- Email: [email protected]
Clinical Genomic Profiling of a Diverse Array of Oncology Specimens at a Large Academic Cancer Center
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
The Journal of Molecular Diagnostics, Vol. -, No. -, - 2016
jmd.amjpathol.org
Clinical Genomic Profiling of a Diverse Array of Oncology Specimens at a Large Academic Cancer Center Identification of Targetable Variants and Experience with Reimbursement Q22
Anthony N. Sireci, Vimla S. Aggarwal, Andrew T. Turk, Tatyana Gindin, Mahesh M. Mansukhani, and Susan J. Hsiao From the Department of Pathology and Cell Biology, Columbia University Medical Center, New York, New York Accepted for publication October 5, 2016.
Q3
Q4
Address correspondence to Susan J. Hsiao, M.D., Ph.D., Department of Pathology and Cell Biology, Columbia University Medical Center, 630 W 168th St, P&S17-414B, New York, NY 10032. E-mail: [email protected].
Large cancer panels are being increasingly used in the practice of precision medicine to generate genomic profiles of tumors with the goal of identifying targetable variants and guiding eligibility for clinical trials. To facilitate identification of mutations in a broad range of solid and hematological malignancies, a 467-gene oncology panel (Columbia Combined Cancer Panel) was developed in collaboration with pathologists and oncologists and is currently available and in use for clinical diagnostics. Herein, we share our experience with this testing in an academic medical center. Of 255 submitted specimens, which encompassed a diverse range of tumor types, we were able to successfully sequence 92%. The Columbia Combined Cancer Panel assay led to the detection of a targetable variant in 48.7% of cases. However, although we show good clinical performance and diagnostic yield, thirdparty reimbursement has been poor. Reimbursement from government and third-party payers using the 81455 Current Procedural Terminology code was at 19.4% of billed costs, and 55% of cases were rejected on first submission. Likely contributing factors to this low level of reimbursement are the delays in valuation of the 81455 Current Procedural Terminology code and in establishing national or local coverage determinations. In the absence of additional demonstrations of clinical utility and improved patient outcomes, we expect the reimbursement environment will continue to limit the availability of this testing more broadly. (J Mol Diagn 2016, -: 1e11; http://dx.doi.org/10.1016/ j.jmoldx.2016.10.008)
The availability and accessibility of next-generation sequencing technologies, combined with the identification of increasing numbers of driver mutations from large-scale cancer sequencing projects, has led to evolving needs in the practice of oncology and precision medicine. Laboratory diagnostic tests that can identify actionable or targetable variants are routinely being incorporated into clinical practice and are moving beyond small panels that can identify well-established targets toward larger cancer panels that can guide eligibility for current and future clinical trials.1e4 Implementing a large cancer panel in a Clinical Laboratory Improvement Amendmentecertified and College of
American Pathologistseaccredited clinical laboratory is challenging in many aspects. Test design and clinical validation, development of clinical expertise, and acquisition of the genomics and bioinformatics resources and infrastructure required are just some of the challenges and hurdles that clinical laboratories face.5e8 Equally important, but less well described, is the challenge of providing a highly demanded test that provides clinically useful information in Supported by institutional funding from the Department of Pathology and Cell Biology, Columbia University Medical Center. Disclosures: None declared.
Copyright ª 2016 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2016.10.008
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 Q2 Q1 123 124
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186
Sireci et al the current uncertain coding, reimbursement, and revenue cycle duration environment.9,10 Multigene panels performed on tumor tissue (solid or hematolymphoid) covering >50 genes inclusive of copy number and translocations are coded using a single Current Procedural Terminology (CPT) code (81455). This CPT code was adopted starting January 1, 2015, by the American Medical Association under Genomic Sequencing Procedures.11 Before this, laboratories performing this testing, including our own, relied on listing codes for specific genes. Adoption of the 81455 code introduced uncertainty with regard to reimbursement because, unlike the prior gene-specific codes, the 81455 code had not yet been valued on the Clinical Laboratory Fee Schedule (CLFS) nor had any coverage determinations been rendered. Therefore, it was it was unclear what, if anything, Medicare and, as a result, commercial payers, would reimburse for this code. Past experience with introduction of new molecular CPT codes (ie, the introduction of tier 1 and 2 codes in 2012) suggested that Medicare administrative contractors and other payers could be inconsistent with payment decisions, leading to inadequate reimbursement and extended time to reimbursement.12 Given these challenges and uncertainties, many clinical laboratories have been cautious to develop and offer large cancer genomics tests. We report herein our experience with the Columbia Combined Cancer Panel (CCCP), in the hopes that our experience can provide some guidance to other clinical laboratories. The CCCP test is a 467-gene cancer panel, developed in collaboration with institutional oncologists and pathologists to interrogate cancer genes implicated in a broad range of solid and hematological tumors. This test was designed, developed, and validated in our Clinical Laboratory Improvement Amendmentecertified and College of American Pathologistseaccredited laboratory and is approved by the New York State Department of Health. We report both on clinical performance and diagnostic yield of this assay, as well as our experience with reimbursement with the 81455 CPT code. We find, despite the ability to detect targetable variants in a large proportion of clinical cases, reimbursements are lower than those for more traditional molecular assays, threatening access to cancer genomic profiling for some patients.
This study was approved by our institutional review board. All consecutive cases from July 2014 through December 2015 were included in this study. Tumor samples included formalin-fixed, paraffin-embedded tissue, cell blocks from fine-needle aspirates, peripheral blood, and bone marrow aspirates. Hematoxylin and eosinestained sections or flow cytometry reports were examined by a pathologist and assessed for tumor cell content. For formalin-fixed, paraffin-embedded tissue, manual macrodissection and
2
jmd.amjpathol.org
Tumor Specimens and DNA Extraction
Q6
Sequencing, Variant Calling, and Interpretation Genomic DNA (50 to 200 ng) was sheared using a Covaris S2 Sonication system (Covaris, Woburn, MA), and targeted sequences from 467 genes were captured using custom Agilent (Santa Clara, CA) SureSelect capture reagents. Sequencing was performed on the Illumina (San Diego, CA) HiSeq2500 as 2 100-bp paired-end reads. Analysis of resulting sequences was performed using NextGENe software (Softgenetics, State College, PA). The Q7 FASTQ files were demultiplexed and filtered on the basis of their quality metrics and converted into FASTA files. Samples with at least 6 Gb of data were used for mapping and variant calling. A minimum average coverage of at least 500-fold, as well as at least 50-fold coverage of >98.0% of coding sequences in the region of interest, was obtained on all samples. The reads were aligned to human genome reference sequence GRCh37, and variants were identified. For all variants, variants were called if the mutant allele was present at a minimum of 10% variant allelic fraction, and seen in a minimum of three variant reads. Single-nucleotide variants, and small insertions and deletions, were annotated by an in-house developed pipeline and were evaluated by a molecular pathologist. Vari- Q8 ants were filtered by several criteria, including whether the variant was a known disease-associated mutation listed in the Catalogue of Somatic Mutations in Cancer database,13 by the effect on protein (synonymous, nonsynonymous, nonsense, canonical splice site, or frameshift variants), and by presence in the 1000 Genomes,14 Exome Variant Server (National Heart, Lung, and Blood Institute Gene Ontology Exome Sequencing Project, Seattle, WA, http://evs.gs. washington.edu/EVS, last accessed), or internal databases. Q9 Prediction of functional effects of missense substitutions was performed using the in silico algorithm Provean.15 Variants (minimum variant allelic fraction of 10%) were reported using the following tiered system (developed in consultation with institutional oncologists and pathologists): tier 1, known actionable mutations in the patient’s tumor type; tier 2, known actionable variants in other tumor types and mutations in well-established cancer genes; tier 3, other pathogenic mutations (ie, mutations that are predicted to be loss of function, such as nonsense mutations, canonical splice site mutations, or frameshift mutations, in genes on the panel in which the role in cancer is not well established); and tier 4, variants of uncertain significance. Variants that were considered to be benign or likely benign were not reported. Actionable mutations were
Materials and Methods
Q5
microdissection was performed on unstained sections to enrich for tumor cells (minimum 30% to 40% tumor cells). Genomic DNA was extracted from paraffin tissue using a QIAcube (Qiagen, Hilden, Germany), and from peripheral blood or bone marrow using a QIASymphony (Qiagen) instrument. DNA quantification was performed using a Qubit 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA).
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
NGS Cancer Targetability/Reimbursement confirmed by an orthologous method (Sanger sequencing or real-time PCR).
Analysis of Reimbursement All claims submitted between January 1, 2015, and December 31, 2015, using the CPT code 81455 and with payment decision information available at the time of writing were included. These claims included cases with dates of service on or after January 1, 2015, and before December 31, 2015. Information on the insurance carrier, primary diagnosis (International Classification of Diseases 9 or 10), total reimbursement per claim, reasons of denial/ rejection, and time from claim submission to first reimbursement decision was collected. For comparison purposes, reimbursement, carrier information, rejection reason, and time-to-reimbursement decision were also collected for cases coded using CPT code 81235 (tier 1 code for EGFR) over the same time period. Reimbursement is reported as a percentage of charges in all cases in observance of contractual obligations with third-party payers. Charges for CCCP were developed to be consistent and competitive with similar commercially available panel assays. For EGFR, laboratory charge was guided by the value of the CPT code on the CLFS. Carriers were coded into three categories for analysis: commercial plans, government plans (Medicare/ Medicaid), and managed government plans (ie, Medicaid health matintenance organizations and managed Medicaid/ Medicare plans). To assess the impact on CCCP reimbursement of other molecular tests billed on the same specimen in the same year (ie, cobilled tests), a manual review of all cases was performed to identify cases with cobilled molecular tests. Molecular tests considered included the following: microsatellite instability testing, MGMT promoter methylation testing, EGFRvIII testing, and single gene analysis of EGFR/KRAS/BRAF/JAK2. Time-to-reimbursement decision was measured in calendar days and was available on a subset of reimbursed cases.
