Utilizing the search strategy described above, 123 records were found. 6 citations were manually scanned. After the removal of duplicate citations, a total of 128 studies were subsequently reviewed. Review of titles and abstracts excluded 66 citations based on their failure to meet inclusion criteria. 62 studies were deemed potentially suitable for inclusion and reviewed in full text. Of these records, 1 did not differentiate outcomes by cohorts, 1 reported different outcomes, 3 were editorials, 3 were guidelines, 5 included less than 10 patients, 14 were about either surgical or non-surgical cohorts of patients, and 17 were review articles. 18 articles were finally included for data extraction. The PRISMA flow diagram is shown in figure 1.
Qualitative and quantitative analysis
The 18 included studies were published between 1993 and 2022, and conducted in Brazil (n = 1) [13], Canada (n = 1) [11], France (n = 2) [14, 15], Israel (n =1) [16], Mexico (n = 1) [17], Poland (n = 1) [6], Portugal (n = 1) 2323[3], United Kingdom (n = 8) 2311,2411,24[2, 4, 5, 7, 8, 18–20], and United States of America (n = 2) [21, 22], with a total of 886 patients: 533 treated with surgery and 353 conservative. 8 records 11111111[2, 4, 11, 17–20, 22] reported outcomes stratifying surgical cohorts by early (n = 228) and late (n = 141) interventions as previously defined. All but one study [13] were retrospective. The mean age of presentation and follow-up were 51.9 and 3.9 years respectively. 13 studies [2, 4–8, 11, 14, 17, 18, 20–22] were consistent stating that patients were more likely to underwent surgery due to the severity of their symptoms, and 2 24[8, 18] considered pre-existing morbidities and anesthetic risks as factors in favor of conservative managements. All studies were assessed and graded as Level 3 according to the guidelines of the OCEBM. A summary of the main characteristics of the included records with patient demographics is presented in Table 1.
Table 1 Summary of the main characteristics of the included studies with patient demographics
Study, year of publication
|
Country
|
Design
|
OCEBM
|
n total
|
%F
|
Age, years
|
n surgery
|
n early surgery
|
n late surgery
|
n conservative
|
Follow-up time, years
|
Almeida, 2019 [11]
|
Canada
|
R
|
3
|
67
|
39
|
57*
|
49
|
22
|
27
|
18
|
2.8*
|
Ayuk, 2004 [5]
|
UK
|
R
|
3
|
33
|
39
|
52*
|
15
|
NR
|
NR
|
18
|
3.7*
|
Bonicki, 1993 [6]
|
Poland
|
R
|
3
|
39
|
NR
|
48*
|
23
|
NR
|
NR
|
16
|
3-10
|
Budohoski, 2022 [2]
|
UK
|
R
|
3
|
160
|
34
|
53*
|
96
|
61
|
35
|
64
|
4*
|
Bujawansa, 2014 [4]
|
UK
|
R
|
3
|
55
|
33
|
58*
|
33
|
23
|
10
|
22
|
7*
|
Cavalli, 2021 [19]
|
UK
|
R
|
3
|
30
|
27
|
54*
|
18
|
10
|
8
|
12
|
2.6*
|
Giritharan, 2016 [20]
|
UK
|
R
|
3
|
31
|
65
|
55*
|
20
|
11
|
9
|
11
|
4.9*
|
Gruber, 2006 3[8]
|
UK
|
R
|
3
|
30
|
23
|
53**
|
10
|
NR
|
NR
|
20
|
4.4**
|
Leyer, 2011 [15]
|
France
|
R
|
3
|
44
|
39
|
53*
|
19
|
NR
|
NR
|
25
|
1.7*
|
Lubina, 2005 [16]
|
Israel
|
R
|
3
|
40
|
32
|
51*
|
34
|
NR
|
NR
|
6
|
4.5*
|
Maccagnan, 1995 [13]
|
Brazil
|
P
|
3
|
12
|
42
|
43*
|
5
|
NR
|
NR
|
7
|
0.2*
|
Marx, 2021 [14]
|
France
|
R
|
3
|
46
|
37
|
47*
|
19
|
NR
|
NR
|
27
|
2.3*
|
Randeva, 1999 [18]
|
UK
|
R
|
3
|
35
|
40
|
50*
|
31
|
12
|
19
|
4
|
6.3*
|
Shepard, 2021 [21]
|
USA
|
R
|
3
|
64
|
55
|
53*
|
17
|
17
|
NR
|
47
|
4*
|
Sibal, 2004 [7]
|
UK
|
R
|
3
|
45
|
38
|
49*
|
27
|
NR
|
NR
|
18
|
4*
|
Singh, 2015 [22]
|
USA
|
R
|
3
|
87
|
34
|
51*
|
69
|
61
|
8
|
18
|
3.7*
|
Teixeira, 2018 [3]
|
Portugal
|
R
|
3
|
23
|
48
|
57*
|
14
|
NR
|
NR
|
9
|
0.5*
|
Vargas, 2014 [17]
|
Mexico
|
R
|
3
|
47
|
28
|
51*
|
36
|
11
|
25
|
11
|
8.15*
|
OCEBM, Oxford Centre for Evidence-Based Medicine grade. n, number of patients. *, mean. **, median. UK, United Kingdom. USA, United States of America. R, retrospective. P, prospective. NR, not reported.
