This article has been updated from a version posted March 14 to include comments from industry stakeholders.
The American Medical Association is seeking feedback on a proposal to create a new set of Current Procedural Terminology codes designed to eliminate the current practice of "stacking" codes for complex molecular tests that are not covered by existing guidelines.
AMA said that the proposed CPT codes, described in detail here, are scheduled to go into effect in 2012, at which time "stacking" codes for such tests will no longer be permitted. However, at least one industry observer believes that timeline is unrealistic.
Patrick Terry, who heads the pricing, reimbursement and market access practice at life sciences consulting firm Scientia Advisors, said that the proposal presents "logistical and structural problems" that many diagnostic firms and other stakeholders will find restrictive. "By the time they get this thing through a review and formalized, you're probably into a 2014 timeframe," he noted.
"I think fundamentally those in the industry can appreciate that the [proposed] coding regime is inadequate for current technology and certainly is going to be inadequate for innovative products to come," Terry said. "We're going to have a long road ahead before we have an adaptive and enlightened system of coding to deal with these challenges."
The new codes were developed by the AMA's CPT Editorial Panel Molecular Pathology Coding Workgroup, which has been working since December 2009 to develop new codes to report molecular pathology. "The goal of the MPCW was to provide more current descriptions of the technology required for these services and to more accurately identify and describe the services for all end users of the CPT code set," AMA said in a statement.
The AMA workgroup said that it is presenting the code descriptions to the public to get feedback from interested stakeholders and to ensure it is not overlooking tests that should be included in the listing. Specifically, it said it is seeking three types of feedback: concerns about the specific wording of any particular code descriptor; concerns that a specific analyte is not currently identified in the code list; and concerns that a currently medically useful multi-analyte molecular pathology assay is not currently identified in the code list.
Comments and suggestions may be submitted by April 15.
The codes focus on laboratory procedures involving nucleic acid analyses to detect variants indicative of germline disorders, somatic conditions, or histocompatibility antigens. The changes do not affect molecular microbiology tests or most cytogenetic assays, AMA said.
MPCW said that the new codes address "greater than 90 percent of medically useful molecular testing currently being performed" and that it will continue to develop new codes "for less common analytes and more complex analyses."
A Two-Tier System
As previously reported by Pharmacogenomics Reporter (PGx Reporter 10/20/2010), the workgroup established a two-tier system to help categorize the multitude of available and emerging molecular tests.
Tier 1 codes represent "the majority of commonly performed clinical molecular analyses," AMA said. In order to eliminate the current practice of code stacking, the new Tier 1 codes include all analytical services performed in the test — such as cell lysis, nucleic acid stabilization, extraction, digestion, amplification, detection, and interpretation. This new format is expected to "furnish providers and payers with greater coding accuracy to communicate the tests that are actually performed." The unit of service for each of these tests is anticipated to be one.
Tier 2 codes, meantime, represent "medically useful procedures that are generally performed in lower volumes than Tier 1 procedures," and are arranged by level of technical resources and interpretive work by the physician or other healthcare professional.
The MPCW lists 92 gene-specific and genomic procedures as Tier 1 tests, including single-gene tests such as BCR/ABL1 translocation analysis; common variants in the CYP2C19, CYP2D6, CYP2C9, and VKORC1 genes; variants in KRAS codons 12 and 13; and variants in UGT1A1.
Other tests are covered by a series of codes. For example, there is one code for full sequence analysis of the BRCA1 and BRCA2 genes; but a separate code for analysis of the 185delAG, 5385insC, and 6174delT variants in those genes; and another code for detection of uncommon duplication/deletion variants in the two genes. There are also separate codes for tests that assess the BRCA1 and BRCA2 genes individually, as well as codes for testing known familial variants in each of those genes.
Likewise, there are 14 different codes to describe HLA typing, five different codes for analysis of the CFTR gene, and three codes each — comprising one for full sequence analysis, a second for known familial variants, and a third for duplications and deletions — for genes such as MLH1, MSH2, MSH6, MECP2, and PMS2.
Other codes include analysis of a number of genes in a single test. The code for Long QT syndrome sequence analysis, for example, lists 12 different genes.
The new codes also address new technologies. For example, there is a code for genome-wide cytogenomic constitutional analysis via comparative genomic hybridization microarrays. The guidelines list a separate code for analysis of genomic regions for copy number and SNP variants for chromosomal abnormalities, and note that this code cannot be reported in conjunction with the array CGH code.