Results Performance Characteristics of the CCCP Clinical Diagnostic Test From clinical launch of the CCCP test (July 2014) through the end of December 2015, a period of 18 months, a total of 255 cases (which includes both solid tumors and hematological malignancies) were submitted for testing. Overall, the success rate was high, with 92% (234/255) of cases successfully sequenced with findings reported in the electronic medical record. The remaining 8% of cases were not successfully sequenced because of either insufficient DNA (6%) or poor quality (low average coverage, poor coverage of the targeted region, and/or large number of lowconfidence variant calls) sequencing results (2%) ½F1 (Figure 1A). The most frequently submitted tumor tissue
Q10 Q11
The Journal of Molecular Diagnostics
-
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 Mutation Detection in CCCP Samples 335 336 The CCCP clinical diagnostic test, covering an exonic 337 sequence of 467 genes, was designed to broadly apply to 338 multiple solid tumor and hematological malignancy types. A 339 diverse array of tumor types was submitted for testing 340 (Figure 2A). These tumors included lung (14%), hemato- ½F2 341 logical malignancies (12%), central nervous system (10%), 342 pancreas (9%), gastrointestinal (luminal gastrointestinal, 343 excluding colon) (8%), melanoma (7%), carcinoma of 344 345 unknown primary (6%), colon (5%), bone and soft tissue 346 (5%), breast (5%), gynecological (4%), genitourinary (4%), 347 and liver (3%). An additional 8% of cases, because of low 348 numbers of individual cases, were categorized together as 349 other and included neuroendocrine tumors, mesothelioma, 350 and meningioma. 351 On average, approximately 7450 single-nucleotide vari352 ants and 1550 insertion/deletion variants were detected in 353 each CCCP sample that was successfully sequenced. These 354 variants were filtered by an in-house developed pipeline, 355 which included several criteria, such as whether the variant 356 357 was a known disease-associated mutation listed in the 358 Catalogue of Somatic Mutations in Cancer database, by the 359 effect on protein (synonymous, nonsynonymous, nonsense, 360 canonical splice site, or frameshift variants), and by pres361 ence in the 1000 Genomes, Exome Variant Server, or in362 ternal databases. After evaluation by a molecular Q12 363 pathologist, variants were selected and categorized into tiers 364 (based on actionability and pathogenicity) for inclusion in 365 the report. When analyzed by tumor type, a range of 4.7 to 366 10.5 variants across tiers 1 to 4 were reported per case 367 (Figure 2A). Breast and central nervous system tumors had 368 369 fewer reported mutations, and the highest rate of variants 370 reported per case was seen in the carcinoma of unknown 371 primary group of tumors. However, this was driven largely 372
was formalin-fixed, paraffin-embedded biopsy material, which accounted for 45% (115/255) of cases. Tumor tissue resulting from surgical resections was the next most common sample type (37% of cases), and bone marrow/blood (11%) and cell blocks from fine-needle aspirate (FNA)/body fluid (7.5%) samples composed the remainder of the samples (Figure 1B). Sequencing success rates, when evaluated by sample type, remained high, but ranged from 78.9% to 93.6%. The highest success rate was seen in resection specimens, and lower success rates were seen in cell blocks of FNA/body fluid samples (Figure 1B). In the FNA/body fluid samples, 21% of the samples showed insufficient amounts of DNA. FNA/body fluid samples typically have more limited tumor tissue than other specimen types and are more likely to yield less DNA. However, although the sample size was small for FNA/body fluid samples, there were no sequencing failures among the FNA/body fluid sample group, suggesting that when sufficient amounts of material are available, the DNA extracted from these samples is of high quality and amenable to sequencing.
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
3
print & web 4C=FPO
373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434
Sireci et al
Figure 1 Sequencing success rates and sample types for the CCCP assay. A: Percentages of cases that were successfully sequenced, failed sequencing, or were insufficient for sequencing. B: Percentage of cases that were successfully sequenced, failed sequencing, or were insufficient for sequencing by specimen type. FNA, fine-needle aspirate.
Q13
by a hypermutated case with 62 variants. Excluding this case from the analysis yields an average of 6.2 reported variants per case for carcinomas of unknown primary. Across all tumor types, an average of 6.9 variants were reported per case. Examination of the 20 most recurrently mutated genes shows frequent mutations in tumor-suppressor genes, such as TP53, APC, NF1, SMAD4, RB1, CDKN2A, and BAP1 (Figure 2C). With the exception of mutations in APC, which were primarily seen in colorectal tumors and lung tumors, the variants in the remaining tumor suppressors were distributed across many of the different tumor types. Activating oncogenic mutations, such as those in KRAS
or EGFR, typically showed a more limited tumor type distribution.
4
jmd.amjpathol.org
Targetable Variants Detected The CCCP clinical diagnostic test yields reportable variants across multiple tumor types; of these reportable mutations, 25.3% of variants were driver mutations. A driver mutation was identified in 82.1% of cases. To address the significance of the reportable variants, the proportion of targetable variants was examined. For this analysis, targetable variants were defined as those that may provide information regarding therapeutic responses, and did not include variants that solely
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 Figure 2 Tumor types and variant detection in the CCCP assay. A: Tumor types tested with the CCCP assay, broken down into the following categories: lung, 596 hematological malignancies, central nervous system (CNS), pancreas, gastrointestinal (GI; luminal GI, other than colon), melanoma, carcinoma of unknown 597 primary, colon, bone and soft tissue, breast, gynecological (Gyn), genitourinary (GU), liver, and other (which includes neuroendocrine tumors, mesothelioma, 598 Q20 and meningioma). B: Average number of variants reported per case, reported by tumor category. C: Twenty most recurrently mutated genes across all tumor 599 categories, with breakdown by tumor type. 600 601 provided information regarding diagnosis or prognosis. Analysis of Reimbursement by Payer 602 Pathogenic or likely pathogenic variants that met one of the 603 The CCCP clinical diagnostic test was universally billed following criteria were included: i) predicted response to 604 605 using the CPT code 81455 (currently not valued on CLFS) approved targeted therapy, ii) predicted response to approved 606 targeted therapies for other tumor types, or iii) predicted from January 1, 2015, through December 31, 2015. A total of 607 sensitivity or resistance of an investigational drug in pre153 of 175 claims submitted to third-party payers (commer608 clinical or clinical studies (Supplemental Tables S1 and S2). cial insurers, managed government plans, or government 609 By this analysis, and across all tumor types, 8.9% of the payers) in 2015 had coverage data available. Of billed cases, 610 reportable variants would currently be considered potentially 52% went to commercial insurers, 30% to Medicare, and the 611 remaining 18% to managed government plans (Figure 4A). ½F4 612 ½F3 targetable (Figure 3A). However, as multiple variants are The average reimbursement for CPT code 81455 across all typically reported per case (on average, 6.9 variants/case), this 613 payer types was 19.4% of billed charges for this code (range, 8.9% of variants that are targetable translates to 48.7% of 614 cases harboring at least one targetable variant. By tumor type, 0% to 100%). Commercial insurers and managed government 615 the percentage of cases harboring at least one targetable plans (ie, Medicaid health maintenance organization) reim616 variant showed a wide range of 21.1% to 77.8%, suggesting bursed at a higher rate (eg, 25% and 37%, respectively) Q14 617 618 that some tumor types are more likely to show targetable compared to 0% reimbursement from government payers 619 variants (Figure 3B). (ie, Medicare) (Figure 4B). 620 print & web 4C=FPO
497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558
NGS Cancer Targetability/Reimbursement
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
5
print & web 4C=FPO
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
Sireci et al
Figure 3
Targetable variants detected by the CCCP. A: Percentage of targetable variants detected (left panel) and percentage of cases with at least one targetable variant (right panel). B: Percentage of cases with at least one targetable variant by tumor type. CNS, central nervous system; GI, gastrointestinal; GU, genitourinary; GYN, gynecological.
For comparison to a more widely accepted CPT code, we compared reimbursement of the 81455 code to the CPT code for EGFR testing (81235). National guidelines and practice recommendations support the testing of EGFR, and this code has a value on the CLFS16 (National Comprehensive Cancer Network Non-Small Cell Lung Cancer Guidelines 2015, http://www.nccn.org/professionals/physic ian_gls/f_guidelines.asp#nscl, last accessed April 26, 2016.). The payer mix for this code was 37% commercial, 21% managed government plans, and 42% government payers (Figure 4C). During the same period of time, the average reimbursement across all payers was 36.8% of charges (n Z 231). Similar to the 81455 code, commercial insurers and managed government plans reimbursed at higher rates (44% and 45%, respectively). Government payers reimbursed 27% of charges (n Z 97) (Figure 4D). Next, we sought to identify the effect (if any) cobilling of a molecular code would have on reimbursement for the CCCP test. A total of 136 CCCP cases had data available on other molecular tests performed on the same specimen within the same year; 40 of the 136 CCCP cases (29%) were cobilled with another molecular test. The average reimbursement for cobilled cases was similar to the average reimbursement in which CCCP was billed alone (21.6% in cobilled cases versus 19.8% in CCCP alone cases); thus, we find no evidence of a negative impact of cobilling on CCCP reimbursement (data not shown).
Rejection Rates and Reasons for Rejection
6
jmd.amjpathol.org
Of the 153 claims submitted to third-party payers, 55% were rejected on the first submission (Figure 5A). The reasons for ½F5 denial are shown in Figure 5B. The major reason for denial, accounting for 82% of cases, was that the coded procedure is a noncovered service. This was driven by one commercial payer and Medicare. In comparison, the rejection rate for the EGFR CPT code (81235) during the same period of time was much lower, at 14% of claims (Figure 5C). The major reason for denial was similar to the 81455 CPT code, with 50% of these rejections being coded as noncovered service (Figure 5D).
