Acta Neuropathol DOI 10.1007/s00401-013-1084-y
ORIGINAL PAPER
A new prognostic clinicopathological classification of pituitary adenomas: a multicentric case–control study of 410 patients with 8 years post-operative follow-up Jacqueline Trouillas • Pascal Roy • Nathalie Sturm • Emmanuelle Dantony • Christine Cortet-Rudelli Gabriel Viennet • Jean-Franc¸ois Bonneville • Richard Assaker • Carole Auger • Thierry Brue • Aure´lie Cornelius • Henry Dufour • Emmanuel Jouanneau • Patrick Franc¸ois • Franc¸oise Galland • Franc¸ois Mougel • Franc¸ois Chapuis • Laurent Villeneuve • Claude-Alain Maurage • Dominique Figarella-Branger • Ge´rald Raverot • The members of HYPOPRONOS
•
Received: 27 September 2012 / Revised: 7 January 2013 / Accepted: 17 January 2013 Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Pituitary adenomas are currently classified by histological, immunocytochemical and numerous ultrastructural characteristics lacking unequivocal prognostic correlations. We investigated the prognostic value of a new clinicopathological classification with grades based on invasion and proliferation. This retrospective multicentric case–control study comprised 410 patients who had surgery for a pituitary tumour with long-term follow-up. Using
pituitary magnetic resonance imaging for diagnosis of cavernous or sphenoid sinus invasion, immunocytochemistry, markers of the cell cycle (Ki-67, mitoses) and p53, tumours were classified according to size (micro, macro and giant), type (PRL, GH, FSH/LH, ACTH and TSH) and grade (grade 1a: non-invasive, 1b: non-invasive and proliferative, 2a: invasive, 2b: invasive and proliferative, and 3: metastatic). The association between patient status at 8-year follow-up
J. Trouillas (&) P. Roy E. Dantony C. Auger E. Jouanneau F. Chapuis L. Villeneuve G. Raverot Universite´ Lyon 1, Universite´ de Lyon, Lyon, France e-mail:
[email protected]
E. Jouanneau Service de Neurochirurgie, Groupement Hospitalier Est, Hospices Civils de Lyon, Bron, France
G. Raverot e-mail:
[email protected] J. Trouillas Centre de Pathologie Est, Hospices Civils de Lyon, Groupement Hospitalier Est, Lyon, France J. Trouillas INSERM, UMR-S1028, Lyon Neuroscience Research Center, Oncoflam Team, Rue Guillaume Paradin, Lyon Cedex 08 69372, France P. Roy E. Dantony E. Jouanneau F. Chapuis L. Villeneuve G. Raverot Hospices Civils de Lyon, Lyon, France C. Auger E. Jouanneau G. Raverot INSERM, UMR-S1028, Lyon Neuroscience Research Center, Oncoflam Team, Lyon 69372, France P. Roy E. Dantony Service de Biostatistique, Hospices Civils de Lyon, Lyon, France P. Roy E. Dantony CNRS; UMR 5558, Equipe Biostatistique Sante´, Pierre-Be´nite, France
F. Chapuis L. Villeneuve Poˆle Information Me´dicale, Unite´ de Recherche Clinique, Hospices Civils de Lyon, Lyon, France G. Raverot Fe´de´ration d’Endocrinologie, Groupement Hospitalier Est, Hospices Civils de Lyon, Bron, France N. Sturm De´partement d’Anatomie et de Cytologie Pathologiques, Centre Hospitalo-Universitaire de Grenoble, Grenoble, France C. Cortet-Rudelli R. Assaker A. Cornelius C.-A. Maurage Universite´ Lille Nord de France, Lille, France C. Cortet-Rudelli R. Assaker A. Cornelius C.-A. Maurage Centre Hospitalo-Universitaire de Lille, Lille, France C. Cortet-Rudelli De´partement d’Endocrinologie, Centre Hospitalo-Universitaire de Lille, Lille, France R. Assaker Service de Neurochirurgie, Centre Hospitalo-Universitaire de Lille, Lille, France
123
Acta Neuropathol
and age, sex, and classification was evaluated by two multivariate analyses assessing disease- or recurrence/ progression-free status. At 8 years after surgery, 195 patients were disease-free (controls) and 215 patients were not (cases). In 125 of the cases the tumours had recurred or progressed. Analyses of disease-free and recurrence/ progression-free status revealed the significant prognostic value (p \ 0.001; p \ 0.05) of age, tumour type, and grade across all tumour types and for each tumour type. Invasive and proliferative tumours (grade 2b) had a poor prognosis with an increased probability of tumour persistence or progression of 25- or 12-fold, respectively, as compared to non-invasive tumours (grade 1a). This new, easy to use clinicopathological classification of pituitary endocrine tumours has demonstrated its prognostic worth by strongly predicting the probability of post-operative complete remission or tumour progression and so could help clinicians choose the best post-operative therapy. Keywords Classification Pituitary tumour Pituitary adenoma
Introduction Pituitary tumours arising from the adenohypophyseal cells have a prevalence of 1/1,500 [7, 10] and are one of the
C.-A. Maurage INSERM U837, Lille 59800, France C.-A. Maurage Laboratoire d’Anatomie Pathologique, Centre Hospitalo-Universitaire de Lille, Lille, France G. Viennet Laboratoire d’Anatomie et Cytologie Pathologiques, Hoˆpital Jean Minjoz, Centre Hospitalier Re´gional Universitaire de Besanc¸on, Besanc¸on, France J.-F. Bonneville Service de Radiologie, Hoˆpital Jean Minjoz, Centre Hospitalier Re´gional Universitaire de Besanc¸on, Besanc¸on, France F. Mougel Service d’Endocrinologie, Hoˆpital Jean Minjoz, Centre Hospitalier Re´gional Universitaire de Besanc¸on, Besanc¸on, France T. Brue H. Dufour D. Figarella-Branger Aix-Marseille Universite´, Marseille, France T. Brue Service d’Endocrinologie, Diabe`te et Maladies Me´taboliques, Hoˆpital de la Timone, Assistance-Publique Hoˆpitaux de Marseille, Marseille, France
123
most frequent of all intracranial tumours. Generally benign, these tumours are associated with endocrine and nonendocrine signs and are clinically classified into functioning and non-functioning tumours. Many of these tumours (30–45 %) are invasive [31, 49] and some are clinically aggressive (review in [30, 37]). Commonly, the aggressive tumours are multirecurrent and resistant to multimodal therapy including surgery and radiotherapy. The early identification of such aggressive tumours may allow a timely intensive treatment that might prevent clinical recurrence or metastasis. In such cases, radiotherapy may prove useful in controlling tumour growth. In addition, some therapeutic agents such as temozolomide have shown potential in the early treatment of pituitary carcinomas and aggressive tumours [34–38]. To date all pathological classifications [2, 24, 25, 47, 51, 54, 56] have classified pituitary tumours by immunocytochemistry into GH, PRL, FSH/LH, ACTH, TSH, plurihormonal and null cell adenomas and into a dozen of ultrastructural subtypes. The most recent 2004 WHO classification [27] classified all benign tumours as typical adenoma (ICD-0 8272/0) while atypical adenoma (ICD-0 8272/1) included all tumours showing borderline or uncertain behaviour. These tumours have ‘‘atypical morphologic features suggestive of aggressive behaviour such as invasive growth. Other features include an elevated mitotic index and a Ki-67 labeling index greater than 3 %, as well as extensive
