Overcoming Barriers to Precision Care and Patient Access

Precision therapies are transforming cancer outcomes, but operational and financial hurdles still stand between patients and optimal care, experts say. Across 4 panels at the Institute for Value-Based Medicine^® (IVBM) event in Seattle, Washington, on July 23, 2026, speakers stressed that closing these gaps will require faster diagnostic workflows, smarter referral pathways, and better collaboration between community and academic oncology practices.

Balancing Precision, Access in Lung Cancer Care

The IVBM event kicked off with “Precision in Practice: Implementing Biomarker Testing in Lung Cancer,” moderated by Rafael Santana-Davila, MD, of Providence Swedish Cancer Institute (PSCI), and featuring panelists Siddhartha Devarakonda, MD, of PSCI; Lei Deng, MD, of Fred Hutchinson Cancer Center (Fred Hutch); and Bin Xie, MD, of PSCI. The panel explored how testing delays, insurance rules, and the limitations of artificial intelligence (AI) are complicating precision treatment decisions in lung cancer.

Devarakonda noted that lung cancer physicians worry about missing a driver alteration, since identifying EGFR or ALK alterations can entirely change first-line treatment. Sequencing immunotherapy ahead of a tyrosine kinase inhibitor (TKI) in driver-positive disease, he warned, is “a dumpster fire waiting to happen.” To help prevent this, Devarakonda said he builds testing into the diagnostic workflow as early as possible, using rapid on-site pathology during bronchoscopy and sending circulating tumor DNA (ctDNA) alongside tissue in patients with a high disease burden.

The panelists also pointed to Medicare’s 14-day rule as a persistent structural barrier. Under the policy, when a genomic test is performed during a hospitalization or within 14 days of discharge, the cost may be bundled into the hospital payment rather than separately reimbursed; Devarakonda described this as “a huge disservice to our patients.” To minimize delays, he said he often orders fast-turnaround assays, including EGFR polymerase chain reaction testing, PD-L1 immunohistochemistry, and ALK/ROS1 fluorescence in situ hybridization, which can return results in 24 to 48 hours. Devarakonda also described ordering ctDNA testing through outpatient channels in urgent cases to avoid inpatient billing restrictions.

Xie, who practices in the community setting, said he takes an even more front-loaded approach, ordering broad ctDNA testing the same day that lung cancer is suspected, often before a tissue diagnosis is confirmed. At the same time, he still sends tissue for next-generation sequencing (NGS) to avoid missing mutations that may not be detectable in ctDNA.

Building on this, the panelists diverged on how much weight ctDNA results alone should carry. Deng said he is comfortable starting treatment on a classic, actionable mutation identified through ctDNA while still pursuing tissue confirmation. Xie took a stricter position, declining to treat without a confirmed tissue diagnosis. Devarakonda landed in between, saying he would start a TKI based on a strongly actionable ctDNA result only under genuine urgency while continuing to pursue tissue confirmation when possible. He also cautioned that ctDNA findings can sometimes reflect clonal hematopoiesis or a nonlung primary tumor rather than the lung cancer itself.

For immunotherapy delivery, the panelists discussed insurers increasingly moving patients from infusion chairs at academic cancer centers to lower-cost third-party sites after treatment has been established. Devarakonda argued that these policies risk fragmenting care coordination and toxicity monitoring across 2 health systems. Santana-Davila raised the payer’s counterargument that the drug itself is identical regardless of the site of administration. Devarakonda, however, maintained that oncologists typically appeal these decisions but ultimately have limited leverage against insurer mandates.

Xie noted that similar site-of-care restrictions already apply to intravenous immunoglobulin, where insurers may require treatment at approved sites without clearly specifying which locations qualify, leaving patients to navigate the ambiguity on their own. He said he expects these challenges to intensify as more therapies shift toward subcutaneous or at-home administration, making them more attractive for payers to redirect. Xie also argued that health systems need proactive, institution-level policies rather than leaving individual oncologists to contest each payer’s decision on a case-by-case basis.

The panelists wrapped up by addressing AI, which Devarakonda considers particularly well-suited for quality control, including flagging incomplete genomic workups, tracking when DNA and RNA results are expected, and cross-checking planned treatments against known alterations. Xie estimated that AI could improve documentation and drafting efficiency by approximately 80% while still requiring physician sign-off.

Deng was more cautious about AI, citing limited access to high-quality, Health Insurance Portability and Accountability Act–compliant clinical data as a challenge for its performance. He added that he had received inaccurate references from OpenEvidence on more than 1 occasion. Patient-facing tools carry their own risk, with Deng emphasizing that oncologists must now vet both the AI tools they use and the ones their patients bring to appointments.

