Molecular Profiles of Papillary Thyroid Tumors Have Been Changing in the Last Decades: How Could We Explain It?

The molecular profile of thyroid tumors can be discovered today in at least 70–80% of cases. The most common alterations are point mutations of BRAF and RAS followed by RET/PTC rearrangements. Other less frequent oncogenes involved are PIK3CA, TP53, TSHR, PTEN, GNAS, and CTNNB1 (1), but taken together they are still not able to account for 100% of cases.

According to the data reported and collected so far (http://www.sanger.ac.uk/genetics/CGP/), BRAFV600E point mutations are present in about 45% of papillary thyroid cancer (PTC), with a higher rate in classical and tall cell variants and a rather low prevalence in follicular variant (2). There are exceptions with higher prevalence of BRAFV600E in PTC found in areas of China with a very high iodine content in the drinking water (3) and PTC found in the volcanic area of Mount Etna in Sicily where the drinking water is particularly rich in several elements, such as boron, iron, manganese, and vanadium, often present in concentrations exceeding usual maximum limits (4).

The observations of Jung et al (5), showing a high prevalence of BRAFV600E mutation in a series of PTC diagnosed in the United States over a 35-year period, from 1974 to 2009, are of great relevance in this regard, associated with a significant increase of its prevalence in the classical variant of PTC over these last decades. The first study that indicated an increase of the prevalence of BRAFV600E mutation was reported in 2005 in a UK series, and the authors hypothesized that some environmental/etiological agent(s) could be responsible for the changing molecular features of PTC over time (6). Similar results were reported 6 years later by Mathur et al (7) in a series of PTC patients reported from California, and the hypothesis in this study was that the increased prevalence of the BRAFV600E mutation could be responsible for the increasing rate of PTC. In 2012, we also reported a significant increasing rate of BRAFV600E mutation in an Italian multicenter study (8), and it was interesting to observe that the increase was similar when the analysis was separately performed in the four different groups of PTC patients (from Pisa, Milan, Perugia, and Catania). Because the reported series are really very heterogeneous from a geographical point of view, it is conceivable that the increased prevalence of BRAFV600E mutation is a worldwide phenomenon, making it plausible that only a common, or very similar, reason could explain this trend. It is known that higher iodine intake has been found to be associated with a higher incidence of PTC (9), and it is notable that programs of iodine prophylaxis appear to be associated with an increase in the prevalence of PTC and a decrease in follicular histology (follicular thyroid cancer [FTC]) (10–12). This evidence and the Chinese report (3) may raise the question of a link between iodine intake, BRAFV600E mutation, and the development of increased incidence of PTC, but up to now, no experimental evidence that iodine can induce somatic gene alterations in follicular cells has been reported. Nevertheless, it is known that the iodine radical (I●), iodinium ion (I+) and the hypoiodous acid intermediate (IO- [IOH]) are commonly formed in the follicular cell and represent reactive elements that together or as an alternative to H2O2 can determine cell oxidative stress, which is required to create conditions for thyroid cell proliferation (13–15). Further experimental studies on this putative relationship between iodine, as well as other more common pollutants, and thyroid cell transformation would be of interest to better clarify the origin of BRAFV600E mutation.

