Study population
A group of 166 patients (93 men, 56%, mean age 58.5 ± 14.8 year) with NF-pitNETs was studied. In approximately one third of patients (n = 62, 37.3%) the diagnosis of adenoma was incidental. In symptomatic patients, the median time from onset of symptoms to diagnosis was 3 (1–12) months. The most prevalent neuro-ophthalmologic symptoms at diagnosis were headaches (n = 82, 49.4%) and visual disturbances (n = 66, 39.8%). Approximately half of the patients (n = 84, 50.6%) had some anterior pituitary hormone deficiency at diagnosis, with complete hypopituitarism in around 10%. The maximal tumor diameter was 22.0 (15–28) mm and the median tumor volume was 3,579 (1,281-6,782) mm3. More than half of the patients presented suprasellar extension with chiasmatic compression (n = 94, 56.6%) and 66.9% (n = 111) showed cavernous sinus invasion (Table 1).
Table 1
Clinical characteristics of patients according to the intensity of pituitary adenoma on T2-weighted MRI.
| Total | Hypointense | Isointense | Hyperintense | Non-isointense (hypo- and hyperintense) |
n (%) | 166 | 25 (15.1) | 57 (34.3) | 84 (50.6) | 109 (65.7) |
Age (yr) | 58.5 ± 14.8 | 59.5 ± 14.8 | 59.1 ± 14.7 | 57.8 ± 15.0 | 58.2 ± 14.9 |
Sex (male) | 93 (56) | 19 (76) | 30 (52.6) | 44 (52.4) | 63 (57.8) |
Incidental | 62 (37.3) | 14 (56) | 23 (40.3) | 25 (29.8) | 39 (35.8) |
Time from onset of symptoms to diagnosis (months) | 3 (1–12) | 2 (1–7) | 3 (1–7) | 3 (1–12) | 3 (1–12) |
Headaches | 82 (49.4) | 14 (56) | 25 (43.9) | 43 (51.2) | 57 (52.3) |
Visual disturbances | 66 (39.8) | 7 (28) | 18 (31.6) | 41 (48.8) | 48 (44) |
Hypopituitarism | 84 (50.6) | 15 (60) | 23 (40.3) | 46 (54.7) | 61 (56) |
Type of hypopituitarisma |
Partial | 66 (39.7) | 13 (86.7) | 14 (60.9) | 39 (84.8) | 52 (85.2) |
Complete | 18 (10.8) | 2 (13.3) | 9 (39.1) | 7 (15.2) | 9 (14.7) |
Maximum tumor diameter (mm)*, b | 22.0 (15–28) | 22 (17–28) | 16 (13–25) | 23 (16.6–29.7) | 23 (16.9–28.5) |
Volume (mm3)*, c | 3,579 (1,281-6,782) | 3,945 (1,915-6,389) | 1,523 (618-5,226) | 4,012 (2,506-8,320) | 4,012 (2,243-7,592) |
Chiasmatic compression*, b | 94 (56.6) | 17 (68) | 22 (38.6) | 55 (65.5) | 72 (66.1) |
Cavernous sinus invasion*, c | 111 (66.9) | 16 (64) | 28 (49.1) | 67 (79.8) | 83 (76.1) |
Knosp grade**, c |
0 | 55 (33.1) | 9 (36) | 29 (50.9) | 17 (17.9) | 26 (23.9) |
1 | 18 (10.8) | 2 (8) | 11 (19.3) | 5 (5.9) | 7 (6.4) |
2 | 41 (24.7) | 6 (24) | 7 (12.3) | 28 (33.3) | 34 (31.2) |
3 | 33 (19.9) | 5 (20) | 8 (14) | 20 (23.8) | 25 (22.9) |
4 | 19 (11.4) | 3 (12) | 2 (3.5) | 14 (16.7) | 17 (15.6) |
Surgery (yes) a | 116 (69.9) | 18 (72) | 34 (59.6) | 64 (76.2) | 82 (75.2) |
Type of surgery |
EETA | 98 (84.5) | 15 (83.3) | 27 /79.4) | 56 (87.5) | 71 (86.6) |
MTS | 17 (14.6) | 3 (16.7) | 7 (20.6) | 7 (10.9) | 10 (12.2) |
TS | 1 (0.9) | 0 | 0 | 1 (1.6) | 1 (1.2) |
Immunophenotypea |
Gonadotroph adenoma | 39 (35.8) | 4 (25) | 9 (30) | 26 (41.3) | 30 (38) |
Null cell adenoma | 35 (32.1) | 6 (37.5) | 8 (26.7) | 21 (33.3) | 27 (34.2) |
Plurihormonal | 13 (11.9) | 2 (12.5) | 8 (26.7) | 3 (4.8) | 5 (6.3) |
Silent lactotroph adenoma | 8 (7.3) | 2 (12.5) | 4 (13.3) | 2 (3.2) | 4 (5.1) |
Silent corticotropinoma | 9 (8.3) | 1 (6.2) | 1 (3.3) | 7 (11.1) | 8 (10.1) |
Silent somatotropinoma | 5 (4.6) | 1 (6.2) | 0 | 4 (6.3) | 5 (6.3) |
Total | 109 (100) | 16 (14.7) | 30 (27.5) | 63 (57.8) | 79 (72.5) |
Ki-67 index | 1 (1–3) | 0 (1-1.7) | 2 (1–3) | 1 (1–3) | 1 (0-2.2) |
Data are expressed as absolute numbers (percentages), mean ± SD or median (interquartile range). |
Abbreviations: EETA, endoscopic endonasal transsphenoidal approach; MTS, microscopic transsphenoidal surgery; TS, transcranial surgery.