Time from Claim Submission to First Reimbursement Decision A prolonged revenue cycle with delayed reimbursement may be an additional hurdle for clinical laboratories to face. To address the question of whether the as-yet unvalued on CLFS 81455 CPT code had a prolonged revenue cycle, we examined the average amount of time from claim submission to communication of an initial decision on reimbursement from the payer in the subset of cases in which the information was available. For the 81455 code, the average amount of time was 25.5 days (n Z 118 for CCCP). By comparison, the average time to reimbursement decision for the more commonly used EGFR code was 27 days (n Z 174). The distribution of lag time to reimbursement is
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744
print & web 4C=FPO
Figure 4 Payer mix and average reimbursements for CPT code 81455 compared to 81235. Distribution of payers for cases coded under code 81455 (A) and the average reimbursement as a percentage of total charges for code 81455 (B). Distribution of payers for cases coded under the epidermal growth factor receptor code 81235 (C) and the average reimbursement as a percentage of total charges for code 81235 (D). n Z 153 (A); n Z 231 (C). ½F6
summarized in Figure 6. These results suggest no significant difference in the amount of time before the first reimbursement decision.
Discussion We show herein the clinical performance and diagnostic yield of our CCCP assay, as well as our experience with
reimbursement over a 12-month period. Our assay performed well, and we were able to successfully sequence 92% of cases, and identify targetable variants in 48.7% of cases. The overall high success rate and low rate of insufficient cases (6%) across the range of tumors and specimens (procured through a variety of methods) highlight the robustness of the assay. A likely contributing factor to our success rate is the involvement of pathologists in selection and review of the material submitted for testing, which
Figure 5 Summary of rejection rates and denial reasons for CPT code 81455 compared to 81235. Payment denial rates under CCCP for 81455 (A) and summary of denial reasons provided by third-party payers for 81455 (B). Payment denial rates under epidermal growth factor receptor for 81235 (C) and summary of denial reasons provided by third-party payers for 81235 (D). n Z 153 (A); n Z 231 (C).
print & web 4C=FPO
745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806
NGS Cancer Targetability/Reimbursement
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
7
807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868
ensures that sufficient material is available for testing, that the most clinically appropriate material is selected when multiple blocks or specimens are available for a patient, and that sufficient tumor cellularity is present to avoid potential false-negative results. Our slightly lower success rate for FNA specimens, which are typically the specimens with more limited tissue, suggests a potential area for improvement. FNA procedures are less invasive, and are more commonly becoming the procedure of choice to obtain material for molecular testing, especially in patients with advanced disease who are not candidates for invasive procedures. Closer collaboration with our cytopathologists and/ or potential optimization of rapid on-site evaluation utilization by interventional radiologists or other practitioners who collect FNA samples may be helpful in further increasing sequencing success rates and identifying targetable variants for more patients, and will be the subject of future studies. Our overall targetability rate of 48.7% is comparable with large cancer gene panels from other institutions or laboratories.17,18 One caveat in comparing targetability/actionability across different cancer assays is the lack of a universally accepted definition or metric of clinical actionability/targetability.19 Various definitions of actionability and targetability have been reported in the literature. Actionability may refer to changes in clinical management from variants that provide diagnostic, prognostic, or therapeutic information, or may refer to variants that predict response to therapy. Targetability may refer to variants that respond to an approved therapy or may refer also to those variants that are predicted to confer response to an investigational therapy. For the sake of this analysis, we chose to use a definition of targetability/ actionability similar to that used by some other institutions and laboratories.17,18 However, standardizing measures of targetability/actionability in addition to clinical trials on clinical outcomes to establish the clinical utility of large, multigene oncology molecular diagnostics will be crucially
helpful in improving the current coverage and reimbursement for this testing.20e22 Uncertainty around coverage and reimbursement for CCCP testing threatens the accessibility of this clinically useful testing to our patient population and limits more general adoption.9,23 The terms coverage and reimbursement are often used interchangeably, but refer to distinct parts of the third payer payment process. Coverage describes the range of testing a payer will pay for and under what clinical scenario. Reimbursement, on the other hand, refers to the amount a payer will pay for a covered service.24 Decisions on coverage and reimbursement made by Medicare affect all insured patients because many commercial carriers and Medicaid will base their coverage policies and rates on the CLFS.9 Our reimbursement for the 81455 CPT code is significantly lower as a percentage of charges than the more established tier 1 code for EGFR (81235) and, in addition, the rejection rate is higher. Cobilling of other molecular tests with CCCP did not appear to affect CCCP reimbursement. The low reimbursement is driven by 0% reimbursement from Medicare for 81455, whereas the same payer covers the EGFR code 81235 at 27% of charges. The high rejection rate is skewed by Medicare’s 100% denial of claims for the 81455 code compared to a 14% denial rate for the EGFR code. The major reason given for denial of the 81455 code is noncovered service. Our experience illustrates problems in coverage and reimbursement resulting from failures in both valuation of the 81455 CPT code and in establishing national or local coverage determinations. The EGFR 81235 code is covered by Medicare, and its reimbursement is valued on the CLFS (Centers for Medicare and Medicaid Services, US Department of Health and Human Services, CLFS 2016, https:// www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ ClinicalLabFeeSched, last accessed April 11, 2016). In contrast, the 81455 code for gene panels >50 genes remains as-of-yet unvalued and a coverage policy had not yet been officially defined on the national or local level by Medicare in 2015. These two factors contribute to our institutional reimbursement experience with 81455, and until the 81455 CPT code is valued and coverage determinations are made, coverage and reimbursement for this CPT code will likely remain poor and highly variable. The typical process for valuation of a new CPT code after it is adopted by the American Medical Association on the Physician Fee Schedule involves a rigorous and transparent valuation process by the Relative Value Scale Update Committee, which makes recommendations on value to the Centers for Medicare and Medicaid Services.25 For codes adopted on the CLFS, such as all tier 1 and 2 and genomic sequencing procedures, there are two options for preliminary valuation: crosswalk and gap fill.25 Under the crosswalk approach, the value of a new CPT is based on the value of an existing and related code. Under gap fill, each individual Medicare administrative contractor sets
8
jmd.amjpathol.org
print & web 4C=FPO
869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930
Sireci et al
Analysis of time from filing to reimbursement decision for 81455 compared to 81235. The distribution of time from invoice generation to first reimbursement decision is divided into 30-day buckets, and both epidermal growth factor receptor (EGFR) and CCCP are depicted.
Figure 6
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992
993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
NGS Cancer Targetability/Reimbursement
Q15 Q16
pricing for a 1-year period on the basis of charges, required resources for a service, and payment by other payers.12 In year 2, the median value across Medicare administrative contractors is used to set the value for the code on the CLFS. The lack of reimbursement by Medicare for the 81455 CPT code before valuation on CLFS may be an example of a breakdown in the gap fill process. Even when the gap-fill mechanism for valuation was applied to molecular testing, such as in 2013, after the American Medical Association adopted tier 1 and tier 2 molecular CPT codes (EGFR is a tier 1 code), this mechanism is imperfect. Reimbursement for tier 1 and 2 molecular CPT codes was incomplete and inconsistent, the industry complained of lack of clarity in the process, and several Medicare administrative contractors simply adopted values from one contractor.12 During this prolonged decision-making process (January 2013, to September 30, 2013), many codes went unpaid.26 In addition to valuation, coverage determination for the 81455 CPT code by Medicare has not been adequately addressed. The determination of coverage by Medicare is predicated on proof that the service is medically reasonable and necessary for the Medicare beneficiary population.27 The determination can proceed at the national level through the transparent and well-documented process known as a national coverage determination, or each individual contractor can establish its own policy via a local coverage determination.28 For the entirety of 2015, Medicare did not have a national coverage determination specifically for molecular profiling using multiplex or next-generation sequencing technology to guide cancer treatment(s), and the local Medicare contractor in New York (National Government Service) had not issued an local coverage determination regarding 81455. However, during preparation of this article, on April 1, 2016, National Government Services (the Medicare administrative contractor covering New York) issued a local coverage determination grouping 81455 in with other genomic sequencing procedures, which would be immediately rejected as not medically necessary [Local Coverage Determination: Molecular Pathology Procedures (L35000), https://www.cms.gov/medicare-coverage-database/details/lcddetails.aspx?LCDIdZ35000&ContrIdZ297, last accessed April 11, 2016]. In contrast, a recent coverage policy by United Health Care allowed for coverage of the 81455 code in the setting of nonesmall-cell lung cancer based mainly on National Comprehensive Cancer Network guidelines and the availability of targeted therapies in this disease (Molecular Profiling to Guide Cancer Treatment, United Health Care Medical Policy Number 2016T0576B, https://www. unitedhealthcareonline.com/ccmcontent/ProviderII/UHC/enUS/Assets/ProviderStaticFiles/ProviderStaticFilesPdf/Tools %20and%20Resources/Policies%20and%20Protocols/Medi cal%20Policies/Medical%20Policies/Molecular_Profiling. pdf, last accessed April 11, 2016).
