Thursday, August 8, 2024

MolDx Finalized LCDs for Prostate Cancer, Lung Nodules

MolDx has finalized LCDs, proposed last year, for prostate cancer management and for indeterminate lung nodule management.  

  • My "Human" article is below (with some AI inserts.)
  • And you get an alternative article in a parallel universe.  
    • The alternate article is written by Chag GPT after I uploaded it with the two PDFs, prostate LCD and lung LCD.  See here.

PROSTATE

See final LCD L39636 here.  Effective on September 22, 2024, it is based on a letter submitted February 2022.   The LCD remarks say minimal changes were made for clarity, and that's true.  A redline against draft DL39636 shows virtually no changes.

There are 11 somewhat complex coverage rules, but I summarize here via Chat GPT 4o.  For exact rules see LCD itself.

Eligibility Criteria:

  1. Patient Condition:

    • Must have castration-resistant or metastatic prostate cancer.
  2. Test Suitability:

    • Must match the intended use and population for which the test was validated.
  3. Management Options:

    • Patient is considering multiple treatment options of varying intensity, and the test aids in decision-making.
    • OR the test predicts response to a specific treatment based on guidelines or FDA labeling.
  4. Previous Testing:

    • Patient has not previously taken a similar test for the same purpose.
  5. Prior Treatments:

    • No prior pelvic radiation or androgen deprivation therapy (ADT) before the biopsy or prostate resection specimen used for the test.
      • Exception: Men who haven't had secondary systemic therapies post-ADT and either:
        • Lack other standard drug-targetable gene alterations.
        • Have drug-targetable alterations but are ineligible for those therapies for another reason.

Test Requirements:

  1. Validity and Utility:

    • Must show analytical, clinical validity, and clinical utility, proving it can stratify patients and guide management effectively.
  2. Evidence:

    • Clinical validity must be supported by peer-reviewed studies.
  3. Algorithm Validation:

    • If using an algorithm, it must be validated separately from its development cohort.
  4. Regulatory Compliance:

    • Testing must comply with CLIA and/or FDA regulations and be performed in an accredited lab.
  5. Lab Responsibilities:

    • Labs must clearly inform physicians of the test's intended population and use.
  6. Technical Assessment:

    • Must pass a MolDX® Technical Assessment confirming all validity and utility criteria

Indeterminate Lung Nodules

MolDx also finalized L39654, Molecular Biomarkers for Risk Stratification of Indeterminate Pulmonary Nodules Following Bronchoscopy.   Find it here.

Note that this is following a non-diagnostic bronchoscopy.   Other tests like Biodesix XL2 are for patients with an indeterminate low dose CT scan (or other CXR), to help guide a decision for biopsy.

L39654 was MAC-initiated and changes to the final included "minor changes" for clarity or related to coverage.  I verified that with a redline.

There is one interesting change at the end of the policy, highlighted text is new.  

Finally, reference to specific tests in this document does not imply coverage by this contractor. ^This policy is applicable to molecular tests that aid in the diagnosis or exclusion of lung cancer in a patient with an indeterminate pulmonary nodule following a non-diagnostic bronchoscopy. This includes all molecular analytes that are within scope of the MolDX Program and is not limited to a single analyte (i.e RNA). Abstracts and preprints of journal articles were not considered, as they were not fully reviewed peer-reviewed publications at the time of drafting this local coverage determination. This coverage decision will be periodically re-evaluated based on evolving literature and national consensus guidelines.
 
I've again asked Chat GPT 4o to read and simplify the criteria.  For coverage rules, see original MolDx text.

Eligibility Criteria:

  1. Patient Condition:

    • Must have an indeterminate pulmonary nodule (IPN) and a non-diagnostic bronchoscopy.
  2. Test Usage:

    • Test results must inform patient management based on recognized guidelines.
    • No alternate method (EBUS, FNA) is available to diagnose the nodule before the test results.
  3. Exclusions:

    • No personal history of lung cancer.
    • No current diagnosis or high suspicion of cancer.
    • Not at very low or very high risk for lung cancer, where test results wouldn’t change management.
  4. Previous Testing:

    • No prior similar test for the same condition.
  5. Test Suitability:

    • Must match the intended use and population for which the test was validated.

Test Requirements:

  1. Validity and Utility:

    • Must show analytical and clinical validity and utility, proving it can guide patient management.
  2. Evidence:

    • Clinical validity must be supported by peer-reviewed studies.
  3. Rule-out Tests:

    • Must have high sensitivity and negative predictive value (NPV) to safely select less aggressive management.
  4. Rule-in Tests:

    • Must have high specificity and positive predictive value (PPV) to safely select more aggressive management.
  5. Algorithm Validation:

    • If using an algorithm, it must be validated separately from its development cohort.
  6. Comparative Performance:

    • Must demonstrate equivalent or superior performance to existing tests for the same use.
  7. Technical Assessment:

    • Must pass a MolDX® Technical Assessment confirming it meets all criteria and is reasonable and necessary.
Coverage Can Be Annoying to Find

Coverage can be difficutl to find.  The covered tests aren't named in the LCD's (to avoid the need to update LCDs).   But the billing and coding articles might not name covered and non-covered tests, either.   (The billing article may just say, "Submit 81479 and Z code").   In theory you can find coverage remarks at the separate MolDx DEX online database, but it's updated at lagging intervals and if you don't have a pre-existing list of what you're shopping for, you can't really even get started.