Pituitary DOI 10.1007/s11102-014-0603-8
Effective time window in reducing pituitary adenoma size by gamma knife radiosurgery Henry Ka-Fung Mak • Shui-Wun Lai • Wenshu Qian • Stanley Xu • Elizabeth Tong • May Lee Vance • Edward Oldfield • John Jane Jr. Jason Sheehan • Kelvin K. W. Yau • Max Wintermark
•
Ó Springer Science+Business Media New York 2014
Abstract Objectives Although the effectiveness of gamma knife radiosurgery (GKRS) in controlling the size of pituitary adenomas has been well demonstrated in many studies, the time period in which significant changes in tumor size occurs has been investigated in a limited fashion. It is important to determine the therapeutic window of GKRS in treating pituitary adenomas, i.e., the effective timeframe during which significant size reduction of these tumors occurs, so that alternative treatments such as further GKRS or microsurgery might be prescribed in a timely manner if clinically indicated. Methods This was a nested sample of an ongoing local cohort study on GKRS for pituitary adenomas at the University of Virginia. Magnetic resonance imaging (MRI) using dedicated sequences was employed. Only patients with a baseline MRI (TP0) and at least 1 follow-up study
H. K.-F. Mak S.-W. Lai W. Qian Department of Diagnostic Radiology, University of Hong Kong, Hong Kong Special Administrative Region, China S. Xu E. Oldfield J. Jane Jr. J. Sheehan Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA E. Tong M. Wintermark (&) Department of Radiology, Neuroradiology Division, University of Virginia, Box 800170, Charlottesville, VA 22908, USA e-mail: [email protected] M. L. Vance Department of Medicine, University of Virginia, Charlottesville, VA, USA K. K. W. Yau Department of Management Sciences, City University of Hong Kong, Hong Kong Special Administrative Region, China
performed in the University Hospital after GKRS were included. The follow-up scans were performed at five timepoints (TP1–TP5) which were 6, 12, 24, 36 and 48 months after GKRS. The dimensional indices of the tumors were measured in three orthogonal planes, i.e., transverse (TR), antero-posterior (AP) and cranio-caudal (CC). The volumes of the tumors were estimated by using the following formula: V ¼ p ðTR AP CCÞ=6. Tumor volume decrease by more than 25 % from baseline was considered as ‘shrinkage’, \25 % tumor size increase or decrease was considered ‘static’, and more than 25 % increase as ‘increment’. Our cohort consisted of 21 patients, with functioning adenomas in 13 subjects i.e. six adrenocorticotrophic hormone (ACTH)-secreting and seven growth hormone (GH)-secreting, and non-functioning (NF) adenomas in eight subjects. Results In 26 adenomas (8 ACTH, 9 GH and 9 NF), tumor control (tumor shrinkage or static) were achieved in 21 tumors (80.8 %); 89, 75, and 78 % for GH-secreting, ACTH-secreting and NF adenomas respectively, at the end of the 4-year follow-up period. Analysis of variance showed significant differences of GKRS margin dose among different types of tumors (p = 0.013), but not of baseline tumor volumes (p = 0.240). Logistic regression analysis showed no significant association of margin dose, baseline volume or tumor type with the tumor control outcome. Comparison of tumor change using dimensional indices relative to the base time point (TP0) showed that in the sample there was an average reduction of 1.290 mm at TP1 (6 months) with p values 0.155 (parametric t test) and 0.098 (non-parametric Wilcoxon signed-ranked test) respectively, showing a moderate reduction in tumor dimensional indices. The change in dimensional indices at later time points (TP2–TP5) showed an average reduction ranging from 1.930 to 2.471 mm. Significant reduction in
123
Pituitary
the mean dimensional indices was firstly observed at TP2 (1 year) with p values 0.013 (t test) and 0.018 (Wilcoxon signed-rank test). Such scale of reduction in the dimensional indices appeared to be maintained along the time axis (from TP2 to TP5). Conclusions Significant decrease in tumor dimensional indices tended to occur at 1 year post-GKRS. Although to a lesser extent, such decrease in dimensional indices continued up to the end of our follow-up period. Keywords Pituitary adenoma Gamma knife radiosurgery Therapeutic time window Magnetic resonance imaging Tumor control Abbreviations GKRS Gamma-knife radiosurgery NPA Non-functioning pituitary adenomas TR Transverse AP Antero-posterior CC Cranio-caudal TP Time point TP0 Baseline TP1 6 months TP2 12 months TP3 24 months TP4 36 months TP5 48 months
Introduction Pituitary adenomas are benign neoplastic lesions of the anterior pituitary gland. Epidemiologic studies reveal that pituitary adenomas are the second most common primary intracranial neoplasm and accounts for about 10–15 % of all intracranial tumor cases [1, 2]. There are currently four treatment modalities that are used to control the tumor size and hormone normalization: medication, surgery, radiation therapy and stereotactic radiosurgery. Transsphenoidal microsurgery is currently considered first-line therapy for most symptomatic pituitary adenomas with the exception of prolactinomas. Surgery provides rapid reduction of tumor size as well as resolution of neurologic symptoms caused by mass effect [3]. However, complete resection of tumors, especially those with cavernous sinus invasion, is not always possible and the recurrence rate after transsphenoidal surgery can range from 20 to 50 % [4–6]. As a result, adjuvant therapies are frequently prescribed in combination with transsphenoidal microsurgery to achieve better tumor control and hormone normalization.
123
Previous studies had evaluated the advantages and disadvantages of pharmacotherapy [7, 8] and radiation therapy [9–12] as adjuvant therapies. Stereotactic radiosurgery using gamma-knife radiosurgery (GKRS) has emerged as a well-accepted second-line modality. It has been demonstrated in many studies that GKRS is efficacious in controlling tumor size in both secretory and non-secretory pituitary adenomas, with a tumor control rate of above 90 % in many studies [4, 13– 18]. However, the effect of GKRS on hormone normalization is variable, with different series reporting a wide range of endocrine remission rates from 17 to 100 % depending on tumor size and type [19–22]. The effectiveness of GKRS in controlling the size of pituitary adenomas has been well demonstrated in many studies. However, the time period during which significant change in tumor size occurs has been investigated in a limited fashion. According to a previous study, the incidence of tumor shrinkage increased from 28 % at 1 year after GKRS to 79 % at 4 years after GKRS [23]. Similar rates have been reported by Kong et al., in which the incidence of an objective response (complete and partial tumor shrinkage) increased from 39.5 % at 2 years after GKRS to 81.8 % at 4 years after GKRS [20]. Shrinkage was achieved in 46.9 % of tumors after 1–2 years, and in 61.1 % of patients 2–5 years after irradiation in another study [24]. Although most studies suggested that GKRS has a short latency period [4, 7, 11, 25], few of these studies investigated the magnitude of change in tumor size at various time points following GKRS [24, 26] or documented the interval at which maximal tumor shrinkage occurred. It is important to determine the therapeutic window of GKRS in treating pituitary adenomas, i.e., the effective timeframe during which significant size reduction of these tumors occurs, so that alternative treatments such as further GKRS or microsurgery [27] can be prescribed in a timely manner if clinically indicated. This pilot study aims, (1) to track the changes in the size of pituitary tumors using dedicated magnetic resonance imaging (MRI) sequences at various time points before and after GKRS, and (2) to identify the precise interval during which significant tumor shrinkage occurs. As demonstrated by several studies, the percentage of patients in which tumor shrinkage was observed increased drastically between the 1st and 5th year [20, 23, 24, 26, 28]; we therefore hypothesize that significant size change would occur at some time between these two time points.