T. Brue Unite´ Mixte de Recherche 7286, Faculte´ de Me´decine de Marseille, Centre de Recherche en Neurobiologie et Neurophysiologie de Marseille (CRN2M), Centre National de la Recherche Scientifique, Aix-Marseille Universite´, 13284 Marseille, France T. Brue Centre de Re´fe´rence des Maladies Rares d’Origine Hypophysaire, Assistance Publique-Hoˆpitaux de Marseille (APHM), Hoˆpital de la Timone, 13005 Marseille, France H. Dufour De´partement de Neurochirurgie, Hoˆpital de la Timone, Assistance-Publique Hoˆpitaux de Marseille, Marseille, France D. Figarella-Branger Service d’Anatomie Pathologique et de Neuropathologie, Hoˆpital de la Timone, Assistance-Publique Hoˆpitaux de Marseille, Marseille, France P. Franc¸ois Service de Neurochirurgie, Hoˆpital Bretonneau, Centre Hospitalier Re´gional Universitaire de Tours, Tours, France F. Galland Universite´ Rennes 1, Rennes, France F. Galland Service d’Endocrinologie Diabe´tologie, Hoˆpital Sud, Rennes, France
Acta Neuropathol
nuclear staining for p53 immunoreactivity’’. The pituitary carcinomas (ICD-0 8272/2) are the only rare malignant tumours (0.2 %), which display systemic or cerebrospinal metastases [17, 21, 50]. Although proliferation, p53 detection and invasion were mentioned as criteria of the ‘‘atypical adenoma’’, due to the lack of histological signs, the invasion was not systematically taken into account in this classification. Some experts have pointed out that the 2004 WHO classification did not account for invasive status [61] and that its numerous ultrastructural subtypes were confusing [11]. In addition, no attempts have yet been made to assess clinical correlations with post-operative results, progression and recurrence [11, 16, 26, 62]. The prediction of pituitary tumour behaviour remains an important challenge for pathologists. To address this issue, we initiated this retrospective, multicentric case–control study in a large number of patients with long post-operative follow-up to evaluate the prognostic value of a new clinicopathological classification of pituitary tumours with five types based on immunocytochemical characteristics (GH, PRL, ACTH, FSH/LH and TSH) and five grades taking into account magnetic resonance imaging (MRI) and histological data concerning invasion and proliferation. The value of this classification in predicting disease-free and recurrence/progression-free status at 8 years post surgery was evaluated by multivariate analyses.
Materials and methods Patients This French collaborative study on prognostic factors of pituitary tumours (PHRC 27-43, HYPOPRONOS) was conducted in nine hospitals with, in each centre, a senior endocrinologist, a neurosurgeon and a pathologist. Patients were identified from the pathological register of each centre. Unfortunately, 45 % had to be excluded due to incomplete clinical and radiological data or absence of long-term follow-up. Moreover, because of the design of the study (a case–control study), the number of controls needed to match the number of cases selected based on the post-operative results. The 410 patients included were operated on between 1987 and 2004 via the transphenoidal route by nine experienced (performing [50 transphenoidal surgical interventions per year) neurosurgeons [3]. Patient information including sex, age at surgery, pre and postoperative hormonal data, post-operative treatments and relevant medical events were recorded. All included patients underwent MRI at the time of diagnosis, prior to
surgery and every year for at least 8 years after surgery. Tumour size was determined by MRI before surgery and the tumours then classified as microadenomas (diameter \ 1 cm), macroadenomas (1–4 cm) or giant adenomas ([4 cm). The tumour volume (height 9 length 9 width/2) was estimated in 379 patients. Tumour invasion was evaluated on the pre-operative MRI for all patients. Invasion of the cavernous sinus was considered when the percentage of encasement of the internal carotid artery by the tumour was 67 % or greater [6], or for grades 3 or 4 of Knosp’s classification [23]. According to Hardy’s neuroradiological classification, tumours with a suprasellar expansion that were frequently lined by non-tumoral pituitary (Fig. 1) were not considered as invasive [1]. Sphenoid sinus invasion was only defined if it was per-operatively confirmed by the surgeon and/or by histology (Fig. 1). Hormone levels were measured before surgery, 1 year after surgery and then every year for at least 8 years. Two hundred and thirty-four (57 %) patients were exclusively treated by surgery; the remaining 176 (43 %) patients received anti-tumoral treatment post-operatively. Case–control study design The principal analysis (disease-free at 8 years post surgery) compared patients with (cases) to those without (controls) evidence of disease during the 8-year follow-up. Patients with disease (increased plasma hormone levels, with or without radiological evidence of a tumour) or presenting recurrence or progression within the 8 years following surgery were considered as ‘‘cases’’. Patients in complete remission who showed no evidence of disease during the 8-year follow-up (no clinical symptoms, normal PRL, GH or IGF1, ACTH and cortisol levels, and no visible radiological tumour remnant) were considered as ‘‘controls’’. Patients with non-immunoreactive tumours (null cell adenoma) or non-functioning FSH/LH tumours were considered as cases only if tumour recurrence or progression was observed on MRI during the follow-up period. The complementary analysis (recurrence/progressionfree at 8 years post surgery) compared patients with recurrence or progression to those without recurrence or progression during the 8-year follow-up. The cases were patients who showed evidence of disease post-operatively and in whom recurrence or progression of the tumour had occurred during the follow-up. The controls were either disease-free patients or patients for whom a stable remnant was observed on MRI. Tumour progression was defined as evidence of an increase in plasma hormone levels and/or the regrowth of the tumour on MRI. Recurrence was defined as an increase in plasma hormone levels with or without
123
Acta Neuropathol
Fig. 1 Radiological and histological criteria for classification of pituitary endocrine tumours. Preoperative MRI of two pituitary adenomas: non-invasive microadenoma (a), macroadenoma with invasion of the left cavernous sinus (d). The suprasellar expansion was not considered as invasion. Coronal sections after gadolinium. Histological characteristics: ACTH adenoma bordered by the normal pituitary (NP). All tumour cells are positive with anti-ACTH antibodies (b). GH adenoma invading the respiratory mucosae (RE respiratory epithelium); 80 % of the cells are positive with anti-GH
antibodies (e). Ki-67 index at 0.4 % (c), Ki-67 index at 4 % (f). Immunoperoxidase reaction, bar 100 lm. For clarification, tumour 1 was classified as follows: ACTH (b) non-invasive microadenoma (a) without proliferation: Ki-67 0.4 % (c); mitosis = 0; p53 = 0 (not shown); Grade 1a. Tumour 2 could be classified as follows: GH (e) invasive macroadenoma (d, e) grade 2b if three markers Ki-67 = 4 % (f), mitoses n = 3 and p53 positive (not shown) are present, or grade 2a if only one marker Ki-67 = 4 % (f) is present and mitoses n = 1 and p53 is negative (not shown)
radiological evidence of a tumour mass, after a previous remission or regrowth of the tumour after total removal. The studies were approved by the ethics committee of Lyon and informed consent was obtained from each patient in accordance with French law.
