NEJM Op Ed Tackles CMS Efforts to Modernize Coverage Policy

 

A policy opinion piece in the November 30, 2023 NEJM, by Kadakia et al., provides a fairly detailed overview of coverage problems at CMS, and the historically unsuccessful attempts to navigate them with either Parallel Review or with CED (Coverage with Evidence Development).  

The authors describe the Trump administration MCIT proposal, its collision (and sinking) with Biden policymakers, and the still-in-progress CMS proposal to create "TCET," temporary coverage for emerging technologies.  A table compares MCIT and TCET. (Below, click to enlarge).  

click to enlarge

While having a somewhat guarded view of TCET - it relies on CED and strongly mirrors the Parallel Review flop in terms of the proposed process - the authors seem to conclude that at least CMS is moving forward with something (TCET).   

While the citation list in their short paper is brief, I'd add they might have cited the Dartmouth/Zeitler paper last year on CED (here).

NEJM includes a 9 minute podcast interview with coauthor Robert Yeh.  Yeh is director of the Center for Outcomes at Beth Israel.

___

AI Corner

See AI summary of Kadakia, here.

New Alzheimer Codes Dominate Lab Agenda for AMA CPT Feb 2024

Registration is open for the February 1-3, 2024, AMA CPT Editorial Meeting in San Diego.  Registration may be both in person or virtual.  Here.

Lab codes were posted in mid-November for an early comment cycle (request Nov 20[*], comment Dec 4).  All other codes will be posted for comment December 1.

Here's the lab agenda link:

https://www.ama-assn.org/system/files/feb-2024-path-lab-mo-path-agenda.pdf

There's a bonanza of codes for Alzheimer's disease.  Note that the codes have "placeholder" codes at this time, which I'm also using here.

• Amyloid Protein in plasma or CSF (82XX0, 82XX1)
• 40 and 42
• Possibly, one code for plasma and one for CSF
• Usually measured as a pair for calculating the ratio, not the absolute amount
• Neurofilament light chain, blood (83XX0)
• Immunoassay Alzheimer Pathology (835XX)
• This is puzzling, possibly tau isoforms??
• Tau protein, Total (8X3XX)
• Tau protein, Phosphorylated (8X3X0)

Some of these tests are represented by PLA codes; the 80,000 series  codes will apply to any user.  The 80,000 series codes do not differentiated FDA and LDT versions.  Different brands or different isoforms may have different clinical accuracy.

There are several other lab codes: 815XX, brain tumor methylation; 8156X, kidney transplant gene expression; 81XX0, fetal recessive conditions (mentions XX28U, a PLA code).

###

[*] As the agenda PDF indicates, someone wanting to review and comment on the code packetts for lab tests is directed to an AMA software portal [select the tab for Interested Party, not the tab for Applicant].  However, AMA locks out requests after 11/20/23, the lab request deadline.  This early request deadline ensures that any lab comments are available to internal subcommittees like MPAG and PCC.  

FDA Posts Agenda and Questions for MCED Panel November 29, 2023

On November 29, 2023, FDA is holding   ^held  an all-day panel on how the FDA should plan for trials and evaluations of MCED, multi cancer early detection.   

See original blog about the FDA workshop, by me, here.

• Update November 30 - I've put several different AI summaries of the 6-hour meeting in a Google doc, here.  
• This includes a 10-page bullet point summary by Otter.AI, my transcription program.
• Also short summaries by Claude.AI and Chat GPT4.  (And then, having got the 2 summaries, I asked them to compare the two summaries, too.) 
• See an excellent pre meeting article by Hyman Phelps here.

# # #

FDA has posted the agenda 9-5, the key questions, and an executive summary of 10 pages.

Find them all at the FDA, here, as "Event Materials," as well as the FDA's link to the streaming video for November 29.

AI CORNER

I've posted a Claude.ai summary of the ten-page executive summary written by FDA.

##

The FDA is convening an advisory panel to get input on regulatory considerations for multi-cancer early detection (MCED) tests. These blood-based tests aim to screen for multiple cancers simultaneously in asymptomatic individuals. Currently no MCED tests have FDA authorization.

The panel will provide recommendations on study design to demonstrate clinical validity of MCED tests. This includes determining appropriate methods to confirm cancer status, diagnostic workup following positive results, and follow-up for negatives. The panel will also discuss use of tissue-of-origin assays to help localize tumors after initial MCED positive, as well as benefit-risk considerations.

Key questions for the panel include [for original FDA text see FDA link above]:

• Should MCED tests be evaluated by cancer type or in aggregate? What minimum sensitivity per cancer?
• How to evaluate MCED versus existing single cancer screening tests? Risk of patients foregoing current methods?
• Definition of "early detection"? Supportive data needed?
• Acceptable false positive rates? Should specificity be fixed to support low false positives?
• Harms of false negatives/positives? Diagnostic odyssey? Overdiagnosis/overtreatment?
• Use of stage shift as surrogate endpoint to show benefit?
• Role of real-world evidence for expanding cancer claims, validating rare cancers, demonstrating stage shift?