Materials and methods This was a nested sample (between September 2004 and April 2007) of an ongoing local cohort study on GKRS for
Pituitary
pituitary adenomas at the University of Virginia. Details on patient population, pre- and post-radiosurgical follow-up, and stereotactic radiosurgery technique were published previously [6, 27]. The study was approved by the local Institutional Review Board of Human Study Research of the University. Inclusion Criteria: Any patient treated during this period was recruited if the person had MRI examination performed in the University hospital, including baseline MRI (TP0) and at least 1 follow-up scan after GKRS. The follow-up scans were performed at 5 different time-points (TP1 to TP5): 6, 12, 24, 36 and 48 months after GKRS. The study cohort consisted of 24 patients, but three subjects showed no identifiable pituitary tumor and were excluded. The final cohort of 21 patients consisted of 8 males and 13 females, with mean age of 49.6 years (standard deviation, SD 15.3 years, range 16–79 years). Functioning adenomas were found to be present in 13 subjects (six ACTH-secreting and seven GH-secreting), and nonfunctioning (NF) adenomas in eight subjects based on clinical, biochemical, radiologic and pathologic criteria. The demographic, clinical profiles, pathological findings, and GKRS margin dose of the patients are summarized in Table 1. The mean and median duration of follow-up was 31.8 and 36 months, respectively. In our study, the median number of follow-up MRI examinations underwent by our patients was 3, which compared favorably to a prior study with a median followup of 2 [29]. Follow-up MRI examinations were performed for 12, 14, 11, 11, and 8 patients at TP1, TP2, TP3, TP4 and TP5 respectively. Standardized pituitary MR (Siemens Magnetom, 1.5 Tesla or 3 Tesla) imaging sequences were used for all baseline and follow-up studies, including T1-weighted images in sagittal, coronal and axial planes, T2-weighted images in coronal and axial planes, and post-gadolinium T1-weighted images in sagittal, coronal and axial planes. All images had slice thickness of 3 mm and no gap in between. A neuroradiologist (HKFM), blinded to the clinical and pathological findings, evaluated the serial MRI imaging studies of each patient and recorded the results on a standard form. A senior neuroradiologist (MW) was being consulted in problematic cases. The dimensional indices of the tumors were measured and recorded in three orthogonal planes, i.e., transverse (TR), antero-posterior (AP) and cranio-caudal (CC). The volumes of the tumors were estimated by using the following formula: V ¼ pðTR AP CCÞ=6 [23]. Tumor volume decrease by more than 25 % from baseline was considered as ‘shrinkage’, \25 % increase or decrease of tumor volume as ‘static’, and more than 25 % increase as ‘increment’.
Primarily, T2-weighted coronal images were used to evaluate the TR and CC dimensional indices (because of clarity of margins), while T1-weighted sagittal images for the AP dimensional index. In addition, post gadolinium T1 weighted images were used to augment the above sequences and to check for any discrepancy. Statistical analysis The margin dose of GKRS and baseline tumor volumes were compared between different tumor types using analysis of variance (ANOVA). Logistic regression analysis was performed, using the tumor status (controlled or not) at the last time point (TPL) as dependent variable, with GKRS dose and baseline volume as covariates and tumor type (NFA, ACTH and GH) as factor. To assess the dimensional indices and volumes reduction of the identifiable tumors of all subjects at different time points (TPi, i = 1, 2, 3, 4, 5) relative to the base time point (TP0), a subset of pairwise comparisons was conducted using both parametric (t test) and non-parametric (Wilcoxon signed-ranked test) methods. Bonferroni correction for multiple pairwise comparisons was taken to adjust the p values. All statistical analyses were conducted using IBM SPSS (Version 19.0.0). Level of significance was set at 0.05.
Results Descriptive statistics Based on the volumes of 26 tumors in our final cohort of 21 patients (three subjects had two separate identifiable tumor masses and one had three separate lesions), mean tumor volume decreased from a baseline value of 3.72 ? 4.52 to 2.53 ± 3.40 cm3 (Table 1). A mean tumor volume decrease of 20.1 % (SD 44.5 %) was observed in the entire cohort of 26 tumors. In the 14 tumors with shrinkage, the volume decreased by 52.4 % (SD 19.0 %), and in the 5 tumors with increment, the volume increased by 50.9 % (SD 20.2 %). Tumor control (tumor shrinkage or static) was observed in 16 (76.2 %) patients and 21 tumors (80.8 %) at the end of the 4-year follow-up period. Figure 1 shows the tumor size changes according to different tumor types after GKRS at last time point. The mean (and standard deviation) of margin dose and baseline tumor volume of different tumor types are shown in Table 2. The number and percentage of tumors (according to different tumor types) with volume shrinkage, and tumor
123
Pituitary Table 1 The demographic, clinical profiles, pathological findings and margin dose of all patients in our cohort. GK- Gamma Knife; TPL- last time point Patient
Tumor type
GK margin dose (Gy)
Volume (cm3)
Age/ gender
Clinical presentation/Tx before GK
1
55/F
Cushing’s disease/2 transphenoid surgery
ACTH-staining adenoma
23.00
16.025
5.039
2
39/F
Cushing’s disease/transphenoid surgery
ACTH-staining adenoma
25.00
0.047
0.079
3
46/F
Elevated growth hormone/transphenoid surgery
No mention
25.00
0.228
0.238
4
41/M
Acromegaly/transphenoid surgery
No mention
25.00
10.329
0.301
5
70/M
Acromegaly/transphenoid surgery
No mention
25.00
0.515
0.205
Mass 2: 0.290
0.076
Baseline
TPL
6
76/M
Acromegaly/placed on Sandostatin LAR
No mention
20.00
1.685
2.292
7
38/F
Acromegaly/placed on Sandostatin LAR, transsphenoid surgery
No mention
25.00
0.101
0.048
8
16/M
Bitemporal hemianopsia/on Levoxyl & prednisone prior to surgery, 2 prior surgical resection
Nonfunctioning
16.50
6.713
8.481
9
59/F
Visual loss/2 transsphenoidal surgery
Nonfunctioning
17.00
1.930
2.129
Mass 2: 0.036
0.030
10
49/F
Progressive vision loss with right temporal field cut/transsphenoidal surgery
Nonfunctioning
16.00
9.839
14.797
11
37/F
Peripheral vision loss and headaches/ craniotomy and resection
Nonfunctioning
25.00
1.619
1.085
12
43/F
Acromegaly/transphenoid surgery
Functioning-GH
25.00
1.349
1.662
Mass 2: 0.505
0.310
13
32/M
Acromegaly/no reduction with Sandostatin, 2 transphenoid surgery
Functioning-GH
20.00
0.218
0.105
14
39/F
Hypothyroidism and Cushing’s, amenorrhea/transphenoid surgery
Functioning-ACTH
25.00
0.105
0.057
15
51/F
Cushing’s/resection
Functioning-ACTH
25.00
2.862
4.956
16
79/F
Sudden headache, found pituitary apoplexy/Transphenoid resection
Silent ACTH staining adenoma
18.00
5.868
2.121
17
39/F
2 transphenoid surgery
Nonfunctioning
15.00
1.473
1.050
18
58/F
Cushing’ syndrome/bilateral adrenalectomy, 5 transphenoid surgery, 1 prior GK
Functioning-ACTH
8.50
2.362
1.644
Mass 2: 2.751 Mass 3: 4.801
2.131 2.384
19
63/M
Non-functioning pituitary adenoma/ transphenoid surgery
Nonfunctioning
15.00
8.190
6.668
20
62/M
Nonfunctioning
16.00
2.817
2.224
21
49/M
Non-secretory adenoma/2 transphenoid surgery Non-secretory adenoma/2 transphenoid surgery
Nonfunctioning
15.00
14.072
5.708
control (tumor shrinkage or static) at different time points are shown (Table 3). Percentage variations between the time points within each tumor type were likely due to different sample sizes under consideration.