|
15 studies [2–5, 7, 8, 11, 13–16, 18, 20–22] reported presenting symptoms after initial examination. From the group that underwent surgery (ie, early and late interventions), 276/444 (62%) patients had visual field deficit, 196/257 (76%) had hypopituitarism and 238/444 (54%) had ophthalmoplegia. Stratifying surgical cohorts, from the early surgery group 125/207 (60%) patients had visual field deficit, 86/111 (77%) had hypopituitarism and 122/207 (59%) had ophthalmoplegia; and from the late surgery group 41/81 (60%) patients had visual field deficit, 36/44 (82%) had hypopituitarism and 36/81 (44%) had ophthalmoplegia. From the conservative cohort, 82/305 (27%) patients had visual field deficit, 152/216 (70%) had hypopituitarism and 127/305 (42%) had ophthalmoplegia. A summary of the presenting symptoms is presented in Table 2.
Table 2 Patient presenting symptoms
Study, year of publication
|
Treatment
|
Visual field deficit, n (%)
|
Hypopituitarism, n (%)
|
Ophthalmoplegia, n (%)
|
Almeida, 2019 [11]
|
S
|
37 (75)
|
31 (64)
|
27 (55)
|
ES
|
17 (77)
|
13 (59)
|
14 (64)
|
LS
|
NR
|
NR
|
NR
|
C
|
7 (39)
|
9 (50)
|
5 (28)
|
Ayuk, 2004 [5]
|
S
|
7 (47)
|
15 (100)
|
8 (53)
|
C
|
6 (33)
|
18 (100)
|
7 (39)
|
Bonicki, 1993 [6]
|
S
|
NR
|
NR
|
NR
|
C
|
NR
|
NR
|
NR
|
Budohoski, 2022 [2]
|
S
|
64 (67)
|
82 (85)
|
46 (48)
|
ES
|
42 (69)
|
51 (84)
|
31 (51)
|
LS
|
22 (63)
|
31 (89)
|
15 (43)
|
C
|
21 (33)
|
50 (78)
|
22 (34)
|
Bujawansa, 2014 [4]
|
S
|
13 (39)
|
NR
|
18 (54)
|
ES
|
10 (43)
|
NR
|
13 (56)
|
LS
|
3 (30)
|
NR
|
5 (50)
|
C
|
5 (23)
|
NR
|
17 (77)
|
Cavalli, 2021 [19]
|
S
|
NR
|
NR
|
NR
|
C
|
NR
|
NR
|
NR
|
Giritharan, 2016 [20]
|
S
|
18 (90)
|
14 (70)
|
10 (50)
|
ES
|
10 (91)
|
9 (82)
|
7 (64)
|
LS
|
8 (89)
|
5 (55)
|
3 (33)
|
C
|
7 (64)
|
8 (73)
|
2 (18)
|
Gruber, 2006 [8]
|
S
|
6 (75)
|
NR
|
3 (30)
|
C
|
4 (25)
|
NR
|
12 (60)
|
Leyer, 2011 [15]
|
S
|
14 (74)
|
NR
|
10 (53)
|
C
|
5 (20)
|
NR
|
12 (48)
|
Lubina, 2005 [16]
|
S
|
21 (61)
|
NR
|
14 (40)
|
C
|
3 (50)
|
NR
|
2 (33)
|
Maccagnan, 1995 [13]
|
S
|
4/5 (80)
|
NR
|
2 (40)
|
C
|
2 (29)
|
NR
|
7 (100)
|
Marx, 2021 [14]
|
S
|
14 (74)
|
8 (42)
|
12 (63)
|
C
|
6 (22)
|
16 (59)
|
13 (48)
|
Randeva, 1999 [18]
|
S
|
25 (80.6)
|
NR
|
23 (74)
|
ES
|
12
|
NR
|
11
|
LS
|
13
|
NR
|
12
|
C
|
0 (0)
|
2 (50)
|
0 (0)
|
Shepard, 2021 [21]
|
ES
|
11 (65)
|
13 (76)
|
10 (59)
|
C
|
9 (19)
|
27 (47)
|
14 (30)
|
Sibal, 2004 [7]
|
S
|
16 (59)
|
21 (78)
|
14 (52)
|
C
|
4 (22)
|
13 (72)
|
8 (44)
|
Singh, 2015 [22]
|
S
|
26 (38)
|
NR
|
41 (59)
|
ES
|
23 (39)
|
NR
|
40 (66)
|
LS
|
3 (37)
|
NR
|
1(12)
|
C
|
3 (17)
|
NR
|
6 (33)
|
Teixeira, 2018 [3]
|
S
|
NR
|
12 (86)
|
NR
|
C
|
NR
|
9 (100)
|
NR
|
Vargas, 2014 [17]
|
S
|
NR
|
NR
|
NR
|
C
|
NR
|
NR
|
NR
|
S, surgery. C, conservative. ES, early surgery. LS, late surgery. NR, not reported or estimable.