The panel's discussion and recommendations will inform the FDA's thinking on evaluation of safety and effectiveness for future MCED test marketing submissions.

NEJM AI: With Complex Cases, GPT-4 Usually Beats Humans (Eriksen et al.)

Last April, NEJM announced it would launch ai.NEJM, a new AI-focused journal (see coverage at MedPageToday).  At the time, it publishes Lee et al., a Microsoft-based article on "benefits, limits, risks of GPT-4" in medicine (here).

In today's news, NEJM publishes Erikesen et al, on the success rate of GPT-4 to "diagnose complex clinical cases."  Based on online data, the authors found GPT-4 (which was right 57% of the time) outperformed 99% of human readers.

See the article here:

https://onepub-media.nejmgroup-production.org/ai/media/000de933-9406-4f17-87b5-8e28c5cf5da7.pdf?

See open-access coverage at MobiHealthNews by Jessica Hagen here:

https://www.mobihealthnews.com/news/gpt-4-outperformed-9998-simulated-human-readers-diagnosing-complex-clinical-cases

Glitch: The article refers to supplemental information, including methods, but at least on the day of release, I couldn't find any at NEJM/

##

Here's a summary from Claude.ai:

Eriksen et al. Use of GPT-4 to Diagnose Complex Clinical Cases. AI.NEJM.ORG.

In this perspective article published on ai.nejm.org, Eriksen and colleagues assessed the performance of the AI system GPT-4 in diagnosing complex real-world medical cases. They compared GPT-4's diagnostic accuracy to that of medical journal readers answering the same questions online.

The authors utilized 38 open-access clinical case challenges with comprehensive patient information and a multiple choice question on the diagnosis. They provided the cases to GPT-4 along with instructions to solve the case and select the most likely diagnosis. They ran GPT-4 on the cases multiple times to assess reproducibility.

GPT-4 correctly diagnosed 21.8 out of the 38 cases on average (57%), with very high reproducibility across runs. In comparison, simulated medical journal readers answering the same cases [based on 248,000 online answers) only diagnosed 13.7 cases correctly on average (36%). 

Based on distributions of reader answers, GPT-4 performed better than 99.98% of a simulated population of readers.

The authors conclude that GPT-4 performed surprisingly well on these complex real-world cases, even outperforming most medical journal readers. However, GPT-4 still missed almost half of diagnoses. More research is needed before considering clinical implementation of such AI systems. Specialized medical AI models, additional data sources beyond text, transparency of commercial models like GPT-4, and further evaluation of safety and validity are still required first. But these early findings indicate AI could become a powerful supportive tool for clinical diagnosis in the future.

MolDxOlogy - Payments for 2022 Category 1 Genomic Codes, by MAC

Header:  Of Category I genomic CPT codes, about 65% MolDx payments, 20% NGS MAC, 15% Novitas.  But, most NGS MAC payments represented Cologuard, an NCD-driven code.  Otherwise, NGS MAC had 4% of payments.

____

In October, I posted some data on national CY2022 spending by CPT code (here).  

This week, in prior blogs, I broke out spending by MAC, first by 81479 miscellaneous code (here) and then by PLA code (here).

In this third and final blog for today, I show Category I CPT code spending by MAC (or MolDx).  I pulled cloud data for all 2022 payments on codes starting with (811, 812, 813, 814, 815).  

I then used an Excel pivot table to sum spending by "carrier" (usually "state") and rolled the results up by MAC and MAC system. The MAC policy systems are MolDx, Novitas/FCSO, and NGS MAC.

Results

Of $1.5B in genomic spending in this code set, 63% was via MolDx, 22% via NGS MAC, and 15% from Novitas/FCSO MACs.

However, I would rush to point out, nearly all the NGS MAC spending in this code set was driven by a single national decision (not local MAC LCDs).  It's the NCD for Exact Sciences Cologuard (state Wisconsin, payments $268M under NCD, code 81528).  Minus that, NGS MAC in all other states and codes altogether paid $64M for Category I codes (811xx etc) in 2022, or 4% of payments of $1.5B.

Click to enlarge.

Summary of 3 Posts

For the following data, I simplified each data set (e.g. dropping lines <$100K), so the numbers vary a tiny bit from exact nationwide numbers.

For Category I codes (811nn-815nn), payments were $1.4B, of which 63% MolDx, 22% NGS MAC, 15% Novitas MAC.   Excluding the NCD for Cologuard, which is close to $300M, NGS MAC payments fell to 4%.

For PLA codes, payments were about $330M of which 81% MolDx, and about 10% each to Novitas/FCSO and NGS MACs.  However, most of the 10% to NGS MACs can be accounted for by FMI payments driven by NCD 90.2 for FDA PMA NGS tests.

For 81479 miscellanous codes, payments were about $470M, of which 98% was paid by MolDx MACs.  90% of that, was via Noridian MAC.  81479 was the national highest-paid genomic CPT code.

Bonus

This fall, I used the Medicare coverage database to tally MAC Articles about genomic codes (811nn to 815nn).   There were 82 MolDx articles, 8 Novitas articles, and 4 NGS MAC articles.

