During the 19th, and for most of the 20th century, malignant tumors were removed by mutilating radical anatomic dissection. Advances such as anesthesia, asepsis, and blood transfusion made possible increasingly more radical operations. There was no scientific rationale for the operations being performed. Surgery in the 20th century was dominated by the principles of William S. Halsted, who contended that the bloodstream was of little significance as a route of tumor cell dissemination; a tumor was autonomous of its host; and cancer was a local-regional disease that spread in an orderly fashion based on mechanical considerations. Halsted believed that both the extent and nuances of an operation influenced patient outcome and that inadequate surgical skill was responsible for the failure to cure. A new surgical era arose in 1957, when cancer surgery began to be influenced by laboratory and clinical research, with results contrary to Halstedian principles. A new hypothesis resulted in a scientific basis for cancer surgery. Clinical trials supported the thesis that operable cancer is a systemic disease and that variations in local-regional therapy are unlikely to substantially affect survival. Complex host-tumor relationships were shown to affect every aspect of cancer and, contrary to Halsted's thesis, the bloodstream is of considerable importance in tumor dissemination. Clinical trials also have shown that less radical surgery is justified. Studies have shown that improved survival can be achieved with systemic therapy after surgery. Such therapy can reduce both the incidence of distant disease and the tumor recurrence at the tumor site after minimal surgery. The use of systemic therapy in patients who have no identifiable metastatic disease is a drastic departure from previous strategies. New technological innovations resulting from engineering research have improved the quality of life of patients by eliminating the need for some surgical procedures. Because cancer is apt to be a systemic disease, however, clinical trials are necessary to determine the effect of these modalities on patient outcome. Although technological developments will continue to play a role in cancer therapy, research in molecular biology and genetics will dictate the future status of cancer treatment and, ultimately, the future of surgery. [Cancer Res 2008;68(24):10007–20]
- breast cancer biology
- radical surgery
- systemic therapy
It is of historic interest that the American Association for Cancer Research (AACR) was founded by four surgeons, five pathologists, one chemist, and a biochemist at the 25th meeting of the American Surgical Association in Washington, DC, on May 7, 1907. The aim of the newly formed organization was to improve cancer research and treatment of cancer and to promote prevention. In recognition of the 100th anniversary of the AACR, this article will examine the role biological research has played in the evolution of cancer surgery.
Four principal aspects with regard to biological research will be addressed: (a) whether research played a part in instigating surgical advances that occurred during the 19th and the first half of the 20th centuries; (b) whether research formed the basis for the concepts advocated by William S. Halsted, which governed the practice of surgery for the first three quarters of the 20th century; (c) a description of findings from my biological research, which I began in 1957, and which was instrumental in replacing Halstedian principles with the guidelines that have been largely responsible for advancing surgical practice to its present status; and (d) a consideration of the way in which biological research is currently being subsumed by the technological advances that have initiated a new era in cancer surgery.
Major Advances in Cancer Surgery (19th and 20th Centuries)
The history of cancer surgery has usually been presented chronologically in a series of timelines that catalogue certain “landmark” events during the 19th and 20th centuries ( 1– 6). This article in the AACR Centennial series also includes such a timeline, with both the advances in surgery and the nonsurgical events that contributed to making performance of radical surgery more feasible ( Fig. 1 ).
To gain a better perspective on surgical advances made in the treatment of tumors that arose at a particular site, however, surgical events from timelines are also listed according to the site of tumor origin and the principal surgeon responsible for the advance ( Table 1 ). In addition, I reviewed articles written by the individuals who “fathered” the landmark events and examined numerous articles written by others about those achievements, with the aim of discovering what part, if any, biological research (science) might have played.
Surgical Advances by Tumor Site
Lung. Surgeons who altered the treatment of lung cancer by performing lobectomy, pneumonectomy, or segmental resection of the lung received considerable recognition. Hugh Davies performed the first lobectomy in 1912 using anatomic dissection ( 7, 8). In reporting the results of Davies' operation, Naef wrote, “If his patient had not died 28 days after the operation, he would have preceded Evarts Graham's first lung resection by 21 years” ( 8). Although Davies also received acclaim for being “… the earliest advocate of interdisciplinary teamwork in thoracic medicine” ( 8), his publication provides no information related to lung cancer other than a technical description of how he performed the operation. A report by Harold Brunn of a single cancer patient who had a one-stage lobectomy provides information only about an operative technique he used ( 9).
Although Rudolph Nissen was credited with performing the first total pneumonectomy in 1931 ( 10), he was recognized for an operation that involved the first stage of a pneumonectomy on a 12-year-old girl who developed a lung abscess following trauma ( 11). It is interesting to note that it was Nissen himself who best described the status of pneumonectomy when he stated, “… the pneumonectomy operation… was far from exhibiting a work of fine or polished technique, and, if merit was attached, it rested on the daring nature of the procedure, and requisite courage reflected more credit upon the patient than the physician” ( 11). The report of Evarts Graham ( 12), who was noted for the successful performance of the first one-stage pneumonectomy for lung cancer, was limited entirely to a single case anecdote that described both the technical procedure and the patient's medical history. Strangely enough, no reference to cancer or its treatment appeared in a 1939 report by Edward Churchill and Ronald Belsey of a segmental resection of the lung. Their report described only the technical aspects of that procedure and indicated only that, “… the bronchopulmonary segment may replace the lobe as the surgical unit of the lung ” ( 13).
