Stream: genomics
Topic: Cytogenomic Result Interpretation in FHIR
SangHo Kim (Feb 09 2022 at 00:50):
Hi,
I would like to seek advice and suggestions from the community and experts to express "interpretations" of a cytogenomic test result.
From my personal analysis of the test result samples I have been given by the lab is that cytogenomic tests have a huge range of possible abnormal results with different interpretations.
For example,
1) 46,XY,t(9;22)(q34;q11.2)[20] ==> translocation of 9 and 22 ==> interpreted as "Philadelphia chromosome" observed
2) 47,XX,t(9;22;16)(q34;q11.2;q12.1),+der(22)t(9;22;16)[20] ==> is of same translocation of 9 and 23 and 16 with addition of abnormal 22 chrom ==> ALSO reported as "Philadelphia Chromosome"
Based on my previous question on FISH I think you can express the result of the Karyotype String.
But in terms of Interpretation of the result, for a Variant report you would have Present/Absent Loinc Codes, are there loinc codes to express these interpretation results like "Pohiladelphia Chrmosome, isodicentric 15 chromosome syndrome, Kleinfelter syndrom, mosaic types, etc)?
Is it also correct to have separate Obx (region studied profile) to express Karyotype and have another obx (interpretation ?) to express interpretation of the result in a cytogenomic tests?
Thanks in advance!
SangHo Kim (Feb 09 2022 at 00:54):
Or can you just input the interpretation result in the "region studied profile" under the "conclusion-string" component?
세표유전-region-studied-profile-component-목록.PNG
SangHo Kim (Feb 09 2022 at 01:02):
Furthermore,
to express a Karyotype as Observation Value, you would express as CodeableConcept...
I searched for loinc code with search term "Karyotype" (https://loinc.org/search/?t=1&s=karyotype)
and got 67 results.
Some codes are specimen type specific and some with test method (FISH, High Res, Cyto, Micro, G-Banded)...
Is there one specific code to express just the general Karyotype?
Thanks!
Bret H (Feb 09 2022 at 15:38):
The interpretation 'Philadelphia Chromosome' would be something to put into a Diagnostic Implication profile (http://build.fhir.org/ig/HL7/genomics-reporting/StructureDefinition-diagnostic-implication.html) . Also there is a slot for OMIM terms to be used. I would not use region-studied for the implication that the karyotype result indicates presence of 'Philadelphia Chromosome.'
The ISCN karyotype information can go into the Variant profile in the component ( component:cytogenomic-nomenclature 0..1 BackboneElement Cytogenomic Nomenclature (ISCN)) http://build.fhir.org/ig/HL7/genomics-reporting/StructureDefinition-variant-definitions.html#Observation.component:cytogenomic-nomenclature
be sure to look at the snapshot views of the profiles as well as the differential view
Bret H (Feb 09 2022 at 15:46):
The component 'cytogenomic-nomenclature' on the Variant profile is meant to be the place to put ISCN nomenclature. For others that read the post, ISNC stands for International System for Human Cytogenomic Nomenclature and is commonly used in FISH and microarray results that deal with cytogenomics, see https://pubmed.ncbi.nlm.nih.gov/27910032/
Link to Variant profile (http://build.fhir.org/ig/HL7/genomics-reporting/StructureDefinition-variant.html)
Bret H (Feb 09 2022 at 16:50):
If you would like share the example as you build it, that would be really awesome for others too learn from, and help develop further guidance.
SangHo Kim (Feb 15 2022 at 00:27):
@Bret H
Thanks for the detailed response and links!
The answer helped me clear out some of the confusions.
After further discussions with the lab, they seem to have reports that are complex.
So, for example, they have Karyotype results that are very long with word "idem" and the like.
After asking, they explained that these "words" are included to reduce the length of the result and is part of the nomenclature standard (ISCN).
I would like to confirm if we can store such complex Karyotype results in the mentioned "cytogenomic Nomenclature" component?
Even though the complex lengthy Karyotype results (with "idem", "der", "mar"... etc) if the result holds so many observations is it correct to just store in one Observation and have that Karyotype result in the said component?
At first glance, I was thinking of storing each Karyotype (they are reported in single string but with "," as separator) as separate observation and have the "Main Stem (the Karyotype that represents the "idem" ) as the center resource. Then have the other results that have "idem" in it have the main stem resource as hasMember? to group them logically. Would this also be a good approach?
Also, some results also have things like a -karyotype or a +karyotype after the main stem karyotype ( {Main Karyotype} , {-part section of main karyotype then second karyotype}, {+some other Karyotype} )... it seems like ISCN is trying to simplify the Karyotype results with words like "idem" or "- / + " but I can't seem to find a way to express these types of result in FHIR :(
I can't seem to logically connect between them.
Thanks!
Kevin Power (Feb 15 2022 at 14:57):
Hi @SangHo Kim - I am afraid you are hitting on an area where we are missing specific guidance. I think the simple approach for now would be to have a single Observation with the entire karyotype string in the cytogenomic-nomenclature component.
I know that @Bret H has spent more time thinking about this so perhaps he can provide better guidance for now.
Kevin Power (Feb 15 2022 at 14:59):
Building out better guidance for cytogenetics is on our list to consider for our next release.
Last updated: Apr 12 2022 at 19:14 UTC
