Study Population and Determination of GNAS Mutations
Our study enrolled 97 acromegalic patients (Fig. 1, Table 1). The average age at diagnosis was 43.3 ± 11.0 years old, and 44.3% (43/97) were male patients. Their median diagnosis delay time was 60.0(24.0, 120.0) months. 9.4% (9/96) were microadenomas, 90.6% (87/96) were macroadenomas, 3.1% (3/96) were giant adenomas, while the tumor maximum diameter of whole cohort was 2.02 ± 0.89cm. All patients received transsphenoidal surgery (TSS) by one neurosurgical team at PUMCH. The median follow-up duration after TSS was 7.0(4.0, 18.0) months. 46.7% (28/60) patients were in biochemical remission at their last follow-up.
Table 1
Clinical characteristics of acromegalic patients in relation to presence or absence of somatic GNAS mutation
| Total (n = 97) | GNAS + group(n = 43) | GNAS -group(n = 54) | P value |
---|
Gender (males) | 43/97(44.3%) | 25/43(58.1%) | 18/54(33.3%) | 0.015* |
Age at diagnosis (year) | 43.3 ± 11.0 | 43.8 ± 9.1 | 43.0 ± 12.4 | 0.726 |
Diagnosis delay (month) | 60.0(24.0, 120.0) | 72.0(48.0, 120.0) | 36.0(21.0, 75.0) | 0.002** |
BMI (kg/m2) | 25.63 ± 3.28 | 26.28 ± 3.39 | 25.12 ± 3.13 | 0.082 |
Baseline characteristics | |
Physical changes | 90/97(92.8%) | 41/43(95.3%) | 49/54(90.7%) | 0.458 |
Headache | 28/97(28.9%) | 12/43(27.9%) | 16/54(29.6%) | 0.852 |
Vision impairment | 18/97(18.6%) | 4/43 (9.3%) | 14/54(25.9%) | 0.036* |
Visual field defects | 7/97(7.2%) | 3/43 (7.0%) | 4/54 (7.4%) | 1.000 |
GH (ng/mL) | 17.90(9.45, 33.65) | 20.80(10.50, 35.60) | 15.05(7.43, 32.40) | 0.194 |
GH-nadir (ng/mL) | 13.65(6.12, 26.14) | 16.40(6.54, 31.10) | 12.50(5.35, 24.20) | 0.292 |
GH/tumor volume (ng/mL/cm3) | 14.20(3.89, 23.74) | 18.93(9.67, 30.12) | 10.91(2.80, 20.40) | 0.005** |
IGF-1 SDS | 6.05 ± 1.68 | 6.14 ± 1.36 | 5.98 ± 1.90 | 0.630 |
Maximum diameter (cm) | 2.02 ± 0.89 | 1.75 ± 0.83 | 2.23 ± 0.89 | 0.008** |
Tumor volume (cm3) | 1.74(0.95, 5.22) | 1.39(0.54, 2.92) | 3.19(1.07, 6.13) | 0.025* |
Macroadenomas | 87/96 (90.6%) | 35/42(83.3%) | 52/54(96.3%) | 0.039* |
Cavernous sinus invasion (Knosp ≥ 3) | 29/85 (34.1%) | 9/36 (25.0%) | 20/49(40.8%) | 0.129 |
Impaired glucose metabolism | 39/97(40.2%) | 19/43(44.2%) | 20/54(37.0%) | 0.476 |
Cardiac diseases | 40/95(42.1%) | 21/42(50.0%) | 19/53(35.8%) | 0.165 |
Dyslipidemia | 44/82(53.7%) | 24/37(64.9%) | 20/45(44.4%) | 0.065 |
Surgery | |
Tumor texture Soft | 63/96 (65.6%) | 29/43(67.4%) | 34/53(64.2%) | 0.716 |
Tough | 26/96 (27.1%) | 12/43(27.9%) | 14/53(26.4%) |
Uneven | 7/96 (7.3%) | 2/43 (4.7%) | 5/53 (9.4%) |
Ki-67 index 1% | 29/83(34.9%) | 18/37(48.6%) | 11/46(23.9%) | 0.042* |
2% | 22/83(26.5%) | 9/37 (24.3%) | 13/46(28.3%) |
3–4% | 26/83(31.3%) | 7/37 (18.9%) | 19/46(41.3%) |
≥ 5% | 6/83(7.2%) | 3/37 (8.1%) | 3/46 (6.5%) |
CSF linkage | 28/97(28.9%) | 9/43 (20.9%) | 19/54(35.2%) | 0.124 |
Immediate postoperative GH (ng/mL) | 1.70(0.85, 3.30) | 1.70(0.80, 4.70) | 1.75(0.98, 3.00) | 0.757 |
Immediate postoperative GH-nadir (ng/mL) | 1.10(0.55, 2.81) | 1.09(0.37, 7.98) | 1.10(0.66, 2.66) | 0.775 |
Immediate postoperative IGF-1 SDS | 4.85 ± 1.84 | 5.32 ± 1.71 | 4.51 ± 1.88 | 0.036* |
Immediate postoperative GH normalization | 38/97 (39.2%) | 19/43(44.2%) | 19/54(35.2%) | 0.367 |
Follow-up | |
