Saturday, May 3, 2014

Lung Cancer Screening: CMS Advisory Panels Splits 180 degrees from USPSTF

Update: Proposed CMS Decision released, November 10, 2014:  See here.

Last winter, the United States Public Services Task Force (USPSTF) issued a Grade B recommendation that low dose CT scans provide a health benefit in the age 50-80 population, who have > 30 pack years of smoking and who currently smoke or quit less than 15 years ago (see report here.)   Under the Affordable Care Act, commercial insurers will be required to offer this benefit, without copays, in a year or two.

In February, Medicare opened a decision analysis on whether to add this benefit to the Medicare program.  Medicare has undertaken this process and concurred with the USPSTF recommendation in numerous NCDs since 2010.   On April 30, Medicare held a public panel - a MedCAC advisory board - to discuss the low dose CT screening benefit for Medicare patients.   The Medicare panelist gave strikingly low votes that they had confidence the benefits would exceed the risks and uncertainties in the Medicare population.

My detailed notes on the meeting can be downloaded from online here. [50 pages including key figures presented, websites, and footnotes for most of the articles cited by speakers.  At the link, see the upper left arrow for 'download.']

(Update 9/24/2014: The CMS transcript of the full session, word for word, is now online here.)

For a brief trade journal review, the MedPageToday summary is here.   The Lung Cancer Alliance immediately expressed its "deep disappointment" with the panel's vote.

* USPSTF Endorsement Usually Viewed as a High Bar

The endorsement of a test or service for preventive medicine by the USPTF has been viewed as a very high bar.  For example, USPSTF recommendations against use of PSA testing in men and against mammography in average risk women age 40-50 were viewed by some stakeholders as examples of the USPSTF setting the bar set "too high."  (Of course, other stakeholders concurred with USPTF on these topics.)

* If Reviewed Favorably, Medicare Can Now Offer USPSTF-endorsed Preventive Services

By law, since 2008, Medicare can offer USPSTF-endorsed preventive services under federal Medicare funding if the agency reviews the service as appropriate for Medicare patients (regulation).  Over the last several years, Medicare has matched the USPSTF numerous times, for topics as diverse as smoking cessation counseling and HIV screening, and Medicare offers the public a frequently revised 32-page booklet explaining its expanding preventive benefits (here).   While I haven't tallied the numerous decisions yet, I don't believe any USPSTF benefit that has been reviewed by Medicare has been rejected.  (Some, such as for pediatric services, wouldn't enter review.)   Most recently, CMS released a proposed decision memo on its coverage for Hepatitis C screening (here.)
There can be quirks to this process. For example, since USPSTF covers primary care prevention services, the Medicare proposal for a Hepatitis C screening benefit defines primary care settings as excluding Medicare patients who are at risk for hepatitis C but are seen in emergency rooms.

* Medicare Opens Discussion on Lung Cancer Screening for Smokers

The favorable USPSTF position was announced in draft form for public comment in summer 2013, and finalized in December 2013.   The favorable decision was most strongly driven by the 53,000 patient National Lung Screening Trial, which compared low-dose CT screening with chest x-ray screening in a two-arm randomized controlled trial.   There was a 20% reduction in lung cancer deaths in the LDCT arm.

While CMS opened the NCD review in mid February 2014, the documentation includes a coverage request letter by Peter Bach MD, a pulmonologist at Sloan Kettering Memorial Cancer Center and a well-known health policy expert, dated September 9, 2013.  Bach noted that the pending final USPTF approval aligned the task force with other credible organizations such as the NCCN and its recommendation.  Bach urged that the coverage be initiated with a registry to ensures answers would be obtained to various unanswered questions. Bach's request included an 18 page dossier.   On October 25, 2013, the Lung Cancer Alliance (LCA) also wrote CMS, also urging adoption of the USPTF recommendation on an expedited basis. (Both request letters online at CMS, here.)

* Medicare announced the April 30 public panel, or MedCAC, on February 24, 2014 (here).

All-day video is online at Youtube (Part One, Part Two), and a full transcript will be published within several months on the CMS website.   The workshop included four main speakers:

  • Paul Pinsky, MD, National Cancer Institute, describing the 53,000-patient National Lung Cancer Screening Trial (NLST)
    • NLST, NEJM, 2011, 365:395-409, here.
    • NLST, NEJM, 2013, 368:1980-1991, here.
  • Peter Bach MD, describing his coverage request, reasons for registry based coverage, and additional reflections on the issue.
  • Laurie Fenton Ambrose, President/CEO of the Lung Cancer Alliance, urging coverage, no additional registries or burdens to access, and describing the LCA's quality assurance programs for nearly 200 centers and growing.
  • Doug Campos-Outcalt MD, American Association of Family Physicians, discussing the AAFP's public disagreement with USPSTF recommendation.
In addition, there were 16 brief, four minute presentations from numerous stakeholders, all favorable to coverage.   I have summarized comments of each group in my full online document (see link above the picture).   Dr. Claudia Henschke presented data from the longstanding international group I-ELCAP (International Early Lung Cancer Action Program).   Quality programs initiated by the American College of Radiology, the Lung Cancer Alliance, and the American Association for Physicists in Medicine were described.  Regional community hospitals presented data on the quality metrics of their programs.  Several speakers urged that coverage not be restricted to academic medical centers that are distant from rural patients.