Tier 2 codes include nine different "levels" of molecular pathology procedures — ranging from Level 1, restriction enzyme digestion or melt curve analysis, to Level 7, analysis of more than 50 exons in a single gene via DNA sequencing. Each level includes several specific analytes that are covered by the code.
The workgroup notes that analytes that are not listed in the Tier 1 or Tier 2 codes should be represented with the "appropriate methodology codes" in the current CPT guidelines.
The MPCW stressed that the codes listed in the proposal are not final, and that they are not yet available for reporting or recognized in any code set.
Further, it noted that the inclusion of an analyte in the CPT code set "does not imply any health insurance coverage or reimbursement policy. While these descriptors describe a service, payment policy will determine appropriate payment for these services."
'Welcome to the 90s'
It's likely that some in the industry will find fault with the proposed changes.
Scientia's Terry noted that the coding scheme is ill-suited for many currently available genetic tests, and not flexible enough to account for emerging tests and technologies.
"It's kind of like, 'Welcome to the 90s, AMA,'" Terry said. "What they're proposing is fundamentally trying to deal with technologies and assays that were developed more than a decade ago."
Terry estimated that the assays listed in the guidelines account for only about one-tenth of the available genetic tests on the market. "For the AMA to assume that that is the codified universe in which they have to create molecular codes, and create this tiered structure based on volume, they're really starting behind the eight-ball from the get-go."
Perhaps even more important is the fact that the proposed changes are "in no way adaptive enough to handle the current diagnostic paradigm that's emerging in genetics and genomics," Terry said.
The proposed CPT codes are currently "focused on the DNA molecule," he said, but "what we're going to be seeing in the short time horizon are RNA-based diagnostics, small interfering RNA sequences, transcriptomics, epigenetic tests, and structural genome tests. So you're going to have this whole new constellation of testing that is going to challenge the proposed framework that the AMA has outlined."
Furthermore, he said, "there is no indication" in the document of how AMA intends to handle "multiplexed, algorithmic assays" such as Genomic Health's Oncotype Dx, Agendia's Mammaprint, and Pathwork Diagnostics' Tumor-of-Origin test — all of which are already on the market.
Terry, a co-founder of Genomic Health, said he believes that AMA will receive negative comments from companies "that are going to be immediately impacted by the changes," though he noted that more established diagnostic players might support the new framework.
"If you're a manufacturer and you want a clear code so that when you develop a test today and launch it tomorrow you can immediately bill for it …you'll support this AMA activity," Terry said. "But if you're an esoteric lab with a highly complex test that doesn't map well to this new coding regime, then obviously this could kill your business and your company."
While he declined to specify any firms that might see their reimbursement rates drop as a result of the new framework, he said that it's likely the changes will impact firms that have pursued "value-based" pricing schemes in order to "navigate around the ineffectual coding system that we currently have."
Such companies, which have to date been able to negotiate reimbursement with payors based on the medical value of their tests, now face "getting stuffed into a pricing band," Terry said.
Genomic Health, which has successfully pursued a value-based reimbursement strategy, could not be reached for comment for this article.
In its most recent annual report, released last week, Genomic Health said that the new codes "could result in higher or lower reimbursement rates for our tests." It added that regardless of whether it obtains a specific CPT code for its tests, "there can be no assurance that an adequate payment rate will continue to be assigned to the tests."
Other industry observers have noted that such firms face uncertainty amid the proposed changes. In a research note on Myriad Genetics published last week, investment bank Jefferies noted that "value-based price negotiation capability might be lost" under the AMA's proposed framework.
Since the AMA has assigned a series of Tier 1 codes for BRCA testing, "we believe that under the new reimbursement rules, [Myriad] will not be able to maintain its current levels of negotiated reimbursement," Jeffries said. However, the analysts also noted that BRCA testing "is not likely to experience as dramatic a cut as can be expected with some other high-complexity tests."
The investment bank estimates that Myriad could see a 10 percent reimbursement reduction for BRACAnalysis beginning in the second half of 2012 — beginning with government reimbursement rates and then rolling through its managed care business through the early part of 2014.
A Myriad spokesperson said that the company is "keeping an eye" on the AMA initiative, but noted that "there are just too many moving pieces to comment on what the impact for Myriad would be."
The ultimate repercussions of the proposed changes for Myriad and other molecular diagnostic companies is "really the million dollar question," the spokesperson said.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at btoner [at] genomeweb [.] com.