Pituitary adenomas, now termed pituitary neuroendocrine tumours (PitNETs) per the World Health Organization (WHO) Classification of Tumors of Endocrine Organs 5th edition (2022),¹ rank third globally among primary intracranial tumours after meningiomas and gliomas.² In Nigeria, local studies report PitNETs as the second most prevalent intracranial tumour.^3,4,5 These neoplasms arise from anterior pituitary cells and, though typically benign, may exhibit aggressive behaviour by means of compression or invasion of the optic chiasm or cavernous sinus, causing visual loss and neurological deficits.^6,7 Despite their clinical significance, pituitary adenomas remain under-investigated in Africa, particularly concerning their immunohistochemical profiles and clinical correlations.

The 2016 European Pituitary Pathology Group,⁸ WHO 2017 (4th edn)⁶ and 2022 (5th edn)¹ classifications emphasise proliferation markers (mitoses > 2/10 high-power field [hpf], Ki-67 ≥ 3%) with or without invasion (defined by pre-operative magnetic resonance imaging [MRI]/endoscopic intraoperative examination) for the identification of PitNETs with aggressive potential (high-risk PitNETs), discouraging ‘atypical adenoma’ terminology. The 2022 WHO classification of PitNETs uses immunohistochemistry for PIT1, TPIT, and SF1 transcription factors to define lineages: PIT1 (somatotroph/lactotroph/thyrotroph), TPIT (corticotroph), and SF1 (gonadotroph), plus anterior pituitary hormones and other markers.^1,7,8,9 This reduces null-cell cases by better identifying gonadotroph tumours.^1,7,8,9 This classification is often integrated with the PitNET’s morphology (e.g. acidophilic densely granulated somatotroph with perinuclear low molecular weight kininogen; chromophobic sparsely granulated with fibrous bodies; basophilic densely granulated corticotroph with diffuse low molecular weight kininogen; Crooke cell with ring-like low molecular weight kininogen).^1,7,8,9 High-risk or aggressive PitNETs, including those formerly termed atypical adenomas or carcinomas (now metastatic PitNETs), have criteria for identification that are still evolving.^1,7 Some of the high-risk histotypes include densely granulated PIT-1+ lactotroph, male lactotrophs, immature PIT1-lineage PitNETs, Crooke cell PitNETs, sparsely granulated somatotrophs, and silent corticotrophs.^1,10 They tend to be non-functioning macroadenomas that exhibit rapid growth, frequently recur after surgical resection and show poor response to conventional therapies.^6,7 Aside from evidence of invasion, proliferative indices – mitotic count > 2/10 hpf, Ki-67 index ≥ 3% (or ≥ 10% for heightened risk), and extensive p53 immunoreactivity (> 10% nuclei or TP53 mutation) – are indicated for identifying proliferative or aggressive PitNETs per the French five-tiered system (grades 1b–3), aiding prognostication beyond WHO morphology alone.^6,7,8,9,10 The WHO 2022 classification eliminated ‘atypical adenoma’ terminology (which required all three proliferative markers) because of poor aggressive behaviour prediction.^1,8,9 The French grading criteria system defines ‘proliferative’ PitNETs (grades 1b/2b) as two or more of the three markers exceeding thresholds: Ki-67 ≥ 3%, mitosis > 2/10 hpf, or p53 > 10% positive nuclei.^9,10 Immunoreactivity for p53 is not yet validated.⁸ Its use is optional, but can strengthen classification when borderline.⁹

Cyclin D1 dysregulation has been implicated in pituitary tumorigenesis and is associated with increased proliferative activity in PitNETs/adenomas. Consequently, CDK4/6 inhibitors, including palbociclib, have emerged as promising therapeutic agents for aggressive pituitary tumors.^11,12,13 Although therapeutic agents targeting cyclin D1 are being developed for several cancers, little is known about its expression in PitNETs, particularly in African populations, and its potential role as a biomarker for tumour aggressiveness.

This study aims to assess the cyclin D1 expression pattern in aggressive PitNETs compared to non-aggressive PitNETs and normal anterior pituitary tissues. By investigating cyclin D1 as a possible marker of aggressive clinical behaviour, this research seeks to fill gaps in understanding the molecular pathology of PitNETs in Nigeria and contribute to improved diagnostic and therapeutic strategies for these tumours.