The Journal of Molecular Diagnostics
-
Coverage policies for new codes are not keeping pace with the adoption of novel diagnostics into clinical practice, and as a result, these policies can be inconsistent or variable. In part, this is because of the disconnect between the pace of new targeted therapy discoveries and development of clinical assays, and the requirements payers have for establishing coverage (ie, rigorous determinations of clinical utility).29 Clinical utility is defined as “the evidence of improved measureable clinical outcomes, and its usefulness and added value to patient management decision-making compared with the current management.”30 Examples of studies establishing clinical utility Q17 are large randomized clinical trials and are impossible for single laboratories to execute. Our study, which showed 48.7% of cases had a targetable variant, together with other similar studies have consistently shown large cancer panels can identify targetable or actionable variants and have the potential to show clinical utility by guiding therapeutic decision making and clinical management in a significant percentage of cases. The level of evidence recommended in a recent guidance document (which is similar to the level of evidence for supporting EGFR testing) will take a significant amount of resources16,21 (National Comprehensive Cancer Network Non-Small Cell Lung Cancer Guidelines 2015). Evidence of clinical utility is a major predictor of payer coverage for genetic testing procedures.21,31 Alternative approaches for establishing utility besides clinical outcomes (eg, decreased resource utilization resulting from fewer invasive procedures required to obtain predictive therapeutic information), therefore, must be explored to prevent the limiting of access to testing that may provide effective targeted therapies in the meantime. A recent publication by the Association of Molecular Pathology’s Framework for the Evidence Needed to Demonstrate Clinical Utility Task Force addresses exactly this need and expands the meaning of clinical utility to be more patient centered, context dependent, and flexible to alternative study design (eg, retrospective studies).32 In addition, the process of CLFS code valuation will require revision and extensive input from stakeholders if new and disruptive technologies are to be equitably integrated into clinical decision making and therapeutics. Finally, collaborative engagement of key stakeholders, such as oncologists, pathologists, and patients with private and government payers, in determining reimbursement policies and rates at the local and national levels will likely be the key to future progress and enhanced access. The Columbia experience with a large targeted cancer panel, along with the experience of other providers of large cancer panels, has proved that genomic profiling of tumors can result in a relatively high rate for identification of targetable variants. To secure access to this powerful testing, molecular pathologists must collaborate with clinician-oncologists and payers to show that the results of this testing affect medical decision making and ultimately result in superior clinical outcomes. Until that time, an
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
9
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178
Sireci et al
1. Abrams J, Conley B, Mooney M, Zwiebel J, Chen A, Welch JJ, Takebe N, Malik S, McShane L, Korn E, Williams M, Staudt L, Doroshow J: National Cancer Institute’s Precision Medicine Initiatives for the new National Clinical Trials Network. Am Soc Clin Oncol Educ Book 2014:71e76 2. Lopez-Chavez A, Thomas A, Rajan A, Raffeld M, Morrow B, Kelly R, Carter CA, Guha U, Killian K, Lau CC, Abdullaev Z, Xi L, Pack S, Meltzer PS, Corless CL, Sandler A, Beadling C, Warrick A, Liewehr DJ, Steinberg SM, Berman A, Doyle A, Szabo E, Wang Y, Giaccone G: Molecular profiling and targeted therapy for advanced thoracic malignancies: a biomarker-derived, multiarm, multihistology phase II basket trial. J Clin Oncol 2015, 33:1000e1007 3. Hyman DM, Solit DB, Arcila ME, Cheng DT, Sabbatini P, Baselga J, Berger MF, Ladanyi M: Precision medicine at Memorial Sloan Kettering Cancer Center: clinical next-generation sequencing enabling next-generation targeted therapy trials. Drug Discov Today 2015, 20: 1422e1428 4. Kim ES, Herbst RS, Wistuba II, Lee JJ, Blumenschein GR Jr, Tsao A, Stewart DJ, Hicks ME, Erasmus J Jr, Gupta S, Alden CM, Liu S, Tang X, Khuri FR, Tran HT, Johnson BE, Heymach JV, Mao L, Fossella F, Kies MS, Papadimitrakopoulou V, Davis SE, Lippman SM, Hong WK: The BATTLE trial: personalizing therapy for lung cancer. Cancer Discov 2011, 1:44e53 5. Luthra R, Chen H, Roy-Chowdhuri S, Singh RR: Next-generation sequencing in clinical molecular diagnostics of cancer: advantages and challenges. Cancers 2015, 7:2023e2036 6. Roy S, LaFramboise WA, Nikiforov YE, Nikiforova MN, Routbort MJ, Pfeifer J, Nagarajan R, Carter AB, Pantanowitz L: Nextgeneration sequencing informatics. Arch Pathol Lab Med 2016, 140: 958e975 7. Ulahannan D, Kovac MB, Mulholland PJ, Cazier JB, Tomlinson I: Technical and implementation issues in using next-generation sequencing of cancers in clinical practice. Br J Cancer 2013, 109: 827e835 8. Xuan J, Yu Y, Qing T, Guo L, Shi L: Next-generation sequencing in the clinic: promises and challenges. Cancer Lett 2013, 340: 284e295 9. Deverka PA, Dreyfus JC: Clinical integration of next generation sequencing: coverage and reimbursement challenges. J Law Med Ethics 2014, 42 Suppl 1:22e41 10. Trosman JR, Weldon CB, Kelley RK, Phillips KA: Challenges of coverage policy development for next-generation tumor sequencing panels: experts and payers weigh in. J Natl Compr Canc Netw 2015, 13:311e318
11. CPT 2016 Professional Edition. Chicago, American Medical Association, 2016 12. Klein RD: Reimbursement in molecular pathology: bringing genomic medicine to patients. Clin Chem 2015, 61:136e138 13. Forbes SA, Beare D, Gunasekaran P, Leung K, Bindal N, Boutselakis H, Ding M, Bamford S, Cole C, Ward S, Kok CY, Jia M, De T, Teague JW, Stratton MR, McDermott U, Campbell PJ: COSMIC: exploring the world’s knowledge of somatic mutations in human cancer. Nucleic Acids Res 2015, 43:D805eD811 14. 1000 Genomes Project Consortium, Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, Korbel JO, Marchini JL, McCarthy S, McVean GA, Abecasis GR: A global reference for human genetic variation. Nature 2015, 526:68e74 15. Choi Y, Chan AP: PROVEAN web server: a tool to predict the functional effect of amino acid substitutions and indels. Bioinformatics 2015, 31:2745e2747 16. Lindeman NI, Cagle PT, Beasley MB, Chitale DA, Dacic S, Giaccone G, Jenkins RB, Kwiatkowski DJ, Saldivar JS, Squire J, Thunnissen E, Ladanyi M: Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. Arch Pathol Lab Med 2013, 137:828e860 17. Frampton GM, Fichtenholtz A, Otto GA, Wang K, Downing SR, He J, et al: Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol 2013, 31:1023e1031 18. Hovelson DH, McDaniel AS, Cani AK, Johnson B, Rhodes K, Williams PD, Bandla S, Bien G, Choppa P, Hyland F, Gottimukkala R, Liu G, Manivannan M, Schageman J, Ballesteros-Villagrana E, Grasso CS, Quist MJ, Yadati V, Amin A, Siddiqui J, Betz BL, Knudsen KE, Cooney KA, Feng FY, Roh MH, Nelson PS, Liu CJ, Beer DG, Wyngaard P, Chinnaiyan AM, Sadis S, Rhodes DR, Tomlins SA: Development and validation of a scalable next-generation sequencing system for assessing relevant somatic variants in solid tumors. Neoplasia 2015, 17:385e399 19. Dienstmann R, Jang IS, Bot B, Friend S, Guinney J: Database of genomic biomarkers for cancer drugs and clinical targetability in solid tumors. Cancer Discov 2015, 5:118e123 20. Kakudo K, Bai Y, Liu Z, Ozaki T: Encapsulated papillary thyroid carcinoma, follicular variant: a misnomer. Pathol Int 2012, 62: 155e160 21. Deverka P, Messner DA, McCormack R, Lyman GH, Piper M, Bradley L, Parkinson D, Nelson D, Smith ML, Jacques L, Dutta T, Tunis SR: Generating and evaluating evidence of the clinical utility of molecular diagnostic tests in oncology. Genet Med 2016, 18: 780e787 22. Liu J, Singh B, Tallini G, Carlson DL, Katabi N, Shaha A, Tuttle RM, Ghossein RA: Follicular variant of papillary thyroid carcinoma: a clinicopathologic study of a problematic entity. Cancer 2006, 107: 1255e1264 23. Mauer CB, Pirzadeh-Miller SM, Robinson LD, Euhus DM: The integration of next-generation sequencing panels in the clinical cancer genetics practice: an institutional experience. Genet Med 2014, 16: 407e412 24. Williams MS: Insurance coverage for pharmacogenomic testing in the USA. Pers Med 2007, 4:479e487 25. Donovan WD: What is the RUC? AJNR Am J Neuroradiol 2011, 32: 1583e1584 26. GenomeWeb. Ray T: Centers for Medicare and Medicaid Services national payment limit for tier 1 codes not as low as before, but challenges remain. New York, NY: GenomeWeb. 2013. Available at http://www.genomeweb.com/clinical-genomics/cms-national-pa yment-limit-tier-1-codes-not-low-challenges-remain (accessed April 11, 2016) 27. Centers for Medicare and Medicaid Services, US Department of Health and Human Services. Publication 100-2, Medicare Benefit Policy
10
jmd.amjpathol.org
expanded definition of clinical utility to drive reimbursement may be the key to improving availability of this testing more broadly.
Acknowledgments We thank Samantha Cano and Vaishali Hodel for technical support; Marissa Pang, Nicholas Rouse, and Stuart Andrews for bioinformatics support; and Melissa Carter for assistance with collection of billing and reimbursement data.
Supplemental Data Supplemental material for this article can be found at http://dx.doi.org/10.1016/j.jmoldx.2016.10.008.
References
Q21
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
Q18
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
NGS Cancer Targetability/Reimbursement
Q19
Manual. Baltimore, MD, CMS, 2014. Available at http://www.cms. hhs.gov/manuals/downloads/bp102c16.pdf (accessed April 11, 2016) 28. Burken MI, Wilson KS, Heller K, Pratt VM, Schoonmaker MM, Seifter E: The interface of Medicare coverage decision-making and emerging molecular-based laboratory testing. Genet Med 2009, 11:225e231 29. Department of Health and Human Services Coverage and Reimbursement of Genetic Tests and Services: Report of the Secretary’s Advisory Committee on Genetics, Health, and Society. Washington, DC, HHS, 2006 30. Teutsch SM, Bradley LA, Palomaki GE, Haddow JE, Piper M, Calonge N, Dotson WD, Douglas MP, Berg AO, Group EW: The
The Journal of Molecular Diagnostics
-
Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Initiative: methods of the EGAPP Working Group. Genet Med 2009, 11:3e14 31. Meckley LM, Neumann PJ: Personalized medicine: factors influencing reimbursement. Health Policy 2010, 94:91e100 32. Joseph L, Cankovic M, Caughron S, Chandra P, Emmadi R, Hagenkord J, Hallam S, Jewell KE, Klein RD, Pratt VM, Rothberg PG, Temple-Smolkin RL, Lyon E: The spectrum of clinical utilities in molecular pathology testing procedures for inherited conditions and cancer: a report of the Association for Molecular Pathology. J Mol Diagn 2016, 18:605e619
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
11
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
The Journal of Molecular Diagnostics, Vol. -, No. -, - 2016
jmd.amjpathol.org
Clinical Genomic Profiling of a Diverse Array of Oncology Specimens at a Large Academic Cancer Center Identification of Targetable Variants and Experience with Reimbursement Q22
Anthony N. Sireci, Vimla S. Aggarwal, Andrew T. Turk, Tatyana Gindin, Mahesh M. Mansukhani, and Susan J. Hsiao From the Department of Pathology and Cell Biology, Columbia University Medical Center, New York, New York Accepted for publication October 5, 2016.
