Pituitary tumor prognostication: WHO is really the best?

A decade after a 5-tiered clinicopathological classification was proposed by Trouillas et al., to predict the outcome and risk of recurrence of pituitary tumors (PTs),¹ there are still areas of controversy, such as the significance of tumor expression of proliferation markers, and the extent to which they can influence tumor behavior. For instance, this aspect does not cover anymore such a crucial importance, as compared to the determination of pituitary tumor subtype based on tumor cell lineage, in the revised version of the World Health Organization (WHO) classification.² Fortunately, a majority of PT exhibits an indolent behavior, but a significant issue concerns the remaining 30% to 50% of cases with invasive features (invasion of adjacent structure), the nearly 20% of cases with a recurrence or a progression after initial surgery and the less than 2% with an aggressive behavior, defined as resistance to conventional therapy and/or a rapid regrowth.^3,4

In the paper published by Lenders et al. in the same issue of European Journal of Endocrinology, the authors evaluated adult patients with operated PT, across 2 international tertiary centers, to determine the clinical utility of assessment of tumor invasion and markers of proliferation and classifying the tumor according to the French (Trouillas') clinicopathological system. Due to the retrospective design of the study, the authors did not (get the chance to) assess by immunohistochemistry, pituitary transcription factors like PIT1, TPIT, SF1, GATA3, and Erα, while the use of these markers is recommended in the WHO 2022 classification to specify the PT subtype, and ultimately, predict the tumor behavior. In our opinion, there is still a need for assessing the pituitary subtype as an independent risk factor of recurrence, namely, independent from the expression of proliferative criteria (like Ki67 and mitotic index), P53, and independent from the invasion status of the PT. On the contrary, the French clinicopathological classification demonstrates at least 2 advantages: Firstly, it is based on both invasion and proliferation, acknowledging the fact that only 1 criteria is not enough to estimate the risk of tumor recurrence.^5–7 It is important to bear in mind, however, that the pathological criteria used in Trouillas’ classification (mitoses, Ki-67, and p53) should be best regarded as a scoring rather than a grading system. In combination with an accurate assessment of tumor invasion, this clinicopathological classification stratifies the PT according to its grade (histological features) and its stage (invasiveness), with the following items: 1a (noninvasive and non-proliferative), 1b (noninvasive and proliferative), 2a (invasive and non-proliferative), 2b (invasive and proliferative), and eventually, 3, which defines PT with metastatic dissemination.¹ In recent years (and in other models as well), grading a lesion progressively became a standard of care as requested by learned societies,^8,9 with the aim of better delineating the behavior of a tumoral tissue, the reason why we found it proper for qualifying the proliferative status of a PT. In combination with invasion, the study conducted by Lenders et al. shows that, in patients originating from Pituitary Centre of Excellence, the French clinicopathological classification succeeds to stratify the risk of recurrence in 1b (hazard ratio [HR] 2.42 [1-5.85], p 0.05), 2a (HR 1.71 [1.13-2.59], p 0.01), or 2b (HR 2.32 [1.06-5.03], p 0.03) PT, as compared to the group of 1a tumors. By showing this, this study is the seventh, including 1 conducted by our group, to consolidate the clinical utility of this classification. If pooled, there is, now, more than 3000 patients that have been assessed so far with this clinicopathological classification system.^10–15 On the contrary, there is no data for at least the same number of patients assessing the prognosis value of histological subtypes according to the WHO 2022 classification. Despite the subtle differences in the pathological and radiological evaluation criteria used in the studies, the ability of the 5-tiered clinicopathological classification that had to predict the risk of progression/recurrence is deemed consistent, with the higher risk of recurrence always observed in 2b PT (HR ranging from 2.32 [1.06-5.03] to 8.62 [2.39-31.3] vs. 1a). Consequently, the latter should, now, be considered as high-risk PT, especially when the Ki67 proliferative index exceeds 10%.¹⁶ Likewise, 1b and 2a PT were sometimes, but not always, associated with a higher risk of recurrence/progression from 2.42 (1.00-5.85) to 3.27 (1.59-6.56) for 1b cases and from 1.71 (1.13-2.59) to 2.98 (1.89-4.70) for 2a cases, respectively. More recently, the study published by Guaraldi et al.¹⁵ suggested an original algorithm involving a combined evaluation of Trouillas’ clinicopathological classification, proliferation markers, and histological subtype, showing that the integrated evaluation of these factors could improve the predictive postsurgical outcomes.