Cancer

Grand Rounds August 1, 2025: Clinical Trial Notifications Triggered by Artificial Intelligence-Detected Cancer Progression (Kenneth L. Kehl, MD, MPH)

Posted on by Sophia Ramirez

Speaker

Kenneth L. Kehl, MD, MPH
Assistant Professor of Medicine and Physician
Dana-Farber Cancer Institute

Keywords

Artificial Intelligence; Cancer; Notification; Enrollment; Patient Identification

Key Points

  • Historically, less than 10% of adults with cancer enroll in clinical trials. At the same time, many trials struggle to reach their accrual goals. One possible contributor is that many trials of novel therapies for cancer have specific molecular criteria.
  • Dana Farber Cancer Institute (DFCI) developed MatchMiner, a computational matching tool, to connect patients to trials. However, identified patients often weren’t at a place in their treatment when information about trials was relevant. The research team was interested in whether they could train an artificial intelligence (AI) model to identify “trial-ready” patients.
  • The team conducted an implementation pilot, providing clinicians and research staff with weekly spreadsheets containing predictions of clinical trial “readiness” as identified by AI. The majority of identified patients were found to be ineligible upon RN review. Of those who were eligible, the majority opted not to move forward with the trial referral. At the end of the 9-month pilot, 6 AI-identified patients had been consented and enrolled in a therapeutic trial.
  • To assess the impact of AI-driven identification of trial-ready patients, the team launched OPTIONS (Optimizing Precision Trials with an artificial Intelligence driven Oncologist Notification System). The primary outcome of the trial was enrollment in any DFCI therapeutic clinical trial.
  • Patients with solid tumors were randomized into either a control group, in which they could be identified by the standard MatchMiner workflow, or 1 of 2 intervention groups. In the intervention arms, treating oncologists for genomically-matched patients with progressive disease and anticipated changes in treatment were contacted via email. In group 3, patients who met the readiness criteria were manually reviewed before the oncologists were contacted.
  • They found that, though the AI models successfully predicted which patients with active or progressive cancer may need treatment changes, sharing the trial information with oncologists did not increase trial enrollment.
  • This intervention addressed 1 barrier to trial participation. Other barriers may include eligibility criteria that goes beyond genomics and recent progression; and factors related to patient or oncologist preference, such as the motivation for participating, the complexity of the trial, and time toxicity.
  • Dr. Kehl concluded with a reminder that while AI can accelerate clinical cancer research by rapidly identifying clinical trial options for patients, impact requires integration. AI must be applied thoughtfully and continuously evaluated, and researchers should be aware of the pitfalls and shortcuts associated with the technology.

Discussion Themes

The DFCI team is currently working on MatchMiner-AI: an open-source tool that they hope will improve the accessibility of clinical trials for all patients by providing a list of relevant clinical trials. They’re running a pilot study focused on incorporating MatchMiner-AI with the historical tool.

It’s easier to train a model than it is to deploy it in a complicated healthcare context. Given that the tool performs as hoped, there are evidently implementation challenges that still need to be worked out.

The study team considered training the model on a more proximal task – i.e., “Predict whether this patient will enroll in a clinical trial.” However, they were concerned that this would introduce biases – a pertinent concern with AI models – based on which patients typically have the opportunity to enroll in clinical trials.

While there may be use cases in which providing the trial information directly to patients would be more efficient, this would need to be done carefully. Information about worsening cancer, for instance, is best contextualized in a conversation with an oncologist.

Grand Rounds February 14, 2025: A Clustered-Randomized Stepped-Wedge Pragmatic Trial to Enhance Goals-of-Care Communication for Older Adults with Cancer (ACP PEACE) (Angelo Volandes, MD, MPH; James A. Tulsky, MD)

Posted on by Sophia Ramirez

Speakers

Angelo Volandes, MD, MPH
Vice Chair of Research, Department of Medicine
Dartmouth Health
Professor of Medicine, Geisel School of Medicine

James A. Tulsky, MD
Poorvu Jaffe Chair, Department of Supportive Oncology
Dana-Farber Cancer Institute
Professor of Medicine, Harvard Medical School