 

MolDxOlogy - Payments for 2022 PLA Codes, by MAC

Header: 80% of PLA codes were paid via MolDx.  About 10% each for Novitas and NGS MAC, but the NGS MAC PLA payments were mostly NCD-driven payments for Foundation Medicine.

___

In the previous blog, I looked at payments for Genomic Miscellaneous Code, 81479, by MAC - here.  I used the CMS cloud data archive for 2022 Part B payments - here.

In this blog, I pulled the CPT codes ending in "U" - the PLA codes - and sorted them in Excel for "Carrier" (usually a state), and sorted those into  the three  MACs for policies - the MolDx MACs, the NGS MACs, and the Novitas/FCSO MACs.

To save manual work, I deleted rows less than $100,000, leaving $330M of PLA payments (97% of the total).   Of this $330M, about $18M were paid out to diverse MACs (such as for a Cepheid IVD, 0241U), which I didn't try to parse by MAC.  

Of the $312M paid to specific MACs for PLA codes, I tallied $252M or 81% to MolDx MACs, and roughly $30M each to Novitas MACs and NGS MACs (about 9-10% each.)

Note, however, that much of the NGS MAC payments come from  PLA codes 0037U and 0239U, Foundation Medicine FDA CDx codes (FFPE and LBX, respectively) - these are controlled by an NCD, NCD 90.2, and are not discretionary payments under the judgment of the NGS MAC.   If you tallied NGS MAC payments NOT controlled by NCD 90.2, it would be very low.

Click to enlarge.

click to enlarge

I previously posted, in October, that national payments for PLA codes were highly concentrated, with the highest-paid codes mostly being FDA approved.  Just 3 of 500 PLA codes, had half the PLA payments.

MolDxOlogy: In 2022, MolDx Captured 98% of 81479 Payments

Header.  Genomic test spending under miscellaneous code 81479 was a third of 2022 Part B spending for Tier 1 lab codes at $470M.   98% was via MOLDX MACs.

___

Back in October 2023, CMS released data for CY2022 Part B payments by CPT code.   (Payments by doctor or lab will be released in summer 2024).  See October blog here.  Molecular spending was about $1.5B for Tier 1 codes, another $340M for PLA codes.   In either code set, the top few codes garnered most of the spending.

I noted that about a third of Category I spending was Unlisted Code 81479.  For a client project, I've gone back to that data and analyzed 81479 spending in CY2022 by MAC.

I analyzed all the states with 81479 over $100K (I had 99.96% of 81479 payments.)  98.3% was in MolDx states, or about $462M of $470M.  More was spent on 81479 than all PLA codes together.  

Also for 81479, $7.5M was in Novitas (PA), and just $500K in NGS MAC states (specifically, Illinois).

70% of 81479 spending was in California (61% northern, 9% southern) and 13% was in Arizona.  

About 90% of 81479 spending was via the Noridian MAC.  

The table below can be clicked to view.

click to view

__

81479 spending at MOLDX was paired with "Z codes," allowing granularity in claims processing. 

However, large amounts of molecular spending fall outside of PAMA and fall outside of public code data that can be used for academic and public health purposes.   For example, one cannot tally up how many Medicare patients had BRCA testing in 2022, if it is recoded as "81479."  (In contrast, you CAN tally up how many FFS patients had prostatectomies, colonoscopies, and so on).  

AMP Hosts FDA LDT Webinar: Monday November 27

 On Monday, November 27, 2 pm ET  11am PT, AMP will host a webinar on FDA LDT, the goal being "to stop this overreaching rule."  Hosts will be Dr Eric Konnick, chair of AMP professional relations, and Annie Scrimenti, Assoc. Director of Public Policy & Advocacy at AMP.

Here's AMP's 6 page PDF briefing on the rule:

https://amp.org/AMP/assets/File/advocacy/Summary%20FDA%20Proposed%20Rule.pdf?pass=75

Here's November 27 ZOOM registration:

https://us06web.zoom.us/webinar/register/WN_P-QgnfjISauUajGoGXP6Bg#/registration

AMP adds,

You can stop this overreaching rule from being finalized by submitting comments on how the proposed rulemaking will affect your ability to practice laboratory medicine and provide critical lifesaving tests to patients. Unique, individual and institutional comments telling the FDA why this rule is bad for patients, bad for medicine, and bad for innovation are critical.

AI Corner: Trying Out Anthropic "CLAUDE"

Since February, I've been a subscription user of Chat GPT.  With their recent management turmoil, I tried a similar competitor service, ANTHROPIC CLAUDE, in its free version.[*]   

I found CLAUDE did a good job of summarizing my recent detailed white paper on FDA proposal for LDTs (October white paper here.)  

In both ChatGPT and CLAUDE, you can now just click a "paperclip" button and upload a PDF.

###

Here's the CLAUDE link (once you're registered): https://claude.ai/chats

Here's home for Anthropic, where I email-registered for CLAUDE: https://www.anthropic.com/product

Free Claude gives you 5-10 questions per 8 hours, I think.  $20/mo Claude gives you much higher bandwidth.

