A new set of nomograms based on easily assessed risk factors accurately predicts a woman’s likelihood of developing lymphedema after axillary lymph node dissection for breast cancer, researchers are reporting at a breast cancer symposium sponsored by the American Society of Clinical Oncology.
The nomograms, which enable risk assessment throughout the course of treatment using data available at the time, had accuracy exceeding 70%, according to results presented in a press briefing before the meeting.
"The accuracy is the same as mammography to detect breast cancer. ... So that’s a very good statistical tool," said lead author Dr. Jose Bevilacqua, a surgical oncologist at the Hospital Sirio-Libanês in São Paulo, Brazil.
Being able to identify high-risk patients has several potential applications, he added. They could be monitored more closely and offered interventions that might prevent or reduce the severity of lymphedema, such as use of compression sleeves. "The earlier you detect [it], the better is the outcome," Dr. Bevilacqua said.
Additionally, high-risk patients with a positive sentinel lymph node could be counseled about the risk of further surgery. Thus, "you have physicians who make closer follow-up and [have] a better discussion with the patients."
Finally, the nomograms could be applied in clinical research. "These tools might help [ongoing] or new studies to select high-risk patients in order to avoid exposing the low-risk patients to unproven therapies," he explained.
Dr. Andrew Seidman, moderator of the press briefing and a medical oncologist at the Memorial Sloan-Kettering Cancer Center in New York, commended the investigators for a study that "has real practical implications for patients."
"It allows us to identify patients who then can be appropriately triaged for early intervention and perhaps for clinical trials aimed at preventing what many consider inevitable, and that is the development of lymphedema."
The effectiveness of early interventions is controversial, according to Dr. Seidman, who did not report any relevant conflicts of interest. "There are believers and there are nonbelievers," he commented. "What I think this nomogram does is [allow] us in a scientific way to segregate out patients in terms of their risk, and by doing that, we can have more uniform cohorts for which early intervention can be studied more rigorously."
The researchers prospectively followed 1,054 women with unilateral breast cancer who underwent breast-conserving surgery or mastectomy with an axillary lymph node dissection in 2001-2002. Median follow-up was 41 months.
"As far as we know, we have established the largest prospective cohort specifically established to study the incidence and factors associated with lymphedema after axillary node dissection for breast cancer," commented Dr. Bevilacqua.
Using Data to Predict Lymphedema
The women had serial arm volume measurements starting before surgery. To facilitate this process, the researchers created a tool that is now available free online (www.armvolume.com).
They then developed and internally validated three multivariate nomograms (or statistical models) for predicting lymphedema using data available at various time points.
The first model, to be used preoperatively, incorporated age, body mass index, and number of cycles of neoadjuvant chemotherapy infusions in the ipsilateral arm.
The second model, to be used within the first 6 months after surgery, incorporated all the factors from the first plus the extent of axillary dissection, the location of radiation therapy field, and the number of cycles of adjuvant chemotherapy infusions in the ipsilateral arm.
The third model, to be used 6 months or later after surgery, incorporated all the factors from the second, plus the development of postoperative seroma and development of early edema.
Study results showed that within 5 years of axillary lymph node dissection, 30.3% of the women developed lymphedema (defined as a difference in volume of at least 200 mL between arms at 6 months or later after surgery).
All of the risk factors used in the models were significantly associated with the development of lymphedema, Dr. Bevilacqua reported.
"One of the novelties of our study is the [finding] that ipsilateral neoadjuvant chemotherapy infusion is as morbid as adjuvant chemotherapy infusion to increase the risk of lymphedema," he pointed out. "This is somewhat intuitive, but it has never been described, and the great majority of clinical oncologists are not aware of this fact."
For predicting the 5-year risk of lymphedema, the first, second, and third models had accuracy of 70.6%, 72.9%, and 73.6%, respectively. The investigators have converted the models into free, user-friendly calculators that are available online (www.lymphedemarisk.com) during the ASCO Breast Cancer Symposium (Sept. 8-10), but will be taken off line until the manuscript is published, after which they are to become available again.