The Past, Present, and Future Statuses of Formerly Classified “Atypical Pituitary Adenomas”: A Clinicopathological Assessment of 101 Cases in a Cohort of More than 1,000 Pure Endoscopically Treated Patients in Single Center

Abstract

Objective  This study was aimed to assess the clinical aggressiveness of pituitary neoplasms that were previously defined as atypical adenomas.

Methods  A total of 1,042 pituitary adenomas were included in the study and 101 of them were diagnosed as atypical adenoma. Demographic characteristics, radiological evaluations, and clinical information were obtained from a computer-based patient database. Cases were categorized as atypical or typical using the criteria listed in 2004 Classification of Tumors of Endocrine Organs.

Results  The cure and reoperation rates did not show any statistically significant difference between the typical and atypical adenomas. However, a higher K _i -67 labeling index was found to be associated with a higher rate of reoperation ( p  = 0.008) in atypical adenomas. Of note, cavernous sinus invasion or parasellar extension was found to be associated with lower cure rates in patients with atypical pituitary adenomas ( p  < 0.001 and p  = 0.001, respectively).

Conclusion  Although atypical pituitary adenomas are known to be more invasive, this study demonstrated that the reoperation and cure rates are the same for typical and atypical adenomas. Our findings advocate for omitting the use of atypical adenoma terminology based solely on pathological evaluation. As stated in the 4th edition of the World Health Organization (WHO) classification, accurate tumor subtyping, evaluation of proliferation by means of mitotic count and K _i -67 labeling index, and radiological and intraoperative assessments of tumor invasion should be taken into consideration in the management of such neoplasms.

Introduction

The fundamentals of classification systems in pituitary adenomas are based on clinical, histopathological, or radiological findings. According to the third edition of World Health Organization (WHO) Classification of Tumors of Endocrine Organs, the pituitary endocrine tumors were classified into three categories, reflecting the malignant potential as adenomas (typical and atypical) and carcinomas. “Atypical pituitary adenoma” category was described in accordance with the proliferation index and p53 staining pattern; the tumors with increased mitotic index (a K _i -67 labeling index [LI] of >3%), and excessive staining with anti-p53 antibody were diagnosed as “atypical pituitary adenoma. ¹ The incidence of such pathology has been reported to range between 2.7 and 14.9% in various publications, ^{2 3 4 5 6 7} and its clinical impact was controversial. While several researchers have attempted to assess the prognosis of this histopathological diagnostic category, ^{3 8 9 10} in general, “grading” by pathological and immunohistochemical findings is inadequate from the aspect of clinical status of the patients. Atypical adenomas tend to behave more aggressive than typical adenomas based on their probable invasive growth pattern. The term, atypical pituitary adenoma, has been used for almost 20 years. However, recent researches produced conflicting results; some demonstrated a correlation between proliferative activity or p53 expression and invasiveness, while others did not. ^{11 12 13 14 15 16 17 18} Thereafter, Trouillas et al published a paper proposing a new prognostic clinicopathological classification scheme for pituitary adenomas which was consisted of size (tumor diameter by magnetic resonance imaging [MRI]), immunocytochemical type, and tumor grade (invasion [histological and/or radiological by MRI] and proliferation [K _i -67, mitotic index, and p53 positivity]). ¹⁹ The idea of this classification system arose from the fact that the tumor's invasiveness should also be considered in predicting treatment outcomes. ²⁰

The striking change in the recent fourth edition of Tumors of Endocrine Organs of WHO is the removal of the atypical pituitary adenoma category. ^{21 22} Even though this latest edition suggests clinicians not to use this diagnosis, it still strongly recommends including some of the morphological prognostic factors (i.e., mitotic count and K _i -67 proliferation index) in the histopathology report.

In the current study, we aimed to report the radiological features, histopathological findings, and clinical outcomes of pituitary adenomas, and to compare the results of typical and atypical adenomas that were diagnosed using the 2004 WHO classification system. Moreover, based on the idea presented in the 2017 WHO Classification of Tumors of Endocrine Organs, this study also sought to determine whether it is necessary to remove the diagnosis of atypical adenoma category from that classification system or not.