Q3
Q4
Address correspondence to Susan J. Hsiao, M.D., Ph.D., Department of Pathology and Cell Biology, Columbia University Medical Center, 630 W 168th St, P&S17-414B, New York, NY 10032. E-mail: [email protected].
Large cancer panels are being increasingly used in the practice of precision medicine to generate genomic profiles of tumors with the goal of identifying targetable variants and guiding eligibility for clinical trials. To facilitate identification of mutations in a broad range of solid and hematological malignancies, a 467-gene oncology panel (Columbia Combined Cancer Panel) was developed in collaboration with pathologists and oncologists and is currently available and in use for clinical diagnostics. Herein, we share our experience with this testing in an academic medical center. Of 255 submitted specimens, which encompassed a diverse range of tumor types, we were able to successfully sequence 92%. The Columbia Combined Cancer Panel assay led to the detection of a targetable variant in 48.7% of cases. However, although we show good clinical performance and diagnostic yield, thirdparty reimbursement has been poor. Reimbursement from government and third-party payers using the 81455 Current Procedural Terminology code was at 19.4% of billed costs, and 55% of cases were rejected on first submission. Likely contributing factors to this low level of reimbursement are the delays in valuation of the 81455 Current Procedural Terminology code and in establishing national or local coverage determinations. In the absence of additional demonstrations of clinical utility and improved patient outcomes, we expect the reimbursement environment will continue to limit the availability of this testing more broadly. (J Mol Diagn 2016, -: 1e11; http://dx.doi.org/10.1016/ j.jmoldx.2016.10.008)
The availability and accessibility of next-generation sequencing technologies, combined with the identification of increasing numbers of driver mutations from large-scale cancer sequencing projects, has led to evolving needs in the practice of oncology and precision medicine. Laboratory diagnostic tests that can identify actionable or targetable variants are routinely being incorporated into clinical practice and are moving beyond small panels that can identify well-established targets toward larger cancer panels that can guide eligibility for current and future clinical trials.1e4 Implementing a large cancer panel in a Clinical Laboratory Improvement Amendmentecertified and College of
American Pathologistseaccredited clinical laboratory is challenging in many aspects. Test design and clinical validation, development of clinical expertise, and acquisition of the genomics and bioinformatics resources and infrastructure required are just some of the challenges and hurdles that clinical laboratories face.5e8 Equally important, but less well described, is the challenge of providing a highly demanded test that provides clinically useful information in Supported by institutional funding from the Department of Pathology and Cell Biology, Columbia University Medical Center. Disclosures: None declared.
Copyright ª 2016 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2016.10.008
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 Q2 Q1 123 124
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186
Sireci et al the current uncertain coding, reimbursement, and revenue cycle duration environment.9,10 Multigene panels performed on tumor tissue (solid or hematolymphoid) covering >50 genes inclusive of copy number and translocations are coded using a single Current Procedural Terminology (CPT) code (81455). This CPT code was adopted starting January 1, 2015, by the American Medical Association under Genomic Sequencing Procedures.11 Before this, laboratories performing this testing, including our own, relied on listing codes for specific genes. Adoption of the 81455 code introduced uncertainty with regard to reimbursement because, unlike the prior gene-specific codes, the 81455 code had not yet been valued on the Clinical Laboratory Fee Schedule (CLFS) nor had any coverage determinations been rendered. Therefore, it was it was unclear what, if anything, Medicare and, as a result, commercial payers, would reimburse for this code. Past experience with introduction of new molecular CPT codes (ie, the introduction of tier 1 and 2 codes in 2012) suggested that Medicare administrative contractors and other payers could be inconsistent with payment decisions, leading to inadequate reimbursement and extended time to reimbursement.12 Given these challenges and uncertainties, many clinical laboratories have been cautious to develop and offer large cancer genomics tests. We report herein our experience with the Columbia Combined Cancer Panel (CCCP), in the hopes that our experience can provide some guidance to other clinical laboratories. The CCCP test is a 467-gene cancer panel, developed in collaboration with institutional oncologists and pathologists to interrogate cancer genes implicated in a broad range of solid and hematological tumors. This test was designed, developed, and validated in our Clinical Laboratory Improvement Amendmentecertified and College of American Pathologistseaccredited laboratory and is approved by the New York State Department of Health. We report both on clinical performance and diagnostic yield of this assay, as well as our experience with reimbursement with the 81455 CPT code. We find, despite the ability to detect targetable variants in a large proportion of clinical cases, reimbursements are lower than those for more traditional molecular assays, threatening access to cancer genomic profiling for some patients.
This study was approved by our institutional review board. All consecutive cases from July 2014 through December 2015 were included in this study. Tumor samples included formalin-fixed, paraffin-embedded tissue, cell blocks from fine-needle aspirates, peripheral blood, and bone marrow aspirates. Hematoxylin and eosinestained sections or flow cytometry reports were examined by a pathologist and assessed for tumor cell content. For formalin-fixed, paraffin-embedded tissue, manual macrodissection and
2
jmd.amjpathol.org
Tumor Specimens and DNA Extraction
Q6
Sequencing, Variant Calling, and Interpretation Genomic DNA (50 to 200 ng) was sheared using a Covaris S2 Sonication system (Covaris, Woburn, MA), and targeted sequences from 467 genes were captured using custom Agilent (Santa Clara, CA) SureSelect capture reagents. Sequencing was performed on the Illumina (San Diego, CA) HiSeq2500 as 2 100-bp paired-end reads. Analysis of resulting sequences was performed using NextGENe software (Softgenetics, State College, PA). The Q7 FASTQ files were demultiplexed and filtered on the basis of their quality metrics and converted into FASTA files. Samples with at least 6 Gb of data were used for mapping and variant calling. A minimum average coverage of at least 500-fold, as well as at least 50-fold coverage of >98.0% of coding sequences in the region of interest, was obtained on all samples. The reads were aligned to human genome reference sequence GRCh37, and variants were identified. For all variants, variants were called if the mutant allele was present at a minimum of 10% variant allelic fraction, and seen in a minimum of three variant reads. Single-nucleotide variants, and small insertions and deletions, were annotated by an in-house developed pipeline and were evaluated by a molecular pathologist. Vari- Q8 ants were filtered by several criteria, including whether the variant was a known disease-associated mutation listed in the Catalogue of Somatic Mutations in Cancer database,13 by the effect on protein (synonymous, nonsynonymous, nonsense, canonical splice site, or frameshift variants), and by presence in the 1000 Genomes,14 Exome Variant Server (National Heart, Lung, and Blood Institute Gene Ontology Exome Sequencing Project, Seattle, WA, http://evs.gs. washington.edu/EVS, last accessed), or internal databases. Q9 Prediction of functional effects of missense substitutions was performed using the in silico algorithm Provean.15 Variants (minimum variant allelic fraction of 10%) were reported using the following tiered system (developed in consultation with institutional oncologists and pathologists): tier 1, known actionable mutations in the patient’s tumor type; tier 2, known actionable variants in other tumor types and mutations in well-established cancer genes; tier 3, other pathogenic mutations (ie, mutations that are predicted to be loss of function, such as nonsense mutations, canonical splice site mutations, or frameshift mutations, in genes on the panel in which the role in cancer is not well established); and tier 4, variants of uncertain significance. Variants that were considered to be benign or likely benign were not reported. Actionable mutations were
Materials and Methods
Q5
microdissection was performed on unstained sections to enrich for tumor cells (minimum 30% to 40% tumor cells). Genomic DNA was extracted from paraffin tissue using a QIAcube (Qiagen, Hilden, Germany), and from peripheral blood or bone marrow using a QIASymphony (Qiagen) instrument. DNA quantification was performed using a Qubit 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA).
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
NGS Cancer Targetability/Reimbursement confirmed by an orthologous method (Sanger sequencing or real-time PCR).
Analysis of Reimbursement All claims submitted between January 1, 2015, and December 31, 2015, using the CPT code 81455 and with payment decision information available at the time of writing were included. These claims included cases with dates of service on or after January 1, 2015, and before December 31, 2015. Information on the insurance carrier, primary diagnosis (International Classification of Diseases 9 or 10), total reimbursement per claim, reasons of denial/ rejection, and time from claim submission to first reimbursement decision was collected. For comparison purposes, reimbursement, carrier information, rejection reason, and time-to-reimbursement decision were also collected for cases coded using CPT code 81235 (tier 1 code for EGFR) over the same time period. Reimbursement is reported as a percentage of charges in all cases in observance of contractual obligations with third-party payers. Charges for CCCP were developed to be consistent and competitive with similar commercially available panel assays. For EGFR, laboratory charge was guided by the value of the CPT code on the CLFS. Carriers were coded into three categories for analysis: commercial plans, government plans (Medicare/ Medicaid), and managed government plans (ie, Medicaid health matintenance organizations and managed Medicaid/ Medicare plans). To assess the impact on CCCP reimbursement of other molecular tests billed on the same specimen in the same year (ie, cobilled tests), a manual review of all cases was performed to identify cases with cobilled molecular tests. Molecular tests considered included the following: microsatellite instability testing, MGMT promoter methylation testing, EGFRvIII testing, and single gene analysis of EGFR/KRAS/BRAF/JAK2. Time-to-reimbursement decision was measured in calendar days and was available on a subset of reimbursed cases.