Rectum. In 1908, W. Ernest Miles performed an abdominoperineal resection to remove a cancer of the rectum ( 14). At that time, he indicated that the reason techniques used for removal of rectal cancers failed was because they did not completely eradicate the upward spread of cancer from the rectum. He stated, “… removal [of the zone of the upward spread] is just as imperative as is the thorough clearance of the axilla in cases of cancer of the breast if freedom from recurrence is to be hoped for” ( 14). Not until the 1950s did Harry Bacon begin to alter Miles's operation when he advocated the use of an anal sphincter-preserving procedure ( 15). Bacon was permitted to conduct his procedure only because of the development of certain technical advances—that is, the successful introduction of anterior resection and end-to-end anastomoses, the introduction of the pull-through excision for cancers of the middle part of the rectum, and the use of a stapling device ( 16).
Thyroid. Theodore Kocher was the first surgeon to receive the Nobel Prize for his “… work on the physiology, pathology and surgery of the thyroid gland.” In fact, his efforts actually related to the surgical treatment of goiter, not cancer of the thyroid ( 17). In an evaluation of Kocher's contribution to thyroid surgery, it has been noted that, despite his meticulous surgical technique, problems did occur. Kocher himself stated, “In technical terms we have certainly learned to master the operation for goiter… But something else has happened… Removal of the thyroid gland has deprived my patients of what gives them human value. I have doomed people with goitre, otherwise healthy, to a vegetable existence. Many of them I have turned to cretins, saved for a life not worth living …” Kocher was devastated by these findings and wrote a candid report in 1883 urging surgeons not to perform “total strumectomy [thyroidectomy]” ( 17).
Pancreas. In 1935, Allen Whipple published information from three patients who had been subjected to a two-stage radical pancreaticoduodenectomy. One patient died postoperatively, one died at 8 months from cholangitis, and a third survived 28 months before dying of liver metastases ( 18). In 1945, Whipple described the first case of a one-stage procedure for removal of a tumor of the pancreatic islet cells ( 19). Over the years, he made numerous anatomic modifications to this operation, and, although it represented a technical feat, there is no evidence that it was supported by any findings from biological research. Whipple's operation continued to be modified and is currently being performed to remove acinar and islet cell tumors of the pancreas.
Head and neck. In 1987 Bobby R. Alford wrote, “The success of the modern head and neck oncologic surgeon in curing patients with head and neck cancer ‘exclusive of the thyroid gland’ is due in large part to the principles that evolved from the astute clinical observations of George W. Crile” ( 20), which were published in 1906. Crile, who later founded the Cleveland Clinic, was one of the four surgeons who established the AACR in 1907 and one of a very few who clearly outlined the concepts that led him to perform radical head and neck operations for cancer. Crile used a radical surgical treatment that incorporated “a block dissection of the regional lymphatics as well as the primary focus [and that] was predicated on exactly the same lines as the Halsted operation for cancer of the breast” ( 21). Thus, it is highly likely that Halsted's principles dictated the efforts of Crile, just as they had governed the work of Miles and other surgeons.
Larynx. The British surgeon Wilfred Trotter pioneered an operation for resection of tumors of the oropharynx, an account of which he published in 1920. Trotter insisted on excision of tumors “regardless of anatomic boundaries” ( 22), a procedure that was, perhaps, based on Halsted's concepts. Because there were no antibiotics available at the time, “he [Trotter] advocated extraction of all teeth before operation, and relied on elective tracheostomy to secure the airway and reduce infection” ( 22).
Esophagus. On March 14, 1913, Franz Torek performed the first successful transthoracic resection of a squamous cell carcinoma in the thoracic portion of the esophagus. Howard Lilienthal, who performed the first extrapleural esophagectomy in 1921, noted that neither Torek nor any other surgeon was able to save any patient by the transthoracic approach. As gleaned from available information, Torek's operation, which he performed in 2 hours, 27 minutes, was a feat of both skill and daring ( 23). It is interesting to note that, at the end of the operation, the patient was given an enema containing hot coffee, whiskey, and strychnine!
Uterus. Although numerous gynecologists performed abdominal hysterectomies at the end of the 19th century, foremost among them was Ernst Wertheim, a Viennese surgeon, whose efforts, in about 1906, gave rise to what was known as the Wertheim operation. Wertheim's landmark accomplishment was the radical hysterectomy, which was based on his premise that, “one had to strive to remove as much as possible the surrounding tissue together with the primary tumor to achieve better results, as is the case with operative procedures for cancer of other organs” ( 24). It seems highly likely that Halsted's views about breast cancer might have influenced Wertheim's thinking. Regarding the regional lymphatics, however, Wertheim questioned that view by stating that, “In opposition to the view that removal of the entire lymphatic system is the foremost requirement is the fact that it is neither possible nor necessary. Even if one admits that in a few cases carcinomatous metastases may be found in nodes slightly enlarged… carcinoma is never found in spindle-shaped nodes normally present in the pelvis” ( 24).
Prostate. In 1904, Hugh Young, who had assisted William Halsted with his operations for cancer of the breast at Johns Hopkins University, was, in turn, assisted by Halsted when he performed the first radical suprapubic prostatectomy. It has been suggested that, “Dr. Halsted's new aggressive and radical operations, such as mastectomy with lymph node dissection en bloc, probably stimulated Young to apply similar concepts when he developed his own urologic operations later on” ( 25).