Methods

Patient Selection

A total of 1,042 cases of pituitary adenomas from a single-center academic tertiary care unit were included in the study. All the patients underwent endoscopic endonasal transsphenoidal surgery performed by the senior author (M.B.) between January 2006 and December 2017. Demographic characteristics and clinical information were obtained from a computer-based patient database. Surgical excision specimens were evaluated by a single pathologist (F.S.), and the cases were categorized as typical or atypical adenoma based on the criteria recommended in the 2004 WHO Classification of Tumors of Endocrine Organs (excessive p53 immunoreactivity, elevated K _i -67 LI [>3%], increased mitotic activity, and atypical morphological features).

Radiological Studies and Evaluation of Invasion

All patients underwent a magnetic resonance imaging (MRI) scan of the pituitary gland and a computerized tomography (CT) scan of the paranasal sinuses to evaluate the relationship between the adenoma and its surrounding structures during the preoperative assessment for endoscopic pituitary surgery. Afterwards, the cases were classified according to Hardy's classification and Knosp's classification. Hardy's classification was used to determine the sellar expansion and the tumor size (microadenoma or macroadenoma), and the Knosp classification was used to evaluate cavernous sinus invasion. ²³ Cavernous sinus invasion was determined by tumor extension beyond the line corresponding to the lateral tangents of the two components of the intracavernous internal carotid artery, as defined by Knosp et al. ²³ Postoperative MRI was performed within 24 hours of surgery, and it was repeated at third and sixth months, twice a year for the first 3 years, and annually thereafter. All radiological evaluations were performed by the neuroradiology department.

Clinical Evaluation

Full endocrinological evaluations were performed in all patients before surgery, and on the first day, at the end of the first week, at the end of the first month, 3 and 6 months, and 1 year after surgery, and then every 6 months following first year. Of note, rather than the aforementioned clinical follow-up schedule, all the tests were repeated, if needed, depending on the patient's clinical status. The criteria for disease control were a gross total tumor resection in nonfunctioning adenomas and biochemical cure along with evidence of clinical remission in functioning adenomas. ^{24 25 26 27} Additionally, formal visual field and visual acuity evaluations during the pre- and postoperative periods were performed by a neuroophtalmologist. For the functional tumors, If the endocrinological cure at postoperative early (first 5 days) and late (first 3 months) was achieved and radiological tumor residues were not seen in the later period, presence of tumor and/or endocrine dysfunction in next period was considered as recurrence. In nonfunctional tumors, if the tumor was radiologically detected in the late postoperative period and tumor was associated with endocrinological and neurologic problems, it was accepted as recurrence.

Histopathological Evaluation

All biopsy samples were fixed in formalin and embedded into paraffin blocks; 4-µm thick sections were cut for hematoxylin and eosin staining. Formalin-fixed paraffin embedded (FFPE) tissues were used for the immunohistochemical studies, and stained slides were evaluated by a single pathologist. Briefly, immunohistochemical staining was performed using the standard streptavidin–biotin complex method with the antisera raised against growth hormone (GH; Neomarkers [California, USA], 1:250), prolactin (PRL; Neomarkers, 1:500), thyrotropin (TSH; Neomarkers, 1:50), β-follicle stimulating hormone (β-FSH; Neomarkers, 1:300), β-luteinizing hormone (β-LH; Neomarkers, 1:500), p53 (Invitrogen, California, United States, Clone: BP53.12, 1/200 dilution), and K _i -67 (DAKO, California, United States, Clone: M7240, 1/200 dilution).

The K _i -67 proliferation index was evaluated by counting the cells that showed nuclear positivity among a total of 1,000 cells. This count was repeated for three separate 1,000 cells, and then the mean values of these three K _i -67 counts were calculated.

Statistical Analysis

Data analysis was performed using SPSS for Windows, version 11.5 (SPSS Inc., Chicago, Illinois, United States). The normality of the distributions of the metric discrete variables was determined using the Kolmogorov–Smirnov test. The metric discrete data were shown as mean ± standard deviation (SD) or median (interquartile range or minimum–maximum), where applicable.

The mean differences were compared using Student's t -test when the number of independent groups was two; one-way analysis of variance (ANOVA) was used to determine the differences among more than two groups. The differences in median values were compared using the Mann–Whitney U -test when the number of independent groups was two; the Kruskal–Wallis test was used to determine the differences among more than two groups. Categorical data were analyzed using Pearson's Chi-square. Cox's regression analysis was used to investigate the effect of several variables on the time; it takes for a specified event (cure or no cure) to occur. A p -value less than 0.05 was considered to be statistically significant.

