Real-World Data and AI Help Improve ILD Management in HER2+ Breast Cancer

3D rendered medically accurate illustration of breast cancer: © SciePro - stock.adobe.com

A significantly lower incidence of interstitial lung disease (ILD) was seen among patients with HER2-positive (HER2+) breast cancer in the real world at 4.17% compared with the prior 10% to 14% reported in published literature (n = 24), according to new findings from Project EVOLVE presented at the 2024 San Antonio Breast Cancer Symposium (SABCS).¹

Within the study, the integration of artificial intelligence (AI)-powered health record analytics with targeted educational interventions was also shown to better ILD management and improve both patient and healthcare provider outcomes.

“This innovative approach presents a promising model for addressing complex adverse events in oncology and beyond, particularly around serious adverse events, paving the way for more personalized and effective cancer care strategies that prioritize patient safety and treatment efficacy,” wrote the authors of the poster presented at 2024 SABCS.

HER2-targeted therapies like trastuzumab (Herceptin) and fam-trastuzumab deruxtecan-nxki (Enhertu) have paved the way for the treatment of HER2+ breast cancer over recent years by helping to improve survival rates. However, these therapies are often linked to ILD, a rare but serious adverse event (AE) that requires careful monitoring and management. Established guidelines for ILD treatment exist; however, experts often find it difficult to recognize in real-world settings.

The real-world data analysis phase of Project EVOLVE included over 500,000 patient records, showing an ILD incidence of 4.17% among respiratory AEs, which was much lower than the rates reported in prior clinical trials. Other respiratory AEs observed included asthma (37.5%), the common cold (20.8%), and COVID-19 (8.3%).

A total of 65.79% of patients lacked COVID-19 vaccination history. In a secondary survey that included 187 patients, fatigue was cited as the most challenging AE among 67% and 38% lacked awareness of cancer subtypes and stages.

Another finding noted in the study was that healthcare providers reported significant improvements in their ability to manage HER2+ breast cancer treatment and associated AEs after participating in the program. The confidence in managing ILD and other respiratory AEs was shown to increase from 66% to 90%, and the ability to correctly identify ILD as a boxed warning for trastuzumab deruxtecan improved from 19% to 47%. Providers also reported greater confidence in tailoring treatment plans to individual patient needs following the program.

In addition, patients benefited from the education with confidence in adhering to treatment and managing AEs at home increased from 55% to 79%. Over half of the patients surveyed (53%) cited the educational modules as instrumental in helping them better understand their condition, treatment options, and side effect management.

The lower ILD incidence observed in this study highlights the value of real-world data in providing unique insights that differ from those seen in clinical trials. This can help potentially reshape AE monitoring protocols.

Further, this data-driven education model offers a scalable, practical approach to managing serious AEs among patients with cancer and shows that as personalized care advances, AI and targeted education can help enhance patient outcomes.

REFERENCE

1. Shankar S, Zalis J, Carter J, et al. EVOLVE: Harnessing the power of AI, real-world data, and targeted education to minimize breast cancer care disruptions. Presented at: San Antonio Breast Cancer Conference; December 10-13, 2024; San Antonio, TX. P4-05-15