 So we're going to talk today about CPT coding and reimbursement. CPT code does not equal reimbursement. It's a catalog and a list of services. And they're two separate items. I'm still getting the hang of where I click. So we're going to start off with CPT codes. And I don't know how many of you remember before January 1st, 2013, is all genetic tests and molecular pathology tests were stacked together to get reimbursed. And they were commonly called the molecular stacking CPT codes. And basically each step of the assay, such as DNA extraction, PCR, probe, hybridization, each of those sort of steps of an assay had different CPT codes. And laboratories would put all those CPT codes together to create a molecular coding stack. So what happened is AMA wanted to sort of create new codes. So the new codes were called Tier 1, Tier 2, and multi-analyte assays with algorithmic analyses, or commonly referred to as MOV. Why did the AMA want to put new CPT codes together? So there was getting, essentially it came down to payers wanted to know what they were paying for, and we needed a clear granular system to do that. So part of my job was to get information from laboratories throughout the country, and usually most of it involves a large reference laboratory and some smaller and medium reference laboratories throughout the U.S. So, and I don't expect you to be able to see this or read this, but so we took all the assays and the relative percentages and put them all together. And what you can see is there's about 100 tests or so that really cover about 95 to 99% of the tests that were performed across the country a few years ago. So now it's probably changed a little bit, but this gives you an idea that most of the testing is done in a relatively small number of tests. For pathology or mole path codes, the Tier 1 was created to be analyte-specific codes. Those are the top 100 that are gene and analysis-specific. And then the Tier 2 codes were levels of complexity lumped together. And then you can go to the AMA website and actually see the most current list of codes. So part of what we had to do is sort of create a whole new format. So in a CP Tier 1 descriptor, we developed that the first part of the name would be the Hugo-approved gene symbol and then the long version of the Hugo-approved gene name in parentheses. Gene names are in italics, so that part's in italics. The second parent includes an example of the disease-stater condition. And then outside of that, it's the gene analysis and the analysis type. One thing about these descriptors is I was at one place and they said, well, if your disease or the condition's not listed in there but it's related to that gene, you can't use that test, but that's not true. The disease and condition is an EG statement and not an all-inclusive list. Often common gene variant names are used and the code includes all analytical services performed in the test. If there's cell lysis, DNA extraction, digestion, amplification, and detection. And then all the analyses are qualitative unless otherwise noted. So just to go over an example here is we have BCR-ABL, which is a gene fusion for chronic myelogenous leukemia. The Hugo is actually two genes that used together, so it's a translocation. So we're using the translocation or the cytogenetic information. And it's a translocation analysis for the major breakpoint, whether it's qualitative or quantitative. And then there's subcodes here for the minor breakpoint and other breakpoints in the fusion. Another example is cystic fibrosis, where we have a genus analysis, a common variant. ACOG and ACMG have some guidelines and we didn't know if that was going to change or stay stable over time. And instead of listing that all, we just put the most ACOG or basically the current guidelines. Then under CF, then you could test for known familial variants, deletion of duplications, full gene sequences, and in some cases, infertility just for the poly T analysis in addition to cystic fibrosis. The array CGH is also commonly performed and at that time a lot of the array CGH codes were oligo or BAC. I think that's mostly gone away. Most people are now, most laboratories are now doing the 81229 for the copy number and single nucleotide or SNP variants arrays. We did put language in there that you couldn't use both of these at the same time. So either one or the other gets used. In Tier 2, approximately 600 different genes or gene type analyses to be divided up. And these were the sort of that long, never-ending tale of that graph I showed you earlier. These tests are like the common ones before, but they still meet Tier 1 or I mean Category 1 status. They're lower volume assays, so these were divided into nine levels of complexity. I'll give you an idea. And I'm not going to end something. There's the examples underneath all of these