Results

Out of 1,042 pure endoscopic endonasal transsphenoidal surgically treated cases, 941 (89.3%) were diagnosed as typical adenomas and 101 (10.7%) were diagnosed as atypical adenomas. The demographic, clinical, and radiological features of the cases are summarized in Table 1 . No patient was lost during follow-up period. Nine patients received medical dopamine agonist therapy in atypical lactrotroph group. Other than that, no patients were taking medications that affected the results. Detailed information about the histopathological classification and distribution of typical and atypical adenomas is shown in Table 2 .

Table 1. Basic demographic, clinical, and radiological characteristics of all adenomas.

	Typical	Atypical
Age (mean, median, [minimum–maximum]) in year	43.35, 43 (14–82)	34.23, 33 (11–68)
Gender (male/female)	439/502	54/47
Follow-up time (mean [minimum–maximum]) in month	79.9 (1–159)	61.3 (3–159)
K i -67 labeling index (mean, median [minimum–maximum])	1.48, 1 (0–3)	5.51, 5 (3–60)
K i -67 ≥ 10%	0	10 (9.9)
Intrasellar (%)	179 (19)	12 (11.9)
Parasellar extension (%)	298 (31.7)	43 (42.6)
Suprasellar extension (%)	464 (49.3)	46 (45.5)
Cavernous invasion (%)	283 (30)	45 (44.6)
Cure (%)	800 (85)	79 (78)
Reoperation (%)	82 (8.7)	10 (9.9)

Table 2. Distribution of all adenomas based on their immunohistochemical hormonal expression status.

Hormonal expression	Typical n (%)	Atypical n (%)
Somatotroph	226 (24.0)	23 (22.8)
Lactotroph	159 (16.9)	26 (25.7)
Corticotroph	159 (16.9)	19 (18.8)
Thyrotroph	6 (0.6)	2 (2.0)
Gonadotroph	196 (20.8)	10 (9.9)
Plurihormonal	125 (13.3)	17 (16.8)
Null cell	70 (7.5)	4 (4.0)
Total	941 (100)	101 (100)

A total of 800 (85%) of the 941 typical adenomas and 79 (78.2%) of the 101 atypical adenomas were cured after surgery. No statistically significant difference was found between the cure rates for typical and atypical adenomas ( p  = 0.074). Further analysis showed that the cure rates for the immunohistochemical subgroups of atypical adenomas were 80.8% for lactotroph adenomas, 73.9% for somatotroph adenomas, 84.2% for corticotroph adenomas, 70% for gonadotroph adenomas, 100% for thyrotroph adenomas, 70.6% for plurihormonal, and 100% for null-cell adenomas. Also, we detected four silent atypical corticotroph adenomas. Two of them had visual loss, while the other two had hypopituitarism ın preoperatively period. After the operation, three of them are in remission but the fourth one is not.

The reoperation rates were 8.7 and 9.9% for typical and atypical adenomas, respectively. This slight difference was not statistically significant ( p  = 0.690).

Among the atypical adenomas, 12 (11.9%) were intrasellar, 46 (45.5%) were suprasellar, and 43 (42.6%) were parasellar. A parasellar extension is associated with significantly lower cure rates ( p  = 0.001). Moreover, cavernous sinus invasion, which was detected in 45 (44.5%) of the atypical adenoma patients, was also found to be a statistically significant factor for poor cure rates ( p  < 0.001).

When K _i -67 LIs were compared between the cured and not cured atypical pituitary adenoma cases, the difference was not statistically significant ( p  = 0.818). When a cut-off value of 10% was set for the K _i -67 LIs and the patients were classified into “≥10%” and “<10%” K _i -67 LI groups, the cure rates did not show any statistically significant difference ( p  = 0.451). However, when the reoperation rates were compared between these two groups (K _i -67 LI ≥10% and <10%), the atypical adenoma cases with K _i -67 LI ≥10% were associated with a higher rate of reoperation ( p  = 0.008). Individual immnuohistochemical subtype evaluation did not show any statistically significant differences between the typical and atypical adenomas ( Table 3 ).

Table 3. The K _i 67 LI > 10 atypical adenoma analysis .