Skin. The contribution of Frederick Mohs to the treatment of basal cell carcinoma has often been noted as a landmark event on various timelines. In brief, the Mohs micrographic surgical procedure is strictly a technical process and has no relationship to the biology of the type of skin tumors that are being removed ( 26). It is currently being used to guarantee that the margins of resection are free of tumor and that surgeons remove only as much tissue as is needed.
Lack of biological rationale. As a consequence of my investigation of the events that were considered to be turning points in the evolution of cancer surgery during the 19th and the first half of the 20th centuries, I concluded that, aside from the desire to eradicate more tumor, there was no biological rationale for the radical surgery that was being performed. My view is in accord with that of Lawrence and Lopez ( 27), who have stated that, “With few exceptions, all of the surgeons who performed [landmark] operations focused only on the feasibility of removing the part of the anatomic structure containing the cancer” ( 27). Lawrence and Lopez also noted that, “all of those procedures were established on the basis of either ‘armchair’ logic or limited clinicopathologic studies” ( 27) and were “purely empirical” ( 27), that is, were unrelated to knowledge obtained from experimentation. Thus, it is evident that laboratory and clinical research played little or no role in instigating most of the landmark events in cancer surgery before the second half of the 20th century.
Nonsurgical Advances That Enhanced Radical Surgery
Anesthesia and asepsis. Some uncertainty exists with regard to who discovered ether anesthesia, as well as to when it was first used. George Hill noted in a commentary ( 1) that John Warren of Boston used ether during an operation that he performed on October 16, 1849. Hill also states, however, that the surgical procedure was performed in 1846, and that Crawford Long used anesthesia during an operation on March 10, 1842. Robert Liston was noted by Hill to have performed a major operation under anesthesia on December 21, 1846. Regardless of who was first to use anesthesia, subsequent developments in its use made it feasible for radical cancer surgery to be performed. Sepsis, however, continued to be a major problem. Not until Joseph Lister applied the concepts of Louis Pasteur to surgery by introducing bactericidal therapy with carbolic acid in 1867 and described the principles of antisepsis was it possible to obtain “success” with the ever increasingly radical surgery that had evolved ( 28).
Prevention of pneumothorax. The greatest advance in thoracic surgery was probably the development of the pressure-differential chamber, which was not an operation but a device for preventing collapse of the lungs due to the entry of air into the thorax during surgery for a thoracotomy. Invented by Ferdinand Sauerbruch in 1908 ( 29), this instrument was supplanted shortly thereafter by the efforts of Samuel Meltzer, who introduced the use of insufflation for intratracheal anesthesia to more efficiently accomplish inflation of the lungs during entry into the pleural cavity. There is no evidence to indicate that either of these developments was instigated specifically for lung cancer surgery.
Blood transfusion. It is unlikely that the era of radical cancer surgery would have taken place without the efforts of Karl Landsteiner, who was awarded a Nobel Prize in 1930 for his discovery of blood types and his advocacy of the use of blood transfusions. Landsteiner received his award 23 years after Reuben Ottenberg performed the first “modern” blood transfusion at the New York Mt. Sinai Hospital, using the testing of blood type, thus making radical cancer surgery more feasible by decreasing the likelihood of postsurgical mortality ( 30).
Cancer and hormone dependence. The landmark event that illustrates the strongest relation between surgery and science during the middle of the 20th century was fathered, in about 1941, by Charles Huggins, a University of Chicago surgeon and a 1966 Nobel Laureate ( 31). Huggins's laboratory and clinical research showed that prostate cancer could be dramatically affected either by castration or by the administration of estrogens, both of which reduced the level of acid phosphatase in the blood. Huggins also showed that it was the synthetic estrogen diethylstilbesterol that caused regression of disseminated prostatic cancer, and in so doing, he initiated the era of cancer chemotherapy ( 32). He also described the activation of prostatic cancer by androgens and its inhibition by the antiandrogenic effects of orchiectomy and estrogens. The antiandrogenic therapy that Huggins proposed was proved effective and was later adopted throughout the world. Huggins's principle of the hormone dependence of tumors has been applied to other human tumors as well. Although his contributions related to the biology of prostate cancer rather than to the surgical treatment of that tumor, they resulted in the improvement of the surgical treatment of prostate and other cancers by functioning as systemic surgical adjuvant therapy.
As a result of his discovery of insulin, the Canadian surgeon and Nobelist (1923) Frederick Banting similarly made possible radical surgery for cancer of the pancreas—that is, total removal of the gland that contained insulin-producing cells. Unfortunately, such radical surgery took place without the availability of insulin, and patients subsequently developed uncontrolled hyperglycemia (diabetes; ref. 33).
In 1896, George Beatson ( 34) made a highly important discovery when he found that removal of the ovaries caused regression of cancer of the breast in women. Beatson's work is classic and is especially noteworthy because it was done before the secretion of endocrine glands had been discovered. Soon it was found that many, but not all, women with mammary cancer benefited from oophorectomy. However, the use of Beatson's procedure subsequently declined and was rarely used as a treatment in cancer of the breast. Huggins, in explaining the reason for that circumstance stated, “The virtual disappearance of a highly useful treatment of disseminated cancer can be attributed in large measure to the empirical nature of the discovery” ( 35).
