Discussion on PerformedGeneticObservationResult.mutationTypeCode

Return to BRIDG

CLOSED - Proposed New Definition and Examples for PerformedGeneticObservationResult.mutationTypeCode

Current Definition

DEFINITION: A coded value specifying whether a genetic variation is inheritable.

EXAMPLE(S): GERMLINE, SOMATIC

OTHER NAME(S):

NOTE(S):

Comment

This should be linked to the LOINC answer list with these values, which are more complete. Proposed wording: “Coded value defining the genomic source of the variant, which can indicate the variant context including inherited germline mutations, acquired tumor specific mutations, mutations within a fetus, or unknown origin.”

Proposed Disposition

Persuasive

Proposed Disposition Comment

Need to make sure this is consistent with PGx SME's understanding and we're not de-supporting their use case - also need to get official LOINC reference info. Are inherited germline mutations, acquired tumor specific mutations, mutations within a fetus, and unknown origin a broader set of example values we should use?

Initial Discussion

Concern was raised that this attribute name may be a misnomer; need to check with PGx SMEs re what was intended with this concept and what is the CDISC use case for it; concern was also raised about example values possibly conflating ideas; acknowledged that some members of CG WG have spent lots of time on this topic and could contribute; might be worth taking a look at ClinGen model, etc. to see how other models represent this concept – need to ensure it’s a model that includes both concepts (germline and somatic) otherwise it might not even have the concept

Outstanding Questions

• PGx SMEs: Does the following definition support your use case? Are the following examples valid?
  • DEFINITION: A coded value specifying the genomic source of the variant which can indicate the variant context.
  • EXAMPLE(S): inherited germline mutations, acquired tumor specific mutations, mutations within a fetus, and unknown origin
• Anyone: Are the original example values conflating ideas? What other models are there that include both germline and somatic mutations and how do they represent this idea?

Responses

Lauren Becnel: The definition is not correct. Germline mutations and even epigenetic changes can be inherited, but somatic mutations like what you see in cancer cannot. Genetic tests are looking for changes in the sample (i.e., cells from whole blood or a piece of a tumor) relative to a reference genome sequence and comparing any changes found by that comparison to databases to determine if the change(s) have been seen before, have biological significance, and have clinical significance.

Propose changing to DEFINITION: A coded value specifying whether a genetic variation is present.

Q from Wendy: I think this definition implies an indicator kind of value rather than a code but Mollie's comment was suggesting there can be more than one kind of genomic variant source, which can indicate variant context, and her values implied that. They were not just GERMLINE and SOMATIC (original ones from PGx team), but germline mutations, acquired tumor specific mutations, mutations within a fetus, or unknown origin. Is this really an indicator or a code?

DISCUSSION 20160225: use the definition Mollie provided: “Coded value defining the genomic source of the variant, which can indicate the variant context including inherited germline mutations, acquired tumor specific mutations, mutations within a fetus, or unknown origin.” Lauren will try to send some more values from the LOINC answer list by Monday afternoon.

Lauren Becnel, 20160226: Wendy, I can’t find anything in LOINC that has germline, somatic, etc. as values. What I could find is from the latest FHIR specification, http://hl7.org/fhir/StructureDefinition/geneticsGenomicSourceClass, that references LOINC’s Allelic State though I don’t find a match in that term, though as warned I’m not a LOINC expert so am probably missing something. Regardless, we can reference FHIR here with these values.

1. Germline
2. Somatic
3. Prenatal
4. Likely Germline
5. Likely Somatic
6. Likely Prenatal
7. Unknown Genomic Origin

UPDATE DONE IN THE MODEL

Return to BRIDG