Monday, February 20, 2017

AMA Endorses Expanded Carrier Screening Through Online CME Course


This blog was written on February 20, 2017, and was based on AMA CME for expanded carrier screening and the fact that a new code for this service appeared on the AMA CPT agenda recently.  

Much of the further discussion, deeper below, is superseded by new ACOG documents released February 22, 2017.  

I leave the original blog below, but see the newer version, February 24, 2017, here.


Original February 20, 2017 blog:

The American Medical Association has a dedicated website for precision medicine issues and resources, here.    

The AMA precision medicine center includes up-to-date CME modules for Expanded Carrier Screening, prenatal cell-free DNA screening, somatic cancer panel testing, and other topics in clinical genomics.

According to an AMA and Scripps joint press release, the first module released in this CME series was Expanded Carrier Screening (press release here).

The Expanded Carrier Screening CME is here, at AMA, and it is also crosslisted as a CME course on the American College of Obstetrics and Gynecology website, here.

Expanded Carrier Screening In Perspective

This CME is timely, because according to the agenda for the February 2017 AMA CPT meeting, a code for Expanded Carrier Screening was under consideration (also called Panethnic carrier screening).

Usage of this clinical service is estimated to be higher than 200,000 cases per year.   An August 2016 publication in JAMA reported the results of testing in over 300,000 individuals of diverse racial and ethnic backgrounds (Haque et al., here.)

AMA CPT created a code for Ashkenazi Carrier Testing effective January 1, 2016.   However, looking into available guidance documents, the introduction of screening codes for germline genetic panel tests seem to raise unusually complex issues in terms of guideline generation and clinical utility assessment - more after the break.


Ashkenazi Carrier Testing Has Preceded Expanded (Panethnic) Carrier Screening 
in the AMA CPT

Ashkenazi carrier screening entered the AMA CPT in 2016 with a very simple vignette: "A 32 year old woman is investigated for carrier testing for autosomal recessive conditions."  

Within the AMA's member associations, the latest guidelines for Ashkenazi screening are almost a decade old - from 2008 (in the case of the American College of Medical Genetics, ACMG) and from 2009 (in the case of the American College of Obstetrics and Gynecology, ACOG).   (The ACOG guideline also bears the imprint, "reaffirmed 2014.")

The ACOG Guideline

The ACOG guideline is open access, online here.

Carrier screening is recommended for Tay-Sachs disease, Canavan disease, cystic fibrosis, and familial dysautonomia.

The 2009 ACOG panel recognized that in 2008 (see below), the AMCG had endorsed a broader panel of genes for Ashkenazi carrier screening.  The ACOG states somewhat obliquely: "Screening options continue to evolve...Because of recent advances in genetics, additional carrier tests are available" for mucolipidosis IV, Niemann-Pick disease type A, Fanconi anemia group C, Bloom syndrome, and Gaucher disease.  Accordingly, ACOG provides the reader with a table of gene frequencies and a clinical summary of each of these additional conditions.

A Medical Guideline Without Clinical Utility or Clinical Actions or Clinical Outcomes

The ACOG carrier screening guideline is unusual in saying almost nothing direct about clinical utility.

The guideline notes that "As a result of carrier screening programs established in the 1970s, the incidence of Tay Sachs disease in the North American Ashkenazi population has decreased by more than 90%."   The guideline states that testing "should be offered before conception, or during early pregnancy, so that that a couple has an opportunity to consider prenatal diagnostic testing" and that "when both partners are carriers of one of these disorders, they should be referred for genetic counseling and offered prenatal diagnosis...[which] can be performed by DNA based testing on cells obtained by chorionic villus sampling and amniocentesis."  

The ACOG guideline states that "informed consent [is] required."  

The ACMG Guideline

The 2009 ACMG guideline is online here.  This guideline uses the term "Jewish Genetic Disorders" (abbreviated JGD), characterizing these as disorders whcih have an incidence from 1/900 to 1/40,000 in the Ashkenazi community.

Like the ACOG guideline, the ACMG guideline notes there has been a substantial drop in the incidence of TSD in the Ashkenazi population, such that "the vast majority of children born with Tay Sachs disease have non-Jewish parents."

The guideline recommends screening in an unusual two-step phrasing.
"We recommend carrier screening for cystic fibrosis, Canavan disease, familial dysautonomia, and TSD," be offered to all Ashkenazi Jews who are (a) pregnant or (b) considering pregnancy.  
The guideline then continues: "in addition, recommend that carrier screening be offered for Fancomi Anemia (Group C), Niemann-Pick (Type A), Bloom syndrome, mucolipidosis IV, and Gaucher disease."  
The ACMG guideline mirrors the ACOG guideline in stating that screening should only be done "for those diseases that have been consented to by the patient."

