Although molecular biomarkers are becoming increasingly relevant in cancer care, community oncologists confront a plethora of challenges in translating research findings into practice.
Although molecular biomarkers are becoming increasingly relevant in cancer care, community oncologists confront a plethora of challenges in translating research findings into practice. Structural changes in payment systems and decision-support tools are among the most critical needs that must be addressed to broaden the use of genomic testing in clinical practice, according to experts.
During the past 5 years, findings from real-world analyses indicate that patients with a range of tumor types are not being tested for biomarkers recommended in guidelines, leading to suboptimal use of targeted therapies.1 Investigators have lamented “pervasive undertesting” for tumor-specific biomarkers and inherited cancer risk.2 They have cited a complex web of factors that are impeding the uptake of genomic testing, such as gaps in knowledge among providers and support for interpreting test results, shortcomings in tissue samples (particularly in lung cancer), a lack of access to tests, and reimbursement challenges.1,2
Overcoming these hurdles will require greater support for oncologists in day-to-day practice so they can obtain molecular testing for their patients and decipher results in an easily digestible format, Howard L. McLeod, PharmD, FASCO, FCCP, said. McLeod is the executive clinical director of precision health at Intermountain Healthcare, a network of hospitals and medical groups based in Utah.
“I remember when PET [positron emission tomography] scans first came out, even when the data were present, there was a reluctance to use it because [oncologists] felt like no one had their back—it was nuclear medicine [specialists] who interpreted it and they didn’t really know those folks,” McLeod said. “Then radiology started interpreting it. Now you see PET scan, or PET/CT, being ordered for many patients because someone had the oncologist’s back. “That’s what needs to happen in the molecular arena to really make this routine,” McLeod added. “It’s not fair that oncologists have to keep up on more things than they’re able to and, just because you’re in a community setting, you should not be punished by not having a good backup for applying these technologies.”
Molecular testing can be challenging even in tumor types such as non–small cell lung cancer (NSCLC), in which somatic alterations in at least 7 genes have been identified as being clinically relevant to the therapy selection for patients with advanced disease.3 Community-based practices must send out tissue specimens for analysis, resulting in turnaround times that could delay the start of therapy, noted Tracey L. Evans, MD, director of thoracic oncology research at Lankenau Institute for Medical Research in Wynnewood, Pennsylvania, and codirector of the thoracic oncology program at the affiliated Main Line Health system.
“There’s a real sense of urgency for patients, especially those with metastatic NSCLC, who can be very symptomatic. You don’t always feel like you have the time to wait to get that genomic testing back, even though it’s critically important,” Evans said in an interview.
The adoption of reflex testing, in which a pathologist automatically orders the required tests based on clinical pathways, would help expand the use of molecular profiling in NSCLC and perhaps other tumor types, Evans said. “This is something that is routinely done in breast cancer, where all patients are tested for HER2, ER [estrogen receptor], and PR [progesterone receptor],” she said.
However, reimbursement barriers for lining up molecular testing can be significant. In addition to obtaining prior authorization, Evans noted, oncologists may also have to contend with the “14-day rule” enforced by the Centers for Medicare & Medicaid Services. Under that policy, physicians may have to wait 14 days after a patient is discharged from the hospital to order certain laboratory tests, a delay that could be problematic for patients with NSCLC who were hospitalized because of emergent symptoms.4
“There’s a myriad of challenges that that go into this, and I’m sympathetic to physicians who have a difficult time doing it,” Evans said.
Strategies for leveraging advances in tumor biology are becoming more important considering the growing number of FDA-approved anticancer therapies for molecularly defined populations, experts have observed. Although many alterations occur at low frequency, research findings have increased understanding of the frequency of mutations, particularly in the most prevalent cancers, which amplifies the potential clinical utility of genomic testing (TABLE 15).
However, there is a lack of clarity when it comes to choosing and ordering genomic sequencing tests and in interpreting results, according to an American Society of Clinical Oncology (ASCO) expert panel. In April, the panel published a provisional clinical opinion with recommendations for somatic genomic testing in patients with advanced or metastatic solid tumors.5
For these patients, genomic sequencing should be conducted on tumors for which genomic alterations have been identified as biomarkers to guide the use of approved therapies, the panel said. If more than 1 biomarker is associated with therapy for a given malignancy, tumors should be tested with multigene panel assays, defined as next-generation sequencing (NGS) tests that analyze a defined list of at least 50 genes.