* MedCAC Panelists Raised Numerous Concerns and Uncertainties

After these presentations, for over two hours the MedCAC panelists raised numerous concerns and reservations about the applicability of the NLST data (age 55-74) to the Medicare population (over 65), the fact there was only one major trial (albeit a 53,000 patient RCT), and whether community practice would enroll unnecessary patients or perform unnecessary follow-up invasive procedures with more harms than risks.  (Several hundred patients would enter a low dose CT screening program for each life saved).  
  • There were concerns that while the whole endeavor still had too much uncertainty, if it is approved, there must be mandatory quality and accreditation standards before reimbursement by CMS.
  • There were concerns that the frail elderly would be exposed to tests which were unlikely to have benefit and would have harms compounded by the patients' frailty.  It was noted that the older members of the cohort, being longterm smokers, had comorbidities and short lifespan, limiting the benefits created by averting a lung cancer specific death.  Data on use of colonoscopy in the very elderly were cited. 
  • It was noted that "patient still die of lung cancer" as the absolute death rate drops from about 2% to 1.5%.   
  • There were objections that some new techniques and protocols should logically improve net outcomes (e.g. better software, better diagnostic rules, better follow-up rules) but that these were "models" and not data.   
  • There were many concerns it was not possible to generalize from the controlled, NCI-funded, monitored and protocol-monitored NLST study to routine community practice.
  • There was considerable discussion of the fact that the NLST study enrolled from age 55-74, but the USPSTF extended the eligible population to age 80 if they had no contraindications to surgery and no other clearly life limiting conditions.   Panelists felt this criteria was vague and unenforceable.  
  • There were concerns about radiation exposure - countered by demonstrations that screening CT can be accomplished with modern equipment at very low doses.  
  • There were concerns that the true-positive rate was low, and most "positives" reflected benign lung nodules.  While these were mostly ruled benign based on follow-up behavior on imaging (no biopsy required), panelists raised concerned that quality of life impact due to anxiety was not published and therefore unknown.  Proponents noted that this data was not published within NLST, but available from other sources.
  • There was mention of concerns that some smokers would continue to smoke, rather than quit, because of the increased safety and assurance of being enrolled in a CT screening program.   Speakers described universally that their centers required simultaneous, though not forced, availability of quit-smoking programs to enrollees in a CT screening program.  

On a scale of 1 to 5, panelists voted low confidence (2.2) that net benefits would exceed harms if the coverage were to be initiated in the Medicare population.

* Next Steps for Medicare

CMS staff noted that the day represented one part of their information-gathering efforts on this topic, and that a draft decision for public comment would be posted by November 10, 2014, and a final decision by February 8, 2015.

Summary of Key References:
MedCAC Webpage, April 30:  here
CMS Coverage Decision, In Process:  here
My 50-page online notes:  here    CMS Full official transcript, here
USPSTF Decision:  here
MedPage Today:  here
Radiology trade journal Aunt Minnie (may require registration): here
Aunt Minnie Europe on European Perspectives: here
Aunit Minnie Opinion Essay by Dr Frederic Grannis: here
Hayes Blog on "overdiagnosis:" here

September, 2014:
New York Times Op Ed, ("A Cancer Battle We Can Win"), here:
JAMA, 9/24/2014:  Should CMS Cover CT Screening for the Fully Informed Patient?  here
JAMA, 9/24/2014: CMS & CT Screening: Last Chance to Get It Right? ["If CMS covers, should do so very carefully..."] here

* Molecular Diagnostics Follow Screening

Wider screening for lung cancer will mean the detection of more small, apparently contained cancers.   Several companies are working on tests to assess which is these actual tumors need more invasive surgeries, radiation, and chemotherapy (e.g. Integrated Diagnostics, here, and Life Technologies, here.)  GeneCentric has launched a molecular assay to help classify the type of lung cancer found, here, available from LabCorp, here, as HistoPlus/SM LungCancer.

*  Economics Studies Presented at ASCO May 2014 - After the MedCAC

On May 15, as advance press for the May 30-June 5 ASCO meeting, some additional economic studies were released early  and featured in major press stories.  For summaries and links, see Kaiser Health News here.  For example, at a 50% penetration rate, there would be 11M CT scans, detecting 55,000 additional and mostly early lung cancers, at a cost of about $9B (5 years).   The lead author was Joshua Roth of the Fred Hutchinson Cancer Research Center in Seattle.  The Medscape story is here.   The current ASCO guidelines, from 2012, favor screening and are here.