The concept of cervical cancer precursors

The concept of cervical cancer precursors dates back to 1886 when Williams noticed, next to invasive cancers, areas of epithelium that he recognized as non-invasive. The term “carcinoma in situ” (CIS) was introduced by Broders in 1932, and this term has been used from its introduction to the present day. Smith and Pemberton in 1934 reported a relationship between CIS and invasive cancer when they found that the changes described as CIS by Broders were present in a retrospective review of biopsies from patients who subsequently developed invasive cancer. This combination of histologic observations and retrospective clinical analysis led to the concept that invasive squamous cell carcinoma develops from precursor lesions that can be identified by the pathologist (Figure 1.1).

Terminology and management (Figure 1.2a)

When the terminology was originally proposed and generally accepted, the prevailing clinical management of patients with dysplasia and CIS was that patients with CIS required a hysterectomy to prevent the development of cancer. Women with a histologic diagnosis of dysplasia, whose clinical course was not well understood but could vary from remission to persistence and progression to CIS, were ignored, followed up, or treated by a variety of means, depending upon the clinician’s understanding and acceptance of the natural history data. However, the distinction between dysplasia and CIS was frequently based upon poorly defined and arbitrary histologic distinctions. It gradually became clear that the enormous differences in approaches to treatment were the result of the unreliability of the diagnoses because of the high interobserver variability.

Biology and natural history

Rationalization of histologic terminology

Low-grade disease (mild dysplasia, cervical intraepithelial neoplasia 1, human papillomavirus/cervical intraepithelial neoplasia 1, low-grade squamous intraepithelial lesion) (Figures 1.2, 1.3, 1.4)

The principal alterations that characterize CIN include changes in the organization within the epithelium and in the cytomorphology of the cells. These changes appear to be triggered by HPV infection and represent the cytopathic effects of the virus in its replicative cycle. The morphologic changes occur because of the ability of the virus to interrupt the cell cycle, to induce polyploidy, and to inhibit functional cytoplasmic differentiation. This produces the characteristic “koilocytotic” change that is seen as enlargement of the bi- or multinucleated cells with a sharp perinuclear halo dispersed in a non-uniform “shotgun” pattern throughout the layers of squamous epithelium. These cells are hyperdiploid cells that have increased amounts of stainable nuclear material and hence are seen as hyperchromatic. Cells die and increased numbers of adjacent cells are recruited into the cycle to replace them. These cells exhibit nuclear enlargement, exaggerating previously unnoticed nuclear envelope irregularities. The most important changes at the tissue level are a retardation of the progressive maturation through the vertical layers of the normal epithelium (Figure 1.2b) and the expansion of the proliferating parabasal or suprabasal zone associated with the presence of koilocytes (the pathognomonic cytopathic change of HPV-infected cells). In low-grade lesions the atypically proliferating suprabasal cells occupy the lower third of the epithelial thickness, though the cytopathic changes of HPV infection are full thickness. The presence or absence of CIN relies on the presence of morphologically recognizable neoplastic cellular features in the suprabasal population (Figure 1.2c). Polyploidy in such lesions is further evidenced by bi-, tri-, or multinucleation, which is another important feature of virally infected epithelium.

The observer should look for the characteristic features of HPV infection in a biopsy and then for the presence or absence of “disturbed” suprabasal layers that might allow the separation of such lesions into HPV alone or HPV and CIN1 (Figures 1.3, 1.4). Within TBS all such lesions are ascribed to LSIL without further differentiation.

High-grade disease (moderate or severe dysplasia, carcinoma in situ, cervical intraepithelial neoplasia 2 or 3, high-grade squamous intraepithelial lesion) (Figures 1.5, 1.6, 1.7, 1.8, 1.9)

A pragmatic approach to treatment exists variably throughout the world in response to whether the precursor lesion is termed “low grade” or “high grade” after biopsy. In countries where CIN terminology is operative there has been considerable variation and difference in the placement of CIN2 into one or other of these categories. If CIN2 is considered high grade, then treatment is largely obligatory. In some countries (e.g., the USA) the grouping of CIN2 lesions with CIN3 in HSIL probably has fewer practical implications since both HPV/CIN1 (LSIL) and CIN2 or 3 (HSIL) are predominantly treated on consideration of the colposcopic extent or findings.

The histologic assessment of whether the epithelial changes involve up to two-thirds of the epithelium or are practically full thickness will allow the designation of CIN2 or 3 accordingly. The recognition of cytopathic HPV-related changes in the upper squamous layers decreases correspondingly. In high-grade CIN (HISL/CIN2 or 3) there is more hyperchromasia; the mitoses occur throughout the full epithelial thickness; the polarity of the parabasal cells is substantially altered; and a high degree of disorganized maturation and cytologic atypia is evident. These changes are thought to correspond with the occurrence of viral integration and/or amplification in cell nuclei and the development of aneuploidy, namely abnormal chromosome numbers and DNA quantities. Along with more severe nuclear chromatin changes come abnormal and bizarre multipolar mitoses. Whether or not these changes are clonal (i.e., arising in a single daughter cell line) has not been determined. Clonality may be essential, however, to terming any lesion “neoplastic,” since its occurrence demonstrates a survival advantage of the proliferating clone and the effacement of less genetically favored populations. Among apparently similar histopathologic high-grade lesions only a certain, but unknown, proportion may be predominantly monoclonal or neoplastic.

Disease may also extend into the cervical crypts that are potentially found under any premalignant lesion within the transformation zone (Figures 1.8a,b, 1.9). Recognition as to the depth of crypt involvement of such lesions is essential for their complete removal.