For the last hundred years conventional wisdom has been that the lymph nodes under the arm (axilla) should be removed as part of the surgery for breast cancer. The development of the lumpectomy in the 1970’s, wherein only a portion of the breast is removed (the tumor plus a rim of normal tissue), did not change the concept of lymph-node removal. Oncologists have felt all along that this was probably a useful operation but no one had ever proven the point. With the advent of effective chemotherapy to prevent cancer spread, the results of the lymph node analysis helped in treatment planning; unanswered was whether that therapy obviated the need for the removal of these very nodes.
About fifteen years ago surgeons discovered that if you injected a dye or a radioisotope into the area of the breast that contained the tumor, you could identify which axillary lymph nodes were the first to drain the breast and catch rogue cancer cells. Those lymph nodes are called “sentinel nodes.” If they are free of cancer, most surgeons today do not remove the rest of the nodes in the axilla because data have shown that these nodes will very likely be normal. If the sentinel nodes are positive, conventional wisdom until now has been to remove the rest of the lymph nodes. Recently surgeons asked whether it was necessary to remove these nodes.
A large study was just published in the Journal of the American Medical Association which attempted to answer this question (JAMA 2011; 305 (6): 569-575). In it patients with small breast tumors and one or two positive axillary nodes were divided into two groups: half received no further surgery and half received removal of additional axillary lymph nodes.
Chemotherapy was given at the discretion of the oncologist. The patients in each group did equally well, calling into question the need for the additional lymph node surgery. The same week in the New England Journal of Medicine (NEJM 2011; 364 (5):412-21) a study was published which looked at the impact of tiny amounts of tumor cells in the sentinel lymph nodes – too tiny to be seen with the conventional microscope. They concluded that tiny amounts of tumor affected prognosis so minimally as to be inconsequential. As a side issue they looked at whether additional lymph node surgery was important, and as in the JAMA study, they concluded it was not.
Are there flaws in this analysis? In the more widely quoted JAMA study chemotherapy was given at the discretion of the oncologist. It was not standardized. Retrospective analysis showed that the two groups got roughly the same chemotherapy, but this approach is not as rigorous as if the chemo had been defined ahead of time. Furthermore in the JAMA trial 7% of the patients in the sentinel only group had negative nodes, even though by their defined entry criteria all should have had at least one positive node Only 1.2% of the full surgery group had negative nodes. The authors of the paper did not address this important discrepancy, nor did they address why these patients got into the trial in the first place!
So, what should the reasonably prudent surgeon do; what should the informed health care consumer insist upon?
I would submit that these studies are not definitive. Medicine changes slowly on purpose. Many times doctors regret after the fact decisions they made based on the latest observations. Fewer than 1000 patients participated in the JAMA trial. The NEJM trial was larger; however, small versus large surgery was not the principal endpoint of the study but rather an incidental finding. While I am certain that in some circles practice patterns will change, I am not certain that this is prudent yet.