*p<0.01; **p<0.001 for comparison between the hypo-, iso-, and hyperintensity groups.
ap<0.05; bp<0.01; cp<0.001 for comparison between iso-, and non-isointense groups.
Clinical, hormonal and radiological data
The main clinical, hormonal and radiological data are summarized in Table 1. Approximately half of the tumors (n = 84, 50.6%) were hyperintense on T2, while 34.3% (n = 57) and 15.1% (n = 25) were iso- and hypointense, respectively. We found no significant differences in age at diagnosis between the different groups analyzed. In relation to sex, although the percentage of males (n = 19, 76%) was higher in the group of patients with hypointense tumors, these differences were not statistically significant when compared with iso- (n = 30, 52%) and hyperintense (n = 44, 52.4%) tumors.
The percentage of incidental tumors was 56%, 40.3%, and 29.8% for hypo-, iso-, and hyperintense adenomas, respectively (p = 0.05). No significant differences in the time of evolution from symptom onset to diagnosis in symptomatic tumors were found. Clinically, the percentage of patients presenting headaches was similar in the three groups of patients. However, there was a greater tendency to present visual alterations in the group of patients with hyperintense tumors (p = 0.05). We found no statistically significant difference in adenohypophyseal hormonal deficiency between the 3 groups. However, when the proportion of partial and complete hypopituitarism was compared between the group of patients with isointense and non-isointense tumors, a higher prevalence of partial hypopituitarism at diagnosis was found in the group of non-isointense tumors, whereas the prevalence of complete hypopituitarism was higher in the group of isointense tumors (p = 0.015) (Table 1).
When the 3 groups were compared in relation to the maximum tumor diameter, statistically significant differences were observed (p = 0.004) (Table 1). The analysis of multiple comparisons (post-hoc tests) showed that maximum tumor diameter of the isointense group [16 (13–25) mm] was significantly lower than that of hyperintense [23 (16.6–29.7) mm] group (p = 0.003), without significant differences between iso- and hypointensive groups.
Similarly, the analysis of tumor volume between the 3 study groups also showed statistically significant differences (p = 0.001) (Table 1).The analysis of multiple comparisons (post-hoc tests) showed that volume tumor of the isointense group [16 (13–25) mm] was significantly lower than that of hyperintense [23 (16.6–29.7) mm] group (p = 0.002), without significant differences between iso- and hypointensive groups.
The percentage of patients presenting adenomas associated with chiasmatic compression was significantly lower in the isointense (n = 22, 38.6%) compared to hypointense (n = 17, 68%) and hyperintense (n = 55, 65.5%) tumors (p = 0.003).
Invasive tumors were significantly more hyperintense (n = 67, 69.8%) compared to hypointense (n = 16, 64%) and isointense (n = 28, 49.1%) tumors (p = 0.001). In addition, hyperintense tumors presented a significantly higher degree of invasion (Knosp grades 3 and 4) than hypo- and isointense tumors (p < 0.001) (Table 1).
The isointense tumor group showed significantly lower maximum tumor diameter (p = 0.001), tumor volume (p < 0.001), percentage of chiasmatic compression (p = 0.001) and cavernous sinus invasion (p < 0.001) than the non-isointense (hypo- and hyperintense) tumor group (Table 1).