|
The 18 included studies reported outcomes evaluated at the last follow-up. The recovery rates in visual field, endocrine function, and ophthalmoplegia, were, for overall surgery 76%, 23% and 81% respectively; early surgery 79%, 0% and 81% respectively; late surgery 73%, 0% and 70% respectively; and conservative management 79%, 23% and 90% respectively. A summary of the recoveries versus presenting symptoms statistics is presented in Table 3.
Table 3.Recoveries versus presenting symptoms statistics
Study, year of publication
|
Treatment
|
Visual field recovery (%)
|
Endocrine recovery (%)
|
Ophthalmoplegia recovery (%)
|
Almeida, 2019 [11]
|
S
|
32/37 (86)
|
4/31 (13)
|
26/27 (96)
|
C
|
4/7 (57)
|
4/9 (44)
|
4/5 (80)
|
Ayuk, 2004 [5]
|
S
|
4/7 (57)
|
2/15 (13)
|
5/8 (63)
|
C
|
6/6 (100)
|
3/18 (17)
|
7/7 (100)
|
Bonicki, 1993 [6]
|
S
|
NR
|
NR
|
NR
|
C
|
All
|
NR
|
NR
|
Budohoski, 2022 [2]
|
S
|
40/64 (62)
|
NR
|
21/46 (46)
|
ES
|
24/42 (57)
|
NR
|
14/31 (45)
|
LS
|
16/22 (73)
|
NR
|
7/15 (47)
|
C
|
14/21 (67)
|
NR
|
13/22 (59)
|
Bujawansa, 2014 [4]
|
S
|
3/13 (23)
|
NR
|
17/18 (94)
|
ES
|
8/10 (80)
|
NR
|
12/13 (92)
|
LS
|
2/3 (67)
|
NR
|
5/5 (100)
|
C
|
4/5 (80)
|
NR
|
17/17 (100)
|
Cavalli, 2021 [19]
|
S
|
NR
|
NR
|
NR
|
C
|
NR
|
NR
|
NR
|
Giritharan, 2016 [20]
|
S
|
18/18 (100)
|
0/14 (0)
|
NR
|
ES
|
10/10 (100)
|
0/9 (0)
|
NR
|
LS
|
8/8 (100)
|
0/5 (0)
|
NR
|
C
|
6/7 (86)
|
0/8 (0)
|
NR
|
Gruber, 2006 [8]
|
S
|
4/6 (67)
|
NR
|
3/3 (100)
|
C
|
2/4 (50)
|
NR
|
12/12 (100)
|
Leyer, 2011 [15]
|
S
|
10/14 (71)
|
NR
|
8/10 (80)
|
C
|
5/5 (100)
|
NR
|
12/12 (100)
|
Lubina, 2005 [16]
|
S
|
17/21 (81)
|
NR
|
10/14 (71)
|
C
|
2/3 (67)
|
NR
|
1/2 (50)
|
Maccagnan, 1995 [13]
|
S
|
2/4 (50)
|
NR
|
2/2 (100)
|
C
|
2/2 (100)
|
NR
|
7/7 (100)
|
Marx, 2021 [14]
|
S
|
11/14 (79)
|
0/8 (0)
|
11/12 (92)
|
C
|
5/6 (83)
|
4/16 (25)
|
10/13 (77)
|
Randeva, 1999 [18]
|
S
|
19/25 (76)
|
NR
|
21/23 (91)
|
ES
|
12/12 (100)
|
NR
|
11/11 (100)
|
LS
|
7/13 (54)
|
NR
|
10/12 (84)
|
C
|
0 (0)
|
1/2 (50)
|
0 (0)
|
Shepard, 2021 [21]
|
ES
|
10/11 (91)
|
NR
|
6/10 (60)
|
C
|
8/9 (89)
|
NR
|
14/14 (100)
|
Sibal, 2004 [7]
|
S
|
15/16 (94)
|
10/21 (48)
|
13/14 (93)
|
C
|
4/4 (100)
|
4/13 (31)
|
8/8 (100)
|
Singh, 2015 [22]
|
S
|
26/26 (100)
|
NR
|
41/41 (100)
|
ES
|
23/23 (100)
|
NR
|
40/40 (100)
|
LS
|
3/3 (100)
|
NR
|
1/1 (100)
|
C
|
3/3 (100)
|
NR
|
6/6 (100)
|
Teixeira, 2018 [3]
|
S
|
NR
|
7/12 (58)
|
NR
|
C
|
NR
|
1/9 (11)
|
NR
|
Vargas, 2014 [17]
|
S
|
32
|
18
|
9
|
ES
|
8
|
7
|
3
|
LS
|
24
|
11
|
6
|
C
|
3
|
6
|
2
|
S, surgery. C, conservative. ES, early surgery. LS, late surgery. NR, not reported or estimable.