The nonsurgical events noted above made it possible for surgeons to perform such “super” radical procedures as extended radical mastectomy, radical head and neck surgery, hemipelvectomy, forequarter and hindquarter amputation, partial evisceration, and even hemicorporectomy ( Fig. 2 ; refs. 36– 39). It was apparent, however, that merely having a knowledge of anatomy and being technically facile, as well as courageous, was not sufficient to advance the frontier of surgery.
The Halsted Era (1894 to ca. 1980)
At the same time that the previously described landmark surgical events were occurring, several experimentalists, through efforts that never achieved landmark status, began to introduce “scientific thought” into the process of cancer surgery. When it was realized that a cancer was not a “foreign body” that could be excised locally, but rather, a tumor that could spread, those investigators became interested in the phenomenon of tumor “metastasis,” a term first used by Claude Recamier in 1822 to describe secondary growths that had occurred in the brain of a woman with mammary cancer ( 40).
Rudolf Virchow, the founder of cellular pathology, who formulated the axiom that “all cells come from cells,” was perplexed about how tumor cells could possibly traverse one organ, such as the lung, and appear in another, for example, the liver. Virchow proposed the idea of a humoral factor to explain that phenomenon. In 1860, he declared that cancer did not spread by the dissemination of tumor cells but, instead, by the elaboration of a parenchymatous juice that passed through the circulation of some organs and affected others ( 41). Virchow maintained that it was that juice from the primary tumor that transformed connective tissue cells at a secondary site into metastases ( 42). Despite T.R. Ashworth's first report of the existence of blood-borne tumor cells in 1869 ( 43), and despite reports by Karl Thiersch (1865; ref. 44) and Wilhelm Waldeyer (1872; ref. 45), who “established” that metastases occurred via cellular emboli, Virchow, throughout his life, contended that “… perhaps the last word in this question of the clinical nature of metastases was not yet said” ( 46).
In 1863, Virchow formulated another theory, which held that lymph nodes effectively trapped particulate matter in the lymph ( 47), and which led to the widely held belief that lymph nodes could provide an effective barrier to the passage of tumor cells. It is remarkable that Virchow was able to reach that conclusion without ever using tumor cells to test the integrity of the lymph node relative to the transmigration of such cells.
Another investigator who made a mark on medical history in the 19th century was Sir James Paget, who introduced the concept of “skip metastases.” In Paget's view, lymphatic metastases in breast cancer appeared in distant rather than in proximal axillary lymph nodes as a result of lymphatic communications and the dynamics of lymph flow ( 48). Because he was skeptical about the purely mechanical explanation for the predilection of certain organs and tissues for developing metastases, Paget proposed, in 1889, that certain organs, such as the liver, possessed a favorable “soil” for such growth, whereas others, such as the spleen, the thyroid, and skeletal muscle, did not. As a result of Paget's concept, two theories for the origin of the metastatic process ultimately arose: One view held that anatomic or mechanical factors were responsible for that process; the other favored the picturesquely labeled “seed-soil” theory of metastasis ( 49).
The concepts proposed by Virchow and Paget, and by several others with a similar “biological” nature, began to set the stage for a new era of cancer surgery that was to be dominated by William S. Halsted. Over the first three quarters of the 20th century, no one was to be more influential in conditioning the minds of generations of surgeons relative to the management of patients with cancer than was Halsted. His name was associated with the operation referred to as the “radical mastectomy,” a procedure that he first described in 1891 and that by 1907, when the meeting to form the AACR was held, was well established as the standard treatment for breast cancer. In an address given at that meeting, on May 8, 1907, Halsted made known his perception of the biology of cancer ( Table 2 ; ref. 50), and consequently, those concepts provided him with the justification for the type of radical surgery he advocated ( 50).
Halsted placed little significance on tumor cells in the bloodstream as a mechanism for the development of metastases, arguing that, “Although it [circulating tumor cells] undoubtedly occurs, I am not sure that I have observed from breast cancer, metastases which seemed definitely to have been conveyed by way of blood vessels.” In quoting from W. Sampson Handley ( 51), he also noted, “In showing that cancer cells in the blood excite thrombosis, and that the thrombosis as it organizes usually destroys or renders them harmless, Goldman and Schmidt seem to have established a fact of primary importance and one which is strongly opposed to the embolic theory as applied to carcinoma.” Halsted further emphasized, “We believe with [W. Sampson] Handley that cancer of the breast in spreading centrifugally preserves in the main continuity with the original growth and before involving the viscera may become widely diffused along surface planes.”
The surgical approach Halsted recommended was based on his view of the disease and is best revealed by the following statement: “Though the area of disease extend[s], from cranium to knee, breast cancer in the broad sense is a local affection, and there comes to the surgeon an encouragement to greater endeavor with the cognition that the metastases to bone, to pleura, to liver, are probably parts of the whole, and that the involvements are almost invariably by process of lymphatic permeation and not embolic by way of the blood.” Halsted stated further, “It must be our endeavor to trace more definitely the routes traveled in the metastases to bone, particularly to the humerus, for it is even possible in case of involvement of this bone that amputation of the shoulder joint plus a proper removal of the soft parts might eradicate the disease. So, too, it is conceivable that, ultimately, when our knowledge of the lymphatics traversed in cases of femur involvement becomes sufficiently exact, amputation at the hip joint may seem indicated.”