Pituitary tumours
123
For each tumour, fragments were fixed in Bouin-Hollande and/or formalin fixative and embedded in paraffin for pathological diagnosis. Other fragments of 120 tumours
Acta Neuropathol
were frozen immediately and stored at -80 °C (Neurobiotec Bank, Lyon, France). The pituitary tumours were provided by the different pathology departments and for each a data set including macroscopic and microscopic features as well as cell proliferation and immunocytological profiles was provided. Standardised immunoperoxidase studies were performed in each pathology laboratory and were duplicated in the reference laboratory in Lyon in a small number (n = 12) of equivocal cases and in particular for tumours classified as null cell adenomas instead of FSH–LH adenomas or as plurihormonal (GH–PRL) due to use of a non-specific PRL antibody in one laboratory. As detailed in a previous publication [57], the immunoprofile was ascertained using ACTH, GH, PRL, beta-FSH, beta-LH and beta-TSH antibodies. The plurihormonal tumours were classified according to the hormone status of the cell component showing the highest percentage immunoreactivity. For each tumour the proliferative rate was evaluated by the detection of mitoses and Ki-67 (Mib1, 1/50, Dako, Glostrup, Denmark). The expression of p53 (clone DO-7, 1/200, Novocastra Laboratories, Newcastle upon Tyne, UK) was also studied. The p53 and Ki-67 immunostaining was performed on new sections after microwave pretreatment [57] in a Benchmark XT automate (Ventana Medical System, Tucson, AZ, USA) in the laboratories in Lille and Lyon. Fifty tumours chosen at random were duplicated in the department of pathology in Marseille to confirm the reproducibility of the automated immunocytochemical detection and the evaluation of the Ki-67 index and p53 detection. Cells from ten representative high power fields (HPF of 0.30 mm2, 4009 magnification) per tumour were counted with an average count of 5,000 nuclei. The Ki-67 index was expressed as a maximum percentage of positive nuclei and the mitoses by their absolute number. Due to the sometimes equivocal detection of p53 and the absence of a validated prognostic cut-off, only positivity or negativity was taken into account [15, 53]. The detection was considered as positive if there were more than 10 strongly positive nuclei per 10 HPF, in accordance with the ‘‘isolated positive scattered cells’’ previously proposed [15]. Ultrastructural subtypes are not included as in 1987, the first year of this retrospective study, it was decided that electron microscopic analyses of pituitary tumours would no longer be performed in pathology laboratories in France. Classification of the tumours As shown in Table 1 and Fig. 1, all tumours were classified based on a combination of their radiological characteristics by MRI (tumour size and invasion) and their histological characteristics by immunocytochemistry (type), by Ki-67
Table 1 tumours
Clinicopathological classification of pituitary endocrine
The classification is based on the three following characteristics: 1: tumour diameter into micro (\10 mm), macro (C10 mm) and giant ([40 mm) by MRI 2: tumour type into GH, PRL, ACTH, FSH/LH and TSH by immunocytochemistry 3: tumour grade based on the following criteria: Invasion defined as histological and/or radiological (MRI) signs of cavernous or sphenoid sinus invasion Proliferation considered on the presence of at least two of the three criteria: Ki-67: [1 % (Bouin-Hollande fixative) or C3 % (formalin fixative) Mitoses: n [2/10 HPF P53: positive ([10 strongly positive nuclei/10 HPF) The five grades are the following: Grade 1a: non-invasive tumour Grade 1b: non-invasive and proliferative tumour Grade 2a: invasive tumour Grade 2b: invasive and proliferative tumour Grade 3: metastatic tumour (cerebrospinal or systemic metastases) HPF high power field (0.30 mm2, 4009 magnification), MRI magnetic resonance imaging
index and mitotic count (proliferation). The p53, a transcription factor involved in the cell cycle, was also taken into account due to its expression in atypical adenomas [27, 47, 62] and pituitary carcinomas [51]. In this study, the rare non-immunoreactive tumours (n = 7) were grouped together with the FSH/LH type and the rare TSH adenomas were excluded. Based on their invasion and proliferation characteristics, the tumours were classified into five grades (grade 1a: non-invasive tumour, grade 1b: non-invasive and proliferative tumour, grade 2a: invasive tumour, grade 2b: invasive and proliferative tumour, grade 3: metastatic tumour). Invasion was defined as radiological and/or histological signs of invasion of the cavernous or sphenoid sinus. The histological invasion of dura mater (n = 16) and of the juxtatumoural pituitary (n = 14) was considered but not taken into account for grading. To evaluate proliferation, the most widely studied markers of the cell cycle in oncology were used: the number of mitoses and the Ki-67 index. Their thresholds had previously been validated in pituitary and/or endocrine tumours: Ki-67 index [1 or C3 % according to the Bouin-Hollande or formalin fixatives [20, 43, 53] and the number of mitoses n [ 2/10 HPF [44]. Since there is no validated prognostic cut-off for p53 detection, only its positivity or negativity was taken into account [15, 53]. Proliferation was defined as the presence of at least two of the three markers: mitoses or Ki-67 index which exceeded the defined thresholds, or positive p53 detection. All tumours were classified by the same
123
Acta Neuropathol
pathologist (JT), according to the criteria shown in Table 1 and illustrated in Fig. 1. The sparsely and densely GH tumours, as well as the plurihormonal tumours, were noted but not taken into account in this study. The number of silent GH–PRL (n = 2) and silent ACTH (n = 16) tumours as well as an ACTH tumour with Nelson syndrome (n = 1) was too small for specific statistical analysis and so these tumours were respectively grouped together with GH and ACTH types. Statistical analysis The main analysis was the disease-free patients at 8-year follow-up. The association between patient status and risk factors was analysed by fitting unconditional logistic regression models [4]. Factors considered were sex, age (treated as a continuous variable), tumour size (microadenoma, macroadenoma or giant adenoma), tumour type as assessed by immunocytochemistry (PRL, GH, FSH/LH or ACTH), and tumour grade. A first multiplicative model, without interaction terms between the type and the grade components (invasion and proliferation), was fitted. A non-multiplicative joint effect of invasion and proliferation was tested by introducing an interaction term between these two components of the grade. A final model was then fitted, to reveal the distinct effects of the grade components among the types, introducing interaction terms between the types and the grade components. Nested models were compared using likelihood ratio tests (LRT). Confidence intervals of odds ratio were calculated using a normal approximation of parameter distributions. In all statistical tests (two-tailed), p values smaller than 5 % were considered significant, except for interaction tests for which a threshold of 10 % was retained. Using predicted ROC-curves, the final model was informally compared with a model that uses Ki-67[cut-off value as the only marker of proliferation. The secondary analysis concerned the recurrence/progression-free status at 8 years. The analyses were similar to those conducted for investigation of disease-free status. All analyses were performed using R software [35].