Follow-up duration (month) | 7.0(4.0, 18.0) | 4.8(3.4, 19.0) | 7.5(4.0, 17.3) | 0.466 |
GH at last visit (ng/mL) | 1.10(0.48, 4.10) | 0.85(0.46, 4.80) | 1.70(0.45, 3.75) | 0.706 |
GH-nadir at last visit (ng/mL) | 0.65(0.17, 2.41) | 0.41(0.11, 1.58) | 0.85(0.21, 2.62) | 0.372 |
GH normalization | 40/63 (63.5%) | 17/26(65.4%) | 23/37(62.2%) | 0.794 |
IGF-1 SDS at last visit | 1.80(0.81, 3.35) | 2.38 ± 1.82 | 1.71(0.56, 2.94) | 0.185 |
IGF-1normalization | 34/60 (56.7%) | 13/25(52.0%) | 21/35(60.0%) | 0.538 |
Remission rate | 28/60 (46.7%) | 10/25(40.0%) | 18/35(51.4%) | 0.382 |
*p < 0.05, **p < 0.01 |
43 out of 97 patients (44.3%) carried somatic GNAS gene mutation, all of which were heterozygous missense mutations. These mutations were: Arg201Cys (n = 33, 34.0%), Arg201His (n = 3, 3.1%), Arg201Ser (n = 2, 2.1%), and Gln227Leu (n = 5, 5.2%). Representative Sanger sequencing results and mutation site location were shown in Figs. 2 and 3. (GNAS gene transcript: NM_000516.7)
Comparison between acromegalic patients with and without GNAS mutations
In terms of demographics, patients in the two groups were similar in age distribution. But there were more males (58.1% vs 33.3%, p = 0.015) and longer diagnosis delay [72.0(48.0, 120.0) vs 36.0(21.0, 75.0) months, p = 0.002] in patients with somatic GNAS mutations.
At the first visits, the symptoms of acromegaly-specific physical changes, headaches and visual field defects happened similarly in two groups. The baseline fasting GH levels and IGF-1 SDS were slightly higher in mutant patients, besides, the GH secreted by per unit of tumor volume was significantly higher [18.93(9.67, 30.12) vs 10.91(2.80, 20.40) ng/mL/cm3, p = 0.005]. According to pituitary MRI, the tumor maximum diameter (1.75 ± 0.83 vs 2.23 ± 0.89cm, p = 0.008), tumor volume [1.39(0.54, 2.92) vs 3.19(1.07, 6.13) cm3, p = 0.025] and proportion of macroadenomas (83.3% vs 96.3%, p = 0.039) were all significantly smaller in patients with GNAS mutations. No significant differences were found in tumor invasiveness (Knosp grade ≥ 3, 25.0% vs 40.8%, p = 0.129). In addition, the incidence of paradoxical GH increase during OGTT was somewhat lower in mutation group (16.7% vs 34.0%, p = 0.059).
Tumor textures were evaluated during operation, and surgeons didn’t find much difference in tumor textures between two groups, both with soft adenoma predominating (67.4% vs 64.2%, p = 0.716). According to operative recordings, intraoperative cerebrospinal fluid (CSF) leakage also happened similarly in two groups (20.9% vs 35.2%, p = 0.124). OGTT and serum GH levels were measured immediately after surgeries (within 3 days postoperatively), there was no predominant difference in postoperative fasting GH levels [1.70(0.80, 4.70) vs 1.75(0.98, 3.00) ng/mL, p = 0.757] or OGTT-nadir GH levels [1.09(0.37, 7.98) vs 1.10(0.66, 2.66) ng/mL, p = 0.775] between two groups, either. The GH levels of 44.2% (19/43) mutated patients and 35.2% (19/54) non-mutated patients decreased to less than 1.0µg/L immediately after the surgery, and could be considered to have achieved biochemical remission (p = 0.367). In terms of Ki-67 indexes, patients in mutant group tend to have lower Ki-67 indexes (p = 0.042, R= -0.225), with only 27.0% (10/37) having a Ki-67 index ≥ 3%, while this proportion in non-mutation group was 47.8% (22/46).