123
Initial increases (\15 % compared to baseline study) in tumor volumes (within 6 months to 1 year) followed by subsequent tumor shrinkage occurred in 3 out of the 26 tumors (11.5 %), likely due to intratumoral edema [26, 29].
Pituitary
Multiple logistic regression with the tumor status (controlled or not) at the last time point as dependent variable, and using dose, baseline volume and tumor types as covariate/factor No significant covariate/factor was found to be associated with the tumor status outcome, probably because of small sample size. Paired comparison in tumor dimensional indices between different time points (Table 4) Fig. 1 Mean tumor volume changes (in percentage) compared with baseline volumes according to different types of tumors after GKRS at the end of the 48-month follow-up period. Shrinkage (decrease by 25 %), static (\25 % increment and \25 % decrease in volume) and increment (increase by 25 %); Error bars represent ±1 SD Table 2 Mean [and standard deviation (SD)] margin dose of GKRS and baseline volumes NF (n = 9)
ACTH (n = 8)
GH (n = 9)
16.94
17.69
23.89
3.13
7.95
2.20
Dose (Gy) Mean SD
Baseline volume (cm3) Mean
5.19
4.35
1.69
SD
4.76
5.13
3.28
Comparison of GKRS margin dose of and baseline tumor volumes between different tumor types using analysis of variance (ANOVA) ANOVA showed significant differences of dose among different types of tumors (p = 0.013), but not of baseline tumor volumes (p = 0.240). With Tukey’s post hoc test, differences of dose between NF versus ACTH and NF versus GH were observed (p = 0.017 and 0.042 respectively).
Comparison of tumor change using dimensional indices, i.e. TR, AP, and CC relative to the base time point (TP0) showed that in the sample there was an average reduction of 1.290 mm at TP1 (6 months) with p values 0.155 (t test) and 0.098 (Wilcoxon signed-ranked test) respectively, showing a moderate reduction in tumor dimensional indices. The change in dimensional indices at later time points (TP2 to TP5) showed an average reduction ranging from 1.930 to 2.471 mm. Significant reduction in the mean dimensional indices was firstly observed at TP2 (1 year) with p values 0.013 (t test) and 0.018 (Wilcoxon signedrank test). Such scale of reduction in the dimensional indices appeared to be maintained along the time axis (from TP2 to TP5), although we observed only a marginal significance at TP3 which was deemed partly due to the conservativeness of the Bonferroni correction in multiple pairwise comparisons. Using 21 tumors with size control The results were similar to the whole group except that significant difference between TP0 and TP1 was also seen i.e. average reduction of 2.007 mm at TP1 (6 months) with p values 0.014 (t test) and 0.027 (Wilcoxon signed-ranked test) respectively.
Table 3 Number (percentage) of tumors of different tumor types after GKRS at different time points with either volume decrease or size control TP1
TP2
TP3
TP4
TP5
Number of tumors with volume decrease (%) NF
2 (33.33)
3 (42.86)
2 (28.57)
4 (50.00)
3 (33.33)
ACTH GH
4 (66.67) 1 (33.33)
4 (57.14) 3 (33.33)
4 (57.14) 6 (66.67)
5 (62.50) 6 (66.67)
5 (62.50) 6 (66.67)
Number of tumors with size control (%) NF
5 (83.33)
6 (85.71)
6 (85.71)
7 (87.50)
7 (77.78)
ACTH
5 (83.33)
6 (85.71)
6 (85.71)
6 (75.00)
6 (75.00)
GH
2 (66.67)
8 (88.89)
7 (77.78)
7 (77.78)
8 (88.89)
123
Pituitary
0.001* 0.007* 1.384
0.004* 0.024*
\0.001* \0.001*
1.560
0.737
0.050* 0.025*
0.002* \0.001*
1.204
1.012
t test
Standard Error (mm)
4.918 16 0.003* 0.009* 0.926 2.936 28 0.023*
4.565
3.806
18
22
0.034* 0.056
\0.001* \0.001* 0.592
0.912 2.176
2.601 33
36 0.063
0.005*
3.501
3.737 30
18 0.027*
0.005* 0.008*
0.014* 0.680
0.768 2.390
2.007 36 0.098
44
Wilcoxon signed-rank test t test
N Wilcoxon signed-rank test
0.018*
Mean difference (mm) Mean difference (mm)
Standard Error (mm)
p value
#
21 tumors with size control
123
* Significant at 5 % level One-sided test with Bonferroni correction #
N number of entries of dimensional indices
0.020* 0.875 31 TP5
2.471
0.075
0.015* 0.623
0.855
39 TP4
1.804
39 TP3
1.930
0.013*
t test
0.155 0.673
0.695 2.024 50 TP2
1.290 45 TP1
N
Mean difference (mm)
Standard Error (mm)
p value
#
All tumors
Table 4 Paired comparisons of tumor dimensional indices relative to the baseline time point (TP0)
Comparison of the dimensional indices relative to the base time point (TP0) showed remarkable reduction as early as at TP1 (6 months). The average reduction in dimensional indices ranged from 3.501 mm (TP1) to 4.918 mm (TP5). Comparison of tumor change using pituitary tumor volumes relative to the base time point (TP0) did not show any appreciable significant difference (results not shown).