Results Performance Characteristics of the CCCP Clinical Diagnostic Test From clinical launch of the CCCP test (July 2014) through the end of December 2015, a period of 18 months, a total of 255 cases (which includes both solid tumors and hematological malignancies) were submitted for testing. Overall, the success rate was high, with 92% (234/255) of cases successfully sequenced with findings reported in the electronic medical record. The remaining 8% of cases were not successfully sequenced because of either insufficient DNA (6%) or poor quality (low average coverage, poor coverage of the targeted region, and/or large number of lowconfidence variant calls) sequencing results (2%) ½F1 (Figure 1A). The most frequently submitted tumor tissue
Q10 Q11
The Journal of Molecular Diagnostics
-
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 Mutation Detection in CCCP Samples 335 336 The CCCP clinical diagnostic test, covering an exonic 337 sequence of 467 genes, was designed to broadly apply to 338 multiple solid tumor and hematological malignancy types. A 339 diverse array of tumor types was submitted for testing 340 (Figure 2A). These tumors included lung (14%), hemato- ½F2 341 logical malignancies (12%), central nervous system (10%), 342 pancreas (9%), gastrointestinal (luminal gastrointestinal, 343 excluding colon) (8%), melanoma (7%), carcinoma of 344 345 unknown primary (6%), colon (5%), bone and soft tissue 346 (5%), breast (5%), gynecological (4%), genitourinary (4%), 347 and liver (3%). An additional 8% of cases, because of low 348 numbers of individual cases, were categorized together as 349 other and included neuroendocrine tumors, mesothelioma, 350 and meningioma. 351 On average, approximately 7450 single-nucleotide vari352 ants and 1550 insertion/deletion variants were detected in 353 each CCCP sample that was successfully sequenced. These 354 variants were filtered by an in-house developed pipeline, 355 which included several criteria, such as whether the variant 356 357 was a known disease-associated mutation listed in the 358 Catalogue of Somatic Mutations in Cancer database, by the 359 effect on protein (synonymous, nonsynonymous, nonsense, 360 canonical splice site, or frameshift variants), and by pres361 ence in the 1000 Genomes, Exome Variant Server, or in362 ternal databases. After evaluation by a molecular Q12 363 pathologist, variants were selected and categorized into tiers 364 (based on actionability and pathogenicity) for inclusion in 365 the report. When analyzed by tumor type, a range of 4.7 to 366 10.5 variants across tiers 1 to 4 were reported per case 367 (Figure 2A). Breast and central nervous system tumors had 368 369 fewer reported mutations, and the highest rate of variants 370 reported per case was seen in the carcinoma of unknown 371 primary group of tumors. However, this was driven largely 372
was formalin-fixed, paraffin-embedded biopsy material, which accounted for 45% (115/255) of cases. Tumor tissue resulting from surgical resections was the next most common sample type (37% of cases), and bone marrow/blood (11%) and cell blocks from fine-needle aspirate (FNA)/body fluid (7.5%) samples composed the remainder of the samples (Figure 1B). Sequencing success rates, when evaluated by sample type, remained high, but ranged from 78.9% to 93.6%. The highest success rate was seen in resection specimens, and lower success rates were seen in cell blocks of FNA/body fluid samples (Figure 1B). In the FNA/body fluid samples, 21% of the samples showed insufficient amounts of DNA. FNA/body fluid samples typically have more limited tumor tissue than other specimen types and are more likely to yield less DNA. However, although the sample size was small for FNA/body fluid samples, there were no sequencing failures among the FNA/body fluid sample group, suggesting that when sufficient amounts of material are available, the DNA extracted from these samples is of high quality and amenable to sequencing.
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
3
print & web 4C=FPO
373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434
Sireci et al
Figure 1 Sequencing success rates and sample types for the CCCP assay. A: Percentages of cases that were successfully sequenced, failed sequencing, or were insufficient for sequencing. B: Percentage of cases that were successfully sequenced, failed sequencing, or were insufficient for sequencing by specimen type. FNA, fine-needle aspirate.
Q13
by a hypermutated case with 62 variants. Excluding this case from the analysis yields an average of 6.2 reported variants per case for carcinomas of unknown primary. Across all tumor types, an average of 6.9 variants were reported per case. Examination of the 20 most recurrently mutated genes shows frequent mutations in tumor-suppressor genes, such as TP53, APC, NF1, SMAD4, RB1, CDKN2A, and BAP1 (Figure 2C). With the exception of mutations in APC, which were primarily seen in colorectal tumors and lung tumors, the variants in the remaining tumor suppressors were distributed across many of the different tumor types. Activating oncogenic mutations, such as those in KRAS
or EGFR, typically showed a more limited tumor type distribution.
4
jmd.amjpathol.org
Targetable Variants Detected The CCCP clinical diagnostic test yields reportable variants across multiple tumor types; of these reportable mutations, 25.3% of variants were driver mutations. A driver mutation was identified in 82.1% of cases. To address the significance of the reportable variants, the proportion of targetable variants was examined. For this analysis, targetable variants were defined as those that may provide information regarding therapeutic responses, and did not include variants that solely
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 Figure 2 Tumor types and variant detection in the CCCP assay. A: Tumor types tested with the CCCP assay, broken down into the following categories: lung, 596 hematological malignancies, central nervous system (CNS), pancreas, gastrointestinal (GI; luminal GI, other than colon), melanoma, carcinoma of unknown 597 primary, colon, bone and soft tissue, breast, gynecological (Gyn), genitourinary (GU), liver, and other (which includes neuroendocrine tumors, mesothelioma, 598 Q20 and meningioma). B: Average number of variants reported per case, reported by tumor category. C: Twenty most recurrently mutated genes across all tumor 599 categories, with breakdown by tumor type. 600 601 provided information regarding diagnosis or prognosis. Analysis of Reimbursement by Payer 602 Pathogenic or likely pathogenic variants that met one of the 603 The CCCP clinical diagnostic test was universally billed following criteria were included: i) predicted response to 604 605 using the CPT code 81455 (currently not valued on CLFS) approved targeted therapy, ii) predicted response to approved 606 targeted therapies for other tumor types, or iii) predicted from January 1, 2015, through December 31, 2015. A total of 607 sensitivity or resistance of an investigational drug in pre153 of 175 claims submitted to third-party payers (commer608 clinical or clinical studies (Supplemental Tables S1 and S2). cial insurers, managed government plans, or government 609 By this analysis, and across all tumor types, 8.9% of the payers) in 2015 had coverage data available. Of billed cases, 610 reportable variants would currently be considered potentially 52% went to commercial insurers, 30% to Medicare, and the 611 remaining 18% to managed government plans (Figure 4A). ½F4 612 ½F3 targetable (Figure 3A). However, as multiple variants are The average reimbursement for CPT code 81455 across all typically reported per case (on average, 6.9 variants/case), this 613 payer types was 19.4% of billed charges for this code (range, 8.9% of variants that are targetable translates to 48.7% of 614 cases harboring at least one targetable variant. By tumor type, 0% to 100%). Commercial insurers and managed government 615 the percentage of cases harboring at least one targetable plans (ie, Medicaid health maintenance organization) reim616 variant showed a wide range of 21.1% to 77.8%, suggesting bursed at a higher rate (eg, 25% and 37%, respectively) Q14 617 618 that some tumor types are more likely to show targetable compared to 0% reimbursement from government payers 619 variants (Figure 3B). (ie, Medicare) (Figure 4B). 620 print & web 4C=FPO
497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558
NGS Cancer Targetability/Reimbursement
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
5
print & web 4C=FPO
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
Sireci et al
Figure 3
Targetable variants detected by the CCCP. A: Percentage of targetable variants detected (left panel) and percentage of cases with at least one targetable variant (right panel). B: Percentage of cases with at least one targetable variant by tumor type. CNS, central nervous system; GI, gastrointestinal; GU, genitourinary; GYN, gynecological.
For comparison to a more widely accepted CPT code, we compared reimbursement of the 81455 code to the CPT code for EGFR testing (81235). National guidelines and practice recommendations support the testing of EGFR, and this code has a value on the CLFS16 (National Comprehensive Cancer Network Non-Small Cell Lung Cancer Guidelines 2015, http://www.nccn.org/professionals/physic ian_gls/f_guidelines.asp#nscl, last accessed April 26, 2016.). The payer mix for this code was 37% commercial, 21% managed government plans, and 42% government payers (Figure 4C). During the same period of time, the average reimbursement across all payers was 36.8% of charges (n Z 231). Similar to the 81455 code, commercial insurers and managed government plans reimbursed at higher rates (44% and 45%, respectively). Government payers reimbursed 27% of charges (n Z 97) (Figure 4D). Next, we sought to identify the effect (if any) cobilling of a molecular code would have on reimbursement for the CCCP test. A total of 136 CCCP cases had data available on other molecular tests performed on the same specimen within the same year; 40 of the 136 CCCP cases (29%) were cobilled with another molecular test. The average reimbursement for cobilled cases was similar to the average reimbursement in which CCCP was billed alone (21.6% in cobilled cases versus 19.8% in CCCP alone cases); thus, we find no evidence of a negative impact of cobilling on CCCP reimbursement (data not shown).
Rejection Rates and Reasons for Rejection
6
jmd.amjpathol.org
Of the 153 claims submitted to third-party payers, 55% were rejected on the first submission (Figure 5A). The reasons for ½F5 denial are shown in Figure 5B. The major reason for denial, accounting for 82% of cases, was that the coded procedure is a noncovered service. This was driven by one commercial payer and Medicare. In comparison, the rejection rate for the EGFR CPT code (81235) during the same period of time was much lower, at 14% of claims (Figure 5C). The major reason for denial was similar to the 81455 CPT code, with 50% of these rejections being coded as noncovered service (Figure 5D).
