Relevance of DNA-ploidy as a prognostic instrument for solid tumors.

Biological and pharmacological translational research has literally 'exploded' in the last two decades. Main aims in the biological field have been the identification of new prognostic markers to use as an alternative to or to integrate traditional pathologic staging and, more recently, predictors of response to systemic treatments. After a first generation of traditional morpho-pathologic factors came a second generation of biological markers, and now a third generation, still open to the area of molecular markers constantly being discovered, has been proposed. Nuclear DNA content, together with other features common to all or some specific tumor types such as cell proliferation and growth factor receptors or hormone receptors, can be historically and conceptually included in the second generation. DNA abnormalities, that is, the deviation from the diploid content of normal cells, represents one of the features most frequently associated with cell transformation and tumor progression. It can be determined by morphometric and image cytometry, but the flow cytometric approach, permitting a fast, automated and highly sensitive determination, is by far the most widely-used method. Ploidy has been investigated in both systemic and solid tumors, and a great deal of information is available for the most frequent pathologies such as breast, colon, lung and ovarian cancers. Unfortunately, the lack of laboratory standardization in terms of type of material, that is, frozen, fresh or paraffin-embedded, and of quality control programs to guarantee the intraand interlaboratory reproducibility of results, has led to discordant results. Moreover, the heretogeneity of case series in terms of stage and type of treatment (loco-regional or systemic), makes it sometimes difficult to define the relevance of ploidy and other biologic markers as prognostic indicators or predictors of response to therapy. Some guidelines for the correct evaluation of prognostic relevance of biologic variables have been proposed, but largely ignored [1-3]. In breast cancers, the frequency of aneuploid tumors ranges from 45% to more than 70% [4, 5] in different studies on large consecutive case series. Aneuploidy is, in general, directly related to poor differentiation, cell proliferation and lack of steroid receptors, but not disease stage. The reviewed studies on patients with node negative or node positive breast cancer, treated with loco-regional therapy alone until the first relapse, demonstrate the prognostic relevance of ploidy, with a better prognosis for patients with diploid tumors at a short follow up (4-5 years) [6,7]. However, this advantage decreases over time and is either lost from the tenth year of follow-up onwards [8-10] or remains evident only in some clinical premenopausal or ploidy subsets [11]. Moreover, the weak predictive potential is confirmed by the loss of prognostic relevance in multivariate analysis including steroid receptors [4] and cell proliferation [12] or more recently proposed molecular markers such as p53 [13] and c-erbB-2 expression [14]. In lung cancers, the frequency of aneuploid tumors reported by different authors ranges from 45%-60% starting from paraffin-embedded samples, to 80%-90% using frozen or fresh tumor material [15]. Available results on the prognostic relevance of DNA content, which virtually all refer to non-small-cell lung cancers, are somewhat discordant. In patients with surgicallyresected tumors, ploidy is an indicator of survival at short follow up periods [16, 17], whereas both positive [18, 19] and negative results for ploidy as a prognostic [15, 20] or independent prognostic factor [21] have been reported at a five-year follow-up. However, ploidy seems to completely lose its predictive relevance on survival at longer follow-up periods [22, 23]. The lack of detailed information available makes it impossible to define whether, as observed for breast cancer, the prognostic relevance of ploidy is really time-dependent, or whether systemic treatments given between relapse and death may have a real impact on the natural history of the tumor.