The study included PitNETs with ages 15–69 years (median 45; interquartile range [IQR] 34.0–54.8 years). Aggressive PitNETs (n = 14) had a median age of 36.5 years (IQR 25.8–43.5), and a male-to-female ratio of 1:1. The median ages of aggressive PitNETs were 39 years (IQR 23.5–42.5) in male patients and 34 years (IQR 25.8–49.0) in female patients (Table 1).

Imaging using MRI and computed tomography identified eight invasive PitNETs, with representative MRI images of one case shown in Figure 1. No surgically or microscopically invasive PitNETs or pituitary carcinomas were identified.

Mitotic counts ranged from 1 to 15, and Ki-67 labeling indices ranged from 0.0% to 9.32% across all PitNETs (Table 2). Using the WHO criteria, eight PitNETs were invasive, and six were proliferative (mitotic activity > 2/10 hpf and Ki-67 index ≥ 3%), totaling fourteen aggressive PitNETs. None was both invasive and proliferative. The remaining 56 adenomas were classified as non-aggressive. Lactotroph PitNETs from PIT-1 lineage (n = 5) were the most frequent aggressive subtypes, accounting for 35.7% of aggressive adenomas (Table 3). Among the six proliferative aggressive PitNETs, the majority (n = 4) were of PIT-1 lineage (three PIT-1 lineage lactotroph PitNETs – two densely granulated and one sparsely granulated; all male patients, and 1 PIT-1 lineage somatotroph PitNET). The remaining proliferative PitNETs lineages were null cell (n = 1) and plurihormonal (n = 1) PitNETs. However, among the eight invasive aggressive PitNETs, half (n = 4) were of the SF-1 lineage. The remaining invasive PitNETs included two null-cell PitNETs, one plurihormonal PitNET, and one PIT-1 lineage lactotroph PitNET. Ten aggressive PitNETs were macroadenomas (1 cm – 4 cm), and four were giant adenomas (> 4 cm). All but one aggressive PitNETs were functional tumours. Examples of Ki-67 immunostaining patterns in normal pituitary, non-aggressive, and aggressive PitNETs are presented in Figure 2.

Fisher’s test with cross-tabulation between cyclin D1 expression and pituitary groups (aggressive PitNETs, non-aggressive PitNETs and normal anterior pituitary tissues) was performed, with the cyclin D1 expression pattern in each group represented in Table 2. Among 29 normal anterior pituitary tissues, 22 (75.9%) showed no cyclin D1 expression, and 7 (24.1%) showed weak expression; none showed moderate or strong expression (Table 2, Figure 3). Of the 56 non-aggressive PitNETs, 5 (8.9%) had no expression, 13 (23.2%) weak, 8 (14.3%) moderate, and 30 (53.6%) strong cyclin D1 expression. Among 14 aggressive PitNETs, 1 (7.1%) had no expression, 1 (7.1%) weak, 2 (14.3%) moderate, and 10 (71.4%) strong nuclear cyclin D1 expression (Table 2, Figure 3).

Moderate and strong cyclin D1 expression was more prevalent in aggressive PitNETs (85.7%) compared to non-aggressive PitNETs (67.9%). No moderate or strong expression was observed in normal tissues. Weak or no expression was observed in 14.2% of aggressive and 32.1% of non-aggressive PitNETs.

Association tests showed a strong positive correlation between cyclin D1 expression and PitNET aggressiveness (Goodman and Kruskal’s gamma = 0.8302; Cramer’s V = 0.5317; Kendall’s tau-b = 0.6006; p < 0.01).

Of the eight invasive PitNETs, five (62.5%) showed strong, two (25.0%) moderate, and one (12.5%) mild cyclin D1 nuclear expression (Table 2). Of the six proliferative PitNETs, five (83.3%) showed strong expression, and one (16.7%) had no expression (Table 2). Cyclin D1 immunostaining of normal pituitary, non-aggressive, and aggressive PitNETs is shown in Figure 4.

Cross-tabulation between cyclin D1 expression with Ki-67 labeling index categories (0% – 1%, > 1% – 2%, > 2% – < 3%, ≥ 3%) and association tests between the two variables showed no statistically significant association (p = 0.071; Gamma = 0.1821; Cramer’s V = 0.2873; Kendall’s tau-b = 0.0941) (Table 2).

Moderate/strong cyclin D1 expression (IRS ≥ 5) yielded a PPV of 24% (12 true positives/50 positives) for aggressiveness among PitNETs (n = 70), with significantly higher prevalence in aggressive cases (85.7% vs 67.9% in non-aggressive; Fisher’s exact test, p < 0.01; Gamma = 0.83).

