Before 1948, in the 1940s, oncology—the science of dealing with the physical, chemical, and biologic properties and features of neoplasms, including causation, pathogenesis, and treatment—was an evolving field in medicine. Tumors were classified and diagnosed according to the latest editions of 2 textbooks in tumor pathology.1, 2 The authors of those textbooks differed substantially on classification and terminology; but they agreed that, if a surgeon considers a lesion malignant and the pathologist's biopsy report states that it is benign, then the surgeon should never accept that report without question. There should be a consultation between the surgeon and the pathologist. They should see the patient together, and then the pathologist may change his opinion. If no agreement is reached, then the patient should be treated on the grounds of clinical and radiologic findings.1, 2 The treatment was either radium and roentgen rays or surgery, ranging from excision, amputation, disarticulation, and hemipelvectomy to pelvic exenteration, radical mastectomy, and neck dissection with commando operation. Chemotherapy virtually did not exist beyond the sporadic application of nitrogen mustard. Regardless of the treatment, cancer was viewed as a fatal disease in most cases, particularly in children. I would not go as far as stating that, before 1948, everything that was done when it came to tumors was improper; however, many patients received inappropriate treatment.

The year 1948 signaled, among other things, the beginning of the end for treating patients without a biopsy-proven diagnosis. Shortly after the Second World War was over, officers and members of the newly formed American Cancer Society and the National Research Council realized that there was a void in communication in oncology. Once an agreement was reached that there was an urgent need for a peer-reviewed journal for communication among physicians who cared for cancer patients, only 1 issue remained unresolved: who should be the Editor of the new journal, Cancer. The choice fell on Fred W. Stewart (1894-1991) at the Memorial Hospital for Cancer and Allied Diseases (Memorial), the present Memorial Sloan-Kettering Cancer Center, in New York City (Fig. 1). The decision was appropriate, because Memorial was the sole hospital in the country that was dedicated exclusively to the care of cancer patients with 6 decades of reputation, and Fred Stewart was a principal figure at Memorial. The M. D. Anderson Cancer Center in Houston, Texas was established in 1947. Stewart was not a newcomer when it came to cancer. He was an associate of the world renowned tumor pathologist, James Ewing (1866-1943), from 1927. He succeeded James Ewing as Chairman of Pathology in 1939. Stewart, in 1948, was a recognized authority in oncologic pathology and was a superb communicator with remarkable writing skills. He possessed a long list of scholarly contributions to the literature on cancer. He believed that physicians who cared for cancer patients must be in constant communication with each other and must be dedicated to render specialized care to their patients. To accomplish these objectives, he was instrumental in the subdivision of the Department of Surgery for various subspecialty services and secured financial support for postresidency fellowships (traineeships) in pathology and surgery from the National Institutes of Health, the American Cancer Society, and private philanthropists. In his own department, in pathology, he established the policy that every unconventional or difficult case must be reviewed by at least 2 attending pathologists (quite often, he was 1 of the pathologists) and fellows in training before the pathologic findings were finalized and communicated to the patient's treating physicians. Having these attributes and known at Memorial, and beyond, as the pathologist of clinicians and pathologists, it did not come as a surprise that, in 1948, Stewart was named the first Editor of the newly inaugurated periodical of the American Cancer Society, Cancer. In many ways, 1948 was a momentous year in history. Harry S. Truman was elected President of the United States in his own right, Dwight D. Eisenhower became President of Columbia University, the Universal Declaration of Human Rights was adopted by the General Assembly of the United Nations, the first computer was put in use, the role of microscopic examination of bronchial secretions in the diagnoses of lung cancer was publicized, it became known that thorotrast (used in radiology as contrast material) produces angiosarcoma of the liver, and it was announced in Boston by Sidney Farber that the remission of acute leukemia in children could be achieved with folic acid antagonists.3

thumbnail image

Figure 1. The first Editor of Cancer, Fred W. Stewart (1894-1991), in action with his old typewriter in his office. Notice the photograph of his mentor, James Ewing, above his desk and the overflown wastebasket on the floor.