Therapy and pathological study
Of the 166 patients, 116 (69.9%) underwent surgery, 32 (19.3%) were followed by active surveillance and 18 (10.8%) were treated with cabergoline as first-line therapy. No differences were found between the type of first-line treatment and intensity in the T2-weighted sequence. Twenty-eight patients (24.1%) were treated with radiotherapy, 24 (20.7%) of them after first surgery and 4 (3.4%) after second surgery. There was also no association between radiotherapy treatment and intensity on the T2-weighted sequence.
The percentage of patients who underwent surgery was higher in hyperintense tumors (n = 64, 76%), followed by the hypointense (n = 18, n = 72%), and isointense groups (n = 34, 59.6%) (Table 1). Although no significant differences were found between these proportions, when we compared the group of patients with isointense tumors with the group of non-isointense tumors (hypo- or hyperintense) grouped together, significant differences were found. In this regard, the percentage of patients with non-isointense tumors who underwent surgery was significantly higher than that of patients with isointense tumors (75.2% vs. 59.6%, p = 0.038) (Fig. 1). No differences were found between the study groups according to T2 intensity and type of surgery (Table 1).
No association was found between the intensity in T2-weighted sequences of the tumors analyzed and the use of active surveillance, cabergoline therapy or radiotherapy.
Immunohistochemical analysis of the anterior pituitary hormones was available in 109 (94%) tumors. The most frequent histological type was gonadotropic adenoma (n = 39, valid percentage 35.8%), followed by null cell adenoma (n = 35, 32.1%), plurihormonal adenoma (n = 13, 11.9%), silent corticotropinoma (n = 9, 8.3%), silent lactotropic adenoma (n = 8, 7.3%) and silent somatotropic adenoma (n = 5, 4.6%) (Table 1).
When the 3 groups (hypo-, iso- and hyperintense) were considered separately, we found no significant differences between the groups analyzed and the pathological study (p = 0.075). However, when the tumors were grouped into isointense and non-isointense, statistically significant differences (p = 0.02) were found (Table 1; Fig. 2). In this regard, the most frequent histological group in both groups of tumors was gonadotropinoma, followed by null cell adenoma; however, plurihormonal tumors and silent lactotroph adenomas were more common in the group of isointense tumors, while silent corticotropinomas and silent somatotropinomas were more frequent in the group of non-isointense tumors (Table 1, Fig. 2). It is noteworthy that of the 9 patients with silent corticotropinomas, 7 belonged to the hyperintense group, while in the hypointense and isointense groups there was only one patient with this diagnosis. Lastly, of the 5 patients with silent somatotropinomas, 4 belonged to the hyperintense tumor group and 1 to the hypointense tumor group.
The median value of the Ki-67 proliferation index was 1% (1–3) and no significant differences in the Ki-67 value between the groups analyzed were found.
Therapeutic responses and clinical follow-up
In those 116 patients who underwent surgery, after a follow-up time of 66 (22.7-118.2) months after the initial surgery, 34.5% (40/116) of the patients showed complete resection and tumor persistence in the remaining patients. No significant differences in the type of response to surgical treatment between the groups analyzed in relation to the intensity in the T2 sequence were found.
In those patients who did not undergo surgery [32 patients under active surveillance and 18 under dopamine agonists (DA)], SD was observed in 75% (24/32) and 61% (11/18), PR in 12.5% (4/32) and 22.2% (4/18), and progression in 12.5% (4/32) and 11.7% (3/18) of the tumors belonging to the active surveillance group and the DA group, respectively (ns). There were also no significant differences in the type of response between the groups analyzed in relation to the intensity in the T2 sequence.
Lastly, in the total patient population, the evaluation of the clinical situation in the last clinical visit showed that the prevalence of hypopituitarism was higher in patients with hypointense tumors (20/24, 83.3%), than in hyperintense (55/83, 66.3%) or isointense (28/57, 49.1%) adenomas (p = 0.009). Gonadotropin deficiency behaved similarly (hypointense 16/24, 66.7%; hyperintense 47/74, 63.5%; and isointense 20/50, 40%; p = 0.018). However, no statistically significant differences were found according to T2 intensity and the presence of headaches, visual disturbances, cranial nerve involvement, cognitive impairment, optic neuropathy, second tumor development, radionecrosis, and mortality.