is that Tier 2 Level 1 is basically one variant or one SNP testing. Tier 2 is 2 to 10 SNPs or a methylated variant or somatic variant. Somatic or tumor testing or oncology testing is usually much more complicated than a single DNA blood test. So that had a little higher level of complexity. Then there was greater than 10 SNPs for Tier 2 Level 3 and Level 4. This is where we start doing sequence analysis. So this is analysis of a DNA sequence of a single exon. Or if you're using some sort of multiplex analysis of 10 or more exons. Or 10 or more variant reactions. In Level 5 we're doing sequence analysis of 2 to 5 exons. And really most of this was developed before the advent of massively parallel or next generation sequencing. Then we have 6 to 10 exons in Level 6. 11 to 25 exons in Level 7. And this is by the number of genes, the gene, how many exons it simply has. Level 8 is 26 to 50 exons. And then in Level 9 it's greater than 50 exons. And actually there's very few in Level 9. There's very few genes that actually fit into Level 9. That came up during this is what do you do if your gene or your favorite gene is not listed? So there is a miscellaneous code called 81479. At the time, during the conversion, a lot of labs said, well, my favorite gene fits in this level. And part of the requirement is that you can't self-assign. If your favorite gene is not in the level, then you have to submit a coding change proposal. Excuse me. We can't get into the WebEx and see the slides. We don't have the meeting... The meeting password is the word genetic with a capital G. Success. Thank you. Use multiples of 81479 because some of the labs said, well, I'm doing lots of genes that aren't in that list. Can I use multiples? And the answer to that was no. The gene is not in the list. Then you can submit a coding change proposal available on the AMA Web site. And you have to be able to have reference to document the clinical validity of that gene with that disorder. You have to draft up. In your submission, you include a clinical vignette and a description of the service. For BRAF, which is a gene test for colorectal carcinoma and the common variant that's usually tested, which is B600E. So in a clinical vignette, a 54-year-old man with metastatic colorectal carcinoma is being considered for targeted therapy with EGFR monoclonal antibodies. Initial molecular studies indicate the tumor does not contain any of the 12 common KRAS mutations. And the tumor-rich sample is submitted for BRAF testing. And then in a description of service, and this is the most common description of service or the most likely description of service. Paraffin is removed. DNA is isolated. It has PCR for the BRAF gene. It goes die-directional sequencing and capillary electrophoresis and pathologists or other qualified healthcare professionals evaluate and provide the report. So once we get a CCP at the AMA, look at it to determine the parameters for analyte assignment. So in Mendelian somatic disorders, we look for a demonstrated relationship between the biomarker and the phenotype. In other words, does it have clinical validity? Do the biomarkers or the SNPs that have an association but not prove, this is more the GWAS type of assays or studies that are done, have not a proven positive effect to the known clinical feed-knife but should have demonstrated clinical usefulness, whether that's such as a high positive predictive value or a high negative predictive value directing therapy and management. We did require at least two U.S. laboratories to perform the analysis, unless it was proprietary or intellectual property issues existed, and that was prior to the Supreme Court decision against Marriott a few years back. Analysis involved greater than 10 variants for unrelated families. So we're really, we're looking for tests that, not the super, super, super rare stuff, but something that you see in multiple unrelated families to prove clinical validity. So in the tier two assignments in the tier two code, not all duplications and deletions got assigned an analysis. It had to be greater than 10% of the disease levels associated with duplications or deletions. We did use professional databases to verify all the information for code assignments. It came along while we were in the middle of this entire process, and then labs started doing multi-gene panels. So the Association for Molecular Pathology submitted a coding change proposal to put multiple genes together into gene panels. They were quantitative genomic sequencing, so some of it included exome testing and whole genome. I don't know how I said genome genomic sequencing testing. It's supposed to be a whole genome sequence analysis. It included report and interpretation, separated the report and interpretation from the analyte, and this is the first time that we did this in the mole