Download figure to PowerPoint

Stewart, as the sole pathology associate of James Ewing, learned from Ewing how to diagnose cancer on smears prepared from aspirates of tumors. His article on aspiration cytology, published in 1933, is a landmark contribution to oncology.4 Stewart was a skilled aspiration biopsy cytologist; and he expected all of his associates, fellows, and the hundreds of pathologists he mentored to master the intricacies of diagnosing cancer on aspiration smears. Decades later, in the 1980s, the aspiration smear technique obtained universal acceptance, largely because of the literary efforts and tenacity of 1 of his associates, Leopold Koss (who later became an eminent cytopathologist), and 1 of his trainees, William Frable, the current Editor of the American Cancer Society's cytology journal, Cancer Cytopathology.

Stewart's interest in understanding tumors better can be traced back to the years he spent with James Ewing. His phenomenal microscopic skills and long hours of reviewing microscopic preparations resulted in publication with his associates and fellows a series of original observations.

Although, during most of the 1940s, invasive breast carcinomas remained classified as scirrhus, medullary, and adenocarcinoma, and only 2 forms of noninvasive carcinomas (papillary and comedo types) were recognized, in 1941, with his associate Frank Foote (1911-1989), who succeeded him as Chairman of Pathology in 1959, Stewart introduced a new term, lobular carcinoma in situ.5 Three decades later, because of the contribution 1 of his associates,6 Robert Hutter (later Professor of Pathology, Officer of the American Cancer Society, and Editor of Cancer from 1990 to 2000), it became clear from long-term follow-up that lobular carcinoma in situ was a precancerous lesion and that, if it is not excised, years later, it will become invasive carcinoma.

In the middle 1940s, Foote and Stewart cautioned pathologists and clinicians about the complexity of the diagnosis and classification of mammary tumors in 2 detailed articles.7, 8 Stewart recognized with his associates for years that medullary carcinoma of the breast, contrary to earlier claims, has a relatively favorable prognosis.9 He also noted that, to the unintroduced,10 sclerosing adenosis can resemble cancer both clinically and microscopically and that cystosarcoma phyllodes, despite the archaic designation, most of the time is a benign neoplasm.11

While all these activities were going on in his department, Stewart found time as a founding member of the Committee on Oncology of the National Research Council to write and publish the first illustrated atlas on breast tumors in 195012 under the auspices of the newly reorganized Armed Forces Institute of Pathology. In 1968, Stewart coauthored the much expanded second edition of the same treatise13 with 2 of his former fellows and associates, Robert McDivitt (later Professor of Pathology in St. Louis) and John Berg (who succeeded Stewart as Editor of Cancer in 1962 and became a recognized cancer epidemiologist at the National Cancer Institute).

Stewart, a student of medical history, came to hold the view with William Osler (1849-1919) that, if you make an original observation, you should publish it. He not only did that by reporting with his associates14 a total of 45 hitherto unrecognized salivary gland tumors with a combination of epidermoid and glandular differentiation, but they coined a new term, mucoepidermoid tumor. Years later, his associates went further by dividing mucoepidermoid tumors into low-grade and high-grade forms,15 as they have been known ever since; and, subsequently, they discovered mucoepidermoid carcinomas in organs other than the salivary glands.16

In 1948, Stewart and his associates published a series of remarkable articles, each of which earned a permanent place in the annals of oncology. The origin and being of carcinomas of the thyroid largely were neglected topics in the literature. Lack of proper classification often led to misdiagnosis and mismanagement of patients. In an attempt to correct the prevailing trend, it was established by Stewart's associates17 that occult carcinoma of the thyroid was far more common than expected and required a less than radical surgical approach. Those authors also concluded that lateral aberrant thyroid cancer was a misnomer, because these tissues represented, with extremely rare exception, benign embryonic thyroid rests.17

Stewart's interest in precancerous lesions extended to every organ. For example, he encouraged his trainees to write and publish18 a series of colonic polyps with malignant transformation that he had diagnosed in past years. He also was aware that the incidence of colonic carcinoma is directly proportional to the number of polyps. His concept was so novel in the 1940s that, at a national pathology meeting, according to the story, a participant asked him from the floor to relate what did he thought about a recent statement by so-and-so eminent professor of pathology of so-and-so famous medical school that he had never seen a colon cancer arising in a polyp of the colon. Apparently, Stewart was in his usual elements, witty and sharp: He got up and replied, ‘I believe him,’ and sat down.