Results Patient and tumour characteristics The study population was composed of 410 patients (246 women and 164 men), with a median follow-up of 11.14 years (range 1.19–23.8 years). The mean age at
123
surgery was 42.7 ± 14.7 years (range 13–77 years). Based on the MRI data, 94 (22.9 %) patients had a microadenoma, 292 (71.2 %) had a macroadenoma and the remaining 24 (5.9 %) had a giant adenoma. Invasion of the cavernous or sphenoid sinus was identified by MRI for 175 (42.7 %) tumours, but could be studied and confirmed by histology in only 16 (9 %). At the end of the 8 years of follow-up, 195 (47.6 %) patients remained disease-free (controls) and 215 (52.4 %) showed evidence of disease (cases). Of the 215 cases, 125 (58.1 %) patients had tumours that recurred or progressed. Based on the immunocytochemical data, this series was composed of 116 PRL, 111 GH, 105 FSH–LH and 78 ACTH tumours. The Ki-67 mean labelling index was respectively 0.40 and 0.68 % with Bouin-Hollande and formalin fixative (range 0–4 %) for the controls and 1.52 and 3.6 % (range 0–23 %) for the cases; the mean number of mitoses was 1.39 (range 0–15) for the controls and 2.81 for the cases (range 0–25). The detection of p53 was positive in 39 controls (20 %) and in 87 cases (40 %). No positive markers were observed in the tumour of 67 % of the disease-free patients (controls), compared to 38 % of the cases, and 76 % of the tumours with two or three markers were from patients showing evidence of disease at 8-year follow-up (cases). Concerning the grade, 194 (47.3 %) tumours were grade 1a; 41 (10.0 %) were grade 1b; 113 (27.6 %) were grade 2a and 62 (15.1 %) were grade 2b. Eight tumours (2.0 %), four PRL and four ACTH, classified as grade 1b (n = 1), 2a (n = 1) and 2b (n = 6) at the time of initial surgery were later classified as carcinomas (grade 3) based on the occurrence of metastases during the follow-up but this was not taken into account in the analyses. The majority of the patients with silent ACTH adenomas were cases (9 out of 16) and five tumours were grade 1a, one was grade 1b, and there were five each of grades 2a and 2b. The two silent GH adenomas were controls and were graded 1a and 2a. The prognostic value of sex, age and the parameters used in the clinicopathological classification Main analysis: disease-free status at 8 years This analysis comprised 215 (52.4 %) patients with evidence of disease (cases) and 195 (47.6 %) disease-free patients (controls) during the 8-year follow-up. As shown in Table 2, when a multiplicative model without interaction terms between the tumour types and the grade components was fitted, patient status at 8 years was found to be associated with age at initial surgery (p \ 0.001), tumour size (p = 0.002) and tumour grade (p \ 0.001) but not with sex (p = 0.742) or tumour type (p = 0.487).
Acta Neuropathol Table 2 Prognostic value of sex, age and the clinicopathological classification of pituitary tumours: disease-free status at 8 years of a case–control study
Total
Controls
Cases
No of patients
410
195
215
Ageb (years), mean ± SD
42.66 ± 14.66
43.84 ± 14.81
41.59 ± 14.47
OR
0.97
95 % CI
p valuea
[0.95; 0.99]
\0.001
Sex
0.742
Male
164
70
94
1.00
Female
246
125
121
1.09
Microadenoma
94
72
22
1.00
Macroadenoma
292
117
175
3.20
[1.64; 6.26]
Giant adenoma
24
6
18
3.03
[0.87; 10.53]
PRL FSH/LH
116 105
56 52
60 53
1.00 0.66
[0.32; 1.37]
ACTH
78
40
38
1.19
[0.59; 2.40]
GH
111
47
64
0.93
[0.47; 1.84]
194
143
51
1.00
[0.65; 1.83]
Tumour size
0.002
Tumour type Multivariate analysis with no interaction between the tumour type and the grade (invasion and proliferation) components OR Odds ratio, CI confidence interval
0.487
\0.001
Grades
p values were calculated from likelihood ratio tests
1a 1b
41
17
24
3.12
[1.72; 5.64]
b
2a
113
28
85
7.97
[4.57; 13.91]
2b
62
7
55
24.84
[10.84; 56.89]
a
Age was centred, the odds ratio was for an increase of 1 year in age
No interaction between the two grade components (invasion and proliferation) was found (LRT = 1.03, p = 0.311). In the model allowing distinct effects of the grade components among the tumour types, no interaction between proliferation and tumour type was found (LRT = 4.37 p = 0.225), leading to the estimation of a common odds ratio (OR) equal to 2.89 for proliferation whatever the tumour type (Table 3). By contrast, a significant interaction was detected between tumour type and the invasion component (LRT = 22.22, p \ 0.001). In this analysis, the tumour type and the grade were found to be strongly associated with the patient status (p \ 0.001). The effect of invasion varied according to the tumour type with the prognostic contribution of invasion being very high for PRL and ACTH tumours (OR = 152.72 and 27.07, respectively) and lower for FSH/LH and GH tumours (OR = 2.60 and 6.59, respectively). In this analysis, no interaction between the two grade components was found. Thus, for any given tumour type, the odds ratio comparing grade 2b to grade 1a is equal to the product of the odds ratios comparing grade 1b to grade 1a and grade 2a to grade 1a. The informal comparison of our model of grading which associated invasion and proliferation components (2 out of the 3 markers: mitosis count, Ki-67 [a cut-off value and p53 detection) with a model considering Ki-67 [a cut-off value as the only marker of proliferation underlined the
major prognostic value of invasion in the disease-free analysis (Fig. 2). By comparison with the area under the curve (AUC) obtained using our model, very similar results were obtained when Ki-67 was included as a binary covariate instead of our proliferation component. However, lower values were obtained when the invasion component was omitted, regardless of how proliferation was considered. Secondary analysis: recurrence/progression-free status at 8 years This analysis compared the 125 (30.7 %) patients displaying tumour recurrence or progression during the 8-year follow-up (cases), with the 282 (69.3 %) disease-free patients (controls) who either showed no evidence of disease (n = 195) or had a stable remnant (n = 87). Three patients who showed evidence of disease post-operatively but without progression were excluded from the analysis because of an insufficient follow-up. When a multiplicative model was fitted (Table 4), age at initial surgery (p = 0.048), tumour type (p \ 0.001) and tumour grade (p \ 0.001) were all found to be associated with patient progression/recurrence status at 8 years, however sex (p = 0.277) and tumour size (p = 0.998) were not. No interaction was found between the two grade components (invasion and proliferation) (LRT = 0.178, p = 0.673).
123
Acta Neuropathol
195
p valuea
215
56 52
60 53
1.00 2.65
[1.00; 7.06]
ACTH
40
38
1.59
[0.70; 3.60]
GH
47
67
1.99
[0.82; 4.82]
0.6
PRL FSH/LH
Sensitivity
\0.001c
0.4
Tumour typeb
\0.001d
Grade according to type
0.2
PRL 1a
52
14
1.00
1b
3
4
2.89
2a
1
23
152.72 [17.45; 1337]
2b
0
19
440.90 [46.72; 4161]
a. Final Model (invasion+proliferation) b. Invasion+Ki−67 ≥ cut−off c. Proliferation without invasion d. Ki−67 ≥ cut−off without invasion
[1.57; 5.29] 0.0
No of patients
95 % CI
0.8
Controls Cases OR
1.0
Table 3 Prognostic value of the clinicopathological classification of pituitary tumours, according to tumour type: disease-free status at 8 years of a case–control study
1.0
FSH/LH
0.6
0.4
0.2
0.0
Specificity
1a 1b
24 7
13 7
1.00 2.89
2a
17
21
2.60
[1.11; 6.09]
2b
4
12
7.51
[2.60; 21.68]
1a
34
10
1.00
1b
4
9
2.89
2a
2
11
27.07
[5.05; 145]
2b
0
8
78.14
[12.65; 483]
1a
33
14
1.00
1b
3
4
2.89
[1.57; 5.29]
2a
8
30
6.59
[2.58; 16.85]
2b
3
16
19.04
[6.26; 57.90]
[1.57; 5.29]
ACTH [1.57; 5.29]
GH
Multivariate analysis allowing distinct effects of the grade components (invasion and proliferation) according to the tumour type, adjusted for age, sex and tumour size CI Confidence interval a
0.8
Fig. 2 Disease-free analysis. Predicted ROC curves of: a the final model with invasion and at least two of the three markers (area under the curve: AUC = 81.4 %); b the model with invasion and Ki-67 [3 % as the only marker (AUC = 81.4 %); c the model with no invasion but with at least 2 of the 3 markers (AUC = 71.3 %); d the model with no invasion but with Ki-67 [3 % as the only marker (AUC = 71.1 %). Curves a and b as well as curves c and d are superimposed
PRL and ACTH tumours (OR = 5.19 and 11.78, respectively) and lower for FSH/LH and GH tumours (OR = 2.54 and 1.54, respectively). In this analysis, no interaction between the two grade components was found. This explains why, for a given tumour type, the odds ratio comparing grade 2b to grade 1a is equal to the product of the odds ratios comparing grade 1b to grade 1a and grade 2a to grade 1a. In this analysis, the tumour type and the grade were found to be strongly associated with the patient status (p \ 0.001).