As a consequence of the above considerations, Halsted believed that clinically recognizable cancer was a local-regional condition that would be more curable if surgeons were more expansive in their interpretation of what constituted the “region.” He also contended that, with the use of better surgical technique using en bloc dissection, the “last” cancer cell could be eradicated and, thus, more cancers would be cured. In Halsted's view, it was the inadequate application of surgical skill that was responsible for local-regional tumor recurrences after operation.
History has shown that Halsted's concepts of the biology of breast cancer unfortunately wrongly influenced subsequent generations of surgeons to use surgical procedures that were scientifically unsound. Although his influence waned a bit when, in 1939, Gray reported that the mode of spread of tumor cells to lymph glands was by lymphatic emboli and that such cells did not remain for any length of time within the lumen of lymphatic vessels ( 52), Halsted's followers dutifully continued to espouse his principles. Moreover, there was little impetus for change because no sufficiently competing ideas arose to challenge Halsted's views. Any challenge that did occur was directed toward questioning whether the radical surgery being performed was sufficiently radical! It has not been possible to determine with certainty what had been accomplished relative to the outcome of patients during the era of radical and super-radical surgery. Systematic record keeping and appropriate data analyses were nonexistent. Technical nuances of surgery, rather than the biological characteristics of a tumor and its host, were deemed responsible for the success or failure of an operation.
Finally, one of the statements that I made in 1970 ( 53) with regard to Halstedian principles deserves repetition. “It seems obvious that the surgical principles laid down by Halsted were properly conceived and in keeping with his concept of tumor spread. As Richard S. Handley ( 54) has written, “To belittle Halsted is to commit the stupidity of judging a man's achievement out of context of his time and his circumstances.” On the other hand, to continue to endorse uncritically a procedure founded on the principles just described is irrational. If, since Halsted's time, as a result of accumulation of new information relative to this phenomenon—and it cannot be denied that this has occurred—new concepts have arisen, then, either the original surgical principles have become anachronistic or, if they are still valid, they were conceived originally for the wrong reasons” ( 53).
Biological Research Influences in Cancer Surgery (1980–present)
Beginning of laboratory and clinical research (1957). In 1957, in association with my brother, Edwin R. Fisher, an experimental pathologist, I began laboratory investigations into the biology of tumor metastasis ( Fig. 3 ). Our laboratory and clinical studies, which continued for more than two decades, ultimately led to the creation of a new hypothesis for cancer management. After that hypothesis was tested by means of randomized clinical trials, there resulted, for the first time, a scientific basis for the surgical treatment of cancer. (For more information about our studies and their findings, the reader is referred to a series of reports summarized in the next section of this article; refs. 53, 55– 65).
When we initiated our laboratory research, the prevailing concept of metastasis was that lymph-borne tumor cells had one destination: the lymph nodes; that tumor cells in the blood vascular system would lodge in the first capillary bed they encountered; and that there was an orderly pattern of tumor spread based on temporal and mechanical considerations. Our experimental findings, however, indicated otherwise. They showed that the blood and lymphatic vascular systems were so interrelated that it was not practical to consider them as independent routes of tumor cell dissemination ( Fig. 4 ; ref. 66). Moreover, we showed via the use of labeled tumor cells that most cells that gained access to an organ via the bloodstream traversed that organ ( 67). Those findings led us to conclude that, contrary to Halstedian principles, there was no orderly pattern to tumor spread and that the phenomenon was not entirely dictated by anatomic considerations but, rather, by intrinsic factors in tumor cells and by a multiplicity of host factors including those in the organs to which tumor cells gained access.
Contrary to the proposals espoused by Virchow and Halsted that regional lymph nodes were effective barriers to tumor cell dissemination, we noted, in another series of investigations, that tumor cells traversed lymph nodes, gaining access to efferent lymph, and that they also gained access to the blood vascular system via lymphatic-venous communications within nodes ( 68– 70). Between 1971 and 1977, our experiments showed that regional lymph nodes were capable of destroying tumor cells and that they likely played a role in the initiation ( 71) and maintenance of tumor immunity ( 72). As a result of our findings ( 73), we proposed that regional lymph nodes might be negative for tumor because of those circumstances and not merely because of removal of a tumor before its dissemination.
Investigations that we carried out using regional lymph nodes from women who had undergone radical mastectomy further suggested that regional lymph nodes had biological significance ( 74). Cells from those nodes expressed immunologic capabilities despite the presence of a growing tumor ( 73). In addition, we noted that nodes varied with regard to that property both between patients and with the location of nodes in the axilla ( 75). We subsequently showed that cells from nodes were instigators of a cascade of events that gave rise to cytotoxic effector cells of both the lymphoid and myeloid series ( 76). Thus, our investigations led us to conclude that biological rather than anatomic factors might be responsible for the appearance of metastasis in certain nodes and the lack of metastasis in others. As a result of those observations, it was my contention that to continue to view lymph nodes as “mechanical receptacles” for trapping tumor cells was an anachronism. Our findings led to the consideration that although more patients with tumor-bearing lymph nodes developed metastases, that did not necessarily mean that the tumor cells in those nodes were instigators of the metastases. Rather, I speculated that the host and tumor factors that permitted tumor growth in nodes might also be responsible for tumor growth at other sites.