Discussion
N
14 tumors with shrinkage
p value#
Wilcoxon signed-rank test
Using 14 tumors with shrinkage only
The study demonstrates a tumor control rate of 80.8 % was found at the end of a 4-year follow-up in our cohort of 21 patients. The effectiveness of GKRS on the control of the size of pituitary tumors has been well documented in different series with most having a tumor control rate of [90 % [4, 13–18, 23, 30]. Subgroup analysis according to tumor types revealed that rates of control of tumor size were 89, 75 and 78 % for growth hormone-secreting, adrenocorticotrophic hormone-secreting and non-functioning adenomas, respectively (Table 3). The slightly lower tumor control rates in this series is likely related to sampling bias, since in larger cohorts treated in the same centre, tumor control rates of 89, 92 and 93 %, respectively, were previously reported [18]. A previous study [5] reported at least a 20 % reduction of tumor volume in 42.3 % of their patients with nonfunctioning pituitary adenomas. Attanasio et al. [23] reported that tumor reduction of [25 % of basal volume occurred in 58 % of their cohort of acromegalic patients 24 months after GKRS, while Kong et al. [20], found an objective response rate (complete being [25 % and partial being [10 % tumor shrinkage) of 39.5 % at 2 years and 81.8 % at 4 years in a large group of patients (including both functional and non-functional adenomas). At the end of a 48-month follow-up period in our cohort, tumor shrinkage (volume decrease [25 %) was achieved in 52.4, 33.3, 62.5 and 66.7 % of all, NF, ACTH, and GH adenomas respectively (Table 3); which was comparable to these prior studies. Contrary to our previous study [6], there was no significant association between tumour control outcome and margin dose. This was probably due to the small sample size. In view of no observed effect of margin dose, baseline tumor volume or tumor type on tumor control outcome in our cohort, tumors of different types were grouped together in current study to increase the statistical power for the analysis of effective timeframe of GKRS. In this study dimensional indices were assessed as an indicator of tumor size change at different time points. These measurements are commonly used in clinical practice
Pituitary
for the evaluation of the efficacy of GKRS [20, 24, 31]. The paired time-point comparisons used in the current study are more accurate than cumulative rate in detection of tumor size change because such a comparison was made within the same subject in each pair of time points and thus variations due to subject differences can be eliminated. In our study, whole group (all subjects) analyses revealed that dimensional reductions of the tumors occur between the baseline and all subsequent time points. Pamir et al. [26] reported in their study that earliest volume decrease after GKRS occurred 6–9 months after the procedure, albeit with higher marginal doses ([27 Gy). Significant reduction in tumor dimensional indices tended to occur 1 year after GKRS in our study. Subsequent reduction in later time points appeared to be not that apparent. In the subgroup analyses of the 21 tumors with size control and 14 tumors with shrinkage [25 %, the results were even more convincing. There are significant differences between baseline and all the subsequent time points, as early as 6 months after GKRS. A further reduction trend was observed in later time points, although in a less drastic manner. Similarly, a review of 82 patients with nonsecretory tumors by Jagannathan et al. [28] revealed that the median time to tumor shrinkage on MR imaging was 9 months (range 6–48 months). Interestingly, no significant difference in tumor volumes across various TPs was found, as compared to dimensional indices. The finding was not surprisingly because the estimated volumes were derived from the product of 3 dimensional index random variables. The observed significant reduction on dimensional indices but along with an insignificant reduction on the estimated ellipsoid volume suggested a non-uniform reduction in the 3 dimensions. Hypothetically, some of the tumor shrinkage only occurred in any 1 or 2 dimensions, thence the volume would not have the same degree of reduction effect size as the individual dimension did, which would give rise to such a phenomenon. Nevertheless, we did not have sufficient data to verify. This may warrant further study in future. Nevertheless, previous studies found that endocrine remission in ‘radioresponsive’ secretory adenomas occurs usually within the first 2 years [4]. However, no direct correlation between tumor volume response and endocrine response following GKRS has been reported so far [4, 7, 32]. Small basal tumor volume was a determinant of volume change [26] after GKRS, and was associated with endocrine remission [6, 13, 25], though not found in one study [24]. In current study, the confounding effect of the antisecretory medication in patients with acromegaly after GKRS was also evaluated. 4 patients were on continual antisecretory medication (Pegvisomant) after GKRS; 2 showed tumor shrinkage, 1 remained static and 1 showed
increment at latest follow-up scan. For the other three patients with medication being stopped after IGF-1 normalization (2 on Sandostatin and 1 on Pegvisomant)), 2 showed tumor shrinkage and 1 remained static at latest follow-up scan. Pegvisomant lowers IGF-1 level but has no proven anti-proliferative effect [33], therefore, co-administration of this drug should not affect the imaging findings of these 5 patients. For the two remaining patients on Sandostatin, since the medication was stopped between 9 months to 1 year, the drug effect should be minimal. In patients with secreting tumors, it would be tempting to have a comparison between anti-tumor and anti-secretory efficacy of GKRS. All patients with ACTH-adenomas had normal laboratory tests except patient 18. All the acromegalic patients (except patient 12) eventually had IGF-1 level normalized (with or without medications after GKRS), 4 with tumor shrinkage, 1 static and 1 with increment. The antitumor and antisecretory efficacies of GKRS could not be compared in those patients with concurrent use of anti-secretory medication. Patient 18 had 3 tumors under control after GKRS (2 with shrinkage and 1 static) but persistently elevated ACTH level (latest— 446 pg/ml). Patient 12 had persistently elevated IGF-1 level, but static tumor size. There is effective tumor control but failure of anti-secretory efficacy of GKRS in these two subjects. Five large tumors with tumor size [8 cm3 have been treated with GKRS in this study. It is generally accepted that GKRS compared with conventional fractionated radiation therapy, provides a decreased time interval between the treatment and the achievement of remission in functioning pituitary adenomas. Furthermore, given the sharp radiation dose fall-off of the GKRS, it is safe and effective to treat large tumors as described above as long as the adenoma is not too close to the optic apparatus. Furthermore, for the five subjects with tumor size increment, no pituitary carcinoma was detected. 2 patients (1 ACTH staining adenoma and 1 non-functioning adenoma) were uneventful and had regular follow-up; 1 patient had non-functioning adenoma but was lost in follow-up; 1 patient had functioning, invasive ACTH adenoma requiring 1 surgical resection and 1 CyberKnife radiosurgery after GKRS, and 1 acromegalic patient with invasive tumor to cavernous sinus was placed on antisecretory medication. Limitations This was a retrospective study to examine the time frame at which pituitary tumor size reduction occurred after the administration of GKRS. Our study is limited by a relatively small sample size and pooled data of different tumor types. Hence, it is exploratory in nature. Prospective
123
Pituitary
studies with a larger sample size and longer follow-up period are needed to confirm our results. Further validation of our results by measuring actual tumor volumes obtained from volume scans is recommended [27, 29, 34]. The time point at which failure of GKRS occurs i.e. pituitary tumors begin to grow, should also be investigated in a larger-scale study. In addition, comprehensive head-to-head comparison with published studies in the literature was difficult due to heterogeneous methodology and study cohorts [23, 25] in analyzing treatment outcome [10].