path coding set, and that was because at least in whole genome and whole exome testing that you may have to go back and reanalyze the data for a different genetic disorder or a different indication. There is a fair amount of work with reanalysis, and we wanted to be able to code for that. So the AMA had multiple open meetings, or at least one open meeting to discuss this with key players or anybody who wanted to come and develop the new codes, which are actually in the 2015 DPD coding book. So here's an example of the new ones that were created. There was an aortic dysfunction, non-syndromic hearing loss, excellent in intellectual disability, inherited colon cancer, fetal chromosome aneuploidy, which is for the non-invasive prenatal testing. We had targeted neoplastic genomic sequences. There was a lot of, nobody was happy with the descriptors when we got done. When we were working on this, I think my email blew up at least four or five times with people making recommendations and suggestions on the neoplastic genomic sequence procedures, and then there was a whole mitochondrial, a whole exome, and a whole genome. The MA codes, CMS announced that the MA codes will be gap filled if the Medicare contract determines that the code is payable under the clinical lab fee schedule. So in the most recent clinical lab fee schedule, there's multiple laws that have been, some of them have been crosswalked and gap filled, and we'll talk about reimbursements. Let's get this question. Why didn't each gene get its own CPT code? And the answer is simply there's not enough numbers in CPT to do that. At the AMA, we did recommend that the gene symbol, the Hugo gene symbol, the Hugo gene be added as a descriptor, especially for the tier two codes. And then another question I often get is, well, can a code be moved from tier one for tier two to tier one? And yes, and that has to be requested by a coding change proposal and approved by the AMA. And actually in the 2016 book, or maybe it wasn't even in the 2015 book, actually I think it is in the 2015 book, there are multiple codes that were actually moved from tier two to tier one. Reimbursement or the pricing process that CMS goes through. I don't know how many people are familiar with that or not, but to give you an idea of what CMS does. So we're available, these new MOLPAS, the new MOLPAS code set was available in the CPT coding book in 2012. But at that, there was really no price or, there was really no price associated with the new code. So labs didn't really want to use them or start using them. So during that sort of, that 2012, that sort of transitional year between 2012 and 2013 when the old stacking codes went away, there was a sort of fight to determine whether the labs are, all those new codes go on the clinical lab fee schedule or the physician fee schedule. The background on that is that all the stacking codes were on the clinical lab fee schedule, but the RUC, which is the Specialty, Society, Relative Value Update Committee, recommended that all the new codes go on the physician fee schedule. When you looked at that, there's federal laws that are related to physician practice, whether it's MD versus PhD, co-pays, anti-kickback rules, and physician signature requirements. So CMS decided to place all the new codes in the Tier 2 code on the clinical lab fee schedule. If you're looking at physician practice, whether it's physician fee schedule or clinical lab fee schedule, actually in the federal law, for physician pathology services, the carrier paid for pathology services burnished by a physician to an individual beneficiary on the fee schedule basis only if the services meet the following conditions. Whether it's a surge path service, a specific cytopath, hematology or bug banking services, a clinical consultation service, and clinical laboratory interpretive services that meet the requirements. Part of the issue is between this is whether sort of the fight between MDs and PhDs in reviewing mole path tests. So a lot of this is done by PhDs and not by MDs. So it was sort of a fight of MDs versus PhDs. So if on the physician fee schedule, labs would have been required to collect a 20% co-pay. At this time, everything on the clinical lab fee schedule, there are no co-pays. So this would have been something completely new for laboratories to do and not to say not sometime in the future that labs may have to do co-pays, but this is not something that's currently being done. There's also special signature rules that are not required of clinical laboratory tests regarding physician kickback and purchasing for clinical laboratory tests. Pathology tests are paid on a different and much lower fee schedule in the outpatient setting than they are on the clinical lab fee schedule. And then indirect costs would be assigned on the basis of all pathology, indirect