Time from Claim Submission to First Reimbursement Decision A prolonged revenue cycle with delayed reimbursement may be an additional hurdle for clinical laboratories to face. To address the question of whether the as-yet unvalued on CLFS 81455 CPT code had a prolonged revenue cycle, we examined the average amount of time from claim submission to communication of an initial decision on reimbursement from the payer in the subset of cases in which the information was available. For the 81455 code, the average amount of time was 25.5 days (n Z 118 for CCCP). By comparison, the average time to reimbursement decision for the more commonly used EGFR code was 27 days (n Z 174). The distribution of lag time to reimbursement is
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744
print & web 4C=FPO
Figure 4 Payer mix and average reimbursements for CPT code 81455 compared to 81235. Distribution of payers for cases coded under code 81455 (A) and the average reimbursement as a percentage of total charges for code 81455 (B). Distribution of payers for cases coded under the epidermal growth factor receptor code 81235 (C) and the average reimbursement as a percentage of total charges for code 81235 (D). n Z 153 (A); n Z 231 (C). ½F6
summarized in Figure 6. These results suggest no significant difference in the amount of time before the first reimbursement decision.
Discussion We show herein the clinical performance and diagnostic yield of our CCCP assay, as well as our experience with
reimbursement over a 12-month period. Our assay performed well, and we were able to successfully sequence 92% of cases, and identify targetable variants in 48.7% of cases. The overall high success rate and low rate of insufficient cases (6%) across the range of tumors and specimens (procured through a variety of methods) highlight the robustness of the assay. A likely contributing factor to our success rate is the involvement of pathologists in selection and review of the material submitted for testing, which
Figure 5 Summary of rejection rates and denial reasons for CPT code 81455 compared to 81235. Payment denial rates under CCCP for 81455 (A) and summary of denial reasons provided by third-party payers for 81455 (B). Payment denial rates under epidermal growth factor receptor for 81235 (C) and summary of denial reasons provided by third-party payers for 81235 (D). n Z 153 (A); n Z 231 (C).
print & web 4C=FPO
745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806
NGS Cancer Targetability/Reimbursement
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
7
807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868
ensures that sufficient material is available for testing, that the most clinically appropriate material is selected when multiple blocks or specimens are available for a patient, and that sufficient tumor cellularity is present to avoid potential false-negative results. Our slightly lower success rate for FNA specimens, which are typically the specimens with more limited tissue, suggests a potential area for improvement. FNA procedures are less invasive, and are more commonly becoming the procedure of choice to obtain material for molecular testing, especially in patients with advanced disease who are not candidates for invasive procedures. Closer collaboration with our cytopathologists and/ or potential optimization of rapid on-site evaluation utilization by interventional radiologists or other practitioners who collect FNA samples may be helpful in further increasing sequencing success rates and identifying targetable variants for more patients, and will be the subject of future studies. Our overall targetability rate of 48.7% is comparable with large cancer gene panels from other institutions or laboratories.17,18 One caveat in comparing targetability/actionability across different cancer assays is the lack of a universally accepted definition or metric of clinical actionability/targetability.19 Various definitions of actionability and targetability have been reported in the literature. Actionability may refer to changes in clinical management from variants that provide diagnostic, prognostic, or therapeutic information, or may refer to variants that predict response to therapy. Targetability may refer to variants that respond to an approved therapy or may refer also to those variants that are predicted to confer response to an investigational therapy. For the sake of this analysis, we chose to use a definition of targetability/ actionability similar to that used by some other institutions and laboratories.17,18 However, standardizing measures of targetability/actionability in addition to clinical trials on clinical outcomes to establish the clinical utility of large, multigene oncology molecular diagnostics will be crucially
helpful in improving the current coverage and reimbursement for this testing.20e22 Uncertainty around coverage and reimbursement for CCCP testing threatens the accessibility of this clinically useful testing to our patient population and limits more general adoption.9,23 The terms coverage and reimbursement are often used interchangeably, but refer to distinct parts of the third payer payment process. Coverage describes the range of testing a payer will pay for and under what clinical scenario. Reimbursement, on the other hand, refers to the amount a payer will pay for a covered service.24 Decisions on coverage and reimbursement made by Medicare affect all insured patients because many commercial carriers and Medicaid will base their coverage policies and rates on the CLFS.9 Our reimbursement for the 81455 CPT code is significantly lower as a percentage of charges than the more established tier 1 code for EGFR (81235) and, in addition, the rejection rate is higher. Cobilling of other molecular tests with CCCP did not appear to affect CCCP reimbursement. The low reimbursement is driven by 0% reimbursement from Medicare for 81455, whereas the same payer covers the EGFR code 81235 at 27% of charges. The high rejection rate is skewed by Medicare’s 100% denial of claims for the 81455 code compared to a 14% denial rate for the EGFR code. The major reason given for denial of the 81455 code is noncovered service. Our experience illustrates problems in coverage and reimbursement resulting from failures in both valuation of the 81455 CPT code and in establishing national or local coverage determinations. The EGFR 81235 code is covered by Medicare, and its reimbursement is valued on the CLFS (Centers for Medicare and Medicaid Services, US Department of Health and Human Services, CLFS 2016, https:// www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ ClinicalLabFeeSched, last accessed April 11, 2016). In contrast, the 81455 code for gene panels >50 genes remains as-of-yet unvalued and a coverage policy had not yet been officially defined on the national or local level by Medicare in 2015. These two factors contribute to our institutional reimbursement experience with 81455, and until the 81455 CPT code is valued and coverage determinations are made, coverage and reimbursement for this CPT code will likely remain poor and highly variable. The typical process for valuation of a new CPT code after it is adopted by the American Medical Association on the Physician Fee Schedule involves a rigorous and transparent valuation process by the Relative Value Scale Update Committee, which makes recommendations on value to the Centers for Medicare and Medicaid Services.25 For codes adopted on the CLFS, such as all tier 1 and 2 and genomic sequencing procedures, there are two options for preliminary valuation: crosswalk and gap fill.25 Under the crosswalk approach, the value of a new CPT is based on the value of an existing and related code. Under gap fill, each individual Medicare administrative contractor sets
8
jmd.amjpathol.org
print & web 4C=FPO
869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930
Sireci et al
Analysis of time from filing to reimbursement decision for 81455 compared to 81235. The distribution of time from invoice generation to first reimbursement decision is divided into 30-day buckets, and both epidermal growth factor receptor (EGFR) and CCCP are depicted.
Figure 6
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992
993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
NGS Cancer Targetability/Reimbursement
Q15 Q16
pricing for a 1-year period on the basis of charges, required resources for a service, and payment by other payers.12 In year 2, the median value across Medicare administrative contractors is used to set the value for the code on the CLFS. The lack of reimbursement by Medicare for the 81455 CPT code before valuation on CLFS may be an example of a breakdown in the gap fill process. Even when the gap-fill mechanism for valuation was applied to molecular testing, such as in 2013, after the American Medical Association adopted tier 1 and tier 2 molecular CPT codes (EGFR is a tier 1 code), this mechanism is imperfect. Reimbursement for tier 1 and 2 molecular CPT codes was incomplete and inconsistent, the industry complained of lack of clarity in the process, and several Medicare administrative contractors simply adopted values from one contractor.12 During this prolonged decision-making process (January 2013, to September 30, 2013), many codes went unpaid.26 In addition to valuation, coverage determination for the 81455 CPT code by Medicare has not been adequately addressed. The determination of coverage by Medicare is predicated on proof that the service is medically reasonable and necessary for the Medicare beneficiary population.27 The determination can proceed at the national level through the transparent and well-documented process known as a national coverage determination, or each individual contractor can establish its own policy via a local coverage determination.28 For the entirety of 2015, Medicare did not have a national coverage determination specifically for molecular profiling using multiplex or next-generation sequencing technology to guide cancer treatment(s), and the local Medicare contractor in New York (National Government Service) had not issued an local coverage determination regarding 81455. However, during preparation of this article, on April 1, 2016, National Government Services (the Medicare administrative contractor covering New York) issued a local coverage determination grouping 81455 in with other genomic sequencing procedures, which would be immediately rejected as not medically necessary [Local Coverage Determination: Molecular Pathology Procedures (L35000), https://www.cms.gov/medicare-coverage-database/details/lcddetails.aspx?LCDIdZ35000&ContrIdZ297, last accessed April 11, 2016]. In contrast, a recent coverage policy by United Health Care allowed for coverage of the 81455 code in the setting of nonesmall-cell lung cancer based mainly on National Comprehensive Cancer Network guidelines and the availability of targeted therapies in this disease (Molecular Profiling to Guide Cancer Treatment, United Health Care Medical Policy Number 2016T0576B, https://www. unitedhealthcareonline.com/ccmcontent/ProviderII/UHC/enUS/Assets/ProviderStaticFiles/ProviderStaticFilesPdf/Tools %20and%20Resources/Policies%20and%20Protocols/Medi cal%20Policies/Medical%20Policies/Molecular_Profiling. pdf, last accessed April 11, 2016).