p values were calculated from likelihood ratio tests
b
Odds ratio (OR) for tumour types was calculated for non-invasive tumours
c
Including interaction of type with invasion
d
Including interaction of invasion with type
In the model allowing distinct effects of the grade components for each tumour type (Table 5), no interaction between proliferation and tumour type was found (LRT = 2.89, p = 0.408), leading to the estimation of a common odds ratio (OR) of 3.88 for proliferation whatever the tumour type (Table 5). However, a significant interaction was detected between tumour type and the invasion component (LRT = 7.13, p = 0.068). The effect of invasion varied according to the tumour type with the prognostic contribution of invasion being very high for
123
Discussion Over the last 40 years, the classification of pituitary endocrine tumours has evolved from a tinctorial classification into three types: acidophilic, basophilic and chromophobic, to an immunocytochemical classification into five types and a dozen ultrastructural subtypes [2, 24, 25, 27, 47, 51, 54, 56]. Electron microscopy, an expensive and time-consuming technique, is now only rarely performed by some specialists. In 2004, the WHO classification [27] confirmed these immunocytochemical and ultrastructural subtypes and identified typical adenoma, carcinoma and atypical adenoma. Until now only two papers have studied the former type. In one series of 121 consecutive patients [62], atypical
Acta Neuropathol Table 4 Prognostic value of sex, age and the clinicopathological classification of pituitary tumours: recurrence/progression-free status at 8 years of a case–control study Recurrence/progression No
OR
282
125
Ageb (years), mean ± SD
42.37 ± 14.81
43.03 ± 14.14
0.98
[0.96; 1.00]
0.048 0.277
112
50
1.00
170
75
1.33
Microadenoma
77
17
1.00
Macroadenoma
190
99
1.02
[0.47; 2.24]
Giant adenoma
15
9
1.03
[0.29; 3.64]
PRL
96
18
1.00
FSH/LH
52
53
7.48
[3.47; 16.10]
ACTH
48
30
4.28
[2.02; 9.09]
GH
86
24
1.33
[0.62; 2.84]
1a
162
31
1.00
1b
23
18
3.77
[2.22; 6.41]
2a 2b
75 22
38 38
3.21 12.09
[1.84; 5.60] [5.66; 25.81]
Female
p valuea
Yes
No of patients Sex Male
95 % CI
[0.79; 2.24]
Tumour size
0.998
\0.001
Tumour type
\0.001
Grades
Multivariate analysis with no interaction between the tumour type and the grade components (invasion and proliferation) CI Confidence interval a
p values were calculated from likelihood ratio tests
b
Age was centred, the odds ratio (OR) was for an increase of 1 year in age
adenomas were defined by a Ki-67[3 %, p53 positivity and increased mitotic activity without a cut-off value. The invasion was not considered in the diagnosis although its frequency was 15 %. Saeger et al. [47] did consider invasion and when associated with Ki-67 [3 % and P53 [5 %, they found only 2.7 % of atypical adenomas among 241 tumours from the German registry. These two examples illustrate Wolfsberger and Knosp’s criticism of the WHO classification 2004 [61] when they stated that ‘‘the definition of invasiveness is needed and should be included in this classification’’ and underline the need to also include proliferation, evaluated by markers of the cell cycle with well-defined thresholds. Here, we have presented our proposal for a new clinicopathological classification with a grading system, such as that used for other endocrine tumours of the foregut [44]. This new classification takes into account tumour size, immunocytochemical type, invasion and proliferation. The invasion was evaluated by MRI. Only unequivocal invasion of the cavernous sinus was considered. Recent anatomical and ultrastructural studies have shown that the medial wall of the cavernous sinus is composed of dura [12]. Any tumour growth into the cavernous sinus would therefore represent a
sign of invasiveness without histological proof, such as the invasion into the sphenoid sinus confirmed by the infiltrated respiratory mucosae on histology. The proliferation is evaluated by the two most commonly used cell-cycle markers in oncology, i.e. Ki-67 index and mitotic count. From a broad spectrum of human neoplasms, the accumulation of immunocytochemically detectable p53 protein has served as a marker of clinical or biological behaviour. In addition, p53 immunoreactivity has been found in all pituitary carcinomas [53] and was one of the criteria to classify ‘‘atypical adenomas’’ [27]. For these reasons, p53 detection is also taken into consideration in this classification. When considering the controversial value of these markers, especially Ki-67 [1, 9, 13, 18, 20, 22, 28, 29, 43, 46, 48, 52, 55] and due to the lack of methodological standards and a validated cut-off for p53 [15, 53], we defined proliferation as the presence of at least two of these markers with cut-off values for Ki-67 index and mitotic count and a defined level of positivity for p53. By predicted ROC curves analyses, our results were similar when Ki-67 [a cut-off value was the only considered marker. However, no positive markers were observed in the
123
Acta Neuropathol Table 5 Prognostic value of the clinicopathological classification of pituitary tumours, according to tumour type: recurrence/progressionfree status at 8 years of a case–control study Recurrence/ OR progression
No of patients
No
Yes
282
125
95 % CI
Tumour typeb
p valuea
\0.001c
PRL
96
18
FSH/LH
52
53
12.84 [3.81; 43.27]
1.00
ACTH
48
30
3.71 [1.29; 10.63]
GH
86
24
2.93 [0.88; 9.79] \0.001d
Grade according to type PRL 1a
61
5
1.00
1b 2a
6 21
1 3
3.88 [2.25; 6.70] 5.19 [1.48; 18.15]
2b
8
9
20.14 [5.42; 74.84]
1a
24
13
1b
7
7
3.88 [2.25; 6.70]
2a
17
21
2.54 [1.09; 5.90]
2b
4
12
9.86 [3.54; 27.47]
1a
36
8
1b
6
7
3.88 [2.25; 6.70]
2a
6
7
11.78 [3.20; 43.32]
2b
0
8
45.75 [10.71; 195]
1a
41
5
1.00
1b 2a
4 31
3 7
3.88 [2.25; 6.70] 1.54 [0.55; 4.30]
2b
10
9
5.99 [1.97; 18.26]
FSH/LH 1.00
ACTH 1.00
GH
Multiplicative analysis allowing distinct effects of the grade components (invasion and proliferation) according to tumour type, adjusted for age, sex and tumour size CI Confidence interval a
p values were calculated from likelihood ratio tests
b
Odds ratio (OR) for tumour types were calculated for non-invasive tumours
c
Including interaction of type with invasion
d
Including interaction of invasion with type
tumour of 67 % of the patients who were disease-free at 8 years (controls) and 76 % of the tumours with two or three markers were from patients showing evidence of disease in the 8 years after surgery (cases). Our findings are in agreement with those recently published by Righi et al. [43] in a series of 166 patients with pituitary tumours followed up for at least 6 years. They showed that invasion is significantly associated with recurrence/progression and
123
that Ki-67 index [3 % has high specificity and low sensitivity. The prognostic value of these two parameters varied with tumour type. To our knowledge, this study which includes 410 patients with a long-term follow-up of at least 8 years is the largest case–control study concerning the classification of pituitary tumours ever published. In the interest of pertinence and statistical strength, a case–control design was retained with similar numbers of patients in the two groups and similar numbers of tumours representing each of four types, with only the exceptional TSH type being excluded. The high percentage of patients with evidence of disease (52.4 %) and the percentage of each grade according to tumour type were related to the design of case–control study with similar number of cases and controls although it does not exactly reflect tumour behaviour. In order to prevent a possible bias linked to the particular behaviour of one tumour type, the fitted models included interaction between the tumour type and grade components, thereby giving specific prognostic value to grade components for each tumour type. As a whole, this series is representative of pituitary tumour behaviour commonly seen in clinical practice with similar percentages of invasive tumours (43 %) [31]. We