Keywords

Palliative Care; Goals of Care; Advanced Care Planning; Cancer; Oncology

Key Points

  • Cancer is one of the leading causes of death for people age 65 and older, and patients have diverse preferences for medical care. However, most patients are unfamiliar with advance care planning (ACP) and clinicians are often unprepared to discuss it. Ultimately, many patients do not have goals-of-care (GOC) conversations with their providers, leading to billions of dollars in unwanted (and painful) interventions.
  • The Advance Care Planning: Promoting Effective and Aligned Communication in the Elderly (ACP PEACE) trial tested a pair of interventions – ACP Decisions and Vitaltalk – to help prepare clinicians and elderly patients to have GOC conversations.
  • ACP Decisions consists of short, accessible video decision aids that aim to help patients share in decision making. They offer a scalable, speedy strategy to improve GOC, and were associated with an increase in GOC conversations in smaller trials.
  • Vitaltalk is a communication training for clinicians discussing goals of care in serious illness. It is evidence-based, designed around the idea that learning new skills requires practice, observation, and feedback. Vitaltalk is associated with documented improvement in conversation quality.
  • ACP PEACE was a large, pragmatic, multicenter stepped-wedge trial that compared an intervention period and a control period, AKA usual care. The trial occurred across three healthcare systems (A, B, and C) and began with 36 clinics, 6 months before the onset of the COVID-19 pandemic. In April 2020, the trial was re-randomized across 29 clinics.
  • Participants were 65 and older with advanced cancer. The primary endpoint was documentation of a GOC conversation.
  • The study team found a 6.8% increase in ACP documentation during the intervention period, driven by a 6.9% increase in documented GOC conversations. There was no statistically significant increase in documentation of referrals to palliative care or to hospice. There was no statistically significant decrease in the limitation of life-sustaining treatments.
  • The results varied significantly by healthcare system. In Healthcare System A, documentation decreased by nearly 8% over the course of the study – whereas in Healthcare Systems B and C, it increased by 11% and 8.3%, respectively.
  • The researchers considered whether GOC conversation documentation was the right target. There has been a conversation in palliative care about how helpful these documents are; goal-aligned medical care (goal concordance) might be a better target.
  • Volandes and Tulsky noted that the approach to pragmatic clinical trials should better reflect the shifting nature of healthcare in the U.S. Researchers in this space must be better stewards of limited resources, so that interventions are sustainable and outcomes are useful to decision-makers.

Discussion Themes

Efforts to enhance rigor in intervention development may prevent effective interventions from getting to scale. Leaders in this space need to make it easier to get implement and test interventions, even if that means they aren’t perfectly optimized.

Right now, researchers are disincentivized to move quickly during the UG3 start-up year. Funders could incentivize research teams to move more quickly, e.g., by encouraging them to use leftover funds to conduct another study.

People in healthcare don’t see 10 years ahead – they see 1 to 2 years ahead. If pragmatic trials are meant to inform decision makers, it’s incumbent on researchers to provide the information healthcare systems need in a manner that is not outdated.

Ongoing partnerships with appropriate health systems can provide a platform for more efficient intervention implementation across different areas of healthcare. For instance, the research team conducted three other trials in Healthcare System C within the time it took them to complete ACP PEACE.

Grand Rounds February 7, 2025: Improving Symptom Control in Pediatric Cancer Patients With SSPedi and SPARK (Lillian Sung, MD, PhD)

Posted on by Sophia Ramirez

Speaker

Lillian Sung, MD, PhD
Canada Research Chair in Pediatric Oncology Supportive Care
Division of Haematology/Oncology
Chief Clinical Data Scientist
The Hospital for Sick Children
Ontario, Canada

Keywords

Cancer; Pediatrics; Cancer Symptoms; Symptom Control

Key Points

  • Symptom control in pediatric cancer patients is very poor; almost all children undergoing treatment for cancer will experience difficult symptoms, including fatigue, changes in hunger, pain, and nausea or vomiting.
  • The Multinational Association for Supportive Care in Cancer issued ambitious goals for care of cancer patients in 2030, including routine symptom screening to facilitate timely individualized care and pairing symptom identification with evidence-based treatment.
  • With this goal in mind, the research team developed Symptoms Screening and Pediatrics tool, or SSPedi. SSPedi is a validated instrument for measuring self-reported symptoms that is simple, quick, and designed for use in clinical care.
  • They integrated SSPedi with care pathways through SPARK (Supportive care Prioritization Assessment and Recommendations for Kids), a web-based application that the team used to enroll patients and remind them to report their symptoms via SSPedi.
  • The researchers conducted two randomized controlled trials (RCTs): A cluster RCT in the United States, from July 2021 to August 2023, and an inpatient RCT in Canada, from July 2018 to September 2023. In both cases, they randomized cancer patients, aged 8 to 18, to either symptom screening or usual care. In the U.S. study, they worked with the sites to adapt 14 care pathways.
  • Their primary outcome was the patient’s symptom severity, measured by their SSPedi score.
  • In both trials, the research team found a reduction in symptom severity for the intervention group. The benefits were more pronounced in the U.S. trial, with statistically significant improvements for 12 out of 15 symptoms (compared to 8 out of 15 in the Canadian trial).
  • In the U.S. study, symptom-specific interventions (i.e., treatment was clearly provided to address a given symptom) were most common for hurt or pain; changes in hunger; constipation; peripheral neuropathy; feelings of disappointment or sadness; and nausea and vomiting. The first four symptoms were also identified as being documented or treated more often in the Canadian study, in addition to feelings of crankiness or anger.
  • The research team concluded, following both the Canadian and U.S. studies, that symptom screening improves symptom control and that there was strong support for its integration into routine care.
  • There were a couple of additional takeaways from the U.S. study: 1) Care pathway use has independent effects on improving symptoms and 2) symptom screening was associated with increased emergency department visits – likely due to identification of symptoms requiring medical attention.

Discussion Themes

Despite launching right before the COVID-19 pandemic and the labor-intensive process of adapting the care pathways for each individual site, the team completed accrual for the U.S.-based-study about a year ahead of schedule. Dr. Sung attributed this to enthusiastic support from the sites around operationalizing supportive care and a study design which allowed patients to be enrolled remotely and report their symptoms remotely.

Pragmatic trials are a team sport. Having a range of expertise and experience can provide a research team with an advantage that makes all the difference. For example, Dr. Sung’s team had a connection to Epic Systems leadership, which facilitated a partnership between the two groups and allowed them to pilot SSPedi in remote care management.