The Journal of Molecular Diagnostics
-
Coverage policies for new codes are not keeping pace with the adoption of novel diagnostics into clinical practice, and as a result, these policies can be inconsistent or variable. In part, this is because of the disconnect between the pace of new targeted therapy discoveries and development of clinical assays, and the requirements payers have for establishing coverage (ie, rigorous determinations of clinical utility).29 Clinical utility is defined as “the evidence of improved measureable clinical outcomes, and its usefulness and added value to patient management decision-making compared with the current management.”30 Examples of studies establishing clinical utility Q17 are large randomized clinical trials and are impossible for single laboratories to execute. Our study, which showed 48.7% of cases had a targetable variant, together with other similar studies have consistently shown large cancer panels can identify targetable or actionable variants and have the potential to show clinical utility by guiding therapeutic decision making and clinical management in a significant percentage of cases. The level of evidence recommended in a recent guidance document (which is similar to the level of evidence for supporting EGFR testing) will take a significant amount of resources16,21 (National Comprehensive Cancer Network Non-Small Cell Lung Cancer Guidelines 2015). Evidence of clinical utility is a major predictor of payer coverage for genetic testing procedures.21,31 Alternative approaches for establishing utility besides clinical outcomes (eg, decreased resource utilization resulting from fewer invasive procedures required to obtain predictive therapeutic information), therefore, must be explored to prevent the limiting of access to testing that may provide effective targeted therapies in the meantime. A recent publication by the Association of Molecular Pathology’s Framework for the Evidence Needed to Demonstrate Clinical Utility Task Force addresses exactly this need and expands the meaning of clinical utility to be more patient centered, context dependent, and flexible to alternative study design (eg, retrospective studies).32 In addition, the process of CLFS code valuation will require revision and extensive input from stakeholders if new and disruptive technologies are to be equitably integrated into clinical decision making and therapeutics. Finally, collaborative engagement of key stakeholders, such as oncologists, pathologists, and patients with private and government payers, in determining reimbursement policies and rates at the local and national levels will likely be the key to future progress and enhanced access. The Columbia experience with a large targeted cancer panel, along with the experience of other providers of large cancer panels, has proved that genomic profiling of tumors can result in a relatively high rate for identification of targetable variants. To secure access to this powerful testing, molecular pathologists must collaborate with clinician-oncologists and payers to show that the results of this testing affect medical decision making and ultimately result in superior clinical outcomes. Until that time, an
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
9
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178
Sireci et al
1. Abrams J, Conley B, Mooney M, Zwiebel J, Chen A, Welch JJ, Takebe N, Malik S, McShane L, Korn E, Williams M, Staudt L, Doroshow J: National Cancer Institute’s Precision Medicine Initiatives for the new National Clinical Trials Network. Am Soc Clin Oncol Educ Book 2014:71e76 2. Lopez-Chavez A, Thomas A, Rajan A, Raffeld M, Morrow B, Kelly R, Carter CA, Guha U, Killian K, Lau CC, Abdullaev Z, Xi L, Pack S, Meltzer PS, Corless CL, Sandler A, Beadling C, Warrick A, Liewehr DJ, Steinberg SM, Berman A, Doyle A, Szabo E, Wang Y, Giaccone G: Molecular profiling and targeted therapy for advanced thoracic malignancies: a biomarker-derived, multiarm, multihistology phase II basket trial. J Clin Oncol 2015, 33:1000e1007 3. Hyman DM, Solit DB, Arcila ME, Cheng DT, Sabbatini P, Baselga J, Berger MF, Ladanyi M: Precision medicine at Memorial Sloan Kettering Cancer Center: clinical next-generation sequencing enabling next-generation targeted therapy trials. Drug Discov Today 2015, 20: 1422e1428 4. Kim ES, Herbst RS, Wistuba II, Lee JJ, Blumenschein GR Jr, Tsao A, Stewart DJ, Hicks ME, Erasmus J Jr, Gupta S, Alden CM, Liu S, Tang X, Khuri FR, Tran HT, Johnson BE, Heymach JV, Mao L, Fossella F, Kies MS, Papadimitrakopoulou V, Davis SE, Lippman SM, Hong WK: The BATTLE trial: personalizing therapy for lung cancer. Cancer Discov 2011, 1:44e53 5. Luthra R, Chen H, Roy-Chowdhuri S, Singh RR: Next-generation sequencing in clinical molecular diagnostics of cancer: advantages and challenges. Cancers 2015, 7:2023e2036 6. Roy S, LaFramboise WA, Nikiforov YE, Nikiforova MN, Routbort MJ, Pfeifer J, Nagarajan R, Carter AB, Pantanowitz L: Nextgeneration sequencing informatics. Arch Pathol Lab Med 2016, 140: 958e975 7. Ulahannan D, Kovac MB, Mulholland PJ, Cazier JB, Tomlinson I: Technical and implementation issues in using next-generation sequencing of cancers in clinical practice. Br J Cancer 2013, 109: 827e835 8. Xuan J, Yu Y, Qing T, Guo L, Shi L: Next-generation sequencing in the clinic: promises and challenges. Cancer Lett 2013, 340: 284e295 9. Deverka PA, Dreyfus JC: Clinical integration of next generation sequencing: coverage and reimbursement challenges. J Law Med Ethics 2014, 42 Suppl 1:22e41 10. Trosman JR, Weldon CB, Kelley RK, Phillips KA: Challenges of coverage policy development for next-generation tumor sequencing panels: experts and payers weigh in. J Natl Compr Canc Netw 2015, 13:311e318
11. CPT 2016 Professional Edition. Chicago, American Medical Association, 2016 12. Klein RD: Reimbursement in molecular pathology: bringing genomic medicine to patients. Clin Chem 2015, 61:136e138 13. Forbes SA, Beare D, Gunasekaran P, Leung K, Bindal N, Boutselakis H, Ding M, Bamford S, Cole C, Ward S, Kok CY, Jia M, De T, Teague JW, Stratton MR, McDermott U, Campbell PJ: COSMIC: exploring the world’s knowledge of somatic mutations in human cancer. Nucleic Acids Res 2015, 43:D805eD811 14. 1000 Genomes Project Consortium, Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, Korbel JO, Marchini JL, McCarthy S, McVean GA, Abecasis GR: A global reference for human genetic variation. Nature 2015, 526:68e74 15. Choi Y, Chan AP: PROVEAN web server: a tool to predict the functional effect of amino acid substitutions and indels. Bioinformatics 2015, 31:2745e2747 16. Lindeman NI, Cagle PT, Beasley MB, Chitale DA, Dacic S, Giaccone G, Jenkins RB, Kwiatkowski DJ, Saldivar JS, Squire J, Thunnissen E, Ladanyi M: Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. Arch Pathol Lab Med 2013, 137:828e860 17. Frampton GM, Fichtenholtz A, Otto GA, Wang K, Downing SR, He J, et al: Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol 2013, 31:1023e1031 18. Hovelson DH, McDaniel AS, Cani AK, Johnson B, Rhodes K, Williams PD, Bandla S, Bien G, Choppa P, Hyland F, Gottimukkala R, Liu G, Manivannan M, Schageman J, Ballesteros-Villagrana E, Grasso CS, Quist MJ, Yadati V, Amin A, Siddiqui J, Betz BL, Knudsen KE, Cooney KA, Feng FY, Roh MH, Nelson PS, Liu CJ, Beer DG, Wyngaard P, Chinnaiyan AM, Sadis S, Rhodes DR, Tomlins SA: Development and validation of a scalable next-generation sequencing system for assessing relevant somatic variants in solid tumors. Neoplasia 2015, 17:385e399 19. Dienstmann R, Jang IS, Bot B, Friend S, Guinney J: Database of genomic biomarkers for cancer drugs and clinical targetability in solid tumors. Cancer Discov 2015, 5:118e123 20. Kakudo K, Bai Y, Liu Z, Ozaki T: Encapsulated papillary thyroid carcinoma, follicular variant: a misnomer. Pathol Int 2012, 62: 155e160 21. Deverka P, Messner DA, McCormack R, Lyman GH, Piper M, Bradley L, Parkinson D, Nelson D, Smith ML, Jacques L, Dutta T, Tunis SR: Generating and evaluating evidence of the clinical utility of molecular diagnostic tests in oncology. Genet Med 2016, 18: 780e787 22. Liu J, Singh B, Tallini G, Carlson DL, Katabi N, Shaha A, Tuttle RM, Ghossein RA: Follicular variant of papillary thyroid carcinoma: a clinicopathologic study of a problematic entity. Cancer 2006, 107: 1255e1264 23. Mauer CB, Pirzadeh-Miller SM, Robinson LD, Euhus DM: The integration of next-generation sequencing panels in the clinical cancer genetics practice: an institutional experience. Genet Med 2014, 16: 407e412 24. Williams MS: Insurance coverage for pharmacogenomic testing in the USA. Pers Med 2007, 4:479e487 25. Donovan WD: What is the RUC? AJNR Am J Neuroradiol 2011, 32: 1583e1584 26. GenomeWeb. Ray T: Centers for Medicare and Medicaid Services national payment limit for tier 1 codes not as low as before, but challenges remain. New York, NY: GenomeWeb. 2013. Available at http://www.genomeweb.com/clinical-genomics/cms-national-pa yment-limit-tier-1-codes-not-low-challenges-remain (accessed April 11, 2016) 27. Centers for Medicare and Medicaid Services, US Department of Health and Human Services. Publication 100-2, Medicare Benefit Policy
10
jmd.amjpathol.org
expanded definition of clinical utility to drive reimbursement may be the key to improving availability of this testing more broadly.
Acknowledgments We thank Samantha Cano and Vaishali Hodel for technical support; Marissa Pang, Nicholas Rouse, and Stuart Andrews for bioinformatics support; and Melissa Carter for assistance with collection of billing and reimbursement data.
Supplemental Data Supplemental material for this article can be found at http://dx.doi.org/10.1016/j.jmoldx.2016.10.008.
References
Q21
-
The Journal of Molecular Diagnostics
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
Q18
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
NGS Cancer Targetability/Reimbursement
Q19
Manual. Baltimore, MD, CMS, 2014. Available at http://www.cms. hhs.gov/manuals/downloads/bp102c16.pdf (accessed April 11, 2016) 28. Burken MI, Wilson KS, Heller K, Pratt VM, Schoonmaker MM, Seifter E: The interface of Medicare coverage decision-making and emerging molecular-based laboratory testing. Genet Med 2009, 11:225e231 29. Department of Health and Human Services Coverage and Reimbursement of Genetic Tests and Services: Report of the Secretary’s Advisory Committee on Genetics, Health, and Society. Washington, DC, HHS, 2006 30. Teutsch SM, Bradley LA, Palomaki GE, Haddow JE, Piper M, Calonge N, Dotson WD, Douglas MP, Berg AO, Group EW: The
The Journal of Molecular Diagnostics
-
Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Initiative: methods of the EGAPP Working Group. Genet Med 2009, 11:3e14 31. Meckley LM, Neumann PJ: Personalized medicine: factors influencing reimbursement. Health Policy 2010, 94:91e100 32. Joseph L, Cankovic M, Caughron S, Chandra P, Emmadi R, Hagenkord J, Hallam S, Jewell KE, Klein RD, Pratt VM, Rothberg PG, Temple-Smolkin RL, Lyon E: The spectrum of clinical utilities in molecular pathology testing procedures for inherited conditions and cancer: a report of the Association for Molecular Pathology. J Mol Diagn 2016, 18:605e619
jmd.amjpathol.org
FLA 5.4.0 DTD JMDI566_proof 23 December 2016 4:28 pm EO: JMD16_0120
11
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376