“Yes, AI does have a lot of potential, but my understanding of AI is that no matter how smart an AI is, they have to train,” he said. “When you don’t have good data, then your model is going to be having issues. I think, yes, it has huge potential, but I think [health care] is going to lag really, really behind compared with other sectors in terms of using AI to improve what we do.”

Navigating New Treatment Frontiers in Leukemia and Lymphoma

The discussion turned to hematologic malignancies when moderator Ryan D. Cassaday, MD, of Fred Hutch, took the stage for “Advancing Frontiers in Leukemia & Lymphoma: Breakthroughs and Barriers,” alongside panelists Sharmila Ahmed, MD, of Providence; Stephanie J. Bardin, PharmD, MA, of PSCI; and Ivan Huang, PharmD, BCOP, of Fred Hutch.

Cassaday opened by framing how leukemias and lymphomas have been transformed by antigen-targeted immunotherapies, namely chimeric antigen receptor (CAR) T-cell therapy and bispecific antibodies. Ahmed called CAR T-cell therapy “transformational,” particularly for patients with diffuse large B-cell lymphoma (DLBCL), turning previously terminal relapses into salvageable cases. Ahmed added that she is equally excited about bispecifics as an earlier-line, outpatient option for patients with lymphoma. Bardin and Huang noted that although CAR T-cell therapy remains centralized to 2 Seattle-area centers, bispecifics are increasingly being extended into satellite locations, allowing patients from outlying areas to receive maintenance therapy closer to home. The panelists also noted the treatment is moving into earlier lines of therapy, changing how relapse is managed.

Within acute myeloid leukemia (AML), Cassaday asked how prior authorization, formulary restrictions, and inpatient-vs-outpatient realities shape treatment decisions. He noted the growing use of venetoclax (Venclexta; AbbVie Inc/Genentech USA, Inc) plus azacitidine as a lower-intensity approach, with targeted agents increasingly being added in triplet regimens.

Illustrating the financial toxicity at stake, Cassaday described oral venetoclax co-pays reaching approximately $1500 a month for Medicare beneficiaries, a burden that can fall on patients with little means to absorb it. Huang and Bardin said financial toxicity remains a major barrier more broadly, and Ahmed agreed that the financial burden can be “overwhelming” for patients already facing the intensive care needs associated with AML. On top of that, some hospitals do not keep venetoclax on their formulary, and payers may push for intensive chemotherapy over venetoclax-based regimens based solely on a patient’s age. Huang also emphasized the need for caregiver and transportation support, as treatment initiation can require frequent monitoring for tumor lysis syndrome (TLS) risk and cytopenias.

Regarding aggressive lymphomas and mantle cell lymphoma, Ahmed said that despite logistical challenges related to caregiving, travel, and follow-up care, she prioritizes CAR T-cell therapy referral for patients with DLBCL who relapse within 12 months. Bardin described how insurer-driven transfers between centers can force late regimen changes midbridging, with PSCI generally reserving bispecifics as bridging or later-line options when CAR T-cell therapy is not feasible.

Huang detailed the complexity of triplet regimens, such as Bruton tyrosine kinase (BTK) inhibitors combined with venetoclax and obinutuzumab (Gazyva; Genentech) for relapsed mantle cell disease. This includes overlapping myelosuppression, TLS risk, and infection management. However, he flagged the recent approval of sonrotoclax (Beqalzi; BeOne Medicines) as a promising new option as availability expands.¹

Before wrapping up, the panel turned to the systemic economic pressures shaping chronic lymphocytic leukemia (CLL) care. Ahmed pointed to the financial toxicity of indefinite continuous BTK inhibitor therapy, saying she is increasingly drawn to fixed-duration venetoclax-BTK combinations that give patients a defined treatment end point. That tension between drug cost and payment structure extends well beyond CLL, noted Cassaday, Bardin, and Huang, as expensive novel agents increasingly collide with inpatient bundled payment models. To work around this, clinicians have resorted to administering drugs in the outpatient setting under 340B pricing before same-day inpatient admission or discharging and readmitting patients specifically to access outpatient drug coverage.

Looking ahead, Cassaday warned that as bispecifics expand into common solid tumors such as prostate and breast cancer, hospitals may need new infrastructure to accommodate short monitoring stays for cytokine release syndrome. This burden, he added, is more difficult to avoid because current FDA trial designs still bake inpatient observation into drug labels.

“All the trials have these sorts of monitoring periods built into the hospitalization, so then when the drug gets approved, what does the package insert say? They have to be hospitalized for the first 48 hours, 72 hours, or the first 2 doses, or whatever,” Cassaday said. “There’s still going to be a chunk of time, I suspect, before the remote monitoring and stuff catches up, but hopefully these will get so routine they’ll be built into the clinical trials, which can then ultimately find their way into the routine treatment.”