The increasing rate of BRAFV600E mutation is not the only molecular alteration reported by Jung et al (5). A significant decrease of RET/PTC rearrangements was also found over the years (from 11 to 2%) as well as a significant increase of RAS mutations (from 3 to 25%), particularly in the follicular variant of PTC (from 18 to 44%). A significant reduction of the RET/PTC rearrangements prevalence was also reported by Smyth et al (6) and by us (8). It is known that RET/PTC rearrangements are present not only after radiation but also in sporadic thyroid cancer (16). The causes inducing RET/PTC rearrangements in sporadic cases are unknown, although some hypotheses have been posed on the basis of some experimental observations (17, 18). It is certain that RET/PTC rearrangements were highly present in the radiation-induced post-Chernobyl tumors (19, 20), especially in those with a short latency period, whereas their prevalence became lower in tumors developed later in the same area (21). On the basis of these considerations, we can explain the reduction of the prevalence of RET/PTC rearrangements with two hypotheses: 1) mostly RET/PTC rearranged PTC are radiation induced and their reduction could be due to a worldwide lower exposure to ionizing radiation in the last several decades; and 2) RET/PTC rearrangements are due to other determinants (17, 18) that have been changing over the last several decades. We are more in favor of the first hypothesis mainly because the Californian and the Sicilian series are both characterized by a low and stable prevalence of RET/PTC rearrangements, and this is consistent with the fact that both of these geographic areas were not affected by the radioactive fallout after the Chernobyl accident in 1986 or nuclear test experiments performed in the United States during the 1970s and 1980s (22, 23). For these reasons, people living in these areas have not been exposed to significant levels of ionizing radiation, which is in agreement with the observation that in both Sicily and in California the RET/PTC-positive cases were relatively few and stable over the years. In agreement with this hypothesis, both in Europe and in central regions of the United States, the above-mentioned events could have been responsible, at least in part, for the incidence of RET/PTC-positive cases that became progressively lower because of the longer latency period from those events that were fortunately reduced and then discontinued. Of course this is a hypothesis that would be rather difficult to be confirmed unless another increase of RET/PTC-positive cases would appear in the future and be clearly correlated with new ionizing radiation exposures.

As far as the increase in RAS point mutation prevalence is concerned, this has been reported only by Jung et al (5), whereas in the series of Mathur et al (7), the rate of these mutations is very low and absolutely stable over the years. A possible explanation for this inconsistency may be due to different methods used in the two studies and a higher sensitivity of Jung's method.

Unfortunately, the data on RAS mutations are missing in the papers of both Smyth et al (6) and Romei et al (8). However, there is evidence that RAS point mutations, although more typical of FTC (24), are also present in a rather high percentage of the follicular variant of PTC, particularly in the encapsulated form (38%) with respect to the nonencapsulated form (10%) (25). According to these findings and to the observation that the follicular variant of PTC has been increased by 173% over the last decades (12), it is conceivable that Jung et al (5) found a significant increase of RAS mutations. No reasons have been postulated for the great “explosion” of this variant but, again, we can identify two possibilities: 1) over the years a revision of the criteria to define the histology of thyroid cancer has been made, and many cases previously diagnosed as FTC have been revised and diagnosed as follicular variant of PTC; and 2) the discovery of a series of less aggressive PTC characterized by the presence of RAS mutation. We are in favor of this last hypothesis because the follicular variant of PTC is mainly found in nodules with a microfollicular cytological pattern, classified today as THY3 according to the new cytological classifications. Currently, there is general agreement to operate on all THY3 cases because of the 20% probability of malignancy; this attitude, which has been applied also for very small nodules, could explain, at least in part, both the increase of the follicular variant of PTC and in general as well the increase of small indolent PTC that likely would have been ignored in the pre-ultrasound and pre-fine needle aspiration era.

Other than a significantly different molecular profile, Jung et al (5) also showed an increased prevalence of small PTC and a significant decrease of extrathyroidal extension and advanced tumor stage. These data are in agreement with those derived from one of the largest series of well-differentiated thyroid cancer in which clinical and pathological features were compared in two groups according to the year of diagnosis (before or after 1989) (26). These findings are strongly suggestive of an improvement over the decades of the healthcare system that likely has determined not only an anticipation of diagnosis but also the discovery of small indolent PTC.

However, the significant older age of the most recently diagnosed cases is difficult to explain with the concept discussed above because we would expect patients of a younger age. One could also suppose that, in the past, older subjects were less prone to see physicians and obtain clinical examination when asymptomatic. In contrast, incidental findings on neck ultrasound for carotid vessel investigation are being routinely performed after age 50, thus incidentally discovering many small asymptomatic nodules. Moreover, the association of BRAFV600E mutation with advanced age, although not fully accepted, is often described, and this could explain the increase of this oncogenic alteration.

It is evident that several clinical, pathological, and molecular features of PTC have changed over the last several decades and have done so all over the world. The study by Jung et al (5) clearly demonstrates that more recent cases of PTC differ from older cases, thus confirming the results of previous studies performed in different countries. Although the causes of these differences can be envisaged, the impact of these differences on the outcome and prognosis of these patients remains to be established.