Conclusion Consistent with the findings in prior studies, GKRS is efficacious in tumor size control. Effective tumor size reduction tended to occur 1 year after the administration of GKRS and further tumor reduction continued up to the end of our follow-up period (4 years post GKRS) especially for those with size control/with shrinkage, although to a lesser extent. If GKRS fails to achieve any significant change during such a period of time, alternative treatments such as further GKRS or microsurgery might be prescribed in a timely manner if clinically indicated. Conflict of interest The authors reported no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Ethical standard We declared that the experiments comply with the current laws of our country.
References 1. Laws ER Jr, Vance ML (1999) Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am 10:327–336 2. Radhakrishnan K, Mokri B, Parisi JE, O’Fallon WM, Sunku J, Kurland LT (1995) The trends in incidence of primary brain tumors in the population of Rochester, Minnesota. Ann Neurol 37:67–73 3. Mortini P, Losa M, Barzaghi R, Boari N, Giovanelli M (2005) Results of transsphenoidal surgery in a large series of patients with pituitary adenoma. Neurosurgery 56:1222–1233; discussion 1233 4. Laws ER, Sheehan JP, Sheehan JM, Jagnathan J, Jane JA Jr, Oskouian R (2004) Stereotactic radiosurgery for pituitary adenomas: a review of the literature. J Neurooncol 69:257–272 5. Losa M, Valle M, Mortini P, Franzin A, da Passano CF, Cenzato M, Bianchi S, Picozzi P, Giovanelli M (2004) Gamma knife surgery for treatment of residual nonfunctioning pituitary adenomas after surgical debulking. J Neurosurg 100:438–444 6. Sheehan JP, Pouratian N, Steiner L, Laws ER, Vance ML (2011) Gamma knife surgery for pituitary adenomas: factors related to radiological and endocrine outcomes. J Neurosurg 114:303–309 7. Sheehan JP, Niranjan A, Sheehan JM, Jane JA Jr, Laws ER, Kondziolka D, Flickinger J, Landolt AM, Loeffler JS, Lunsford LD (2005) Stereotactic radiosurgery for pituitary adenomas: an intermediate review of its safety, efficacy, and role in the neurosurgical treatment armamentarium. J Neurosurg 102:678–691
123
8. Marko NF, LaSota E, Hamrahian AH, Weil RJ (2012) Comparative effectiveness review of treatment options for pituitary microadenomas in acromegaly. J Neurosurg 117:522–538 9. Barkan AL, Halasz I, Dornfeld KJ, Jaffe CA, Friberg RD, Chandler WF, Sandler HM (1997) Pituitary irradiation is ineffective in normalizing plasma insulin-like growth factor I in patients with acromegaly. J Clin Endocrinol Metab 82:3187–3191 10. Brada M, Ajithkumar TV, Minniti G (2004) Radiosurgery for pituitary adenomas. Clin Endocrinol (Oxf) 61:531–543 11. Prasad D (2006) Clinical results of conformal radiotherapy and radiosurgery for pituitary adenoma. Neurosurgery Clin N Am 17:129–141, vi 12. Rim CH, Yang DS, Park YJ, Yoon WS, Lee JA, Kim CY (2011) Radiotherapy for pituitary adenomas: long-term outcome and complications. Radiat Oncol J 29:156–163 13. Castinetti F, Nagai M, Morange I, Dufour H, Caron P, Chanson P, Cortet-Rudelli C, Kuhn JM, Conte-Devolx B, Regis J, Brue T (2009) Long-term results of stereotactic radiosurgery in secretory pituitary adenomas. J Clin Endocrinol Metab 94:3400–3407 14. Feigl GC, Bonelli CM, Berghold A, Mokry M (2002) Effects of gamma knife radiosurgery of pituitary adenomas on pituitary function. J Neurosurg 97:415–421 15. Franzin A, Spatola G, Losa M, Picozzi P, Mortini P (2012) Results of gamma knife radiosurgery in acromegaly. Int J Endocrinol 2012:342034 16. Kim M, Paeng S, Pyo S, Jeong Y, Lee S, Jung Y (2006) Gamma knife surgery for invasive pituitary macroadenoma. J Neurosurg 105(Suppl):26–30 17. Pollock BE, Cochran J, Natt N, Brown PD, Erickson D, Link MJ, Garces YI, Foote RL, Stafford SL, Schomberg PJ (2008) Gamma knife radiosurgery for patients with nonfunctioning pituitary adenomas: results from a 15-year experience. Int J Radiat Oncol Biol Phys 70:1325–1329 18. Sheehan JP, Jagannathan J, Pouratian N, Steiner L (2006) Stereotactic radiosurgery for pituitary adenomas: a review of the literature and our experience. Front Horm Res 34:185–205 19. Kobayashi T (2009) Long-term results of stereotactic gamma knife radiosurgery for pituitary adenomas. Specific strategies for different types of adenoma. Prog Neurol Surg 22:77–95 20. Kong DS, Lee JI, Lim DH, Kim KW, Shin HJ, Nam DH, Park K, Kim JH (2007) The efficacy of fractionated radiotherapy and stereotactic radiosurgery for pituitary adenomas: long-term results of 125 consecutive patients treated in a single institution. Cancer 110:854–860 21. Pollock BE, Brown PD, Nippoldt TB, Young WF Jr (2008) Pituitary tumor type affects the chance of biochemical remission after radiosurgery of hormone-secreting pituitary adenomas. Neurosurgery 62:1271–1276; discussion 1276–1278 22. Vik-Mo EO, Oksnes M, Pedersen PH, Wentzel-Larsen T, Rodahl E, Thorsen F, Schreiner T, Aanderud S, Lund-Johansen M (2007) Gamma knife stereotactic radiosurgery for acromegaly. Eur J Endocrinol 157:255–263 23. Attanasio R, Epaminonda P, Motti E, Giugni E, Ventrella L, Cozzi R, Farabola M, Loli P, Beck-Peccoz P, Arosio M (2003) Gamma-knife radiosurgery in acromegaly: a 4-year follow-up study. J Clin Endocrinol Metab 88:3105–3112 24. Jezkova J, Marek J, Hana V, Krsek M, Weiss V, Vladyka V, Lisak R, Vymazal J, Pecen L (2006) Gamma knife radiosurgery for acromegaly–long-term experience. Clin Endocrinol (Oxf) 64:588–595 25. Castro DG, Cecilio SA, Canteras MM (2010) Radiosurgery for pituitary adenomas: evaluation of its efficacy and safety. Radiat Oncol 5:109 26. Pamir MN, Kilic T, Belirgen M, Abacioglu U, Karabekiroglu N (2007) Pituitary adenomas treated with gamma knife
Pituitary
27.