Here's the test:

 Please summarize in detail this complicated white paper about a new policy proposed by FDA for regulating laboratory tests [upload it]. I am the author of the white paper, and I would like to see how you summarize it. I expect a detailed summary to be in the area of 500 words long. Thank you.

ANTHROPIC CLAUDE:

Here is a 475-word detailed summary of the white paper about the FDA's proposed policy for regulating laboratory tests:

The white paper analyzes a new proposal released on September 29th, 2023 by the FDA to regulate laboratory-developed tests (LDTs). The proposal would phase out the FDA's enforcement discretion and subject LDTs, including those solely used within a single lab, to FDA oversight as medical devices. This would require LDTs to comply with requirements like registration, quality systems, premarket review pathways, and labeling rules.

The proposal lays out a 5-year timeline for phasing in compliance. However, the white paper author, Dr. Bruce Quinn, argues that the dramatic costs of compliance, estimated by FDA to be around $50 billion over the first 5 years, make the timeline wholly unrealistic and the proposal “simply impossible to execute.” The FDA also projects over $300 billion in benefits over 20 years, but the white paper contends these are based on speculative assumptions about LDT accuracy improvements.

The paper critiques how the high costs were obscured in the publicly released documents and summary. It analyzes the underlying FDA assumptions and calculations, including estimating over 200,000 regulatory expert-years and thousands of FDA reviewers would be needed in Years 3 and 4 alone to handle the influx of LDT applications. It compares this to the total yearly profits of less than $4 billion for LabCorp and Quest combined.

The white paper further examines the FDA's claimed legal authority to regulate LDTs as devices, noting this remains a disputed view. It contrasts the FDA arguments with analyses suggesting labs conducting testing may fall outside device definitions and commerce clause restrictions on federal power over intrastate activities.

In conclusion, the paper argues the proposed rule is essentially unworkable as written. It expects court challenges if finalized unchanged, as well as continued debate over the appropriateness and legality of FDA reaching into lab operations. But it suggests there may be alternative approaches, short of the complex VALID Act legislation, that could address FDA’s concerns on LDT claims and LDT accuracy without imposing such dramatic bureaucratic and financial burdens.

[*] Both ChatGPT and Anthropic Claude have good free versions, while a $20 subscription to either one gets you more bandwidth and priority when servers are overworked.

FDA Holds Panel on Multi Cancer Early Detection: November 29, 2023

On Wednesday, November 29, 2023, FDA will hold a molecular advisory panel on the topic of Multi Cancer Early Detection.  Here's a clipping:

• On November 29, 2023, 9-5 ET, the Molecular and Clinical Genetics Panel will discuss and make recommendations on the design of multi-cancer detection (MCD) in vitro diagnostic devices (tests) as well as potential study designs and study outcomes of interest that could inform the assessment of the probable benefits and risks of MCD screening tests. The committee’s discussion and recommendations from this meeting will help inform future Agency regulatory efforts for these novel tests.
• FDA intends to make background material available to the public no later than 2 business days before the meeting.  NOTE: POSTED BY NOV 28; KEY QUESTIONS; 10 P BACKGROUND.
• The meeting presentations will be heard, viewed, captioned, and recorded through an online teleconferencing platform.

The web page is here:

https://www.fda.gov/advisory-committees/advisory-committee-calendar/november-29-2023-molecular-and-clinical-genetics-panel-medical-devices-advisory-committee-meeting

The page includes a Webcast Link.

See FED REG and comments submission portal:

https://www.federalregister.gov/documents/2023/11/13/2023-24896/molecular-and-clinical-genetics-panel-of-the-medical-devices-advisory-committee-notice-of-meeting

It's published as 88 Fed Reg 77588-90, November 13, 2023.

Check yourself for updates but the YouTube link appears to be:

https://youtu.be/xyOrADJQj4U

ARUP Hosts Webinar on FDA-LDT: Wednesday November 29

Heads up:

Live Webinar, November 29: What the FDA’s Proposed Rule to Regulate Lab-Developed Tests Means for Hospitals and Health Systems

An ARUP expert will answer questions about a proposed FDA rule to regulate laboratory-developed tests (LDTs) as medical devices, which will potentially limit patient access to essential medical care.

https://www.aruplab.com/webinar-fda-proposed-rule-further-regulate-lab-developed-tests-hospitals-health-systems

Webinar Description (10 PT, 11 MT, 12 CT, 1 ET)

Join ARUP’s Jonathan Genzen, MD, PhD, chief medical officer and senior director of governmental affairs, on Wednesday, November 29, at 11 a.m. MST for a conversation about the FDA’s proposed rule to regulate LDTs as medical devices. He’ll address patient safety, test availability, and the impact on innovation. ARUP’s compliance officer, 

Jonathan Carr, JD, will moderate a Q&A during the webinar.

CMS Posts Final Prices for New Lab Codes CY2024

CMS has posted final prices on new lab codes for CY2024.  .

What Happened?

By my informal tally, there are 84 agenda items.  13 changes.   

10 changed from a crosswalk recommendation to a Gapfill, and 3 changed from gapfill to crosswalk or from one crosswalk to a different one.  WIth the 10 new Gapfills and a couple gapfill "losses," I tally 30 codes for the 1H2024 gapfill process, out of 84.