Additional clinical scenarios involving tumor-agnostic biomarkers such as tumor mutational burden (TMB), mismatch repair deficiency (dMMR), or NTRK fusions also create a strong rationale for conducting genomic testing on all patients with advanced or metastatic solid tumors, the panel said. “For treatment planning, the clinician should consider the functional impact of the targeted alteration and expected efficacy of genomic biomarker–linked options relative to other approved or investigational treatments,” the ASCO authors noted.
Although such broad-based strategies may offer the best option for ensuring that all patients who are candidates for molecularly selected therapies receive testing, findings from several studies suggest that the oncology field is far from reaching that goal.
Only 26.7% of patients with localized or metastatic gastrointestinal stromal tumors (GISTs) received testing for KIT mutations, according to an analysis of information extracted from the Surveillance Epidemiology and End Results (SEER) database from 2010 to 2015.6 Knowledge of KIT mutational status is an important predictive and prognostic biomarker for the treatment of GISTs.7
In a retrospective review of electronic health data from 2013 through 2017 for patients with metastatic colon cancer, investigators found that less than 50% of patients had guideline-recommended testing for RAS alterations, BRAF mutations, and microsatellite instability/dMMR.8
More recent results show an increased uptake of molecular testing among patients in some clinical settings but not in others. Additionally, testing rates continue to lag guideline recommendations.
In NSCLC, the proportion of patients with newly diagnosed advanced disease who were tested for EGFR mutations before the initiation of first-line therapy increased from 45.2% to 56.6% from January 1, 2015, through November 30, 2020, according to an analysis of information for 22,726 patients in the Flatiron Health database presented during the 2021 ASCO Quality Care Symposium.9
In another study, the use of NGS testing increased significantly from 33% to 45% (P < .0001) of patients with metastatic NSCLC initiating first-line systemic therapy between April 1, 2018, and March 31, 2020, at community oncology practices in The US Oncology Network. The retrospective review was conducted by MYLUNG Consortium, a collaboration between The US Oncology Network, pharmaceutical companies, and patient advocacy groups.10
Among 3474 patients, 90% were tested for at least 1 of 5 biomarkers but only 46% were tested for all of them. Testing rates were 70% for EGFR mutations and ALK translocations, 68% for ROS1 rearrangements, 55% for BRAF mutations, and 83% for PD-L1 protein expression.10
“Our real-world study showed that most patients with metastatic non–small cell lung cancer had at least 1 biomarker test result available prior to initiation of first-line therapy,” Makenzi Evangelist, MD, a medical oncologist and hematologist at New York Oncology Hematology in Albany and Clifton Park, New York, said in presenting the results at the 2021 ASCO Annual Meeting. “However, only approximately 50% of patients had all 5 tests available prior to first-line therapy. NGS testing was low but increased over time. This may provide a better opportunity for comprehensive biomarker testing.”
In contrast to the improvements observed in lung cancer testing, there is a wide chasm between guideline recommendations and clinical practice concerning germline genetic testing in ovarian cancer. Patients diagnosed with ovarian, fallopian tube, or primary peritoneal cancer should receive genetic risk evaluation as well as germline and somatic mutation testing, according to National Comprehensive Cancer Network (NCCN) guidelines.11 For genetic risk assessment, the NCCN recommends testing for BRCA1/2 mutations and other high-penetrance cancer susceptibility genes for patients with a history of ovarian cancer.12
However, only 34.3% of patients diagnosed with ovarian cancer in California or Georgia from 2013 to 2017 (N = 14,689) received multigene panel testing, according to an examination of SEER data reported in the Journal of Clinical Oncology in February 2021. Investigators were dismayed to find that although the use of multigene assays increased compared with testing only for BRCA1/2 mutations, the overall uptake of genetic testing increased by only approximately 3% from 2013 to 2014 levels.13
“There is urgent need to further define the patient, clinician, and health care system factors that limit testing of patients with ovarian cancer and to develop interventions that surmount these barriers,” Allison W. Kurian, MD, MSc, and colleagues wrote.
In another study, investigators analyzed germline BRCA (gBRCA) testing patterns among 3603 women treated for ovarian cancer between 2008 and 2018 using claims data from a large national commercial insurer. During the 10-year period, only 33.9% of patients received gBRCA testing. The data show that testing rates were lower for women 65 years and older compared with those younger than 50 years (adjusted difference, −20.8%; 95% CI, −25.8% to −16.4%). There also was a difference in testing according to practice setting, with a lower rate (32.5%) in community practices compared with academic centers (36.2%) and National Cancer Institute centers (39.8%).14
When viewed on a year-by-year basis, the proportion of patients tested increased from 14.7% in 2008 to 46.4% in 2018. The median time from diagnosis to germline BRCA testing decreased from 280.0 days in 2008 to 72.5 days in 2018.