Based on these results, the null hypothesis that ‘Cyclin D1 expression does not increase in aggressive PitNETs compared to non-aggressive PitNETs and normal tissues’ was rejected. This shows a positive association between moderate to strong cyclin D1 expression in PitNETs with increased aggressive potential. However, the low PPV of 24% suggests that a larger sample size may be required for the assessment of the use of moderate to strong cyclin D1 expression in predicting high-risk or aggressive PitNETs.

This study aimed to characterise the expression pattern of cyclin D1 in PitNETs, including both aggressive and non-aggressive cases, with the goal of evaluating its potential use as a marker to identify aggressive PitNETs for possible future individualised management strategies. About 20% of the PitNETs (14 of 70) were classified as aggressive, comparable to the proportion of pituitary adenomas exhibiting increased proliferative activity reported by Salami et al.,¹⁶ who identified 21% of the PitNETs in their study as ‘atypical adenomas’ (currently termed high-risk or aggressive PitNETs) defined by Ki-67 ≥ 3%.

Our finding that no cases exhibited both radiologic invasion and proliferative markers reflects the inconsistent relationship between Ki-67 expression and tumour invasiveness reported in the literature. While some studies have demonstrated a significant association between increased Ki-67 proliferation index and radiological tumour invasion, others have found Ki-67 to be non-predictive of tumour invasiveness.^17,18 However, Salami et al.¹⁶ reported a higher frequency of microscopic dural invasion (38.6%) with concomitant high Ki-67 expression in a subset of invasive tumours. The radiologically invasive rate of 11.4% observed in our study is comparable to MRI-based series reporting invasion rates of approximately 10%.¹⁹ However, it is considerably lower than the higher invasion rates reported in studies using intra-operative or histopathological criteria, including the 40% surgically determined (intra-operative) invasion reported by Selman et al.²⁰ and the 45.5% microscopic dural invasion reported by Meij et al.²¹ These differences are likely attributable to variations in the definition and assessment of tumour invasion across studies. Whereas our study defined invasion using radiological criteria (MRI-based Knosp grades 3–4 and Hardy grades C–E), other investigators employed direct intra-operative assessment or histopathological examination of resected dura, both of which may identify tumour invasion that is not evident on imaging alone, thereby resulting in higher reported invasion rates.²²

Most aggressive tumours were PIT-1 lineage prolactinomas, consistent with Liu et al.’s²³ work linking higher cyclin D1 expression to invasive prolactinomas. The rarity of pituitary carcinomas in our sample aligns with the known low incidence of 0.1% to 0.2% of pituitary carcinomas among pituitary tumours worldwide.²⁴

Regarding cyclin D1, nuclear staining being confined to tumour cells matches previous international studies.^6,25,26,27 Our study demonstrates a sequential increase in cyclin D1 expression from normal pituitary to non-aggressive and then aggressive PitNETs, corroborating findings by Liu et al.,²³ Jordan et al.,²⁵ Turner et al.,²⁶ and Hewedi et al.²⁷ While normal anterior pituitary tissues mostly lacked strong cyclin D1 expression in our study, weak staining in about 24% of cases is similar to findings of reports by Jordan et al.²⁵ Lack of any nuclear expression of cyclin D1 in normal anterior pituitary tissues reported by Hewedi et al.²⁷ might be because of their use of smaller normal anterior pituitary tissue sample size.²⁷

The higher prevalence of strong or moderate cyclin D1 expression in aggressive PitNETs (85.7%) compared to non-aggressive ones (67.9%) supports a significant association between cyclin D1 expression and tumour aggressiveness, consistent with multiple studies across different populations.^{18,23,25,26,27} These observations suggest that deregulation of cell cycle control, potentially through cyclin D1-related pathways affecting the G1-S phase transition, may contribute to PitNET tumourigenesis. The moderate-to-strong cyclin D1 nuclear expression (IRS ≥ 5) yielded a PPV of 24% for identifying high-risk or aggressive PitNETs in this study, indicating low specificity as a standalone diagnostic predictor since 76% of positive cases were non-aggressive tumours. This modest PPV underscores the need for cyclin D1 to be integrated with established markers such as Ki-67 index and radiological invasion for improved risk stratification, while highlighting its greater value as a sensitive adjunct rather than a definitive diagnostic tool for PitNETs. Although cyclin D1 is not a specific marker, its overexpression could serve as a sensitive indicator for identifying PitNETs with aggressive potential.