have validated the prognostic value of this new clinicopathological classification. First of all, we confirmed the poor prognostic value of young age in predicting disease-free outcome and also the recurrence/progression-free status at 8 years (review in [45]). However, tumour size was associated with disease-free outcome [28] but not with the recurrence/progression-free status. Patient sex had no prognostic value, as suggested previously [45]. By multivariate analysis, this new clinicopathological classification of pituitary tumours into grades, taking into account tumour invasion and proliferation, displayed highly significant prognostic value for predicting postoperative disease-free outcome or recurrence/progressionfree status, across all tumour types and for each type of tumour. At 8-year follow-up, the probability of a patient showing evidence of disease or showing tumour progression was 25- and 12-fold higher, respectively, if he had an invasive and proliferative tumours (grade 2b) than if he had an non-invasive and non-proliferative tumours (Grade 1a). No grade 3 tumours were identified at the time of initial surgery, the metastasis appearing post-operatively during the 8 years of follow-up. Six out of the eight carcinomas identified were classified as grade 2b at the initial surgery. These results confirm those obtained in our preliminary study on 94 PRL tumours [39] that were classified into three groups, i.e. non-invasive, invasive and aggressive-invasive, corresponding respectively to the grades 1a, 2a and 2b of the present classification. In this case–control study, the percentage of grade 2b (15 %) is similar to the percentage
Acta Neuropathol
of the atypical adenoma in the Law’s surgical series in which 83 % were invasive tumours [62]. Our case–control study comprising 410 patients with a long post-operative follow-up demonstrates that invasion and proliferation, the two grade components, were independently [18, 55] and strongly associated with both disease-free and recurrence/ progression-free status at 8-year follow-up. Using a multivariate statistical model allowing distinct effects of the grade components according to tumour type, and a predicted ROC curves analysis comparing this model with another that considered proliferation without invasion, we demonstrated invasion as major prognostic factor in predicting the disease-free status which varied according to tumour type. These results confirm the well-known prognostic value of invasion in predicting the disease-free status following surgical removal of pituitary tumours [61]. This new clinicopathological classification should be used in conjunction with clinical findings, especially those from MRI which is needed to evaluate the invasion. Indeed in the present study, confirmation of invasion by histology was only possible for 9 % of invasive tumours in cases of invasion of the sphenoid sinus. Data from pituitary imaging and surgical findings are included in the patient database, which is available to the pathologist when establishing the tumour diagnosis. In our opinion, these data should be added to the recently published synoptic checklist for pituitary lesions [32]. Adapted from classifications of endocrine tumours with hypersecretion and tumours incurring problems defining the histological criteria of malignancy, such as pancreatic and adrenal tumours, this classification forms a more representative framework onto which other clinical factors and pathological signs can be incorporated. These include male sex in PRL adenomas [8], subtypes such as sparsely granulated GH adenoma [33], silent ACTH adenoma (review in [40]), silent GH–PRL or subtype 3 adenomas [19] which must also be taken into consideration, due to their suspected aggressiveness. The expressions of PSA-NCAM [55], endocan [5], PTTG [63], FGFR4 [36], and certain molecular markers [14, 39, 58, 59] as well as genetic alterations [34, 41, 42, 60] have all been suggested to be correlated to invasion and aggressiveness, however they need confirming by clinical correlations in large series. In conclusion, this new clinicopathological classification of pituitary endocrine tumours into five grades presented here is easy to implement, although it does require the multidisciplinary collaboration between pathologist, endocrinologist and neurosurgeon. It has proven prognostic value in predicting disease-free and recurrence/progression-free status and has highlighted tumours with signs of invasion and proliferation (grade 2b) as being those most likely to show aggressive behaviour and metastasize. We hope these findings will convince pathologists to adopt this
clinicopathological classification and that clinicians will consider it helpful when choosing the appropriate therapeutic strategy. Acknowledgments We thank Emily Witty from Angloscribe for help with the translation into English, A. Reynaud for her excellent technical assistance, P. Ge´rardi for typing the manuscript, Dr David Forman and Dr Alexandre Vasiljevic for helpful criticisms. Co-authors of HYPOPRONOS are Barlier A., Bernier M., Bonnet F., Borson-Chazot F., Brassier G., Caulet-Maugendre S., Chabre O., Chanson P., Cottier JF., Delemer B., Delgrange E., Di Tommaso L., Eimer S., Gaillard S., Jan M., Girard JJ., Lapras V., Loiseau H., Passagia JG., Patey M., Penfornis A., Poirier JY., Perrin G., Tabarin A. This work was supported by grants from the Ministe`re de la Sante´ (Programme Hospitalier de Recherche Clinique National no 27-43, HYPOPRONOS) and research contracts with the Institut National de la Sante´ et de la Recherche Me´dicale and the Ligue Contre le Cancer Rhoˆne-Alpes. Conflict of interest
The authors have nothing to disclose.
References 1. Amar AP, Hinton DR, Krieger MD, Weiss MH (1999) Invasive pituitary adenomas: significance of proliferation parameters. Pituitary 2(2):117–122 2. Asa SL (2011) Tumors of the Pituitary Gland AFIP Fascicle 15/4. 1st edn. p 275. ISBN: 9781933477152 3. Barker FG 2nd, Klibanski A, Swearingen B (2003) Transsphenoidal surgery for pituitary tumors in the United States, 1996–2000: mortality, morbidity, and the effects of hospital and surgeon volume. J Clin Endocrinol Metab 88(10):4709–4719 4. Breslow NE, Day NE (1980) Statistical methods in cancer research. In: The analysis of case-control studies, vol I, issue number 32. IARC Sci Publ, Lyon, pp 5–338 5. Cornelius A, Cortet-Rudelli C, Assaker R, Kerdraon O, Gevaert M-H, Pre´vot V, Lassalle P, Trouillas J, Delehedde M, Maurage C-A (2012) Endothelial expression of endocan is strongly associated with tumor progression in pituitary adenoma. Brain Pathol 22(6):757–764 6. Cottier J-P, Destrieux C, Brunereau L, Bertrand P, Moreau L, Jan M, Herbreteau D (2000) Cavernous sinus invasion by pituitary adenoma: MR imaging. Radiology 215(2):463–469 7. Daly AF, Rixhon M, Adam C, Dempegioti A, Tichomirowa MA, Beckers A (2006) High prevalence of pituitary adenomas: a cross-sectional study in the province of Liege. Belgium. J Clin Endocrinol Metab 91(12):4769–4775 8. Delgrange E, Trouillas J, Maiter D, Donckier J, Tourniaire J (1997) Sex-related difference in the growth of prolactinomas: a clinical and proliferation marker study. J Clin Endocrinol Metab 82(7):2102–2107 9. Dubois S, Guyetant S, Menei P, Rodien P, Illouz F, Vielle B, Rohmer V (2007) Relevance of Ki-67 and prognostic factors for recurrence/progression of gonadotropic adenomas after first surgery. Eur J Endocrinol 157(2):141–147 10. Fernandez A, Karavitaki N, Wass JA (2010) Prevalence of pituitary adenomas: a community-based, cross-sectional study in Banbury (Oxfordshire, UK). Clin Endocrinol (Oxf) 72(3):377–382 11. Figarella-Branger D, Trouillas J (2006) The new WHO classification of human pituitary tumors: comments. Acta Neuropathol 111(1):71–72 12. Francois P, Travers N, Lescanne E, Arbeille B, Jan M, Velut S (2010) The interperiosteo-dural concept applied to the perisellar
123
Acta Neuropathol
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24. 25.