Again, there is no reference to specific clinical utility, only that "individuals [take] time to make appropriate reproductive decisions based on their own personal choices and cultural backgrounds."  A similar approach is found in U.S. guidelines for NIPT (here).  This latter NIPT guideline has a section defining "analytical validity, clinical validity and clinical utility" but says nothing about clinical utility.  Toward the end, it states obliquely that "clinical utility evidence is beginning to emerge" without further discussion or references.

The ACMG guideline differs from the later ACOG guideline in stating specifically that "one Jewish grandparent is sufficient to offer testing."  Although it was written later, the ACOG guideline chose to neither endorse, nor disagree with, this position.

The ACOG Guideline on Carrier Frequency over Multi Ethnic Groups

The authors state that, except for TSD and cystic fibrosis, "the prevalence of these disorders in non Jewish populations is unknown."  This seems a remarkable position to take, for a guideline re-endorsed as recently as 2014.

Frequency and Severity Vary Among Screened Conditions

Frequency of carriers varies considerably among the genes recommended; from 1 in 18 (for Gaucher disease Type I) to as rare as 1 in 127 (for Mucolipidosis IV).

Severity also ranges considerably, from fatal diseases of infancy to cystic fibrosis, where patients survive into the adult years.

Under the ACMG guideline,  testing is endorsed in one-quarter Ashkenazi person, so the de facto frequency range would extend to about 1 in 500 (for Mucolipidosis IV).   The frequency of affected offspring can be calculated as [carrier rate squared, e.g. 1/30 x 1/30] times one-fourth (for recessive genes).   The frequency of a particular couple being a match for any recessive gene would be the carrier rate squared (e.g. 1/900 of couples in the case of a carrier rate of 1/30).

Unusual Call-Out to Religious Groups

The ACMG document carriers an unusual call-out to religious groups, which should be seen in light of an accompanying editorial, which had specific discussions of the positions of African-American and Latino groups (Pletcher et al., here.)

Gross et al. write (here):
It is important to take into account the needs of the community in which screening is intended.  Although there is no definite mechanism whereby any individual or organization can speak for the population as a whole, experience both in the United States and Israel points to overall acceptance of carrier screening in the Jewish population...
The ACMG authors cite two examples:
The ultra-orthodox Dor Yeshorim program has advocated for broadbased testing for decades.  The Central Conference of American Rabbis, the rabbinic arm of the Reform movement, North America's largest Jewish denomination, has passed a resolution urging all Reform Rabbis to counsel prospective couples on the availability of testing.
Guidelines Bodies Face Panethnic Testing

The term "panethnic testing" was used at least as far back as 2008, appearing in the article by Pletcher et al. (here) that accompanied the ACMG guidelines for Ashkenazi person.  Pletcher et al. noted that CF screening was recommended in 2001 on a panethnic basis by a joint ACOG/ACMG guideline, and that spinal muscular atrophy already (in 2009) was close to qualifying for a panethnic screening recommendation.

After the 2008 and 2009 Ashkenazi-focused guidelines from ACOG and ACMG, American guideline production appears to have stopped.

Broader National Use of Panethnic Carrier Testing

Regarding broader use of panethnic testing, ACOG, ACMG, NSGC, and SMFM  produced a formal consensus based "joint statement" (aka not a guideline) on expanded carrier screening, characterized as "an approach for health care providers and laboratories who are currently offering expanded carrier screening to their patients."  The joint statement is here.  It finds that "assessment of ancestry" has a limitation "in our increasingly multi ethnic society" and "the recognition that genetic conditions do not occur solely in specific ethnic groups."  The document also includes tables showing the variability among existing (and now often decade-old) genetic screening guidelines.  

The joint statement states that individuals undergoing "expanded carrier screening" should undergo consent, but this is exactly the same statement found in both the ACMG and ACOG guidelines for Ashkenazi panel testing.

Ethnicity and Genetics Is Complex

Ethnicity is now more complex than ever - with inter-racial marriages rising towards 15%, up from 5% twenty to thirty years ago.   I'm particularly aware of this myself, since my Iowa farmboy Irish-Catholic self has twice been on the bima of a synagogue in California with my African-American daughters becoming Bat Mitzvah.

But more broadly, we are not only a genetic melting pot, but we have substantial populations from every continent and country in the world.

As shown in the 2016 Haque et al. publication looking at over 300,000 persons in over a dozen self-stated ethnic or racial groups, there is an enormous range of gene frequency for a set of profound and severe recessive genetic diseases.  Simplytesting cystic fibrosis and spinal muscular atrophy - as if we were all Caucasian or Ashkenazi - falls extremely short for large swaths of many populations, even using the same cutoff criteria (e.g. carrier rates as uncommon as 1/500 as disorders as severe as cystic fibrosis and Gaucher disease) as used in the 2008 ACMG criteria.  See in particular numerous charts in the Supplement to Haque et al.