Early detection and diagnosis of uterine cervical carcinoma was a continuous topic at Memorial because of the large number of cases diagnosed every year. The matter turned from important to urgent in 1941 after George Papanicolaou (1883-1962), working across the street from Memorial at the New York Hospital-Cornell Medical School, published his original article on the detection of cervical carcinoma by using a smear technique (which, later on, became known as the Pap smear). Stewart was familiar with the term carcinoma in situ, which was introduced by Albert Broders (1885-1964) in 1932, and promoted its use with minor modification. Stewart preferred the name intraepithelial epidermoid carcinoma for noninvasive epidermoid (squamous) carcinomas of the cervix. Foote and Stewart, in their classic article on the anatomic distribution of intraepithelial carcinomas of the cervix (published in 1948 and illustrated with color drawings), reasoned that the lesions were confined to specific areas in the cervix.19 Two decades later, we had the opportunity to continue in a large prospective study, going a step beyond where Foote and Stewart stopped. We traced the site of origin of in situ epidermoid carcinomas of the cervix to the transformation zone (squamocolumnar junction).20 Consequently, we advanced the notions that properly carried out cold-knife conization by excising the transformation zone can be a diagnostic as well as therapeutic procedure and that hysterectomy is not required for the treatment of carcinoma in situ of the cervix,21 a view that was shared by Foote and Stewart. Foote and Stewart were 2 of my teachers and mentors with whom I had the distinct honor of working, first as a fellow and later on as an Attending Pathologist at Memorial for several years before their retirement to Florida in the early 1970s.

Although Stewart was not a gynecologic cytologist, he recognized talent when he saw it. His enthusiasm for the early detection of cervical cancer by Pap smear continued from the time he appointed Papanicolaou as a consultant to Memorial in the 1940s until he retired.22 In 1963, Stewart collaborated with Koss (who was then Chief of the Cytopathology Service of the Department of Pathology at Memorial and newly published author of the first comprehensive text in cytology) in a long-term, prospective study of the behavior of precancerous lesions of the uterine cervix.23 Stewart also conceived the idea of a prospective mass screening program to curb the cervical cancer epidemic. His former fellow, William Christopherson (then Professor and Chairman of Pathology at the University of Kentucky), and his associates reported for the first time in 1962 that invasive cervical carcinoma could be prevented by the early detection and treatment of precancerous lesions.24

Perhaps it is important to remember that, as late as the 1940s, pathologists and clinicians had no clear understanding of the histogenesis and nature of melanoma.1, 2 In general, pigmented skin lesions in adults were under-diagnosed and under-treated as nevi. In children, conversely, pigmented lesions were overdiagnosed and overtreated as melanomas. In 1948, an associate of Stewart, Sophie Spitz, introduced the concept of juvenile melanoma.25 Three years later in a follow-up study, she called attention to the disparity between the worrisome histologic appearance and benign clinical behavior of pigmented skin lesions in children.26 The accuracy of her detailed analysis of a specific group of pigmented lesions soon was recognized and was named eponymically as Spitz nevus.

Having no acceptable classification for melanomas, Foote and Stewart began to devise a brilliant scheme for the prognostication of melanomas in the late 1950s. It consisted of masterful delineation of the thickness and the linear (surface) extent of melanomas with the aid of the microscope. The results obtained were expressed in millimeters and were incorporated into the pathology reports issued by Foote and Stewart or their associates. Although their design remained unpublished until 1976,27 the system was used routinely at Memorial and in the laboratories of Stewart and Foote's former trainees across the country and beyond. Needless to say, Foote and Stewart's scheme preceded the presently used microscopic designs by at least 10 years.

From the outset, when Stewart joined with James Ewing in 1927, he watched with concern the indiscriminate use of radiation for benign and malignant lesions, a practice that continued into the 1950s despite Stewart's and his associates' warning, in 1948, that radiation causes sarcomas in bone.28 After retirement, Stewart kept himself occupied for a while with radiation-induced tissue changes. The last article he coauthored with his former trainee and associate, Andrew Huvos, was on postirradiation sarcomas.29 (Andrew Huvos remained on the staff at Memorial as a bone tumor pathologist, par excellence, for 37 years, until his untimely death in 2006.)