26.
27.
28.
compartment: a microanatomical and electron microscopic study. J Neurosurg 113(5):1045–1052 Fusco A, Zatelli MC, Bianchi A, Cimino V, Tilaro L, Veltri F, Angelini F, Lauriola L, Vellone V, Doglietto F, Ambrosio MR, Maira G, Giustina A, degli Uberti EC, Pontecorvi A, De Marinis L (2008) Prognostic significance of the Ki-67 labeling index in growth hormone-secreting pituitary adenomas. J Clin Endocrinol Metab 93(7):2746–2750 Galland F, Lacroix L, Saulnier P, Dessen P, Meduri G, Bernier M, Gaillard S, Guibourdenche J, Fournier T, Evain-Brion D, Bidart JM, Chanson P (2010) Differential gene expression profiles of invasive and non-invasive non-functioning pituitary adenomas based on microarray analysis. Endocr Relat Cancer 17(2):361–371 Gejman R, Swearingen B, Hedley-Whyte ET (2008) Role of Ki-67 proliferation index and p53 expression in predicting progression of pituitary adenomas. Hum Pathol 39(5):758–766 Grossman AB (2006) The 2004 World Health Organization classification of pituitary tumors: is it clinically helpful? Acta Neuropathol (Berl) 111(1):76–77 Heaney AP (2011) Clinical review: pituitary carcinoma: difficult diagnosis and treatment. J Clin Endocrinol Metab 96(12):3649– 3660 Honegger J, Prettin C, Feuerhake F, Petrick M, Schulte-Monting J, Reincke M (2003) Expression of Ki-67 antigen in nonfunctioning pituitary adenomas: correlation with growth velocity and invasiveness. J Neurosurg 99(4):674–679 Horvath E, Kovacs K, Smyth HS, Cusimano M, Singer W (2005) Silent adenoma subtype 3 of the pituitary—immunohistochemical and ultrastructural classification: a review of 29 cases. Ultrastruct Pathol 29(6):511–524 Jaffrain-Rea ML, Di Stefano D, Minniti G, Esposito V, Bultrini A, Ferretti E, Santoro A, Faticanti Scucchi L, Gulino A, Cantore G (2002) A critical reappraisal of MIB-1 labelling index significance in a large series of pituitary tumours: secreting versus nonsecreting adenomas. Endocr Relat Cancer 9(2):103–113 Kaltsas GA, Nomikos P, Kontogeorgos G, Buchfelder M, Grossman AB (2005) Clinical review: diagnosis and management of pituitary carcinomas. J Clin Endocrinol Metab 90(5):3089–3099 Knosp E, Kitz K, Perneczky A (1989) Proliferation activity in pituitary adenomas: measurement by monoclonal antibody Ki-67. Neurosurgery 25(6):927–930 Knosp E, Steiner E, Kitz K, Matula C (1993) Pituitary adenomas with invasion of the cavernous sinus space: a magnetic resonance imaging classification compared with surgical findings. Neurosurgery 33(4):610–617 (discussion 617–618) Kovacs K, Horvath E, Ezrin C (1977) Pituitary adenomas. Pathol Annu 12(Pt 2):341–382 Kovacs K, Scheithauer B, Horvath E, Lloyd RV (1996) The WHO classification of adenohypophysial neoplasms. A proposed five-tier scheme. Cancer 78:502–510 Laws ER Jr, Lopes MB (2006) The new WHO classification of pituitary tumors: highlights and areas of controversy. Acta Neuropathol 111(1):80–81 Lloyd RV, Kovacs K, Young Jr WF, Farrel WE, Asa SL, Trouillas J, Kontogeorgos G, Sano T, Scheithauer B, Horvath E (2004) Pituitary tumours : introduction. In: DeLellis RA, Lloyd RV, Heitz PU, Eng C (eds) Tumours of the pituitary, chapter 1. Pathology and genetics of tumours of endocrine organs. World Hearth Organization Classification of Tumours. IARC Press, Lyon, pp 10–13 Losa M, Mortini P, Barzaghi R, Ribotto P, Terreni MR, Marzoli SB, Pieralli S, Giovanelli M (2008) Early results of surgery in patients with nonfunctioning pituitary adenoma and analysis of the risk of tumor recurrence. J Neurosurg 108(3):525–532
123
29. Madsen H, Borges TM, Knox AJ, Michaelis KA, Xu M, Lillehei KO, Wierman ME, Kleinschmidt-DeMasters BK (2011) Giant pituitary adenomas: pathologic-radiographic correlations and lack of role for p53 and MIB-1 labeling. Am J Surg Pathol 35(8):1204–1213 30. McCormack AI, Wass JA, Grossman AB (2011) Aggressive pituitary tumours: the role of temozolomide and the assessment of MGMT status. Eur J Clin Invest 41(10):1133–1148 31. Meij BP, Lopes MB, Ellegala DB, Alden TD, Laws ER Jr (2002) The long-term significance of microscopic dural invasion in 354 patients with pituitary adenomas treated with transsphenoidal surgery. J Neurosurg 96(2):195–208 32. Nose V, Ezzat S, Horvath E, Kovacs K, Laws ER, Lloyd R, Lopes MB, Asa SL (2011) Protocol for the examination of specimens from patients with primary pituitary tumors. Arch Pathol Lab Med 135(5):640–646 33. Obari A, Sano T, Ohyama K, Kudo E, Qian ZR, Yoneda A, Rayhan N, Mustafizur Rahman M, Yamada S (2008) Clinicopathological features of growth hormone-producing pituitary adenomas: difference among various types defined by cytokeratin distribution pattern including a transitional form. Endocr Pathol 19(2):82–91 34. Pack SD, Qin LX, Pak E, Wang Y, Ault DO, Mannan P, Jaikumar S, Stratakis CA, Oldfield EH, Zhuang Z, Weil RJ (2005) Common genetic changes in hereditary and sporadic pituitary adenomas detected by comparative genomic hybridization. Genes Chromosomes Cancer 43(1):72–82 35. R Development Core Team (2009) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. http://www.R-project.org 36. Ramirez C, Cheng S, Vargas G, Asa SL, Ezzat S, Gonzalez B, Cabrera L, Guinto G, Mercado M (2012) Expression of Ki-67, PTTG1, FGFR4, and SSTR 2, 3, and 5 in nonfunctioning pituitary adenomas: a high throughput TMA, immunohistochemical study. J Clin Endocrinol Metab 97(5):1745–1751 37. Raverot G, Castinetti F, Jouanneau E, Morange I, FigarellaBranger D, Dufour H, Trouillas J, Brue T (2012) Pituitary carcinomas and aggressive pituitary tumours: merits and pitfalls of temozolomide treatment. Clin Endocrinol (Oxf) 76(6):769–775 38. Raverot G, Sturm N, de Fraipont F, Muller M, Salenave S, Caron P, Chabre O, Chanson P, Cortet-Rudelli C, Assaker R, Dufour H, Gaillard S, Francois P, Jouanneau E, Passagia JG, Bernier M, Cornelius A, Figarella-Branger D, Trouillas J, Borson-Chazot F, Brue T (2010) Temozolomide treatment in