|
12 studies [2, 5–7, 11, 15, 17–22] reported tumor growth/recurrence statistics, with 49/420 (12%) and 31/262 (12%) of tumoral growth/recurrences in the overall surgical and conservative management cohorts, respectively. A summary is presented in Table 4.
Table 4 Tumor growth/recurrence
Study, year of publication
|
Treatment
|
Tumor growth/recurrence, n (%)
|
Almeida, 2019 [11]
|
S
|
4 (8)
|
C
|
2 (11)
|
Ayuk, 2004 [5]
|
S
|
1(7)
|
C
|
1(5)
|
Bonicki, 1993 [6]
|
S
|
0 (0)
|
C
|
5 (31)
|
Budohoski, 2022 [2]
|
S
|
19 (20)
|
C
|
10 (16)
|
Bujawansa, 2014 [4]
|
S
|
NR
|
C
|
NR
|
Cavalli, 2021 [19]
|
S
|
1(5)
|
C
|
1 (8)
|
Giritharan, 2016 [20]
|
S
|
2 (20)
|
C
|
0 (0)
|
Gruber, 2006 [8]
|
S
|
NR
|
C
|
NR
|
Leyer, 2011 [15]
|
S
|
0 (0)
|
C
|
4 (16)
|
Lubina, 2005 [16]
|
S
|
NR
|
C
|
NR
|
Maccagnan, 1995 [13]
|
S
|
NR
|
C
|
NR
|
Marx, 2021 [14]
|
S
|
NR
|
C
|
NR
|
Randeva, 1999 [18]
|
S
|
1 (3)
|
C
|
1 (25)
|
Shepard, 2021 [21]
|
S
|
1 (6)
|
C
|
1 (2)
|
Sibal, 2004 [7]
|
S
|
1 (4)
|
C
|
4 (22)
|
Singh, 2015 [22]
|
S
|
7 (10)
|
C
|
0 (0)
|
Teixeira, 2018 [3]
|
S
|
NR
|
C
|
NR
|
Vargas, 2014 [17]
|
S
|
12 (33)
|
C
|
2 (19)
|
S, surgery. C, conservative. NR, not reported or computable.
|
Comparing the effects of overall surgical and conservative managements, the OR for visual field, hypopituitarism, and ophthalmoplegia recoveries were 0.88 (CI 95% 0.32, 2.42 p = 0.81), 0.96 (CI 95% 0.33, 2.76 p = 0.94) and 1.8 (CI 95% 0.67, 4.87 p = 0.25) respectively. The OR for tumor growth/recurrence was 0.87 (CI 95% 0.3, 2.53 p = 0.8) (figure 2).
Comparing the effects of early surgical and conservative managements, the OR for visual field and ophthalmoplegia recoveries were 1.2 (CI 95% 0.46, 3.18 p = 0.71), and 2.78 (CI 95% 1.09, 7.09 p = 0.03) respectively (figure 3).
Comparing the effects of late surgical and conservative managements, the OR for visual field recovery was 0.75 (CI 95% 0.24, 2.34 p = 0.62) (figure 4).
When comparisons were performed between the effects of early and late surgical managements, the OR for visual field and ophthalmoplegia improvements were 0.84 (CI 95% 0.36, 1.95 p = 0.69) and 0.82 (CI 95% 0.29, 2.34 p = 0.72) respectively (figure 5).