Optimizing Referral, Treatment Pathways in Skin Cancer

Moderator Shailender Bhatia, MD, of Fred Hutch, led “Developing Optimal Treatment Referral Patterns for Skin Cancer Patients,” featuring panelists Jay P. Lopez, MD, of PSCI; Katherine M. Jackson, MD, of the University of Washington School of Medicine; and Meng Zhao, MD, PhD, of PSCI. They examined melanoma referral and triage challenges, the potential for AI to improve care coordination, and the ongoing debate over the value of sentinel lymph node biopsy (SLNB) and gene expression testing in guiding melanoma staging.

Jackson explained that most patients reach the Fred Hutch Skin Oncology Clinic through outside dermatologists or a nurse navigation pathway, with triage determining whether cases are routed to dermatologists, surgical oncologists, or ear, nose, and throat specialists, depending on complexity. Patients with melanoma who may need SLNB are generally routed to a surgeon. Once a diagnosis is made, her team aims to see patients within 72 hours, she said, although delays can occur earlier in the referral process, particularly when patients are waiting to see a dermatologist.

Zhao said most melanoma referrals to medical oncology come from surgical oncologists or in-system dermatologists. Insurance coverage can split a single patient’s care across facilities, she added, with surgery approved at one system and medical oncology only covered at another. In community practice, referrals for cutaneous malignancies typically arrive from dermatology, surgery, or radiation oncology, Lopez said, with complex cases discussed at a local tumor board.

The panel identified several recurring pain points. Lopez cited insurance-mandated “preferred” dermatology groups that do not always match patient needs, adding friction to referrals. Zhao noted that in community settings without melanoma-focused surgeons, patients sometimes go straight to surgery and miss the window for neoadjuvant systemic therapy that a specialized multidisciplinary system might have identified. Jackson described imaging delays, saying a PET/CT or MRI marked as urgent can still take approximately 2 weeks, compounded by insurance denials citing guideline mismatches.

Bhatia shared a case in which a patient with stage IA melanoma traveled 3 hours to see him for medical oncology when they really needed surgical care. He attributed the mis-triage to a 72-hour access initiative that prioritized speed without sufficient safeguards to ensure patients were routed to the appropriate specialist; Bhatia caught the error the night before and redirected the referral. Audience members added that schedulers often lack the clinical context to triage correctly, while documentation systems can be cluttered with low-value, auto-populated content that does not aid decision-making.

The panelists largely saw AI as a promising but underused solution to these gaps. Jackson said that with nurse navigators increasingly scarce, she would welcome AI that could triage patients to the appropriate clinic, flag needed imaging, and even draft insurance appeal letters. Zhao envisioned AI embedded directly in the electronic health record to interpret documentation, determine stage, automatically flag National Comprehensive Cancer Network (NCCN)–recommended referrals, and prescreen patients for clinical trial eligibility. She said tools such as OpenEvidence have improved substantially and now aid evidence review quickly.

Lopez, describing himself as an AI skeptic outside medicine, said he has become an advocate for AI in oncology, calling OpenEvidence “revolutionary” for structuring information and cross-checking his clinical reasoning. Audience members countered that health care remains more conservative about adopting AI than other industries, citing privacy and policy barriers.

The panel closed on a technical debate over SLNB in melanoma. Jackson explained that SLNB is a staging, rather than therapeutic, procedure, with NCCN guidelines recommending it for patients with more than a 10% predicted risk of nodal positivity and generally against it for those with less than a 5%, leaving a 5% to 10% range in which the procedure can be considered.

Bhatia presented data from the observational MERLIN_001 trial (NCT04759781), which evaluated the Merlin test, a commercial 8-gene expression profile (CP-GEP) intended to better identify low-risk patients who could skip SLNB.² The test’s low-risk group still showed approximately 7% to 8% node positivity, missing its target of less than 5%.

Jackson and Zhao were skeptical of the test’s added value, noting that existing free nomograms and standard clinicopathologic factors already provide strong predictive power. They questioned whether CP-GEP adequately resolves the clinical uncertainty surrounding SLNB decisions, particularly because some of its strongest predictive information overlaps with factors already used in clinical practice.

Bhatia argued that any new test would need to clearly outperform existing free tools to justify its cost, cautioning that prior gene expression tests have been marketed heavily to dermatologists despite questions about their clinical utility.

“The other thing to keep in mind is, since this is a value-based forum, to find those few people where the test may actually have value, you have to order this test on thousands of patients, and there’s going to be a substantial charge,” Bhatia said. “If we are going to subject our precious health care resources [to] that direction, we better have much more believable data.”