6 New CGP Codes Get Gapfilled, Not $600

In likely the most-watched codes, CMS had proposed that 6 new AMA CPT comprehensive genomic profiling codes (CGP) be priced at about $600 (CW 81445).  All 6 of those $600 crosswalks met with strong complaints, and CMS has split the difference by turfing them to the 2024 gapfill process.   (This story got a 12/14/23 article in Genomeweb.)

3 Slide-AI Codes All $700

There were 3 codes for AI-enhanced slide-based imaging, 0376U (Artera AI Prostate), 0414U (former X084U, LungOI Imagene), 0418U (former X088U, PreciseDX Breast Bx).  Requested prices were as high as the $2500 range (Artera request).   

The expert panel, the CMS proposed  price, and the CMS final price were all crosswalk to 0220U (PreciseDx Breast) at about $700.    (I've noted earlier that if you views "slides + AI" as a category, prices have reached as high as $5435 for ADLT pricing of the Prelude DCISioniRT test 0295U).  Labs in this group can appeal, meaning the price would be revisited next summer.

Crosswalk = Crosswalk to 1 Code

Of about 56 codes crosswalked, all but 4 were crosswalked to a single target code.  One was crosswalked to a 0.5 fraction, and three were crosswalked to addition of 2 or 3 codes.

Read For Yourself

Find the new final spreadsheet at this link.  Scroll down for Agenda and Announcements, and find a Zip file link for CY2024 final prices.

https://www.cms.gov/medicare/payment/fee-schedules/clinical-laboratory-fee-schedule-clfs/annual-public-meetings

Stanford Biodesign Posts Webinar on FDA LDT Proposal

 Last week, Stanford Biodesign hosted a webinar on the FDA proposal to regulate LDTs.

They've now posted the one-hour video online at their YouTube channel.  Find it here:

https://www.youtube.com/watch?v=9zVlwbWp2Lo

In a conversation moderated by Stanford Biodesign Policy Program director Kavita Patel, ACLA president Susan Van Meter, physician executive Bruce Quinn, and Stanford Biodesign's director of policy research Sandra Waugh Ruggles discuss the Food and Drug Administration's proposed language on laboratory developed tests (LDTs), and the potential impact on clinicians, patients, and innovators.

My white paper on the FDA proposal is here:

https://www.discoveriesinhealthpolicy.com/2023/10/online-ahead-of-print-white-paper-on.html

Not mentioned in the webinar, a former FDA attorney's guide to the FDA's legal weaknesses:

https://www.discoveriesinhealthpolicy.com/2023/11/a-lawyers-view-potential-weakness-of.html

###

AI CORNER

I've included an AI summary below from ChatGPT [some manual edits].

This webinar from the Stanford Biodesign Center focused on the FDA's regulation of Laboratory Developed Tests (LDTs). The discussion centered on the FDA's proposed rule which appeared following Congress's failure to pass the VALID Act.  (VALID hadsought to establish a new framework for the regulation of LDTs.)

The panelists expressed disappointment with the proposed rule, seeing it as a blunt instrument that could stifle innovation and patient access to essential tests. They highlighted the need for a regulatory framework that acknowledges the unique characteristics of LDTs, as opposed to applying the legacy  1976 medical device regulatory framework.

The proposed rule's timeline was criticized for being too short to accommodate the industry and FDA's needs adequately. There were concerns about the ability to maintain access to high-risk LDTs due to the short timeline for submission requirements. The panelists also expressed legal concerns, doubting the FDA's authority under current law to regulate LDTs as medical devices.  The FDA's somewhat irregular approach to laying out a cost-benefit argument was critiqued.

Significant points of contention included the lack of a grandfathering policy for existing LDTs, the absence of provisions for rare disease testing, and the potential negative impacts on innovation, particularly for small-scale tests and those from academic medical centers. The discussion highlighted the potential chilling effect on new diagnostics development, increased capital requirements, and the need for more strategic planning to navigate the regulatory landscape.

The panelists suggested that the FDA should seek a more nuanced, goal-oriented approach, with a focus on ensuring test safety and controlling claims without imposing excessive costs. They stressed the importance of legislative involvement to create a diagnostic-specific regulatory framework and advised stakeholders to submit comments during the open period to influence the final rule. There was a consensus that the FDA's rule, as it stands, may not be implementable and that any effective regulation must balance patient safety, access, and the facilitation of innovation.

GRAIL Announces Investigational Device Study (IDE) of GALLERI Test

GRAIL announces its IDE (investigational device exemption) for the GALLERI test in Medicare-age patients.   See a detailed press release accompanied by a two-minute YouTube video. 

Find the press release here:

https://grail.com/press-releases/grail-to-initiate-reach-study-to-evaluate-clinical-impact-of-galleri-multi-cancer-early-detection-mced-test-among-the-medicare-population/

Find the video here:

https://www.youtube.com/watch?v=nksTA6B2w-o

Find a news report at Clinical Trials Arena here.  (By the way, CTA is an interesting news source).