26 patients with GNAS mutations and 37 patients without GNAS mutations who didn’t receive any other postoperative treatments were followed up. The median follow-up duration was 4.8 (3.4, 19.0) months and 7.5(4.0, 17.3) months, respectively (p = 0.466). At the last visits, random GH levels, OGTT-nadir GH levels, IGF-1 SDS and the proportions of GH and IGF-1 SDS normalization were all similar in two groups. Overall, 40.0% (10/25) patients of mutation group and 51.4% (18/35) of non-mutation group met the acromegaly remission criteria (p = 0.382). In addition, 4 GNAS mutated patients and 5 non-mutated patients received postoperative SRLs and (or) radiotherapy therapies, but none of them achieved biochemical remission at last follow-up.
Clinical characteristics according to GNAS mutations in patients with different genders
We also analyzed the clinical characteristics according to GNAS mutations in patients with different genders (Table 2). In male patients, the baseline GH levels [24.40(14.40, 36.30) vs 10.55(5.25, 16.95) ng/mL, p = 0.002] and OGTT-nadir GH levels [20.00(9.36, 32.30) vs 10.04(4.01, 14.55) ng/mL, p = 0.010] of patients with GNAS mutations were significantly higher than those in non-mutation group. Whereas, the tumor maximum diameter, tumor volume, GH secreted by per unit tumor volume and IGF-1 SDS didn’t differ in above two groups. In female patients, the tumor maximum diameter (1.55 ± 0.55 vs 2.32 ± 0.85cm, p < 0.001) and volume [1.08 (0.51, 1.68) vs 3.58(1.45, 6.07) cm3, p = 0.003] of GNAS mutant patients were significantly smaller than those in non-mutation group, and thus their GH level of per unit tumor volume was much higher [20.83(12.74, 29.93) vs 11.26(2.87, 18.91) ng/mL/cm3, p = 0.013]. The diagnosis delay time was longer [60.0(42.0, 120.0) vs 36.0(15.0, 60.0) months, p = 0.027] and Ki-67 index was lower in female GNAS mutant patients (p = 0.012, R=-0.299). However, baseline GH levels and IGF-1 SDS of female mutation group were not significantly different from female non-mutation group.
Table 2
Clinical characteristics of acromegaly patients in relation to the status of GNAS gene in different genders
| Male | Female | P3 | P4 |
---|
| GNAS + group (n = 25) | GNAS- group (n = 18) | P1 | GNAS + group (n = 18) | GNAS- group (n = 36) | P2 |
---|
Age at diagnosis (year) | 42.8 ± 9.1 | 44.4 ± 13.6 | 0.667 | 45.1 ± 9.1 | 42.3 ± 11.9 | 0.388 | 0.436 | 0.550 |
Diagnosis delay (month) | 84.0(54.0, 120.0) | 60.0(21.0, 120.0) | 0.243 | 60.0(42.0, 120.0) | 36.0(15.0, 60.0) | 0.027 | 0.285 | 0.207 |
Impaired glucose metabolism | 11/25 (44.0%) | 3/18 (16.7%) | 0.059 | 8/18 (44.4%) | 17/36 (47.2%) | 0.847 | 0.977 | 0.028 |
Dyslipidemia | 12/21 (57.1%) | 3/15 (20.0%) | 0.026 | 12/16 (75.0%) | 17/30 (56.7%) | 0.220 | 0.260 | 0.020 |
Baseline GH (ng/mL) | 24.40(14.40, 36.30) | 10.55(5.25, 16.95) | 0.002** | 18.75(8.00, 40.25) | 17.30(10.70, 61.73) | 0.620 | 0.301 | 0.018* |
Baseline GH-nadir (ng/mL) | 20.00(9.36, 32.30) | 10.04(4.01, 14.55) | 0.010* | 13.05(4.99, 23.68) | 14.40(7.24, 33.53) | 0.477 | 0.158 | 0.084 |