28.
29.
30.
radiosurgery: volumetric analysis of 100 cases with minimum 3 year follow-up. Neurosurgery 61:270–280; discussion 280 Starke RM, Williams BJ, Jane JA Jr, Sheehan JP (2012) Gamma knife surgery for patients with nonfunctioning pituitary macroadenomas: predictors of tumor control, neurological deficits, and hypopituitarism. J Neurosurg 117:129–135 Jagannathan J, Yen CP, Pouratian N, Laws ER, Sheehan JP (2009) Stereotactic radiosurgery for pituitary adenomas: a comprehensive review of indications, techniques and long-term results using the gamma knife. J Neurooncol 92:345–356 Wowra B, Stummer W (2002) Efficacy of gamma knife radiosurgery for nonfunctioning pituitary adenomas: a quantitative follow up with magnetic resonance imaging-based volumetric analysis. J Neurosurg 97:429–432 Mingione V, Yen CP, Vance ML, Steiner M, Sheehan J, Laws ER, Steiner L (2006) Gamma surgery in the treatment of nonsecretory pituitary macroadenoma. J Neurosurg 104:876–883
31. Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD (2002) Radiosurgery for residual or recurrent nonfunctioning pituitary adenoma. J Neurosurg 97:408–414 32. Losa M, Gioia L, Picozzi P, Franzin A, Valle M, Giovanelli M, Mortini P (2008) The role of stereotactic radiotherapy in patients with growth hormone-secreting pituitary adenoma. J Clin Endocrinol Metab 93:2546–2552 33. Higham CE, Chung TT, Lawrence J, Drake WM, Trainer PJ (2009) Long-term experience of pegvisomant therapy as a treatment for acromegaly. Clin Endocrinol (Oxf) 71:86–91 34. Jagannathan J, Sheehan JP, Pouratian N, Laws ER Jr, Steiner L, Vance ML (2008) Gamma knife radiosurgery for acromegaly: outcomes after failed transsphenoidal surgery. Neurosurgery 62:1262–1269; discussion 1269–1270
123
Effective time window in reducing pituitary adenoma size by gamma knife radiosurgery Henry Ka-Fung Mak • Shui-Wun Lai • Wenshu Qian • Stanley Xu • Elizabeth Tong • May Lee Vance • Edward Oldfield • John Jane Jr. Jason Sheehan • Kelvin K. W. Yau • Max Wintermark
•
Ó Springer Science+Business Media New York 2014
Abstract Objectives Although the effectiveness of gamma knife radiosurgery (GKRS) in controlling the size of pituitary adenomas has been well demonstrated in many studies, the time period in which significant changes in tumor size occurs has been investigated in a limited fashion. It is important to determine the therapeutic window of GKRS in treating pituitary adenomas, i.e., the effective timeframe during which significant size reduction of these tumors occurs, so that alternative treatments such as further GKRS or microsurgery might be prescribed in a timely manner if clinically indicated. Methods This was a nested sample of an ongoing local cohort study on GKRS for pituitary adenomas at the University of Virginia. Magnetic resonance imaging (MRI) using dedicated sequences was employed. Only patients with a baseline MRI (TP0) and at least 1 follow-up study
H. K.-F. Mak S.-W. Lai W. Qian Department of Diagnostic Radiology, University of Hong Kong, Hong Kong Special Administrative Region, China S. Xu E. Oldfield J. Jane Jr. J. Sheehan Department of Neurosurgery, University of Virginia, Charlottesville, VA, USA E. Tong M. Wintermark (&) Department of Radiology, Neuroradiology Division, University of Virginia, Box 800170, Charlottesville, VA 22908, USA e-mail: [email protected] M. L. Vance Department of Medicine, University of Virginia, Charlottesville, VA, USA K. K. W. Yau Department of Management Sciences, City University of Hong Kong, Hong Kong Special Administrative Region, China
performed in the University Hospital after GKRS were included. The follow-up scans were performed at five timepoints (TP1–TP5) which were 6, 12, 24, 36 and 48 months after GKRS. The dimensional indices of the tumors were measured in three orthogonal planes, i.e., transverse (TR), antero-posterior (AP) and cranio-caudal (CC). The volumes of the tumors were estimated by using the following formula: V ¼ p ðTR AP CCÞ=6. Tumor volume decrease by more than 25 % from baseline was considered as ‘shrinkage’, \25 % tumor size increase or decrease was considered ‘static’, and more than 25 % increase as ‘increment’. Our cohort consisted of 21 patients, with functioning adenomas in 13 subjects i.e. six adrenocorticotrophic hormone (ACTH)-secreting and seven growth hormone (GH)-secreting, and non-functioning (NF) adenomas in eight subjects. Results In 26 adenomas (8 ACTH, 9 GH and 9 NF), tumor control (tumor shrinkage or static) were achieved in 21 tumors (80.8 %); 89, 75, and 78 % for GH-secreting, ACTH-secreting and NF adenomas respectively, at the end of the 4-year follow-up period. Analysis of variance showed significant differences of GKRS margin dose among different types of tumors (p = 0.013), but not of baseline tumor volumes (p = 0.240). Logistic regression analysis showed no significant association of margin dose, baseline volume or tumor type with the tumor control outcome. Comparison of tumor change using dimensional indices relative to the base time point (TP0) showed that in the sample there was an average reduction of 1.290 mm at TP1 (6 months) with p values 0.155 (parametric t test) and 0.098 (non-parametric Wilcoxon signed-ranked test) respectively, showing a moderate reduction in tumor dimensional indices. The change in dimensional indices at later time points (TP2–TP5) showed an average reduction ranging from 1.930 to 2.471 mm. Significant reduction in
123
Pituitary
the mean dimensional indices was firstly observed at TP2 (1 year) with p values 0.013 (t test) and 0.018 (Wilcoxon signed-rank test). Such scale of reduction in the dimensional indices appeared to be maintained along the time axis (from TP2 to TP5). Conclusions Significant decrease in tumor dimensional indices tended to occur at 1 year post-GKRS. Although to a lesser extent, such decrease in dimensional indices continued up to the end of our follow-up period. Keywords Pituitary adenoma Gamma knife radiosurgery Therapeutic time window Magnetic resonance imaging Tumor control Abbreviations GKRS Gamma-knife radiosurgery NPA Non-functioning pituitary adenomas TR Transverse AP Antero-posterior CC Cranio-caudal TP Time point TP0 Baseline TP1 6 months TP2 12 months TP3 24 months TP4 36 months TP5 48 months
Introduction Pituitary adenomas are benign neoplastic lesions of the anterior pituitary gland. Epidemiologic studies reveal that pituitary adenomas are the second most common primary intracranial neoplasm and accounts for about 10–15 % of all intracranial tumor cases [1, 2]. There are currently four treatment modalities that are used to control the tumor size and hormone normalization: medication, surgery, radiation therapy and stereotactic radiosurgery. Transsphenoidal microsurgery is currently considered first-line therapy for most symptomatic pituitary adenomas with the exception of prolactinomas. Surgery provides rapid reduction of tumor size as well as resolution of neurologic symptoms caused by mass effect [3]. However, complete resection of tumors, especially those with cavernous sinus invasion, is not always possible and the recurrence rate after transsphenoidal surgery can range from 20 to 50 % [4–6]. As a result, adjuvant therapies are frequently prescribed in combination with transsphenoidal microsurgery to achieve better tumor control and hormone normalization.