Collaborative Strategies to Improve Cancer Screening and Trial Access

Veena Shankaran, MD, MS, of Fred Hutch, led the evening’s final discussion, “Collaborations in Today’s Environment for Successful Clinical Trials,” which examined gaps in cancer screening and barriers to clinical trial access. Shankaran was joined by Rachel Issaka, MD, MAS, of Fred Hutch; Osler Andres, PharmD, BCOP, BCSCP, of Providence; and Scott D. Ramsey, MD, PhD, of Fred Hutch.

Issaka outlined 3 core screening failure points: inadequate follow-up after abnormal tests, poor posttreatment surveillance, and incomplete family histories that can limit risk-based screening for relatives. Patients who do not complete diagnostic follow-up after an initial screening test may ultimately present with later-stage disease, she noted, and inadequate surveillance can delay detection of recurrence. Issaka also emphasized opportunities to use population health tools to identify patients due for screening and to conduct targeted outreach, including through home-based testing.

Andres highlighted how access barriers can vary sharply across the urban, rural, and tribal communities served by his practice. Although the Seattle area has extensive health care resources, he noted that even patients in well-resourced areas can face lengthy waits for gastrointestinal (GI) appointments. Andres added that rural and tribal communities may experience additional challenges related to income, health literacy, and access to specialty care. Ramsey and Issaka agreed that cancer screening rates, particularly for colorectal cancer, have not fully recovered to pre–COVID-19 pandemic levels. Issaka attributed much of the lag to pandemic-era delays that have since compounded into longer wait times for endoscopies and other GI procedures. Andres pointed to a related, less-documented factor: the lasting workforce effects of COVID-19, describing a “Great Migration” of medical assistants, nurses, and other support staff out of the field during the pandemic. That turnover, he said, has left practices chronically short-staffed for the follow-up work that screening depends on, including tracking results and arranging next steps.

Regarding financial barriers to screening, Ramsey noted that although the Affordable Care Act eliminated co-pays for many cancer screening services, patients can still face costs associated with transportation, parking, and time away from work. When a screening test requires diagnostic follow-up, patients may encounter additional cost-sharing, he added, potentially creating another barrier to completing the evaluation. Andres emphasized the importance of social workers and trusted relationships with patients in identifying and addressing practical barriers, including transportation and other support needs. Issaka agreed, saying that patient navigators could play a larger role across the cancer continuum, helping patients access screening, complete follow-up, and maintain surveillance after treatment.

Ramsey also addressed the emerging role of multicancer early detection (MCED) blood tests, highlighting their appeal given that approximately half of cancers lack an approved screening test. However, he cautioned that evidence supporting widespread use remains limited, citing results from the large United Kingdom–based NHS-Galleri trial (NCT05611632), which fell short of its primary end point despite a reduction in stage IV cancer diagnoses.³ Ramsey said ongoing studies will be important in determining whether MCED testing has a role in population-level cancer screening, particularly as commercially available tests continue to be offered directly to patients.

The discussion then turned to clinical trial access, with Ramsey highlighting that only about 5% to 10% of adults with cancer enroll. He identified several barriers, including limited awareness of available trials, the challenge of manually determining patient eligibility, and unreimbursed travel costs for patients who must travel to academic centers. Ramsey suggested that AI could eventually help identify potential trial matches by continuously evaluating clinical information as it becomes available.

Andres advocated moving away from a model in which patients are transferred entirely from community practices to academic centers, toward collaborative care models. Under this approach, academic institutions could manage trial-specific procedures while community practices provide appropriate supportive care closer to patients’ homes. Such a model, he said, could reduce travel burdens while helping clarify the distinction between care covered by insurance and services provided through a clinical trial.

The panel concluded by emphasizing the need for stronger collaboration among academic institutions, community oncology practices, and other stakeholders to address persistent gaps in cancer screening and clinical trial access and ensure that eligible patients can participate in research without unnecessarily leaving their communities.

“I don’t think we have the answers here, but I think we want to make sure that if we have an eligible patient who wants a trial, we can figure out a way to support them,” Shankaran concluded.

References

1. Caffrey M. Sonrotoclax granted accelerated approval for R/R mantle cell lymphoma. AJMC. May 13, 2026. Accessed September 11, 2026. https://www.ajmc.com/view/sonrotoclax-granted-accelerated-approval-for-r-r-mantle-cell-lymphoma
2. Melanoma research lymph node prediction implementation national_001 (MERLIN_001). ClinicalTrials.gov. Updated April 23, 2026. Accessed September 11, 2026. https://clinicaltrials.gov/study/NCT04759781
3. Does screening with the Galleri test in the NHS reduce the likelihood of a late-stage cancer diagnosis in an asymptomatic population? a randomized clinical trial (NHS-Galleri). ClinicalTrials.gov. Updated July 10, 2026. Accessed September 11, 2026. https://clinicaltrials.gov/study/NCT05611632