A new tool for specifically assessing the risk of toxicity in older, early-stage breast cancer patients who are candidates for chemotherapy performs significantly better than an earlier iteration of the tool for a range of solid tumors and better than the widely used Karnofsky Performance Status method, new research indicates.
“This tool could be considered as a part of adjuvant treatment decision-making,” concluded study presenter Allison Magnuson, DO, of the James Wilmot Cancer Center, University of Rochester in New York.
The tool’s primary purpose is not to reduce or forgo chemotherapy in breast cancer patients, she said, but to illuminate domains associated with impairment, allowing targeted supportive care and helping patients “get through the treatment they need.”
She spoke here during an oral presentation at the San Antonio Breast Cancer Symposium (SABCS) 2018.
Magnuson was the second author of the study, but took the place of admired and respected first author Arti Hurria, MD, of the City of Hope Comprehensive Cancer Center in Duarte, California, who died last month in an automobile accident.
The new study “truly exemplifies [Hurria’s] passion for improving our knowledge about caring for older adults with cancer,” said Magnuson, holding back tears and her voice quivering.
The new breast cancer tool, known as the Cancer and Aging Research Group-Breast Cancer (CARG-BC), is in keeping with lead author Hurria’s articulated vision for geriatric oncology, which she had once described at an event at her home institution.
“The dream” and her mission, said Hurria, was “that all older adults with cancer will receive personalized, tailored care, utilizing evidence-based medicine with a multidisciplinary approach.”
Magnuson explained how CARG-BC offers such an approach. There a lot of tools in breast cancer to evaluate the benefits of treatment, she said. But older adults also need their toxicities/risks gauged. There are “few” tools available to assess both risks and benefits of adjuvant treatment in these older patients, Magnuson noted, and yet they are at increased risk of toxicities compared to younger, fitter patients.
Meghan Karuturi, MD, of the University of Texas MD Anderson Cancer Center in Houston, agreed: “Our ability to predict side effects in older breast cancer patients is very limited.”
Karuturi, who was not involved with the study, spoke with Medscape Medical News at the San Antonio meeting and said the field of geriatric oncology is “very enthusiastic” about the results and called the new study “a major contribution in the field.”
Karuturi said that these breast cancer patients are “generally going to do great.” Oncologists don’t want to undertreat but they don’t want to debilitate patients, either, she said.
She added that geriatric oncology greatly misses Hurria: “The impact this woman made is astounding.”
The CARG-BC presentation in San Antonio is a “beautiful story,” Karuturi said, “because one her mentees [Magnuson] is presenting her final work.”
In order to develop the new predictive tool, the team enrolled 473 stage I to stage III patients with a median age of 70 years scheduled to receive standard chemotherapy regimens at 16 US institutions: 283 patients in a development cohort and 190 in a validation cohort.
The enrolling physicians assessed their patients before and after chemotherapy. “Treatment toxicity was common,” said Magnuson, with 46% of all patients having grade 3 to 5 toxicities. Also, among other measures, 24% discontinued treatment and 23% were hospitalized.
Next, in the data-crunching phase of the study, the authors first evaluated patients with the original CARG tool, which was intended for patients with various solid tumors, and found the older tool was significantly associated with grade 3 to 5 toxicities, with an area under the curve [AUC] of 0.64 (95% confidence interval [CI], 0.57-0.70).
They then added breast cancer tumor and geriatric assessment variables to the older model. For grade 3 to 5 toxicities, this new CARG-BC tool had an improved AUC of 0.76 (95% CI, 0.70-0.82).
The CARG-BC score ranged from 0 to 19, (low risk 0-5, middle risk 6-9, high risk 10+) and was significantly associated with grade 3 to 5 toxicities (P < .001) but, comparatively, the Karnofsky Performance Status was not (AUC 0.53; 95% CI, 0.48-0.59; P = .21).
The external validation AUC was not statistically different from the development AUC, the authors report in their abstract.
A higher CARG-BC score was associated with dose delay, dose reduction, hospitalization, and relative dose intensity (RDI) <85% (all P-value < .001).
The team also reports that 79% of patients who had a high, pretreatment CARG-BC score went on to have grade 3 to 5 toxicities vs 45% of those with a medium score and 21% of those with a low score (P < .001)
“We have developed and validated a chemotherapy toxicity risk score (CARG-BC) that predicts toxicity in older adults with stage I-III breast cancer,” summarized Magnuson.