aggressive pituitary tumors and pituitary carcinomas: a French multicenter experience. J Clin Endocrinol Metab 95(10):4592–4599 39. Raverot G, Wierinckx A, Dantony E, Auger C, Chapas G, Villeneuve L, Brue T, Figarella-Branger D, Roy P, Jouanneau E, Jan M, Lachuer J, Trouillas J (2010) Prognostic factors in prolactin pituitary tumors: clinical, histological, and molecular data from a series of 94 patients with a long postoperative follow-up. J Clin Endocrinol Metab 95(4):1708–1716 40. Raverot G, Wierinckx A, Jouanneau E, Auger C, Borson-Chazot F, Lachuer J, Pugeat M, Trouillas J (2010) Clinical, hormonal and molecular characterization of pituitary ACTH adenomas without (silent corticotroph adenomas) and with Cushing’s disease. Eur J Endocrinol 163(1):35–43 41. Rickert CH, Dockhorn-Dworniczak B, Busch G, Moskopp D, Albert FK, Rama B, Paulus W (2001) Increased chromosomal imbalances in recurrent pituitary adenomas. Acta Neuropathol 102(6):615–620 42. Rickert CH, Scheithauer BW, Paulus W (2001) Chromosomal aberrations in pituitary carcinoma metastases. Acta Neuropathol 102(2):117–120 43. Righi A, Agati P, Sisto A, Frank G, Faustini-Fustini M, Agati R, Mazzatenta D, Farnedi A, Menetti F, Marucci G, Foschini MP
Acta Neuropathol
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
(2012) A classification tree approach for pituitary adenomas. Hum Pathol 43(10):1627–1637 Rindi G, Kloppel G, Alhman H, Caplin M, Couvelard A, de Herder WW, Erikssson B, Falchetti A, Falconi M, Komminoth P, Korner M, Lopes JM, McNicol AM, Nilsson O, Perren A, Scarpa A, Scoazec JY, Wiedenmann B (2006) TNM staging of foregut (neuro)endocrine tumors: a consensus proposal including a grading system. Virchows Arch 449(4):395–401 Roelfsema F, Biermasz NR, Pereira AM (2012) Clinical factors involved in the recurrence of pituitary adenomas after surgical remission: a structured review and meta-analysis. Pituitary 15(1):71–83 Saeger W, Ludecke B, Ludecke DK (2008) Clinical tumor growth and comparison with proliferation markers in non-functioning (inactive) pituitary adenomas. Exp Clin Endocrinol Diabetes 116(2):80–85 Saeger W, Ludecke DK, Buchfelder M, Fahlbusch R, Quabbe HJ, Petersenn S (2007) Pathohistological classification of pituitary tumors: 10 years of experience with the German Pituitary Tumor Registry. Eur J Endocrinol 156(2):203–216 Salehi F, Agur A, Scheithauer BW, Kovacs K, Lloyd RV, Cusimano M (2009) Ki-67 in pituitary neoplasms: a review—part I. Neurosurgery 65(3):429–437 (discussion 437) Scheithauer BW, Kovacs KT, Laws ER Jr, Randall RV (1986) Pathology of invasive pituitary tumors with special reference to functional classification. J Neurosurg 65(6):733–744 Scheithauer BW, Kurtkaya-Yapicier O, Kovacs KT, Young WF Jr, Lloyd RV (2005) Pituitary carcinoma: a clinicopathological review. Neurosurgery 56(5):1066–1074 Solcia E, Klo¨ppel G, Sobin L (2000) Histological classification of adenohypophyseal tumours. Histological classification of endocrine tumours. Springer, Berlin Thapar K, Kovacs K, Scheithauer BW, Stefaneanu L, Horvath E, Pernicone PJ, Murray D, Laws ER Jr (1996) Proliferative activity and invasiveness among pituitary adenomas and carcinomas: an analysis using the MIB-1 antibody. Neurosurgery 38(1):99–106 Thapar K, Scheithauer BW, Kovacs K, Pernicone PJ, Laws ER Jr (1996) p53 expression in pituitary adenomas and carcinomas: correlation with invasiveness and tumor growth fractions. Neurosurgery 38(4):765–770 Trouillas J, Cure M, Lheritier M, Guichard Y, Girod C (1974) Essai de classification cytofonctionnelle de 19 ade´nomes hypophysaires
55.
56.
57.
58.
59.
60.
61. 62.
63.
d’apre`s des donne´es histologiques, immunocytochimiques, biome´triques et ultrastructurales. Rev Oto-Neuro-Ophtalmol 47(3):223– 236 Trouillas J, Daniel L, Guigard MP, Tong S, Gouvernet J, Jouanneau E, Jan M, Perrin G, Fischer G, Tabarin A, Rougon G, Figarella-Branger D (2003) Polysialylated neural cell adhesion molecules expressed in human pituitary tumors and related to extrasellar invasion. J Neurosurg 98(5):1084–1093 Trouillas J, Girod C (1996) Pathology of pituitary adenomas. In: Landolt AM, Vance ML, Reilly PL (eds) Pituitary adenomas. Churchill Livingstone, New York, pp 27–46 Trouillas J, Labat-Moleur F, Sturm N, Kujas M, Heymann MF, Figarella-Branger D, Patey M, Mazucca M, Decullier E, Verges B, Chabre O, Calender A (2008) Pituitary tumors and hyperplasia in multiple endocrine neoplasia type 1 syndrome (MEN1): a casecontrol study in a series of 77 patients versus 2509 non-MEN1 patients. Am J Surg Pathol 32(4):534–543 Wierinckx A, Auger C, Devauchelle P, Reynaud A, Chevallier P, Jan M, Perrin G, Fevre-Montange M, Rey C, Figarella-Branger D, Raverot G, Belin MF, Lachuer J, Trouillas J (2007) A diagnostic marker set for invasion, proliferation, and aggressiveness of prolactin pituitary tumors. Endocr Relat Cancer 14(3):887–900 Wierinckx A, Raverot G, Nazaret N, Jouanneau E, Auger C, Lachuer J, Trouillas J (2010) Proliferation markers of human pituitary tumors: contribution of a genome-wide transcriptome approach. Mol Cell Endocrinol 326(1–2):30–39 Wierinckx A, Roche M, Raverot G, Legras-Lachuer C, Croze S, Nazaret N, Rey C, Auger C, Jouanneau E, Chanson P, Trouillas J, Lachuer J (2011) Integrated genomic profiling identifies loss of chromosome 11p impacting transcriptomic activity in aggressive pituitary PRL tumors. Brain Pathol 21(5):533–543 Wolfsberger S, Knosp E (2006) Comments on the WHO 2004 classification of pituitary tumors. Acta Neuropathol 111(1):66–67 Zada G, Woodmansee WW, Ramkissoon S, Amadio J, Nose V, Laws ER Jr (2011) Atypical pituitary adenomas: incidence, clinical characteristics, and implications. J Neurosurg 114(2):336–344 Zhang X, Horwitz GA, Heaney AP, Nakashima M, Prezant TR, Bronstein MD, Melmed S (1999) Pituitary tumor transforming gene (PTTG) expression in pituitary adenomas. J Clin Endocrinol Metab 84(2):761–767
123
