Yale Data Informs Plans for a Lymphedema Prediction Tool

Siba Haykal, MD, PhD, is determined to reduce the percentage of patients suffering with lymphedema—a chronic, debilitating side effect of breast cancer surgery—through a more robust prediction tool that could lead to earlier interventions.

As chief of Yale’s Reconstructive Oncology section within the Yale School of Medicine’s Department of Surgery, Haykal has seen too many women with the often-painful swollen limbs that typify lymphedema. Although not life-threatening, lymphedema is frustrating and disruptive to the lives of millions. It also is incurable.

Haykal and her team are finalizing a new predictive tool to identify those at risk after surgery, another step in shifting the power from lymphedema to surgeons and patients. Earlier intervention based on better probability could head off lymphedema through preventative measures, including exercise and physical therapy.

“Our tool represents an improvement over the nearly 40 other models currently available because it adds new predictors and can forecast the time-to-diagnosis of lymphedema in individual patients,” she says. “This will help us develop better strategies for enhancing patient care.”

The new model is drawn from a deep patient data base. Haykal and her team analyzed a large dataset of 15,666 patients at Smilow Cancer Hospital at Yale New Haven Health system from 2013 to 2024. The researchers say it is the largest single-center investigation of its kind to date with three times the cases studied at other cancer centers, resulting in a more statistically defensible evidence base.

The data review was focused specifically on patients who underwent axillary lymph node dissection (ALND) during breast cancer surgery. In ALND, surgeons remove and test lymph nodes from the underarm area (axilla) to determine if a patient’s cancer has spread. Among these patients, about 15 percent, or 2,345 individuals, developed lymphedema, with an average onset of 20.5 months post-surgery, the data review found.

After narrowing the data to patients who had undergone ALND, Haykal and her team analyzed a wide range of potential factors, including age, high Body Mass Index (which strains the lymphatic system), race, and types of cancer treatments, to assess their impact on lymphedema development and timing after ALND. They also examined diabetes status and HbA1c (blood glucose) levels, given the role of diabetes in poor circulation and inflammation, which may worsen lymphedema.

“Our study highlights the need for a multidisciplinary, data-driven approach to predictive tools for lymphedema,” Haykal says. “Integrating our findings into algorithms enables personalized risk assessments to guide early interventions, such as exercise and physical therapy, which improve lymphatic flow and muscle strength. Recognizing disparities in risk, based on race and co-morbidities such as diabetes, can also help promote equitable care.”

The data showed that patients with diabetes, who had chemotherapy or radiation treatments, who have elevated HbA1c levels, who have a Body Mass Index over 30, and who are Black/African American were all at risk for developing lymphedema. In a novel finding, the research team is the first to identify elevated HbA1c levels as an independent predictive marker for lymphedema, even in patients without diabetes.

Unlike most other studies, the Yale team also pinpointed predictors for both earlier and later than average emergence of lymphedema. Race (Black/African American and Asian) and radiation treatment were augurs of early presentation while diabetes, BMI over 30, and elevated HbA1c levels were indicative of later presentation.

The lymphatic system, the body’s internal waste-removal network, comprises tiny tubes and filters called lymph nodes. Sometimes dubbed “the highway of health,” the tubes convey a clear liquid called lymph, which vacuums up waste, toxins, and germs from the body.

When the lymphatic system is damaged, often from breast cancer treatments such as lymph node removal or radiation therapy, fluid can accumulate, causing swelling, usually in the arm or hand on the side where the cancer was treated. The swelling, pain, and heaviness in the affected limb can make everyday tasks difficult, from buttoning a shirt to carrying groceries.

Daily management of lymphedema often involves wearing compression garments, undergoing manual lymphatic drainage therapy, and taking precautions to prevent infections or injuries that could lead to dangerous flare-ups.

Many women see lymphedema as nearly as bad as breast cancer—or worse in certain ways. Unlike breast cancer, which has a defined treatment plan and the possibility of remission, lymphedema requires constant management.

The search for relief could account for the 38 different prediction models for lymphedema. A review of the existing predictive models found there was a need for a model that was robust where others fell short because:

• Many were based on small patient cohorts, resulting in insufficient statistical power and which did not, or could not, assess important variations within sub-populations of patients including race, ethnicity, and co-morbidities.
• The tools generally failed to predict the timing of lymphedema diagnosis, potentially hampering early intervention.
• Many have short follow-up periods and lack laboratory-based predictive markers that could provide more precise risk assessments.

Haykal and team are finalizing their findings in preparation for submitting them for publication.