Parenthetically, it should be noted that Stewart and his associates were the first, in 1938, to define the histologic and radiologic differences between benign and malignant giant cell tumors of bone.30 He also elucidated the microscopic characteristics of benign and malignant peripheral nerve tumors and recognized fasciitis as a benign reactive lesion decades ahead of others. After all is said, Stewart perhaps is remembered best for coining the term chordoid sarcoma31 and the term postmastectomy lymphangiosarcoma,32 which is known to every student in oncology as Stewart-Treves syndrome.

Stewart was a bachelor and city dweller with many interests, but his life centered around us: his trainees and associates. Having the occasional opportunity to be his driver out to Long Island when I was going home and he was going for a long weekend with his friends, I had the privilege of spending long hours with him in my car. I regard myself as a beneficiary of Stewart's intelligence as a tumor pathologist, linguist, and medical historian.

Stewart's personal contributions to oncology were many, as mentioned above. However, his influence on oncology continues through the teachings, diagnostic skills, and scholarly contributions of hundreds of his pupils. Some of those who attained the highest distinction as leaders in oncopathology in the United States and abroad are listed here in alphabetic order: Arthur Allen (Professor and Chairman of Pathology, Lutheran Medical Center, Brooklyn, NY), Mario Alvizouri (Professor of Pathology and leading tumor pathologist in Mexico), Javier Arias-Stella (Professor and Chairman of Pathology, Lima, Peru), Darryl Carter (Professor and Chief of Pathology at Yale University), Bevan Coombes (Leading Surgical Pathologist in Australia), Ramzi Cotran (Professor and Chief of Pathology at Harvard University, Boston, Mass), Joseph Eggleston (Professor and Chief of Surgical Pathology at Johns Hopkins Hospital, Baltimore, Md), Mattoshinho Franca (principal tumor pathologist in Brazil), Patrick Fitzgerald (Professor and Chairman of Pathology at the University of New York in Brooklyn, NY; he returned to Memorial Sloan-Kettering Cancer Center as Chairman of Pathology in 1972 upon the official retirement of Stewart and Foote), Alvan Foraker (Professor of Pathology at the State University of Florida, Gainesville, Fla), Claude Gompel (Professor and Chairman of Pathology at the University of Brussels, Belgium), William Hartman (Professor of Pathology at Johns Hopkins and later Executive Vice President of the American Board of Pathology), Philip Lieberman (Professor of Pathology and Chief of Surgical Pathology at Memorial Sloan-Kettering Cancer Center, New York, NY), Myron Melamed (became the Chief of Cytology at Memorial Sloan-Kettering Cancer Center in 1972 after the departure of Koss and subsequently became Professor and Chairman of Pathology at Memorial Sloan-Kettering and at the New York Medical College, Valhalla, NY), Charles Miles (First Director of Cytogenetics at Memorial Sloan-Kettering and later Professor at the University of Utah), Anthony Nicastri (Professor and Chief of Pathology at the University of New York, Brooklyn, NY), Kunio Oota (Chief of Pathology at the Tokyo Cancer Institute, Tokyo, Japan), Elias Pantangco (Professor of Pathology and Chief Oncologic Pathologist in the Philippines), Arthur Patchefsky (Professor and Chairman of Pathology at Fox Chase Cancer Center, Philadelphia, Pa), Paul Rosen (Professor of Pathology and Chief of Breast Pathology at the Cornell University School of Medicine, NY, NY), Roderick Ross (Professor of Pathology at McGill University, Montreal, Quebec, Canada), Fredrich Shipkey (Director of the Electron Microscopy Laboratory at Memorial Sloan-Kettering Cancer Center and later Professor and Chief of Pathology at the University of Mississippi), Steven Silverberg (Professor and Chief of Pathology at the University of Maryland), Thomas Simon (Professor and Chief of Pathology and Cytology at the University of Syracuse, Syracuse, NY), Stephen Sternberg (Professor of Pathology and Attending Pathologist at Memorial Sloan-Kettering Cancer Center and Founding Editor of the American Journal of Surgical Pathology), Louis Thomas (subsequently the Director of the National Cancer Institute, Bethesda, Md), Cyril Toker (Professor and Chief of Pathology, Mount Sinai School of Medicine, NY, NY), Neda Voutsa (Professor of Pathology and leading tumor pathologist in Greece), and Stanislaw Woyke (Professor and Chairman of Pathology in Szeczin, Poland).


  1. Top of page