In another series of investigations, we were the first to show the reality of dormant tumor cells ( Fig. 5 ; ref. 77) and showed that perturbation of the host by a variety of means could produce lethal metastases from those cells ( 78). We suggested that “cancer cells alive to begin with may be enduringly capable of growth if conditions are favorable.” We also considered that local recurrences following operation were not the result of inadequate surgical technique, as Halsted had believed, but, rather, could be due to systemically disseminated cells lodging and growing at a trauma site ( 79).
The experimental studies we conducted over a period of 15 years showed that a tumor and its metastases were not autonomous of the host but were, instead, related to a variety of host factors, such as blood flow alterations, trauma, biological changes, diet, reticuloendothelial activity, endocrine interrelationships, and others ( 55, 80). We concluded that because many of the host factors we investigated exhibited opposing effects on the growth of metastases, the possible combinations and permutations that might occur could be responsible for the difficulty in defining host-tumor relationships.
A series of clinical trials we conducted concurrently with our laboratory studies, and which I considered to be an extension of my work in the laboratory, provided new information that raised questions about the concepts on which treatment for breast cancer, as well as other cancers, was based. Recurrence and survival were found to be independent of the number of axillary nodes removed ( 81) and examined. Tumor location also failed to influence prognosis ( 82), as was observed in patients who had radical mastectomies without removal of internal mammary nodes in the mediastinum. In those women, a similar rate of treatment failure and survival occurred regardless of whether the tumors were located in the inner or outer half of the breast.
During the late 1960s, tumor size also began to be viewed as a marker of breast cancer heterogeneity. Neoplastic surgery was based on the concept that the time of a tumor's existence, as measured by size, determined surgical success ( 83) and that the earlier the operation (thus, the smaller the tumor), the better the chance for a cure. In an effort to determine the validity of that concept, we obtained information that led us to conclude that size did not necessarily relate to either “earliness” or “lateness” of a tumor and that outcome was related to tumor and/or host factors ( 53).
Formulation and testing of the Fisher alternative hypothesis (ca. 1968). From a historical perspective, it is interesting to note that in 1865 the French physiologist Claude Bernard, who is considered to be the father of experimental medicine, had introduced his “scientific process” ( 84) and promulgated the thesis that, “A hypothesis is… the obligatory starting point for all experimental reasoning.” According to Bernard, a hypothesis was of value only if it could be tested. Others had noted that a hypothesis constructed without an adequate program for estimating its worth “is a burden to science and to the world” and emphasized that any scientist who put forth hypotheses that could not be tested “is a purveyor of rubbish” ( 85). There is no evidence to indicate that Halsted had formulated a hypothesis. Halstedian principles had their origins in anecdotalism, and their validity was accepted without being tested.
In the late 1960s, it became apparent that the results from our laboratory and clinical studies all had the same characteristic: They did not conform to Halstedian principles. Instead, our findings provided the basis for the formulation of a hypothesis that was biological in concept ( Table 3 ). Its components were completely antithetical to those that had dictated Halstedian surgery ( 60). On the basis of Bernard's thesis, I realized that the validity of my hypothesis needed to be tested because only with scientific confirmation would it be permissible for a new generation of surgeons to abandon the surgical heritage that had been passed on to them by William Halsted.
The opportunity to evaluate the tenets of my hypothesis occurred in 1971, when I initiated a randomized clinical trial involving almost 2,000 women. The aim of that trial was to test the hypothesis by comparing the outcomes of those women in three treatment groups. One group was treated with a Halsted radical mastectomy; two other groups received less extensive operations. One of the latter two groups also received postoperative breast irradiation. Despite the differences in treatment, during 25 years of follow-up, the outcomes of the groups were not significantly different ( 86). Those findings supported the hypothesis by corroborating my previous contention that variations in local-regional treatment were unlikely to substantially affect survival and, most important, eliminated any considerations that might have contradicted evaluating breast-conserving operations by means of a second randomized trial. In 1976, I initiated another trial to reevaluate the alternative hypothesis and, at the same time, determine whether breast-conserving surgery—lumpectomy—was as effective as breast removal. In that study, nearly 2,000 women were randomly distributed among three treatment groups: One group underwent mastectomy, a second received lumpectomy alone, and a third received lumpectomy followed by breast irradiation. Findings through 20 years of follow-up showed no significant difference in the outcome of the three groups ( Fig. 6 ; ref. 87). Those findings yielded additional support for my hypothesis and provided further confirmation that there was neither a biological nor a clinical rationale for opposition to breast-conserving surgery.
As a result of the alternative hypothesis and its validation by appropriately designed randomized clinical trials that produced data that were subjected to rigorous biostatistical analysis, after three quarters of a century (1975), Halstedian concepts and the expansive operations fostered by them reached their zenith and gradually began to be relegated to the dustbin of history. Those concepts, which had resulted in the mutilating radical and super-radical surgical procedures that had been used for treating not only breast cancer but also melanoma, colorectal cancer, soft tissue sarcoma, thyroid and other head and neck cancers, and lung cancer, could now be viewed merely as markers against which progress in the understanding and treatment of cancer could be measured.
As a consequence of the tenets of the alternative hypothesis, it became apparent that only with the use of systemic therapy could improved survival after surgery be achieved, and that further manipulation of local-regional treatment by means of surgery and/or radiation therapy, another modality governed by Halstedian principles, was likely to be futile. Fortunately, at the same time that the rationale for the surgical treatment of cancer was being redefined, the era of systemic adjuvant therapy was beginning to evolve. That circumstance was also to have a profound effect on the role of surgery for the treatment of cancer.
