/17-1484_4f57.pdf

This case turns on whether the government’s 2014 announcement established or changed a “substantive legal standard.” The government suggests the statute means to distinguish a substantive from an interpretive legal standard and thus tracks the Administrative Procedure Act (APA), under which “substantive rules” have the “force and effect of law,” while “interpretive rules” merely “advise the public of the agency’s construction of the statutes and rules which it administers,” Perez v. Mortgage Bankers Assn., 575 U. S. 92, ___. Because the policy of counting Part C patients in the Medicare fractions would be treated as interpretive rather than substantive under the APA, the government submits, it had no statutory obligation to provide notice and comment before adopting the policy. The government’s interpretation is incorrect because the Medicare Act and the APA do not use the word “substantive” in the same way. First, the Medicare Act contemplates that “statements of policy” can establish or change a “substantive legal standard,” §1395hh(a)(2), while APA statements of policy are not substantive by definition but are grouped with and treated as interpretive rules, 5 U. S. C. §553(b)(A). Second, §1395hh(e)(1)—which gives the government limited authority to make retroactive “substantive change[s]” in, among other things, “interpretative rules” and “statements of policy”—would make no sense if the Medicare Act used the term “substantive” as the APA does, because interpretive rules and statements of policy—and any changes to them—are not substantive under the APA by definition. Third, had Congress wanted to follow the APA in the Medicare Act and exempt interpretive rules and policy statements from notice and comment, it could have simply cross-referenced the exemption in §553(b)(A) of the APA. And the fact that Congress did cross-reference the APA’s neighboring good cause exemption found in §553(b)(B), see §1395hh(b)(2)(C), strongly suggests that it “act[ed] intentionally and purposefully in the disparate” decisions, Russello v. United States, 464 U. S. 16, 23. Pp. 5–12. (b) The Medicare Act’s text and structure foreclose the government’s position in this case, and the legislative history presented by the government is ambiguous at best. The government also advances a policy argument: Requiring notice and comment for Medicare interpretive rules would be excessively burdensome. But courts are not free to rewrite clear statutes under the banner of their own policy concerns, and the government’s argument carries little force even on its own terms. Pp. 13–16

2021 - Labcorp Buys VECTRA 81490 (13th Highest Test in 2020; $22M)

81490 - Active policy at Palmetto (for North Carolina, A53110)
None, and no history shown, at Noridian

Vectra

 https://www.labcorp.com/newsroom/labcorp-acquire-myriad-autoimmunes-vectra-testing-business-myriad-genetics

Labcorp to Acquire Myriad Autoimmune's Vectra Testing Business from Myriad Genetics

BURLINGTON, N.C. – May 3, 2021– Labcorp (NYSE: LH), a leading global life sciences company, today announced that it has entered into a definitive agreement to acquire select operating assets and intellectual property (IP) from Myriad Genetics’ autoimmune business unit, including the Vectra® rheumatoid arthritis (RA) assay. More than one million Vectra tests have been completed since the product’s launch in November 2010, and a meaningful portion of testing volume currently flows through Labcorp.

“Labcorp has consistently been a major player in rheumatology and continues to focus on providing medical professionals with the data they need to best treat their RA patients,” said Brian Caveney, M.D., chief medical officer and president of Labcorp Diagnostics. “The addition of the Vectra testing capabilities to our in-house products offers tremendous potential for us to expand the test’s availability and make Labcorp a single-source diagnostics solution for RA providers. We look forward to welcoming the team to Labcorp.”

Vectra is a non-invasive, blood-based test that analyzes 12 biomarkers to measure RA disease activity. It combines those measures to generate an easy-to-understand score, which indicates the severity of RA inflammation and how well current treatments are working. It also can predict potential, future joint damage. This enables the 5,000 practicing rheumatologists in the U.S. to provide targeted treatment and adjust existing treatments to better manage RA symptoms. 

Rheumatoid arthritis, an autoimmune disorder and one of the most common forms of arthritis, causes painful inflammation and tissue damage in the knees, wrists, and other joints. The U.S. Centers for Disease Control and Prevention project that, by 2040, roughly 26% of the country’s adult population will be living with doctor-diagnosed arthritis such as RA.

The acquisition of the Vectra test and related IP and other assets complements Labcorp’s prior business activity aimed at bolstering its scientific leadership in RA testing and treatment.

The transaction is expected to close by the end of the third quarter, subject to customary closing conditions and regulatory approvals, including under the Hart-Scott-Rodino Antitrust Improvements Act of 1976, as amended. Specific terms of the transaction were not disclosed. Hogan Lovells acted as legal advisor to Labcorp.