costs and hospital costs. The concern around physicians not being able to, who do actually review and sign out mole path test codes, the mole path test codes. CMS did create a 4-5-2 for the 2013 code book, and this allowed physicians or MDs to bill for the interpretation of reporting services that go beyond the technical reporting of the test. It cannot be billed by non-physicians, geneticists or other lab personnel. And in 2013, and I'm not sure that it's changed at this time, it was reimbursed at 1871 per test. This is to determine reimbursement. One is crosswalk and gap fill. And I have to tell you, I was not that familiar on gap fill before mole path happened. So for crosswalking, if the test is comparable to an existing test, CMS sets reimbursement of the new test to the existing test. And then it's assigned a local fee and corresponding national limitation amount. So while everybody was fighting in 2012 when the new mole path codes were in the CPT code book along with the stacking codes, CMS could have crosswalked the stacking codes to the mole path codes. But there was too much fighting whether it went on the physician fee schedule or the CLIN lab fee schedule. So it wasn't crosswalked at that time. So in 2013, then, because they were new codes, then CMS determined that they should be gap filled. And in gap filling, CMS determines that there's no adequate comparable. Before the new mole path code set came along, and it was around 110, 109 CPT codes, the contractors had never done that many gap fills at one time. So when Medicare carriers or contractors are instructed to gap fill, it's an empirical process based on local pricing patterns. And the medical directors among the contractors can meet and share information, but they can't reach consensus. The gap filling actually, the process takes a year. So particularly during the first three months, the contractors get the amount. And then around April 30, CMS posts an interim contract-specified amount online. It allows a 60-day, by law, you have a 60-day comment period on interim amounts, which essentially may in June. And then CMS posts the final contractor and limitation amounts online, like mid-summer time. Usually it ends up being around August. Then it allows for reconsideration for about 30 days, which goes into October-ish tongue-in-line. And then the final national limitation amounts are made effective January 1st for the entire country. So it's a long process. I'm not going to go in this in too much detail, but I will. I provided it more for your information. So here's the gap fill to your one code. But if you're really super familiar with the MOLPAS CPT code, there's actually genes missing. So one of the genes I had mentioned previously was cystic fibrosis. Cystic fibrosis is not in the Tier 1, is not in Tier 1 here, along with some various other genes. And that is because primarily cystic fibrosis does not affect the Medicare population or the 65 and over crowd. So CMS didn't set a price for cystic fibrosis and various other codes, or just codes in the MOLPAS tier 1 code set. So this is sort of an example. Again, more of the codes. We have some Lynch syndrome codes, more Lynch syndrome, TMB cell gene rearrangements, and HLA. In the 2015 book, as I mentioned previously, the Genomic Sequencing Procedures, AMA created the Genomic Sequencing Procedure codes or the GSP codes. In all of 2015, CMS had instructed the contractors to gap-fill those. And that process actually was just, I think, last week, CMS announced the, and this was between once prior to giving this slide set to Blue Cross. Last week, CMS just announced the gap-filler rights for the GSP codes. So for your favorite area in the United States where you are, I'm here in Indiana, so my Medicare contractor is Wisconsin Physician Services. And then I have a nice, this was the most recent chart of the regions that the Medicare administrative contractors cover. This does change at times, depending on the contracts. The contracts are renewable and does change over time. Sort of in summary is many of the MACs in the gap-fill rates seem to have coordinated their gap-fill rates for Mulpath, though some of the MACs, such as Palmetto, have established rates for all the codes including Tier 2 codes, but CMS did not include them in their release. I do know that I just saw a local coverage decision proposal from WPS for some Tier 2 analyses. And this is one of the first contractors I've seen do Tier 2 descriptors besides Palmetto. CMS hasn't finalized any reimbursement for any of the Tier 2 codes, and each individual Medicare contractor will continue to establish pricing for that. I don't know, I'm sure for Blue Cross Blue Shield as well as the Mulpath community, the complete revision has had a huge impact on reimbursement for molecular pathology. That is the end of my talk.