UPDATE:  See also press from ACS here.

The press release is pretty long, at 1100 words.

Find a short summary of the GRAIL information here:

On November 20, 2023, GRAIL announced the initiation of the REACH/Galleri-Medicare study, which aims to evaluate the clinical impact of its multi-cancer early detection (MCED) test, Galleri®, among Medicare beneficiaries. 

This pivotal study, greenlit by both the FDA and CMS, will assess the potential of Galleri® to improve cancer outcomes and reduce healthcare costs by detecting multiple cancer types early. The study will compare the health outcomes of 50,000 Medicare beneficiaries using Galleri® alongside standard cancer screenings against a control group receiving usual care. With a focus on inclusivity, the study will emphasize recruitment from historically under-represented groups, ensuring broad access to this innovative technology. 

The study’s findings will contribute to a growing body of evidence supporting the use of MCED tests in diverse populations, aiming to advance early cancer detection and address health disparities.

In General: What's a CMS IDE Study vs Para Review Coverage vs TCET?

CMS manages an intake portal for applicants with FDA-approved IDE studies and a roster of current studies, here.  The FDA-CMS-IDE program should be kept separate from the Parallel Review program (here) and the proposed but not yet enacted  TCET program (here).  

• The CMS FDA IDE program is for products or services fully engaged in the FDA clinical trials for FDA labeling, which studies are called "IDE" studies.  Hundreds of such studies have had CMS coverage of the device under investigation, on the basis that the device is providing a service for each Medicare patient receiving it.   
• FDA FAQ here.
• The Parallel Review program engages CMS in the writing of an NCD during the last phase of  FDA approval, and has been used only a very few times in a decade. 
• Article on Para Review here.  
• Finally, the TCET program is only a proposal for a future program, a main part of which would be sheparding the new device into an NCD with CED quickly and efficiently.    
• Law firm's update on TCET here (Holland Knight).  Academic article on CED here. 

____

 The 1995 rule establishing the IDE coverage  was an agreement  between FDA and CMS and was very well written, still very lucid and readable today (1995, 60 FR 48417) (here).  Compared to the clarity of the logic, writing, and justification,in the 1995 FDA rule, the FDA's more recent "FDA LDT" rule proposal was just a train-wreck.  

Antibiotic Resistance in Crisis: New Report from Sepsis Alliance

The Sepsis Alliance has released a new 11 page white paper on the crisis of antibiotic resistance, featuring surveys of health management professionals conducted by sage Growth Partners.

Find the report's website here, with a summary and with links-through to the PDF report.

https://www.sepsis.org/news/sepsis-alliance-releases-groundbreaking-new-report-to-mark-world-antimicrobial-awareness-week/

(Separately, on December 6, Sepsis Alliance will have an registration-access webinar on the association's new initiatives and goals.  Here.)

Here's an AI-supported highlights of the 11-page report.

• The "2023 Sepsis Alliance AMR Market Report" emphasizes the growing threat of antimicrobial resistance (AMR), identifying it as a pressing societal issue that could potentially lead to another pandemic. 
• The report reveals that 90% of healthcare C-suite executives see AMR as a societal threat, 88% believe the problem is worsening, and 85% are concerned about another pandemic. 
• Key findings include 
• the lack of public awareness about AMR, 
• the inappropriateness of antibiotic use exacerbated by COVID-19, and 
• the urgent need for improved rapid diagnostics and antibiotic development. 
• The report underscores the importance of education in combating AMR and highlights the potential of AI in revolutionizing antibiotic development and patient care. It also points to the PASTEUR Act as a significant legislative step towards curbing this public health crisis.

What the Hey? PLA Deadlines Are Now 3 Weeks Early Every Quarter

For several years, AMA PLA Code submission deadlines have been at the start of a quarter (circa January 3, circa April 3, etc). 

Beginning shortly in December 2023, AMA CPT has pulled the submission deadline forward, it is now 2-3 weeks *before* the start of a quarter.

Coming AMA PLA deadlines are December 11 (for January 2024 quarter) and March 11 (for April 2024 quarter). Read more at the AMA PLA website.

https://www.ama-assn.org/about/cpt-editorial-panel/pla-editorial-panel-meetings-calendar

__

Pic: Dalle 3 via ChatGPT4.

Journal Club: With Comprehensive Sequencing, Many Cancers Still Lack "Actionable Mutations"

Debate over how "actionable mutations" are defined, and how common they are, has gone on for a decade.  A major new article by Suehnholz et al. updates us with data from 47,000 annotated solid cancers tested with the MSK IMPACT assay.   

From 2017 to 2022, just five years, the rate of actionable mutations (defined year to year by the database OncoKB) has risen from 9% to 32%.   (The other 68% of solid tumors assessed by MSK IMPACT included those with "no" actionability as well as "limited" actionability.  Such tumors most often had a TP53 mutation.)    

The article suggests that the development of new therapies  against tumors driven by common tumor suppressor genes or transcription factors, will be crucial to further progress, while acknowledging this has been challenging to date.  (For a 2023 review on drugging p53 see Hassin & Oren, open access, Nature Drug Discovery, here.)