Nevertheless, the data show that gBRCA testing remains underutilized despite well- established guidelines recommending universal testing for women with ovarian cancer, investigators noted.
Some cancer centers are incorporating comprehensive genomic profiling (CGP) into routine practice, such as Intermountain Healthcare, which operates in Utah, Idaho, Nevada, and other Western states. McLeod said all patients with metastatic disease undergo 3 types of analysis: somatic tumor testing, germline testing for heritable cancers, and germline pharmacogenomics testing. Whereas testing for somatic and germline aberrations has dominated attention in the field, the importance of pharmacogenomics in identifying toxicity risks from anticancer therapy is being increasingly recognized.15
The somatic testing analyzes approximately 600 genes in DNA and RNA extracted from the tumor and, like the other assays, was developed in house, enabling greater control over turnaround times, McLeod said. “We have the expertise for not only interpreting the molecular piece, but also for doing a molecular tumor board with medical interpretation,” he said. “Our oncologist can quickly see the results and act on them. Many places do the molecular tumor and the molecular pathology piece, and give you a nice report, but the oncologists don’t really have the medical backing that they need. That’s something that’s really overlooked as we’re getting into this precision medicine era.”
Moving forward, some oncology thought leaders expect CGP to become standard practice across tumor types. In December 2017, the FDA approved the FoundationOne CDx genomic test, the first such assay for commercial use.16,17
The oncology field is at “an inflection point,” Bruce A. Feinberg, DO, said during an American Journal of Managed Care® Peer Exchange program in December 2021.18 Feinberg is vice president and chief medical officer at Cardinal Healthy Specialty Solutions.
“We’re just 4 years from the approval of the f irst comprehensive genomic profile, which was approved by the FDA as a test for patients with solid-tumor cancer,” Feinberg said. “After 4 years, we may be at the point where comprehensive genetic profiling is no longer the last thing you do when you’ve exhausted all standard of care. Rather, it has become the first thing you do before you initiate any systemic therapy in a patient with advanced cancer.”
For that to happen, oncologists will need decision-support tools that help identify appropriate testing that could then be conducted reflexively, said Kenna R. Mills Shaw, PhD, who participated in the Peer Exchange program. Shaw is executive director of the Sheikh Khalifa Bin Zayed Al Nahyan Institute for Personalized Cancer Therapy at The University of Texas MD Anderson Cancer Center in Houston. This information can be incorporated into electronic health programs.
“The more that we can put into these electronic health records to provide clinicians with what would be essentially the best practice alert kind of language would help us all do better,” Shaw said in an interview. “[Our systems need to] tweak that balance between notifying a clinician so they don’t have alert fatigue while making sure that [alerts] that truly can benefit patient care and outcomes are taken care of.”
At the same time, Shaw noted that charges continue to be a barrier to more widespread adoption of CGP. The “$1000 genome” has become emblematic of the plummeting costs of sequencing, but Shaw noted that the often-quoted price point refers to a research environment.19,20 In clinical practice, CGP can result in charges ranging from $5000 to $10,000 when all expenses are considered, including charges for a high-quality NGS test, a molecularly trained pathologist, and the decision-support team, Shaw said.
Coverage and reimbursement for the cost of NGS testing alone is highly variable, according to findings from a study of medical database claims reported during the 2021 ASCO Annual Meeting. The average allowed amounts ranged from $1269 to $2058 per test for NGS testing conducted during the study period (2016-2019). The average allowed amounts varied from $438 to $3700 per test for NGS-derived tests that included TMB and from $1722 to $2249 per test for hereditary cancers.21
In NSCLC, broader use of CGP could be achieved for a modest increase in costs per patient, according to recent study findings. Investigators analyzed direct and associated costs using data that were inflation adjusted to 2018 dollars (TABLE 222) and incorporated the information into a model they developed for a hypothetical health plan. They concluded that CGP for all patients with advanced NSCLC would expand the number of patients who receive CGP testing and improve outcomes at an additional cost of only $205 per patient.22
Ultimately, changes are needed in reimbursement procedures for oncologists to use CGP more extensively, Evans said. Oncologists need “federal payment schemes that recognize the importance of this and are consistent with this, so you don’t have to jump through so many hoops to do something that’s the right thing for the patient,” she said.
"Broadening the utilization is good for patient care and for research,” Evans said. “The downside, of course, is the expense. Health care is already very, very expensive. But I do think of the things that we spend money on, this is definitley up there in importance."