Source: www.medpagetoday.com
Author: Michael Smith, North American Correspondent, MedPage Today

In early oral squamous cell carcinoma, a sentinel node biopsy correctly predicted an absence of lymphatic metastasis in all but 4% of patients, researchers said.

For T1 and T2 lesions that were clinically node-negative, the procedure — combined with additional sectioning and immunohistochemistry — yielded a negative predictive value of 96%, according to Francisco Civantos Jr., MD, of the University of Miami, and colleagues.

For T1 lesions, the value was 100%, while for T2 cancers it was 94%, the researchers reported online in the Journal of Clinical Oncology.

The finding may position the procedure as an intermediate option between watchful waiting and selective neck dissection, the researchers said, asserting that it’s now “reasonable” to conduct a head-to-head trial of sentinel node biopsy and neck dissection.

The procedure has significantly increased the sensitivity for detecting lymphatic metastasis in melanoma and breast cancer patients, Civantos and colleagues noted.

But in oral cancer, many surgeons prefer a completion neck dissection, they added, despite the “measurable morbidity” that’s associated with the procedure. On the other hand, because of that morbidity, other specialists prefer watchful waiting and elective neck irradiation.

To investigate the issue, Civantos and colleagues conducted a multicenter trial in which patients with early invasive oral cancers were treated with both procedures — a sentinel node biopsy, followed by completion selective neck dissection.

The primary goal was to see if a negative hematoxylin and eosin finding on the sentinel node biopsy accurately predicted the negativity of the other cervical lymph nodes removed in the neck dissection.

All told, 140 patients qualified and had the dual procedures, the researchers reported.

The sentinel nodes were identified using a radioactive gamma probe. The primary tumor was removed transorally, followed by the sentinel node biopsy through a small incision within the area of the planned incision for the neck dissection.

Staining of the sentinel nodes at the various trial sites resulted in 106 that were negative. Of those, 100 were also negative by hematoxylin and eosin staining of the neck dissection specimens.

That yielded a negative predictive value of 94%, the researchers said.

Additional step sectioning and immunohistochemistry at a central pathology lab increased the negative predictive value to 96%, they said.

Both findings were significant, they reported, with a one-sided P-value of P<0.0001.

One limitation of the study, the researchers noted, is that the dual procedures may have interfered with each other, in that sentinel lymph biopsy might have changed the way the neck dissection was performed or the other way around.

But that “may actually lead to underestimation of the accuracy of this technique,” they said, since the neck dissections were guided by information gleaned from nuclear imaging and the gamma probe used in the sentinel node procedure.

The study was also limited, the researchers said, because many surgeons involved were only moderately experienced and none was experienced “at levels currently considered appropriate for surgeons caring for breast cancer or melanoma.”

Nonetheless, they said, the negative predictive value found in the study was “higher than anticipated for a multi-institutional setting with relatively inexperienced surgeons.”

They added that only a clinical trial in which outcomes after a negative sentinel node biopsy are simply observed for several years would yield a true negative predictive value for the procedure.

Notes:
1. Reviewed by Zalman S. Agus, MD; Emeritus Professor University of Pennsylvania School of Medicine and Dorothy Caputo, MA, RN, BC-ADM, CDE, Nurse Planner
2. Source: Civantos FJ, et al “Sentinel lymph node biopsy accurately stages the regional lymph nodes for T1-T2 oral squamous cell carcinomas: Results of a prospective multi-institutional trial” J Clin Oncol 2010; DOI: 10.1200/JCO.2008.20.8777.