Advent of systemic therapy as an adjunct to surgery (1972). Although, in the mid-1950s, it had been observed in experimental animals that chemotherapy had a cytotoxic effect on tumor cells in the blood ( 88), it was not until the early 1970s that I initiated the first trial to evaluate the use of postoperative chemotherapy for the treatment of cancer. Kinetic principles of tumor growth elucidated from animal studies conducted during the 1960s and early 1970s had resulted in the formulation of a hypothesis that could be tested in clinical trials ( 89). The initial results of that study, which were reported in 1975, showed that such therapy could alter the natural history of patients with primary breast cancer ( 90). A series of studies in which antiestrogens (tamoxifen) were used postoperatively, either alone or in combination with or without chemotherapeutic agents, firmly established the worth of such therapy ( 91– 95). The treatment of patients who had no identifiable metastatic disease with systemic adjuvant therapy because of the possibility that in the future they might develop distant disease was a revolutionary departure from previous strategies.
As a result of those findings, I concluded, in 1980, that two independent paradigms governed the management of cancer. The first was related to the use of surgery to eradicate local and regional disease; the second was related to the eradication of systemic disease. When it was discovered that systemic adjuvant therapy—that is, chemotherapy and/or tamoxifen—decreased local and regional recurrence as well as distant disease after minimal surgery, there was reason to conclude that the two independent paradigms had converged into one unified paradigm. It was no longer possible to consider the surgical management of breast cancer as an independent therapeutic strategy, and as a consequence, there arose the need to determine how other therapeutic strategies and their mode of administration influenced surgical treatment and vice versa. The use of preoperative therapy related to that circumstance.
The hypotheses that I formulated from biological and clinical information obtained during the 1980s led me to initiate the first randomized clinical trial to evaluate the role of preoperative chemotherapy ( 96, 97). A finding of particular importance in that study was that reduction in the size of a large tumor after preoperative therapy permitted more patients to be treated with breast-conserving surgery (lumpectomy and breast irradiation). Those findings showed that in patients who received preoperative therapy, the outcome with regard to disease-free survival and survival was the same as when that therapy was given after mastectomy. The advent of the use of systemic therapy resulted in further reducing the role of surgery in cancer treatment to the extent that, today, surgery might be considered to be the adjuvant therapy.
The Technological Era of Surgery (2000–present)
As surgeons began to be free from the constraints imposed on them by Halstedian concepts, radical cancer surgery began to be replaced by less extensive operations. The use of regional and systemic surgical adjuvant therapy in the form of radiation, chemotherapy, and/or hormonal therapy and the ability to diagnose progressively smaller tumors also played a role in further decreasing the need for contemplating the use of radical surgery. While those aspects of the evolution of cancer surgery were occurring, research in engineering was making feasible new technological advances. Those accomplishments opened the door to a new surgical era that is still in its infancy—an era in which the scalpel is no longer the symbol ( Table 4 ). In fact, it is becoming increasingly more difficult to separate what we commonly think of as surgery from other forms of therapy that are evolving for the local-regional treatment of cancer. It has even been predicted that, in the near future, robotic surgery will replace the surgery currently in vogue ( 98).
Treatment According to Tumor Site
Lung. Since the onset of the 21st century, there has been an increase in the number of publications that describe drastic changes in the management of patients with lung cancer, particularly those with small tumors. Thoracotomy, which previously used long incisions, and which involved cutting muscles and spreading ribs to remove a lobe of the lung, is rapidly being replaced by minimal or noninvasive procedures. The use of video-assisted thoracic surgery has achieved considerable popularity in that regard. A videoscope that is inserted through an intercostal space and then connected to a TV monitor provides guidance for removal of both a lobe of lung and the mediastinal lymph nodes. Comments about the worth of video-assisted thoracic surgery by surgeons who have reported their experience with the procedure are variable ( 99– 103). It was recently stated, “It is more likely that video-assisted thoracic surgery will find a defined, specialized role for a specific patient population rather than being used on every patient, by every surgeon at every hospital” ( 103).
Other techniques, such as those that use thermal energy, are also being used to remove lung cancers without opening the chest. For example, in patients with small tumors or in those who are at poor risk, radiofrequency ablation is being used ( 103). When radiation therapy is used in conjunction with radiofrequency ablation, the value of the latter is enhanced ( 104). Another technical advance, the CyberKnife, which can be used to remove lung cancer, consists of a miniature linear accelerator that is attached to a robotic arm so as to deliver highly concentrated beams of radiation that converge at a single site in the lung—in the tumor itself ( 105).
Colon and rectum. The first report of the use of laparoscopic (LAP) colectomy for the removal of colon cancer appeared in 1990 ( 106). In 2004, the results from a large multi-institutional randomized trial, after 4 years of follow-up, concluded that, “the rates of recurrent cancer were similar after laparoscopically assisted colectomy and open colectomy, suggesting that the laparoscopic approach is an acceptable alternative to open surgery for colon cancer” ( 107). Data reported from other clinical trials confirmed those findings and emphasized the worth of laparoscopic surgery in that patients made a more rapid recovery and had fewer complications as well as a shorter hospital stay ( 108– 111). It was stated that, “Finally LAP colectomy may be a first step toward a newer, less invasive, operative approach, namely Natural Orifice Transluminal Endoscopic Surgery” ( 110). Whereas it has been stated, “There is no reason not to offer laparoscopic colectomy in all stages of colon cancer in contemporary clinical practice” ( 112), credible data have not yet become available to advocate its use in rectal cancer resection outside of a clinical trial.
