The tall cell subtype of papillary thyroid carcinoma (TC-PTC) has long been considered a poor prognostic category of PTC [1,2,3,4,5,6,7,8,9,10,11,12,13]. In general, the diagnosis of TC-PTC requires that a significant proportion of the neoplastic cells be taller than they are wide. However, the precise criteria applied have varied over time and in different institutions. In their entity defining description, Hawk and Hazard originally described TC-PTC as a PTC in which at least 30% of the neoplastic cells are at least twice as tall as wide [14]. During the following two decades, most publications kept the twice as tall as wide criteria [15,16,17,18]. However by 1996, some publications began to use height criteria of three times as tall as wide [19]. In the WHO 2004 system, the 3:1 ratio was endorsed [20]. Subsequent publications used either the 3:1 or 2:1 cut-off [5, 21,22,23,24] and in the WHO 2017 system the tall cell component was required, somewhat ambiguously, to be ‘two to three times as tall as they are wide’ [25]. Similarly, the proportion of cells which must be ‘tall’ has varied over time with cut-offs of 10%, 30%, 50%, and 70% being proposed and used by different authors [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32].

The current World Health Organization (WHO) 2022 criteria unambiguously defines TC-PTC as a tumour with more than 30% of cells being 3 times tall as wide [1, 12]. However, given the very different criteria used by different groups, it is not surprising that the historical reported incidence of TC-PTC in different series has varied widely—from 1.3% to 12% of PTCs in different institutions [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32]; and there is strong evidence that the pathological diagnosis of TC-PTC is subject to poor interobserver concordance [32]. A confounding feature in assessing the clinical significance of tallness is that, because it is so well established as an adverse prognostic factor; once the diagnosis of TC-PTC is made, patients may be offered more aggressive therapy and this may make it hard to retrospectively assess the independent clinical significance of different criteria for TC-PTC.

Our experience confirms that different institutions certainly have different tendencies to invoke the diagnosis of TC-PTC. Historically (that is prior to 2012), at our two tertiary referral centres in Sydney, Australia, we had only rarely used the diagnosis of TC-PTC. Whilst this was out of step with many centres, it reflected our experience that TC-PTCs commonly already show a variety of other features associated with aggressive behaviour. Therefore, we, in agreement with at least some others, felt that the presence of tall cells commonly does not alter the treatment options for our patients because most TC-PTCs are also associated with other pathological or clinical features suggesting a high risk of aggressive behaviour [32, 33].

As a result of our institutions’ historical bias against making a pathological diagnosis of TC-PTC, when we review our database of PTCs treated in our institution from 1985 to 2012, only 14 of 2053 (0.68%) of resected PTCs were recorded as TC-PTC—much lower than in any other published series. Of course, this can only be because during that historical period we had not labelled as tall cell subtype many tumours that would be considered TC-PTC by many, if not most, pathologists by current criteria; and if they were re-reviewed by different criteria many would be considered TC-PTC.

Regardless of the relative merits of applying a low or high threshold to the diagnosis of TC-PTC, this large cohort of patients with PTCs that may show tall cell features (but were not treated differently to other tumours as they were not prospectively identified as TC-PTC) offers the unique opportunity to retrospectively assess the validity of different criteria for the diagnosis of TC-PTC and investigate questions such as what cut-offs for both height and proportion should be employed to make the diagnosis of TC-PTC and whether the presence of TC-PTC is so strongly associated with other adverse features that it may not have independent clinical significance in all circumstances.

We therefore sought to use this unique cohort to critically assess criteria in current use for the diagnosis of TC-PTC.