No.	Age (y)	Tumor type	PS/SS/IS extension	Cavernous invasion	Reoperation	Cure
1	30	Somatotroph	PS	−	−	+
2	31	Somatotroph	SS	−	−	+
3	62	Corticotroph	IS	−	−	+
4	33	Corticotroph	PS	−	−	+
5	53	Corticotroph	IS	−	−	+
6	25	Corticotroph	PS	+	+	+
7	34	Corticotroph	PS	+	+	−
8	24	Plurihormonal	PS	+	+	+
9	24	Plurihormonal	PS	+	+	−
10	40	Plurihormonal	PS	+	+	−

Abbreviations: IS, intrasellar; LI, labeling index; PS, parasellar; SS, suprasellar.

Note: “ + ” denotes “yes” and “ − ” denotes no.

Discussion

Ever since the atypical pituitary adenoma diagnostic category was first introduced, it has been a controversial issue for both neuropathologists and neurosurgeons. After the initial study have shown the invasive behavior of these mitotically active tumors, ^{28 29 30} researchers have focused on identifying this unclear entity; some of them have found out opposing results ^{31 32 33} ; however, some have returned to their initial findings. ^{3 34}

The criteria of atypical pituitary adenoma was based on the results of two major studies conducted by Thapar et al that stressed the utility of a K _i -67 LI threshold of 3% and the expression of p53. ^{28 29} Afterwards, the term atypical pituitary adenoma was included in the 2004 WHO Classification of Tumors of Endocrine Organs. ¹

In the absence of any other morphological parameters, it is problematic to assign grades to tumors by means of p53 and K _i -67 immunohistochemistry. Even though K _i -67 LI correlates with grades and clinical behavior in several tumor types, using K _i -67 LI, as the only conclusive parameter, may cause skepticism among pathologists. ³⁵ It is well known that there is heterogeneity in immunohistochemical staining and subjectivity in its interpretation. ³⁶ We also argue the value of using p53 immunostaining because the threshold was not clearly stated in the 2004 WHO classification. Additionally, the caveats of immunohistochemistry that are listed above also apply to this marker. All these reasons may explain the variations in the atypical pituitary adenoma frequency in the literature (2.7–18.7%). ^{5 11 34} In the current series, the K _i -67 LI cut-off value of 3% and excessive p53 immunostaining, as defined by the 2004 WHO classification, were used in a relatively large series handled by a single surgeon and a single pathologist; the atypical pituitary adenoma rate was determined to be 9.3%, which fits within the range reported in the literature.

In the current study, which consisted of 1,042 cases (101 atypical and 941 typical adenomas), we have compared invasiveness as assessed by radiological studies, reoperations, and cure rates of typical and atypical pituitary adenomas that were stratified according to the 2004 WHO classification system. We have found that atypical pituitary adenomas tend to be more invasive than typical adenomas, but without any statistical significance based on reoperation and cure rates.

Although the reoperation rates were not affected by K _i -67 LI when a cut-off value is set as 10%, a statistically significant difference was detected in atypical adenoma patients based on reoperation rates. Moreover, we were unable to find statistical evidence for this cut-off in our database and no cut-off value has been published in the English literature. Based on the authors' clinical and histopathological experiences, a K _i -67 LI > 10% could be associated with the aggressive behavior of atypical adenomas. That was why we set a cut-off value of 10% and performed further analysis. It is important to note that the number of atypical adenoma patients (>10% K _i -67 LI) is limited (10/101); most of the cases had parasellar extension (7/10) and only three cases were not cured in the present case series ( Table 3 ).