GH/tumor volume (ng/mL/cm3) | 14.70(7.41, 32.32) | 10.75(1.93, 23.14) | 0.121 | 20.83(12.74, 29.93) | 11.26(2.87, 18.91) | 0.013* | 0.513 | 0.811 |
IGF-1 SDS | 6.34 ± 1.16 | 6.30 ± 1.67 | 0.925 | 5.86 ± 1.58 | 5.82 ± 2.01 | 0.940 | 0.257 | 0.387 |
Cavernous sinus invasion (Knosp ≥ 3) | 4/20(20.0%) | 5/16(31.3%) | 0.470 | 5/16 (31.3%) | 15/33(45.5%) | 0.343 | 0.470 | 0.343 |
Maximum diameter (cm) | 1.89 ± 0.96 | 2.05 ± 0.96 | 0.612 | 1.55 ± 0.55 | 2.32 ± 0.85 | < 0.001** | 0.185 | 0.282 |
Tumor volume (cm3) | 1.62(0.61, 5.10) | 1.29(0.48, 7.50) | 0.961 | 1.08(0.51, 1.68) | 3.58(1.45, 6.07) | 0.003** | 0.191 | 0.233 |
Ki-67 index | | | | | | | | |
1% | 9/21(42.9%) | 4/14(28.6%) | 0.809 | 9/16(56.3%) | 7/32(21.9%) | 0.012* | 0.370 | 0.709 |
2% | 5/21(23.8%) | 5/14(35.7%) | 4/16(25.0%) | 8/32(25.0%) | | | |
3–4% | 6/21(28.6%) | 4/14(28.6%) | 1/16(6.3%) | 15/32(46.9%) | | | |
≥ 5% | 1/21 (4.8%) | 1/14(7.1%) | 2/16(12.5%) | 2/32(6.3%) | | |
Immediate postoperative GH (ng/mL) | 1.90(0.80, 6.10) | 1.25(0.73, 1.88) | 0.175 | 1.25(0.48, 3.13) | 1.95(1.13, 4.28) | 0.098 | 0.349 | 0.017 |
Immediate postoperative GH-nadir (ng/mL) | 1.31(0.28, 13.95) | 0.88(0.52, 1.10) | 0.356 | 0.81(0.37, 2.27) | 1.97(0.90, 3.02) | 0.149 | 0.598 | 0.053 |
Follow-up duration (month) | 4.8(3.4,11.5) | 9.5(4.8, 22.9) | 0.072 | 11.0(3.3, 24.3) | 7.0(3.5, 13.0) | 0.613 | 0.378 | 0.272 |
GH at last visit (ng/mL) | 0.55(0.30, 1.43) | 0.45(0.18, 1.13) | 0.381 | 1.60(0.58, 7.20) | 3.10(1.40, 5.00) | 0.677 | 0.110 | 0.001** |
GH-nadir at last visit (ng/mL) | 0.19(0.06, 1.01) | 0.21(0.15, 0.78) | 0.691 | 0.96(0.35, 9.18) | 1.51(0.53, 4.27) | 0.868 | 0.077 | 0.010* |
IGF-1 SDS at last visit | 2.49 ± 1.75 | 1.25(0.47, 2.17) | 0.060 | 2.23 ± 1.98 | 2.12 ± 2.09 | 0.883 | 0.870 | 0.178 |
Remission rate | 7/14 (50.0%) | 9/14 (64.3%) | 0.445 | 3/11(27.3%) | 9/21(42.9%) | 0.465 | 0.414 | 0.214 |
P3: comparison between males and females in mutation group, P4: comparison between males and females in non-mutation group, *p < 0.05, **p < 0.01 |
We further studied whether sexes could affect clinical features of acromegaly patients. Among our 97 patients, no significant difference was found between two genders on baseline hormone levels, adenoma volumes, or biochemical remissions, except the GH levels at patients’ last visits, as females have higher random GH [2.80(0.80, 5.70) vs 0.50(0.30, 1.28) ng/mL, p < 0.001] and GH-nadir [1.41(0.46, 5.01) vs 0.20(0.10, 0.75) ng/mL, p = 0.002] levels at last visits.
However, in patients having no somatic GNAS mutations, females had much higher baseline GH levels [17.30(10.70, 61.73) vs 10.55(5.25, 16.95) ng/mL, p = 0.018, more common abnormal glucose metabolism (47.2% vs 16.7%, p = 0.028) and dyslipidemia (56.7% vs 20.0%, p = 0.020). What’s more, their serum GH levels were also much higher than males when immediately after surgery [1.95(1.13, 4.28) vs 1.25(0.73, 1.88) ng/mL, p = 0.017] and at their last visits [3.10(1.40, 5.00) vs 0.45(0.18, 1.13) ng/mL, p = 0.001]. (Table 2)
In patients carrying somatic GNAS mutations, no gender difference on GH levels or metabolic complications was found.