Although in American literature the establishment of the concept of carcinoma in situ is attributed to A. C. Broders, a pathologist of the Mayo Clinic, who published a paper on this topic in 1932, [3] he was not the first person to use this term. In fact this term came from European investigators who had worked especially on cervical epithelium nearly 30 years before Broders' paper of 1932. In a paper published in 1908 by W. Schauenstein, a gynecologist from Graz, Austria, he addressed the remarkable morphologic similarities between carcinoma in situ and squamous cell carcinoma of the cervix. [4] He also emphasized that the abnormal surface epithelium should be called cancer because the cells that constituted the lesion were very similar to those of carcinoma and considered as the origin of the latter. Other authors such as Pronai [5] in 1909 and Rubin in 1910 [6] supported Schauenstein's concept and provided additional examples of such lesions. In a textbook published in 1912, Schottlander and Kermauner for the first time coined the term carcinoma in situ to describe the lesion confined in the epithelium of the cervix and thought to be the origin of carcinoma. [7] However, the term carcinoma in situ began to gain recognition in the medical field and enter into the medical lexicon only after Broders' paper published in 1932.

The concept of carcinoma in situ implies that all carcinomas originate from epithelium. When a single epithelial cell acquires a genetic alteration and becomes a neoplastic cell, it and subsequently its daughter cells will proliferate first within the epithelium of origin for some time (the time period varies and is unpredictable) before they enter into the underlying stroma. In other words, every carcinoma at its beginning must have an in situ growth phase of variable duration. In this sense, carcinoma in situ is nothing more than a very early stage of development of carcinoma in the epithelium of origin.

Broders originally defined carcinoma in situ with these words: "Carcinoma in situ is a condition in which malignant epithelial cells and their progeny are found in or near positions occupied by their ancestors before the ancestors underwent malignant transformation. At least they have not migrated beyond the juncture of the epithelium and connective tissue or the so-called basement membrane; such migration would be manifested by entering the connective tissue interstices or any part of the blood or lymph vascular systems." [3] According to Broders, histopathologic diagnosis of carcinoma in situ was based solely on cellular features seen in carcinomas that were described in classic textbooks of pathology. The criteria for diagnosing carcinoma in situ were set forth by Broders with these words: "The diagnosis of carcinoma in situ, as of carcinoma in general, is based chiefly on altered cellular characteristics in contradistinction to the cellular situation. The nuclei, as compared with those of normal cells, have an increased avidity for the basophilic dyes, are usually increased in size, both actually and in relation to the cytoplasm, are frequently irregular, are not infrequently in a state of atypical or pathologic mitosis and are occasionally observed in groups in the form of tumor giant cells." [3] In other words, Broders defined carcinoma in situ as a malignant epithelial neoplasm that is still confined in the epithelium of origin with no stromal invasion. In sum, Broders crystallized the term carcinoma in situ 75 years ago with a lucid and understandable definition and repeatable diagnostic criteria, which we are still using today.

Once the concept, definition, and criteria for carcinoma in situ are established, the histopathologic diagnosis of carcinoma in situ is actually straightforward and repeatable. Every pathologist is familiar with the cytologic features of carcinoma. Diagnosis of carcinoma in situ uses the exact same cytologic features. The only difference is that in carcinoma in situ, all the neoplastic cells are still confined in the epithelium of origin with no evidence of stromal invasion.

As our histopathologic assessment demonstrated well, in all the examples of so-called colonic adenomas we studied by conventional microscopy, all were epithelial neoplasms and regardless of different patterns of growth, all neoplastic cells were confined in the epithelium. They showed cytologic features of carcinoma such as crowding, stratification, overlapping of large hyperchromatic nuclei with loss of maturation, increased mitotic figures, and single cell necrosis (Figs. 2–5). As is self-evident, these cytologic and architectural features are really those seen in and used to diagnosis carcinoma. The histopathologic findings of so-called colonic adenomas fulfill the criteria of carcinoma in situ. Thus, we are compelled by evidence and logic to conclude that the so-called colonic adenoma is not adenoma, namely, a benign neoplasm of glandular tissue or glandular epithelium, but carcinoma in situ, obviously a well-differentiated one!

As for all carcinomas that can be well or poorly differentiated, carcinoma in situ, too, can be well or poorly differentiated. However, most pathologists (with the exception of breast pathologists, where the correct concept of carcinoma in situ is applied; this will be explored further in Part IV of this series) restrict the term carcinoma in situ only to those poorly differentiated lesions (Fig. 6). We disagree with this usage conceptually. As well-differentiated carcinoma is still a carcinoma; well-differentiated carcinoma in situ is still carcinoma in situ.

Just as carcinomas are heterogeneous, we believe that carcinoma in situ is also not a single uniform disease. Not to mention the different morphologic appearances, such as polypoid vs. flat type; tubular vs. tubulovillus vs. villus type; and well vs. poorly differentiated, we speculate that what we diagnose as carcinoma in situ histopathologically is a group of molecularly heterogeneous lesions. We all know well that any microscopic or morphologic study has its limitations. It is not possible to tell the exact molecular make-up of a neoplasm when viewing the lesion by conventional microscopy. At the molecular level, we speculate that carcinoma in situ may be divided into at least two types. Carcinoma in situ type I is the one that has all the molecular make-up of invasive carcinoma at outset. The neoplastic cells of this type usually break basement membrane, enter stroma, and become invasive carcinoma shortly after onset. We believe the so-called de novo carcinoma [8] derives from this type of carcinoma in situ. It has been reported that significant proportions of colorectal carcinomas are de novo carcinomas. [9–11] Carcinoma in situ type II is the other type that does not have all the molecular make-up of invasive carcinoma; more specifically, it lacks the molecular machinery for stromal invasion. This type will continue its in situ growth until one of its cells gains additional genetic changes that provide it with invasive capability, then, this cell and its daughter cells become carcinoma in situ type I (full molecular make-up of invasive carcinoma) and follow the pass as described above for carcinoma in situ type I, to wit, breaking basement membrane and entering stroma. Is it possible to distinguish carcinoma in situ type I from type II under a microscope? The answer is no. However, for practical purposes, most cases of carcinoma in situ that we see in a routine pathology practice are probable carcinoma in situ type II. The reason is that carcinoma in situ type I usually becomes overt carcinoma shortly after onset. When a carcinoma in situ grows to grossly detectable size (for example 5 mm and larger) and shows no evidence of stromal invasion, then it is most likely carcinoma in situ type II.

Most believe that carcinoma in situ if left untreated will certainly progress into invasive carcinoma. This is really a misconception. First of all, more and more data indicate that significant numbers of cases of carcinoma in situ do not progress into invasive carcinoma. [12,13] Second, as we proposed conceptually above, most cases of carcinoma in situ diagnosed in routine pathology practice are probable carcinoma in situ type II, which will not progress to invasive carcinoma unless additional genetic changes are gained and occurrence of these additional genetic changes would not be predictable. When a diagnosis of carcinoma in situ is made, the reasonable treatment is to eradicate the lesion completely via endoscopic resection. There is no need to treat carcinoma in situ aggressively. Conservative treatment should be the rule. The current clinical management of so-called colonic adenomas, namely, endoscopic polypectomy in toto with clinical follow-up appears to be appropriate and sufficient. [14] We see no point of treating these lesions more aggressively than the current clinical management just because we have realized that the true nature of these lesions is really carcinoma in situ.