Skip to end of metadata
Go to start of metadata

This document will be used to track future topics for our weekly meetings:



LOINC: Level of Evidence changes

See these notes from Swapna: LOINC Significance vs Evidence and TMB code proposal.pdf

See these previous call notes for earlier discussion:  CG-2019-08-27

Need to consider the following two new motions:

  • Motion A: Use 53037-8 for both germline and somatic variant clinical significance reporting, and add information to the Term description about the different guidelines for somatic and germline variants;  Keep the Answer list the same, but update the Update the type from Preferred to Example

          o  1st/2nd -

          o  Discussion -

          o  Abstain/Nay/Yea -  / /

          o  Result -

 

  • Motion B: Consider creating new codes for diagnostic, therapeutic, and prognostic significance (see Quest screenshot, LabCorp report) and/or type of evidence (see Baylor report)

          o  1st/2nd -

          o  Discussion -

          o  Abstain/Nay/Yea -  / /

          Result -  



LOINC: TMB changes


See these notes from Swapna (starting on page 8): LOINC Significance vs Evidence and TMB code proposal.pdf

Here are a set of example reports: TMB Examples.zip


This was discussed at Sep 2019 WGM, and here are the notes:

https://docs.google.com/document/d/1dPlTu4yqqJRD4nO_8Chk3G4S1jkIDetIcTGNdOueAdk/edit

        • Question over synonymous inclusion vs only non-
        • Values are extrapolated, will be very difficult to compare values between disparate methods
        • Bob F: best to not hardcode the methodology in the concept as it hasn’t solidified yet, suggest keeping the description of the code


It was also discussed at the following WG call - it seems a proposal was made, but no vote:

https://confluence.hl7.org/display/CGW/CG-2019-06-25

CG IG should support: MSI, TMB, PD-L1

18988

Issue: Currently MSI, TMB, PD-L1 can't be captured with the CG IG. As these a relevant observations in oncology CG reports these values should be supported

  • Examples and info provided in gforge discussion

Discussion:

Patrick - there is not a good fit for these fields in current profiles

May - mCODE struggles with this as well, don’t think we need to make these too specific, would prefer more abstract containers for these data

Patrick - these data are not related to variants, but must be supported

Jamie - this would be a way to attach other observations (biomarkers) to a genetics report

Clem - +1

Patrick - need example for this (see oncology bundle from Alex)

This tracker issue will be completed on a future call

Persuasive with mod - add “Other observations” to figure 1 on “general reporting” as report results. Add textual guidance under the figure for sending other test results.

Motion to accept recommendation / 2nd: 

Discussion: 

Vote (Abstain/Nay/Yea): 

Result: 

Future proposals:

  1. Create LOINC codes for results being reported clinically (not research only) at this time
  2. Create separate LOINC codes for quantitative (muts/Mb) and ordinal (interpretation) results
  3. Quantitative codes for muts/Mb –
    a. Create separate codes for muts/Mb quantitative values for assays targeting different mutation types (non-synonymous vs. non-synonymous+synonymous)
      i. Question – what about insertions & deletions (indels) and copy number gains/losses?

  b. Alternate proposal – create one code for Muts/Mb (without specifying synonymous/non-synonymous) since additional information about the assay will be provided with the result. (Note that a quantitative muts/Mb TMBresult value cannot be compared with another muts/Mb result without information about each assay.)
4. If agree with #2 and #3a above, model terms in LOINC as shown below:

Non-synonymous mutations/Megabase [# Ratio] in Tumor
  Result: # of muts/Mb

Non-synonymous+synonymous mutations/Megabase [# Ratio] in Tumor
  Result: # of muts/Mb

Tumor mutation burden [Interpretation] in Tumor
Results: High, Low


<Insert new item here, use style = Heading 2 as a header>



<Insert new item here, use style = Heading 2 as a header>



<Insert new item here, use style = Heading 2 as a header>



  • No labels