It was in 1891 that attention turned toward the malignant potential of some colonic polyps and in that regard Bardenheuer presented what he thought was "irrefutable histologic evidence" for what he called "malignant transformation of an adenomatous polyp." [16] So-called malignant degeneration of the adenomas was also described in 1896 by Port [17] and in 1903 by Zahlmann. [18] Their concept was accepted very soon by many colleagues such as Lockhart-Mummery and W. J. Mayo who worked on the subject of adenomatous polyps of the colon. [19–20] In 1925, Lockhart-Mummery stated that adenomas should be distinguished from inflammatory polyps, and that multiple adenomas have a marked tendency to undergo malignant change. [21] However, there were others such as Earle, in 1911, [22] and Pennington, in 1923, [23] who stated that the evidence that adenomas of the colon could become malignant was not compelling to them. Earle wrote that nonmalignant tumors were, in fact, less likely to cause malignant disease than the normal morphologic elements of the body itself. Pennington also pointed out that there was absolutely no evidence that an adenoma becomes malignant and changes into adenocarcinoma.

In 1899, Quenu and Landel, in the French literature of pathology, [24] and Rotter and Goebel, in articles published in German, [25] described adenomatous polyps which they deemed to have become malignant and named them as "papillomata" (because of finger-like projections).

Adenomatous polyp of the colon was addressed very briefly in Allingham's textbook, under the heading "The diagnosis and treatment of diseases of the rectum, in 1882. [26] In another textbook written by James P. Tuttle in 1906 titled Diseases of the Anus, Rectum, and Pelvic Colon, [27] however, adenomatous polyps were described extensively and most information on them came from the work by Quenu and Landel. Goebel, in 1914, collected and discussed 17 cases of adenomatous polyps in his textbook, [25] but not much reference was made in the text regarding the nature of these lesions.

As mentioned above, Quenu and Landel wrote the first extensive monograph on papillomata of the colon in 1899. [24] They classified neoplastic polyps as adenomata and papillomata. Although these two types of neoplastic polyps were deemed to be benign, the ones termed papillomata were considered by them to be "precancerous." Accordingly, they attempted to separate papillomata from adenomata and they stated that all papillomata are types of "epithelioma cylindriques" and correspond to the slowly growing epithelioma [basal cell carcinoma] of the skin of the face in elderly people. They pointed out that the cells of papillomata have multiple, hyperchromatic nuclei with few goblet cells, and correspond to the cell seen in papillary cancer of the rectum.

Goebal described very carefully two cases in which he had observed the appearance of benign papillomata of the rectum grossly, whereas microscopic examination showed evidence of carcinoma, to wit, penetration of malignant epithelial proliferations in the submucosa. He collected 57 cases from the literature up to 1914, of which he believed 8 of the lesions to be malignant, including the two of his own. Based on these observations, he concluded that all papillomata of the large bowel were suspicious for malignancy. [25] Lockhart-Mummary, in his textbook in 1923, reported one case of papillomata of the rectum and sigmoid and said that both papillomata and adenomata of the large bowel eventually become malignant. [28]

David, in 1925, continued to separate papillomata from adenomata; in his opinion, however, the relation of papillomata to adenomata was very close. He reported on nine tumors of the large bowel which he believed should be classified as papillomata or adenopapillomata because of their pattern of growth. [29] He described morphologic features of colonic papillomata in detail and compared clinical and histopathologic features of those lesions in the colon with similar ones of bladder and larynx. David acknowledged that the true nature of the lesions of the colon was not clear to him: "The question of the malignancy or pre-cancerous nature of these tumors was constantly in mind." For differentiation between benign and malignant tumors, he used a combination of criteria including clinical features, gross inspection, and microscopic features. He stressed the "infiltrative character" as an important criterion of malignancy on microscopic examination. This is what he wrote: "While deeply staining hyper-chromatic nuclei undergoing occasional mitosis are findings in a benign tumor that are suspicious of malignancy, they are changes not uncommonly found in rapidly growing epithelial tumors. While such microscopic changes are not to be neglected in weighing the evidence as to the malignancy of the tumors, it seems of much greater importance to observe regular arrangement and form of the epithelial cells on the basement membrane which everywhere presented an unbroken continuity in a benign tumor. In other words, the infiltrative character of the tumor is of greatest significance in the microscopic study of a malignant tumor. Couple with this, and of almost equal importance, is the gross pathological appearance and "feel" of the tumor as well as the clinical history of its growth." David believed that some of the lesions studied by him probably were malignant tumors from the beginning because in one of his cases first diagnosed as benign papilloma microscopically another biopsy from a depressed area showed carcinoma infiltrating the submucosa. He commented on that as follows: "Whether the tumor was malignant from the start, with papillomatous overgrowth on the surface of surrounding bowel, as I believe it was, or whether it represents the early malignant degeneration of papillomata is not easy to decide." Accordingly, he suggested that papillomata should be removed completely, unless clinical evidence showed a very slow growing and soft, nonindurate tumor.