Brain. Technologic advances are drastically altering the surgical treatment of brain tumors. Minimally invasive procedures that remove such tumors without injuring normal tissue are being developed. During the 1980s, neurosurgeons began using the microscope for that purpose. The extent of their vision of the operative site, however, was limited. With the advent of the solid-lens or fiberoptic endoscope that is connected to a high-resolution camera, a xenon light source, a video monitor, and compatible instrumentation, the surgeon now has a wide-angle view of the operative field that is being traversed. Such a technique has been used for the removal of tumors in the pituitary and for intraventricular tumors ( 113).
Another technological advance related to the development of robotics and to the use of the “Smart Scalpel” could further change brain surgery as we have known it. The Smart Scalpel has a microlaser attached to it that is capable of precisely detecting and destroying cancer cells during surgery. Because cancer cells contain more protein and, consequently, are denser, the Smart Scalpel avoids damaging cells that are noncancerous. As a result, surgeons can limit their efforts to removing only those cells that have been flagged as cancerous. Another new innovation is the “Gamma Knife,” which can eradicate brain tumors by administering a concentrated radiation dose from a cobalt-60 source. Multiple beams of radiation intersect to focus on a targeted area of abnormal brain tissue while adjacent normal tissue is spared.
Breast. The surgical treatment of breast cancer is also being profoundly altered as a consequence of the technological revolution. Whereas radical and super-radical mastectomy have been replaced during the last two decades by the use of breast-conserving surgery, more minimally invasive and noninvasive procedures are currently being evaluated. Endoscopic breast-conserving surgery for tumor and sentinel-node removal has been used to both improve cosmesis and shorten the time of operation ( 114). Stereotactic removal of small tumors, their ablation by radiofrequency or interstitial laser beams, magnetic resonance imaging, guided focused ultrasound, and cryosurgery are all undergoing evaluation for the treatment of breast cancer ( 115, 116).
Other sites. Conventional surgical methods that have been used for the treatment of patients with laryngeal cancer are being replaced by the use of endoscopic laser surgery ( 117). Similarly, the removal of urologic tumors of the prostate and kidney by means of endoscopic surgery with the Da Vinci robotic system and the Zeus-associated endoscope for pelvic lymph node dissection have attracted attention ( 98).
This brief description of the ways in which current technological advances are revolutionizing the surgical treatment of cancer is likely to be only the prelude to greater change, for, while the use of endoscopic surgery is gaining acceptance, new tools to overcome deficiencies in it are being considered ( 118). Computer-based methods for planning, monitoring, and controlling surgical procedures are becoming a reality. An interventional magnetic resonance imaging–guided robot-assisted system and a magnetic resonance imaging–compatible endoscope are considered likely to play an important future role in the conduct of surgery. Similarly, it is anticipated that nanorobots will be able to precisely define and map cancer tissue and, thus, provide the surgeon with precise information about target areas that require dissection ( 119).
As occurred with the use of radical and super-radical surgery during the Halstedian era, anecdotal reports of the use of these new technologies are abundant. It is evident that quality of life is better following their use. However, there is need for more information obtained from appropriate scientific methodologies, such as clinical trials before it can be known whether freedom from tumor recurrence and survival are the same or better than that being obtained with current surgical treatment. Because these new instruments are used to eliminate localized tumor, and because cancer is a “systemic” disease, the need for that information is imperative.
For nearly 50 years it has been my contention that progress in the surgical treatment of cancer would occur only from research that resulted in a better comprehension of the biology of the disease. There is little evidence, however, to indicate that today's technological innovations related to surgery have been the product of biological principles. Advances in engineering research and the resulting empiricism have led to the use of these innovations. Thus, before and after the founding of the AACR, the evolution of cancer surgery has been influenced by surgical technique to a greater extent than by knowledge related to the biological nature of the disease. Nonetheless, it has been the results from laboratory and clinical research that have provided justification for opening the door to many of the changes during the past 30 years.
Although current and future technological developments will continue to play a role in the surgical treatment of cancer, it is ultimately the findings from research in molecular biology and genetics that will dictate the future status of cancer treatment and, ultimately, the fate of surgery. By the time the AACR celebrates its sesquicentennial, it is almost a certainty that cancer surgery, as it was performed during the first 100 years of the AACR, will be supplanted by other modalities that have resulted from scientific investigation.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
The author wishes to thank Tanya Spewock for editorial assistance and Mary Perez Lopez, Judith Eckert, and Jean Coyle for their aid in preparation of the manuscript.
It must be acknowledged that numerous other investigators, both in the United States and in other countries, have made significant contributions with regard to the aspects of cancer that I have discussed. Space limitations, however, have prevented the inclusion of their efforts in this article.
Note: B. Fisher is a Distinguished Service Professor at the University of Pittsburgh.
This article is dedicated to Edwin R. Fisher, M.D., an experimental pathologist, who died on March 13, 2008. His laboratory and clinical research during the past 50 years played a significant role in the evolution of cancer surgery.
- Received January 15, 2008.
- Revision received July 18, 2008.
- Accepted August 12, 2008.
- ©2008 American Association for Cancer Research.