A few general concepts about cancer make this question interesting. In order to surgically remove “all” the cancer, how much surgery is required, surgeons and pathologists speak in terms of the “margins” of normal tissue around resected cancers. In the 150 years since pathologists and surgeons started talking about this, we still have too little evidence about how much margin optimizes cure and minimizes mutilation. So even for surgery, “how close is close enough?” is a question still in play.

First, “cancer,” the word, comes from the latin word for crab. This highlights the characteristic of primary cancers that lies at the heart of the question. Cancer spreads into adjacent tissues with multimple “arms and fingers”. There is no precise “edge” as there often is with benign growths.

Second, cancers are usually defined operationally by their ability to skip or jump from their site of origin to more distant sites, either by the lymphatic system or the bood. This property for creating metastases can make treatment of the primary cancer irrelevent to a patient’s outcome. On the other hand, cancers can be cured if they can be completely removed surgically. The trick, of course, is to remove the cancer completely before it metastasizes. Some cancers spread before they are detectable. Others can grow to many centimeters before they spread.

Take breast cancer as an example. Because the breast is an external organ, it is inherently more easily examined, screened and treated than an internal organ like the prostate. One hundred years ago Dr. Wiliam Halsted (William Osler's surgical counterpart among the founding faculty at Johns Hopkins Medical School) reasoned that if he could remove the entire breast along with the most proximate lymph nodes where the cancer may have spread, he could cure some breast cancers. The “radical mastectomy” (itself a imprecise term since there are many variations of the actual operation) became the standard treatment for breast cancer for the first 75 years of the last century.

Halsted’s patients generally had large and locally advanced cancers.  A big surgery was often required just to clear the macroscopic tumor.  Over the past 25 years, as cancers were found earlier and radiation treatment became widely available (in the U.S. at least), smaller operations (lumpectomy) followed by radiation to the remaining breast and lymph nodes largely replaced the mastectomy. (Ironically, in the past few years more sensitive imaging such as breast MRI has shown that tiny other breast lesions often exist, and the mastectomy is making a come back as a result (though not based on much evidence)).

So hewing to the cry of “evidence based medicine”, quality in the ‘90’s became defined by a reduction in mastectomies. This was so because a large cooperative trail group (NSABBP) “proved” that lumpectomy plus radiation was as good as mastectomy. (These trials had their problems, including a notorious episode of fudged randomization by a surgeon who “knew” lumpectomy was better before the trial was done).

But here is the rub, and it applies to breast cancer surgery in the same general way as radiation. Some breast cancers have such a good prognosis that the choice of local treatment is unlikely to make much difference. Some breast cancers have such a poor prognosis that the choice of local therapy is also unliklely to make much difference (since the disease has already metastasized, whether or not that fact can be known at the time).

Have we excluded the possibility that there is a group of patients in the middle where the “right” local treatment is crucial? I don’t think so. If such a group exists, how can we identify it? Anatomic staging alone (tumor size, lymph node status) probably won’t work…still too much heterogeneity in prognosis. Tumor characteristics ranging from hormone receptor status to profiles based on hundreds of genes might work. Time will tell, if the proper studies are designed and well executed.

Here I have focused on surgery as it relates to “How close is close enough?” I did so because surgeons need little knowledge of physics. Next comes an examination of the question as it relates to the Proton Antiproton Conundrum in cancer therapy.