123
Previous studies had evaluated the advantages and disadvantages of pharmacotherapy [7, 8] and radiation therapy [9–12] as adjuvant therapies. Stereotactic radiosurgery using gamma-knife radiosurgery (GKRS) has emerged as a well-accepted second-line modality. It has been demonstrated in many studies that GKRS is efficacious in controlling tumor size in both secretory and non-secretory pituitary adenomas, with a tumor control rate of above 90 % in many studies [4, 13– 18]. However, the effect of GKRS on hormone normalization is variable, with different series reporting a wide range of endocrine remission rates from 17 to 100 % depending on tumor size and type [19–22]. The effectiveness of GKRS in controlling the size of pituitary adenomas has been well demonstrated in many studies. However, the time period during which significant change in tumor size occurs has been investigated in a limited fashion. According to a previous study, the incidence of tumor shrinkage increased from 28 % at 1 year after GKRS to 79 % at 4 years after GKRS [23]. Similar rates have been reported by Kong et al., in which the incidence of an objective response (complete and partial tumor shrinkage) increased from 39.5 % at 2 years after GKRS to 81.8 % at 4 years after GKRS [20]. Shrinkage was achieved in 46.9 % of tumors after 1–2 years, and in 61.1 % of patients 2–5 years after irradiation in another study [24]. Although most studies suggested that GKRS has a short latency period [4, 7, 11, 25], few of these studies investigated the magnitude of change in tumor size at various time points following GKRS [24, 26] or documented the interval at which maximal tumor shrinkage occurred. It is important to determine the therapeutic window of GKRS in treating pituitary adenomas, i.e., the effective timeframe during which significant size reduction of these tumors occurs, so that alternative treatments such as further GKRS or microsurgery [27] can be prescribed in a timely manner if clinically indicated. This pilot study aims, (1) to track the changes in the size of pituitary tumors using dedicated magnetic resonance imaging (MRI) sequences at various time points before and after GKRS, and (2) to identify the precise interval during which significant tumor shrinkage occurs. As demonstrated by several studies, the percentage of patients in which tumor shrinkage was observed increased drastically between the 1st and 5th year [20, 23, 24, 26, 28]; we therefore hypothesize that significant size change would occur at some time between these two time points.
Materials and methods This was a nested sample (between September 2004 and April 2007) of an ongoing local cohort study on GKRS for
Pituitary
pituitary adenomas at the University of Virginia. Details on patient population, pre- and post-radiosurgical follow-up, and stereotactic radiosurgery technique were published previously [6, 27]. The study was approved by the local Institutional Review Board of Human Study Research of the University. Inclusion Criteria: Any patient treated during this period was recruited if the person had MRI examination performed in the University hospital, including baseline MRI (TP0) and at least 1 follow-up scan after GKRS. The follow-up scans were performed at 5 different time-points (TP1 to TP5): 6, 12, 24, 36 and 48 months after GKRS. The study cohort consisted of 24 patients, but three subjects showed no identifiable pituitary tumor and were excluded. The final cohort of 21 patients consisted of 8 males and 13 females, with mean age of 49.6 years (standard deviation, SD 15.3 years, range 16–79 years). Functioning adenomas were found to be present in 13 subjects (six ACTH-secreting and seven GH-secreting), and nonfunctioning (NF) adenomas in eight subjects based on clinical, biochemical, radiologic and pathologic criteria. The demographic, clinical profiles, pathological findings, and GKRS margin dose of the patients are summarized in Table 1. The mean and median duration of follow-up was 31.8 and 36 months, respectively. In our study, the median number of follow-up MRI examinations underwent by our patients was 3, which compared favorably to a prior study with a median followup of 2 [29]. Follow-up MRI examinations were performed for 12, 14, 11, 11, and 8 patients at TP1, TP2, TP3, TP4 and TP5 respectively. Standardized pituitary MR (Siemens Magnetom, 1.5 Tesla or 3 Tesla) imaging sequences were used for all baseline and follow-up studies, including T1-weighted images in sagittal, coronal and axial planes, T2-weighted images in coronal and axial planes, and post-gadolinium T1-weighted images in sagittal, coronal and axial planes. All images had slice thickness of 3 mm and no gap in between. A neuroradiologist (HKFM), blinded to the clinical and pathological findings, evaluated the serial MRI imaging studies of each patient and recorded the results on a standard form. A senior neuroradiologist (MW) was being consulted in problematic cases. The dimensional indices of the tumors were measured and recorded in three orthogonal planes, i.e., transverse (TR), antero-posterior (AP) and cranio-caudal (CC). The volumes of the tumors were estimated by using the following formula: V ¼ pðTR AP CCÞ=6 [23]. Tumor volume decrease by more than 25 % from baseline was considered as ‘shrinkage’, \25 % increase or decrease of tumor volume as ‘static’, and more than 25 % increase as ‘increment’.
Primarily, T2-weighted coronal images were used to evaluate the TR and CC dimensional indices (because of clarity of margins), while T1-weighted sagittal images for the AP dimensional index. In addition, post gadolinium T1 weighted images were used to augment the above sequences and to check for any discrepancy. Statistical analysis The margin dose of GKRS and baseline tumor volumes were compared between different tumor types using analysis of variance (ANOVA). Logistic regression analysis was performed, using the tumor status (controlled or not) at the last time point (TPL) as dependent variable, with GKRS dose and baseline volume as covariates and tumor type (NFA, ACTH and GH) as factor. To assess the dimensional indices and volumes reduction of the identifiable tumors of all subjects at different time points (TPi, i = 1, 2, 3, 4, 5) relative to the base time point (TP0), a subset of pairwise comparisons was conducted using both parametric (t test) and non-parametric (Wilcoxon signed-ranked test) methods. Bonferroni correction for multiple pairwise comparisons was taken to adjust the p values. All statistical analyses were conducted using IBM SPSS (Version 19.0.0). Level of significance was set at 0.05.