Despite cytologic features of adenomatous polyps being those seen in malignant tumors, most students of the subject considered adenomatous polyp to be a benign tumor because those lesions, in most instances, followed a benign course, growing slowly and well circumscribed in polypoid fashion. Nevertheless, it was well known already at that time that the course of the so-called adenoma is not always benign. Manheim and Druckerman described a case of an adenoma that they felt had existed for at least 10 years. The patient was monitored by endoscopy and both the clinical appearance and pathologic results were those of a benign adenomatous polyp. Eight weeks later, the clinical appearance had changed to such an extend that the diagnosis of carcinoma was made clinically and then confirmed by the pathologist. [30] Martin believed that most of polypoid lesions of the colon and rectum show abnormal changes that over time will become frankly malignant. [31] Swinton also thought that adenomas are premalignant lesions. [32]"The fact that in 14 percent of all our cases of carcinoma of the colon and rectum the malignant condition can be demonstrated to have arisen from mucosal polyps, that in our series it has been possible histologically to demonstrate all stages in the sequence of changes from normal mucosa to adenocarcinoma and that the distribution of polypoid disease in the colon and rectum parallels the distribution of malignant growths in this region are sufficient evidence to justify consideration of all polyps of the colon and rectum as premalignant lesions." The evidence that led workers on this subject to believe that adenomatous polyps are "precancerous lesions" and may convert or transform to carcinoma resulted later in the generation of the so-called "adenoma-carcinoma sequence" (Table 1), [33] a hypothesis that proposed and advocated by Morson in 1978. [34]

Colvert and Brown studied 235 cases of rectal polyps and in a 5-year follow-up identified patients who developed carcinoma. They compared 167 patients in whom the lesions were removed with 68 in which the lesions were left in the patient. Five years later, in both groups an equal percentage of carcinomas had developed. The authors concluded that some adenomatous polyps were either malignant from the beginning or became malignant very early, but others seemed to follow a different course and grow very slowly. [35] A similar perception is also illustrated in the work of Helwig who designated these lesions as "histologically malignant but clinically benign." [36]

In 1956, Womack, in an essay about genesis of colonic cancer, strictly emphasized that adenomatous polyp (adenoma) is really cancer. [37] She stated that "adenomatous polyps are not premalignant lesions, but actually are cancer in an early stage of progression." [37]

Sunderland and Binkley studied 48 cases of papillomata for which they had suggested using the term "papillary adenoma" instead of papilloma. [38] The authors thought that papilloma is not a separate lesion and should be considered a type of adenomatous polyp with papillary features or an example of the variation of growth. Sanderland and Binkley believed that all adenomatous polyps are a type of adenoma, some with papillary configuration and some without. This is what is said about papillary adenoma: "While the characteristic surface growth and spread clearly differentiates papillomas from adenomas, there is evidence that they are of common origin . . . Papillary adenoma or villous tumors can be considered to be a growth variant of the common mucosal adenoma, characterized by 1- disproportionately increased growth of the surface epithelium in comparison to growth of the epithelium of the glands, 2- marked tendency to recur after local removal, and 3-lateral spread by replacement of the covering epithelium of the surface epithelium." They thought all adenomatous polyps, in spite of cytologic features of cancer, do not behave as cancer biologically and clinically: "It is difficult to argue with those who maintain that those tumors that subsequently develop infiltrative properties must have been carcinomatous from the beginning rather than examples of malignant change in an originally benign tumor. Since it was impossible to do serial sections on these tumors, no absolute conclusion can be drawn. Let it suffice, that the behavior of the group as a whole is not that of primary carcinoma. If a papillary adenoma is soft to palpation and shows only atypical epithelial change, there is then no other morphological or clinical criterion by which one can determine whether it will behave as an invasive carcinoma or as a benign tumor. Furthermore, it would be unreasonable to maintain that a tumor that was carcinoma from the beginning would recur superficially at irregular intervals over a six-year period, and then suddenly invade metastases and cause dead." Sanderland and Binkley thought that adenoma is a benign or premalignant lesion that may convert to carcinoma. They acknowledged, however, that the boundary between adenoma and carcinoma could not be defined clearly: " . . . no gradual or uniform chronological transition from benign to malignant structure was demonstrated . . . no uniform, gradual progression in degree of abnormality could be traced in frank carcinomatous changes. In other words, there was no chronological, orderly transition of the tumor as a whole from benign structure through increasing degrees of epithelial abnormality to a culmination in metastasing carcinoma. The changes was purely focal and sequence of events so variable that the time and extend of malignant transformation could not be predicted."

Sanderland and Binkley emphasized the indistinct boundary between adenoma and carcinoma, and other publications showed also that it was the major problem for many workers on the subject to forge morphologic criteria that separate carcinoma from "adenoma" in repeatable manner. Ortmayer undertook a study for evaluation of reproducibility of histologic diagnosis of carcinoma in an adenoma and demonstrated the confusion among pathologists by sending specimens of 19 polyps to three different pathologists, asking that they classify them only as benign, malignant, or suspicious. The results were interesting; they were in complete agreement on only 3 of the 19 tumors. They were in total disagreement on 6 of them, and in partial disagreement on remaining 10. [4]

In fact, there are very contradictory opinions about the criteria of malignancy in an adenoma in the literature, and in a number of publications criteria are not mentioned, a problem persisting to date. Swinton used the following criteria for identifying carcinoma in an adenoma: (1) anaplasia, (2) irregularity of architecture, and (3) invasion. [32] He mentioned that it is necessary for at least two of these three criteria to be met before making a diagnosis of malignant growth. The first two of these criteria are highly subjective and, therefore, many lesions were "signed out" as so-called early or "transitional forms" that were difficult to classify. Olsen and Davis suggested the criterion of invasion of the muscularis mucosa for diagnosis of carcinoma. For them, all ranges of malignant cytologic features that were seen in the mucosa, if not proven to penetrate to muscularis mucosa had to be named benign. [39] In short, the same lesion that was considered as malignant or pre-malignant by others, was benign for Olsen and Davis. As the reader will see in part two of this series this lack of agreement still exists to date and the lesion that is diagnosed "dysplastic" or in most part 'adenoma' by Western pathologists is considered 'carcinoma' by Japanese pathologists. [40]