Open access:

https://aacrjournals.org/cancerdiscovery/article/doi/10.1158/2159-8290.CD-23-0467/729589/Quantifying-the-Expanding-Landscape-of-Clinical?sf183128105=1

Quantifying the Expanding Landscape of Clinical Actionability for Patients with Cancer, in Cancer Discovery (2023).   Suehnholz et al. (senior author, Chakravarty).

Coverage and interviews at Precision Oncology News here.

Clinical Utility is Not New

While I hope I'm not grumpy, I'm old enough (65) to be annoyed by long-established things being referred to as new.   At a conference today someone will say, "Until 5 years ago, coverage did not require clinical utility," or "Until 10 years ago, no one ever asked about clinical utility of a test."  

This is nonsensical, and could probably be refuted with quotes from William Osler (d. 1919).  But I certainly remember when I was a med student in the mid-1980s it was routine for residents and med students to be grilled about "How will that diagnostic test improve the patient's outcome?"

Here's a paper I ran across, from 2022, but it's a medical history study of the period 1989-2000.   It's called, "Framing utility: Regulatory reform and genetic tests in the USA" (Steve Sturdy, Edinbough, here).

He writes, "Over the decade of the 1990s, however, regulatory reformers in the United States began to argue that genetic tests, specifically, should also be assessed to determine whether or not they actually benefit those undergoing testing—whether they possess 'clinical utility'...[tests should only be introduced] if they had been shown to be beneficial to those tested."   

##
As I wrote about in 2016 here, I've read histories of the approval of PSA in 1986 (diagnostic) and 1994 (screening) and they were quite stormy over the patient benefit (clinical utility) that would result.  One source was a NYT article from 30 years ago, 1993, here.

FOIA Made Easy: An Example CMS FOIA Request (NCDs)

Bullet:  Example FOIA Request, regarding CMS NCD Data.

###

I submit five or six FOIA request to CMS each year, and people sometimes ask, how hard is it?   I find it's not too hard, and the replies can be fairly fast (a couple months), but there's a huge range of success vs failure, speed vs delay.  I have one FOIA topic in the appeals process right now.

One tip I've learned, it's usually better to submit a FOIA request for MAC documents or email to the CMS FOIA office and let THEM deal with the MAC.   Otherwise your request "direct-to-MAC" may fizzle or be brushed off.  (Voice of experience).  CMS has dedicated FOIA staff, and logs the FOIA request and tracks it diligently.  MACs may have hit or miss performance.

Recently CMS updated its dashboard of request letters it has accepted, for future NCDs.   I wondered how old these letters were (6 months?  2 years?) and so I submitted a FOIA request from them.  It's only been a few weeks, so I didn't expect a reply this fast, and I haven't gotten one.  But I thought the example might be a "case study" for readers interested in some FOIA topic of their own.

###

October 15, 2023

FOIA_Request@cms.hhs.gov 

 

CMS FOIA Officer
Centers for Medicare & Medicaid Services
Mailstop N2-20-16
7500 Security Blvd
Baltimore MD 21244

 

Dear FOIA Officer:

 

Request

Under the Freedom of Information Act, Subsection 552, I am requesting access to the following records:

a.      Background:  The CMS Coverage and Analysis Group is part of Center for Clinical Standards and Quality.

b.      Background:  For several years, CMS CAG has regularly updated a public list of “pending” future National Coverage Decisions.   CMS updated the “NCD DASHBOARD” including a “WAIT LIST” on August 23, 2023.

c.      LINK:  https://www.cms.gov/files/document/ncd-dashboard.pdf

d.      Background: the WAIT LIST currently has 7 topics which are “Accepted by CMS” for potential NCDs  This is a cut/paste from the link:

1)      Diaphragm Pacemakers for Neuromuscular Disease (request for non-coverage)

2)      Hepatitis C Screening

3)      Infusion Pumps - Continuous subcutaneous insulin infusion (CSII)

4)      Power Standing Systems

5)      Pulmonary Pressure Sensor for Heart Failure Management

6)      Screening for Cervical Cancer with Human Papillomavirus (HPV) Testing

7)      Ventilators for Chronic Obstructive Pulmonary Disease (COPD)

e.      REQUEST:   I request from CMS copies of the accepted “letters of request” for the NCDs above. 

To determine my status for the purpose of fees, I am an individual seeking information for personal use and not commercial use.   The documents have important public policy value understanding CMS’s current examples of acceptable request letters and durations of the “wait” process.  The letters are normally, eventually, published if and when an NCD is opened in the future.   

I am willing to pay fees to a maximum of $300.  Potential contact staff include Dr Joseph Chin and colleagues in the Coverage and Analysis Group. 

 

CONTACT

If you have any questions you may please telephone me at 1 323 839 8637.   Distribution of your findings by PDF Email or other format is fine.