Are US Guidelines Now Paralyzed by Their Silence on Clinical Utility?  
Where Does Evidence Based Medicine Take Us?

Unlike the ACMG and ACOG clinical guidelines on carrier screening, two other readily available public sources are quite clear on the clinical utility of recessive carrier screening.

First, the Jewish Genetic Diseases Consortium (here - which by the way has discussions of both 19-gene and 38-gene panels) provides this discussion.  Couples may consider:
  • A carrier couple may choose not to marry (screened before engagement).
  • A carrier couple may choose to become pregnant and test the fetus early.
  • A carrier couple may choose in vitro fertilization with pre implantation diagnostics.
  • A carrier couple may use a sperm (or egg) donor.
  • A carrier couple may choose to adopt.
  • A carrier couple may choose not to have children.
This is essentially the same as the discussion at a second source, the secular medical website UpToDate (subscription, here):  
When carrier mutations are identified preconceptionally in both parents, they may choose to avoid pregnancy, use gametes from a donor who is a noncarrier, employ preimplantation genetic diagnosis with transfer of only unaffected embryos, or conceive naturally and undergo prenatal diagnosis if desired.  If pregnancy is already established, fetal DNA testing can be performed on chorionic villi at 10 to 14 weeks of gestation 
Another example could be a male gay or a lesbian couple which is similarly using sperm or egg donors, surrogates, etc, and given the great deal of effort and cost involved wants to be especially cautious about severe recessive disorders.  Additional clinical utility publications are available, e.g. Azimi et al. 2016 (here).   When a couple is positive for the same carrier gene, the chance of an affected offspring is always exactly one in four - it doesn't matter if the prior base rates for being a carrier were 1 in 50 or 1 in 100.

Clearly, if one asks, "We need more evidence to write or update an evidence based guideline" - no one is ever going to randomize couples to (a) adopt versus (b) use a sperm donor, or anything or the sort.

Having lived through a wide range of evidence-based medicine discussions in the past decade, and written a paper on the topic (here), I would propose that the major outcomes listed above do not require randomized controlled trials - they are simply self evident.   This fact fits well into the range of diverse genomic tests requiring, or offering, diverse types of evidence, as discussed in an important Association of Molecular Pathology publication in summer 2016 (Joseph et al., here).  Couples do not need to "know more" about the difference between adopting and using a sperm donor based on RCTs.

Remember that the ACMG and ACOG guidelines only recommend that genetic testing be offered or discussed - not that it be undertaken.  Such decisions are highly personal, and clearly, such choices will vary by ethnic group.   (To make this point once: in Israel, amniotic testing for trisomy at advanced maternal age has less than 1% uptake in ultra-conservative women, but 94% uptake in other Israeli women, e.g. conservative or secular [Raz, 2009, here]).

The ACMG/ACOG guidelines captured in the existing AMA CPT code miss the Sephardic/Mizrahi populations, which have a largely different panel of disorders (see here).   These would either need to be testing under the nomenclature of the Ashkenazi CPT code (which would then include some irrelevant genes) or under stack coding or under an unlisted code.

Without a New Code, Coding Is A Mess 

Pretty alarmingly to me as a former payer, in August 2016 the AMA CPT Assistant formerly endorsed stack-coding of unlimited numbers of Tier 2 and Tier 1 genetic codes in cases where there was not a specific GSP panel code.  This is how you get claims that are 40 lines long for $30,000 for tests that have a market panel rate closer to $1000 or $2000.   A panethnic code would solve such chaos.


Sometimes a single representational ethnic authority is available - such as the American Rabbinical Council cited favorably in the guideline of the ACMG a decade ago.

But it is unreasonable - in a nation where thousands of physicians in current years are ordering panethnic screening for hundreds of thousands of their patients - for a medical body to wait for separate positions (as the ACMG did for Ashkenazi tesitng) from the Lutheran Church Missouri Synod, the Catholic Archdiocese of Denver, and so on.

The acceptance in the US population (on a voluntary and consented basis) is already shown by the widespread clinicians and patients using carrier screening.   Nor need we continually bring finer religious and ethnic distinctions into the AMA CPT coding system, or else the code for vasectomy (55250) would have a parenthetical (not for ethnic Catholics) and the code for circumcision (54150) would have a parenthetical (not for ethnic Muslims).


Comment - I have done consulting work for multiple genetic laboratories, but the analysis and opinions are solely my own.

Scripps precision medicine education website, including panethnic carrier screening, here.