In the WHO 2017 classification, the codes of International Classification of Diseases for Oncology (ICD-O) are given separately for each lineage or hormonally differentiated adenoma. The other practical change in diagnostic methodology is a novel approach to classify pituitary adenoma according to the cell lineages. Again in the latest edition, routine immunohistochemistry (IHC) for transcription factors is required and prioritized to classify adenomas. In addition, “atypical adenoma” term was fully eliminated, primarily due to a lack of sufficient evidence to predict poor prognosis, only with pathological markers. ²¹ However, the aggressive adenoma term has not been clarified in this new classification. It has been generally accepted by the surgeons and even by the pathologists for some of the adenomas with rapid growth, significant invasion, large and irregular configuration, and resistance to conventional treatment. High K _i -67 LI invasion and rapid growth are considered to represent clinical aggressiveness and such adenomas with these features are defined as “high risk” pituitary adenomas in the recent WHO classification. Despite elimination of the term “atypical adenoma,” assessment of markers of tumor proliferation is still recommended as an important histological marker to predict recurrence, particularly following incomplete adenoma resection. However, no specific cut-off value is recommended. ^{21 37} Local invasion assessed by MRI and the intraoperative visualization still recommend in follow-up and predicting aggressive adenoma by many neurosurgeons. ^{19 37 38} In the present study, we compared the invasion status of cured and uncured atypical pituitary adenomas. After running Cox's regression analysis that is including K _i -67 LIs, cavernous invasion, intrasellar, suprasellar extension, parasellar extension, and K _i -67 LIs ≥10%, only cavernous invasion was found to have a negative effect on treatment results (hazard ratio [HR] = 14.481; 95% confidence interval [CI]: 2.622–79.981; p  = 0.02; Fig. 1 ). None of the other factors showed any statistically significant correlation in terms of cure rates or recurrence for atypical pituitary adenomas.

Fig. 1.

After running Cox's regression analysis that is including K _i -67 LIs, cavernous invasion, intrasellar, suprasellar extension, parasellar extension, and K _i -67 LIs ≥10%, only cavernous invasion was found to have a negative effect on treatment results (HR = 14.481; 95% CI: 2.622–79.981; p  = 002). None of the other factors showed any statistically significant correlation in terms of cure rates or recurrence for atypical pituitary adenomas. CI, confidence interval; HR, hazard ratio; LI, labeling index.

According to the classifications of pituitary tumors, we did not exactly know the place of these aggressive tumors in the classification schemes. Due to the complex anatomical structure of the sellar region, many critical neurovascular structures are closely related to these lesions which might make the surgical intervention challenging. In addition, the more accurate the preoperative clinical and radiological evaluation, the more clearly the surgeon could be able to visualize the tumor location during surgery, which could be resulted in better surgical and endocrinological outcomes. Classification or subtyping of pituitary adenomas should provide insights for clinicians. Thus, the obsolete atypical adenoma terminology is only based on histopathological evaluation, no anatomical relationship between clinical condition of the patient and tumor morphology was assessed, and it does not provide a prediction of clinical follow-up. Although “high risk” adenomas are identified, the “aggressive adenoma” term has not been clarified in the recent classification. This experience shows us that future research on pituitary adenoma classification should combine the clinical status of the patient, the radiological parameters including anatomical relationships with critical neurovascular structures, and the immunohistopathological evaluations.

Limitations

This study has several limitations. The retrospective study design in a single center may limit the ability to generalize the findings. Because of the available data, formerly classified atypical adenomas were not compared with their typical counterparts in depth, and the study was failed to identify a new classification scheme. Homogeneous distribution of statistical groups was not obtained either. However the current study is adequate to highlight the elimination of the diagnostic category, “atypical pituitary adenoma,” and should force the researches to find out the critical points for pituitary endocrine tumor classification.

Conclusion

The results of this study have suggested that the most important parameter affecting the cure rate in bygone atypical pituitary adenoma pathology is cavernous sinus invasion. However, based on our clinical and histopathological experiences, a K _i -67 LI > 10% could be associated with the aggressive behavior of the tumor. Consequently, as the definition of “atypical adenoma” did not contribute to the patients' follow-up and cure criteria, we advocate for the rationale of eliminating the use of atypical adenoma terminology. As stated in the fourth edition of the WHO Classification of Tumors of Endocrine Organs, accurate tumor subtyping, and evaluation of proliferation by means of mitotic count and K _i -67 LI is critical; however, radiological features and intraoperative assessment of tumor invasion should be taken into consideration when determining treatment, follow-up of the patients, as well as the future research, on pituitary adenoma classification.

Highlights

1. The most important parameter to affect the cure in atypical adenoma pathology is cavernous sinus invasion.

2. A K _i -67 IL > 10% is associated with the aggressive behavior of atypical adenomas.

3. Definition of the atypical pathology does not contribute to the patient follow-up and cure criteria.

4. Evaluation of proliferation by means of mitotic count and K _i -67 LI, as well as radiological features and intraoperative assessment, of tumor invasion should be taken into consideration when determining treatment and follow-up for patients with pituitary adenomas.

Funding Statement

Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.