麻豆传媒

AMSTERDAM -- Patients with early breast cancer had significantly less lymphedema if they received axillary radiotherapy instead of surgical lymph node dissection, results of a randomized trial showed.

The incidence of clinically significant lymphedema ranged from 21% to 25% over 5 years in patients who underwent surgical dissection of axillary lymph nodes as compared with 10% to 15% among patients who had radiotherapy.

Patients who had both lymph node dissection and radiotherapy had the highest rates of lymphedema.

A previous analysis of the data showed that surgery and radiation therapy achieved similar 5-year disease control, Mila Donker, MD, of Netherlands Cancer Institute in Amsterdam, reported here at the European Cancer Congress.

"Axillary lymph node dissection (ALND), compared to axillary radiotherapy (ART), is associated with a 2.5 times higher baseline surgical complication rate and a two times higher rate of lymphedema at 5 years," said Donker. "Axillary lymph node dissection plus axillary radiotherapy is associated with a lymphedema rate that is five times higher than axillary radiotherapy at 5 years.

"The type of axillary treatment is the strongest risk factor for both lymphedema and shoulder function."

Sentinel lymph node biopsy has supplanted ALND as the preferred approach to management of patients who have early breast cancer and clinically negative axillary lymph nodes. For patients with positive sentinel lymph nodes, ALND and ART provide similar axillary disease control.

ALND carries a well-recognized risk of side effects, including potentially severe lymphedema. In contrast, few studies have examined side effects associated with ART for patients with positive sentinel lymph nodes, said Donker.

Study Details

The randomized AMAROS trial compared ALND and sentinel lymph node biopsy (SNB) in patients with early breast cancer. Additionally, patients with positive sentinel nodes were randomized to surgical axillary dissection or radiotherapy. The trial involved 4,800 patients, 1,400 of whom had positive sentinel nodes.

As reported earlier this year, the trial showed similar low rates of axillary recurrence with the two treatment strategies. The 5-year disease-free and overall survival did not differ significantly between the groups.

The AMAROS results confirmed those from the ACOSOG Z0011 trial, which also showed no difference in recurrence rates with surgical dissection or radiation therapy to the axilla.

Updated AMAROS results with follow-up beyond 10 years showed 5-year axillary recurrence rates of 0.43% with surgical dissection and 1.19% with radiation therapy, Donker reported. She also reviewed complication rates for patients who underwent upfront ALND versus SNB.

ALND was associated with significantly higher rates of hemorrhage (3.1% versus 1.7%, P<0.001), infection (10.7% versus 3.8%, P<0.001), persistent seroma (10.4% versus 1.3%, P<0.001), and early lymphedema (1.6% versus 0.2%, P=0.007). The total complication rates were 22.6% with ALND and 9.0% with SNB (P<0.001).

Lymphedema was assessed at 1, 3, and 5 years and compared among patients who underwent ALND, ART, or both forms of treatment. The assessment included clinical observation and measurement of arm circumference at multiple points.

By clinical observation, lymphedema rates with ALND were 25.6% at 1 year, 21% at 3 years, and 20.8% at 5 years. That compared with rates of 15%, 13.4%, and 10.3% at the same time points for ART.

Patients who underwent both ALND and ART had lymphedema rates of 59.3%, 44.8%, and 58.3%.

Lymphedema also was defined as an increase in arm circumference ≥10%. By that definition, the rates at 1, 3, and 5 years were 7.2%, 9.2%, and 11.7% with ALND, 5.9%, 6.2%, and 5.7% with ART, and 14.8%, 24.1%, and 29.2% with ALND plus ART.

Assessment of shoulder function comprised anteversion/retroversion and abduction/adduction function. Values did not differ significantly between ALND and ART at 1, 3, or 5 years.

Multivariate analysis identified four factors associated with lymphedema: menopausal status, body mass index, treatment on the dominant side, and type of axillary treatment. Predictors of impaired shoulder function were type of axillary treatment, supraclavicular radiation therapy (in the absence of ART), and level of ALND (I + II versus I + II + III).

Discussants Weigh In

Reviewing results of AMAROS and other trials, invited discussant Peter Dubsky, MD, of the Medical University of Vienna, said the weight of the evidence supports the conclusion that "ART is the recommended treatment in breast cancer patients with positive sentinel nodes."

However, session moderator , questioned whether the data represented a fair comparison of ALND and ART. He noted that two-thirds of the ALND group had level III dissections, whereas rates in the single digits have become the standard throughout Europe. The infection rate of almost 11% also seemed out of line with contemporary practice.

"If you compare what you said is the 'new preferred treatment' to something that is hampered by these factors, the conclusion might be a little different," said Gnant, also of the Medical University of Vienna.

Donker said the high rate of level III dissection reflected contemporary practice when the trial began. However, she argued, the lymphedema is not associated with the extent of dissection. With respect to the infection rate, she said participating centers throughout Europe applied local definitions of infection, which might have confounded calculation of the overall rate.

A surgical oncologist from the U.S. revisited the primary results of AMAROS and focused on the assumption that ALND and ART lead to similar disease control. In the presentation of the primary outcome data, AMAROS investigators acknowledged that the low event rates in both arms left the trial underpowered to demonstrate non-inferiority.

"So the major question remains whether the two treatments are equivalent in terms of cancer outcomes," , of Magee-Women's Hospital and the University of Pittsburgh, told 麻豆传媒 in an email. "I believe that they likely are, but this trial had the same pitfalls as ACOSOG Z0011: low event rate of axillary recurrence and use of a non-inferiority test to establish the two treatments to be equivalent.

"Statisticians tell us that a non-inferiority test is inferior to a superiority test. It requires fewer patients, but can often fail to detect small differences.

"The data definitely support the theory that radiation is equivalent to surgery in the node- positive patient, but both AMAROS and Z0011 will benefit from longer follow up. For some patients, this will be very important as they may be willing to accept a possibly increased, but still very low risk of recurrence to decrease their risk of lymphedema significantly."

Comment