My email is bruce@brucequinn.com 

Bruce Quinn MD

Address

Journal Watch: In Unusual Op Ed, NEJM Asks About Tissue Biopsies in Healthy People

 

Journal watch:

Biopsies from Healthy Volunteers to Advance Precision Medicine

NEJM:  Paul L. Kimmel, M.D., and David S. Wendler, Ph.D

https://www.nejm.org/doi/full/10.1056/NEJMp2307451

##

The New England Journal of Medicine has published an unusual Op-Ed (Kimmel & Wendler) discussing the ethical and scientific considerations of obtaining tissue biopsies (even renal biopsies) from healthy volunteers to advance precision medicine. This approach aims to compare healthy tissue with tissue from patients to gain insights into various diseases. 

The article highlights the importance of ethical safeguards and IRB (Institutional Review Board) approval when considering such research. While obtaining research tissue from patients is more common, the article argues that obtaining tissue from healthy volunteers may be ethically acceptable, particularly when risks are low and the study has substantial social value. It emphasizes the need for careful risk assessment, informed consent, and compensation for research-related injuries in high-risk studies. 

Overall, and not surprisingly, the article suggests that research procedures involving healthy volunteers require very careful evaluation and ethical analysis.

CMS Retires NCD Blocking Amyloid PET Scans; Now Left Up to MACs

In the past year or two, there's been great debate over CMS coverage of Alzheimer drugs like Aduhelm.  Ten years ago, in 2013/2014, there was debate over whether CMS should cover the then-brand-new PET tracers for Alzheimers/Amyloid.

CMS decided such tracers would only be covered within CED clinical trials.  Patients geting PET scans in those trials were almost entirely White people

With the new surge of interest in drugs, the old NCD blocking PET scan coverage in Alzheimer's has been more problematic.  CMS canceled the NCD effective October 13, 2023 (there had been a comment cycle earlier in the year.)  Since there is no NCD about PET Alzheimer tracers, their coverage is up to local MACs now.

Coverage at ACR here, Applied Radiology here, Alzforum here, Forbes here.  

The decision memo may be of interest for its discussion of when and whether and how to drop an NCD in favor of local coverage.  There are only infrequent examples of CMS proposing a new NCD topic, and deferring to local coverage right away in the final version, or enacting an NCD, and later dropping it in favor of local coverage.  See discussion alone, here.

Find the October 13, 2023 final decision discussion here:

https://www.cms.gov/medicare-coverage-database/view/ncacal-decision-memo.aspx?proposed=N&ncaid=308

Find the NCD cancelation notice here (brief):

https://www.cms.gov/medicare-coverage-database/view/ncd.aspx?ncdid=356&ncdver=2&

Find the claims manual update, effective December 19 but claims payable from October 13 forward.

https://www.cms.gov/files/document/r12364cp.pdf

Nerd Note 1

Sometimes, many months elapse before CMS issues NCD transmittals, or updates instructions to its NCDs.  Here, CMS acted to issue the transmittal to MACs in a few weeks, because MACs aren't responsible for new actions until such a "transmittal" occurs. 

Nerd Note 2

Most CMS NCD decisions come out at their latest allowed due date.  This one came out at 3 months (6 months allowed)

AI CORNER

Here is an AI summary of the "Analysis" part of the NCD, the decision-making part.

In summary, CMS is removing the NCD for beta amyloid PET scans due to advances in medical care, the importance of patient selection for treatment, evolving clinical research, stakeholder and patient feedback, the need for local MAC discretion, and the aim to address health disparities in AD diagnosis and treatment.

In more detail, the main reasons why CMS has decided to remove the national coverage determination (NCD) for beta amyloid Alzheimer PET scans are as follows:

Advances in Medical Care and Treatments: There have been significant advances in medical care and treatments for Alzheimer's disease (AD), including lifestyle changes, risk factor management, and the development of antiamyloid drug treatments. These advancements have rendered the once-in-a-lifetime limit on amyloid PET scans outdated.

Importance of Patient Selection: The development of antiamyloid drug treatments requires the confirmation of the presence of brain amyloid pathology, which can be achieved through amyloid PET scans. Ensuring that patients have brain amyloid is crucial to determining the benefits and avoiding potential harms of these treatments.

Evolving Clinical Research: Clinical research is now focused on new treatments directed against amyloid, which necessitate the use of tests like amyloid PET scans to confirm brain amyloid pathology.

Stakeholder and Patient Feedback: Stakeholders and patients have expressed concerns about the limitations on the choice of treatments and the appropriate management of antiamyloid treatments due to the once-in-a-lifetime limitation on amyloid PET scans.

Local MAC Discretion: Allowing Medicare Administrative Contractors (MACs) to make coverage determinations for amyloid PET scans based on individual patient factors and local clinical environments is seen as more flexible and responsive to evolving medical care and treatment standards.

Health Disparities: Recognizing health disparities in AD prevalence among racial and ethnic groups, CMS aims to address these disparities by allowing local MACs to determine coverage for amyloid PET scans for patients with Mild Cognitive Impairment (MCI) due to AD or mild AD dementia.

##

Those are reasons CMS used.  

In a sidebar, Chat GPT asks if CMS's decision was unwise, here.

In a sidebar, Chat GPT 3.5 and 4 Classic, each summarize the four trade journal articles. Consolidated summary here.