Results Descriptive statistics Based on the volumes of 26 tumors in our final cohort of 21 patients (three subjects had two separate identifiable tumor masses and one had three separate lesions), mean tumor volume decreased from a baseline value of 3.72 ? 4.52 to 2.53 ± 3.40 cm3 (Table 1). A mean tumor volume decrease of 20.1 % (SD 44.5 %) was observed in the entire cohort of 26 tumors. In the 14 tumors with shrinkage, the volume decreased by 52.4 % (SD 19.0 %), and in the 5 tumors with increment, the volume increased by 50.9 % (SD 20.2 %). Tumor control (tumor shrinkage or static) was observed in 16 (76.2 %) patients and 21 tumors (80.8 %) at the end of the 4-year follow-up period. Figure 1 shows the tumor size changes according to different tumor types after GKRS at last time point. The mean (and standard deviation) of margin dose and baseline tumor volume of different tumor types are shown in Table 2. The number and percentage of tumors (according to different tumor types) with volume shrinkage, and tumor
123
Pituitary Table 1 The demographic, clinical profiles, pathological findings and margin dose of all patients in our cohort. GK- Gamma Knife; TPL- last time point Patient
Tumor type
GK margin dose (Gy)
Volume (cm3)
Age/ gender
Clinical presentation/Tx before GK
1
55/F
Cushing’s disease/2 transphenoid surgery
ACTH-staining adenoma
23.00
16.025
5.039
2
39/F
Cushing’s disease/transphenoid surgery
ACTH-staining adenoma
25.00
0.047
0.079
3
46/F
Elevated growth hormone/transphenoid surgery
No mention
25.00
0.228
0.238
4
41/M
Acromegaly/transphenoid surgery
No mention
25.00
10.329
0.301
5
70/M
Acromegaly/transphenoid surgery
No mention
25.00
0.515
0.205
Mass 2: 0.290
0.076
Baseline
TPL
6
76/M
Acromegaly/placed on Sandostatin LAR
No mention
20.00
1.685
2.292
7
38/F
Acromegaly/placed on Sandostatin LAR, transsphenoid surgery
No mention
25.00
0.101
0.048
8
16/M
Bitemporal hemianopsia/on Levoxyl & prednisone prior to surgery, 2 prior surgical resection
Nonfunctioning
16.50
6.713
8.481
9
59/F
Visual loss/2 transsphenoidal surgery
Nonfunctioning
17.00
1.930
2.129
Mass 2: 0.036
0.030
10
49/F
Progressive vision loss with right temporal field cut/transsphenoidal surgery
Nonfunctioning
16.00
9.839
14.797
11
37/F
Peripheral vision loss and headaches/ craniotomy and resection
Nonfunctioning
25.00
1.619
1.085
12
43/F
Acromegaly/transphenoid surgery
Functioning-GH
25.00
1.349
1.662
Mass 2: 0.505
0.310
13
32/M
Acromegaly/no reduction with Sandostatin, 2 transphenoid surgery
Functioning-GH
20.00
0.218
0.105
14
39/F
Hypothyroidism and Cushing’s, amenorrhea/transphenoid surgery
Functioning-ACTH
25.00
0.105
0.057
15
51/F
Cushing’s/resection
Functioning-ACTH
25.00
2.862
4.956
16
79/F
Sudden headache, found pituitary apoplexy/Transphenoid resection
Silent ACTH staining adenoma
18.00
5.868
2.121
17
39/F
2 transphenoid surgery
Nonfunctioning
15.00
1.473
1.050
18
58/F
Cushing’ syndrome/bilateral adrenalectomy, 5 transphenoid surgery, 1 prior GK
Functioning-ACTH
8.50
2.362
1.644
Mass 2: 2.751 Mass 3: 4.801
2.131 2.384
19
63/M
Non-functioning pituitary adenoma/ transphenoid surgery
Nonfunctioning
15.00
8.190
6.668
20
62/M
Nonfunctioning
16.00
2.817
2.224
21
49/M
Non-secretory adenoma/2 transphenoid surgery Non-secretory adenoma/2 transphenoid surgery
Nonfunctioning
15.00
14.072
5.708
control (tumor shrinkage or static) at different time points are shown (Table 3). Percentage variations between the time points within each tumor type were likely due to different sample sizes under consideration.
123
Initial increases (\15 % compared to baseline study) in tumor volumes (within 6 months to 1 year) followed by subsequent tumor shrinkage occurred in 3 out of the 26 tumors (11.5 %), likely due to intratumoral edema [26, 29].
Pituitary
Multiple logistic regression with the tumor status (controlled or not) at the last time point as dependent variable, and using dose, baseline volume and tumor types as covariate/factor No significant covariate/factor was found to be associated with the tumor status outcome, probably because of small sample size. Paired comparison in tumor dimensional indices between different time points (Table 4) Fig. 1 Mean tumor volume changes (in percentage) compared with baseline volumes according to different types of tumors after GKRS at the end of the 48-month follow-up period. Shrinkage (decrease by 25 %), static (\25 % increment and \25 % decrease in volume) and increment (increase by 25 %); Error bars represent ±1 SD Table 2 Mean [and standard deviation (SD)] margin dose of GKRS and baseline volumes NF (n = 9)
ACTH (n = 8)
GH (n = 9)
16.94
17.69
23.89
3.13
7.95
2.20
Dose (Gy) Mean SD
Baseline volume (cm3) Mean
5.19
4.35
1.69
SD
4.76
5.13
3.28
Comparison of GKRS margin dose of and baseline tumor volumes between different tumor types using analysis of variance (ANOVA) ANOVA showed significant differences of dose among different types of tumors (p = 0.013), but not of baseline tumor volumes (p = 0.240). With Tukey’s post hoc test, differences of dose between NF versus ACTH and NF versus GH were observed (p = 0.017 and 0.042 respectively).
Comparison of tumor change using dimensional indices, i.e. TR, AP, and CC relative to the base time point (TP0) showed that in the sample there was an average reduction of 1.290 mm at TP1 (6 months) with p values 0.155 (t test) and 0.098 (Wilcoxon signed-ranked test) respectively, showing a moderate reduction in tumor dimensional indices. The change in dimensional indices at later time points (TP2 to TP5) showed an average reduction ranging from 1.930 to 2.471 mm. Significant reduction in the mean dimensional indices was firstly observed at TP2 (1 year) with p values 0.013 (t test) and 0.018 (Wilcoxon signedrank test). Such scale of reduction in the dimensional indices appeared to be maintained along the time axis (from TP2 to TP5), although we observed only a marginal significance at TP3 which was deemed partly due to the conservativeness of the Bonferroni correction in multiple pairwise comparisons. Using 21 tumors with size control The results were similar to the whole group except that significant difference between TP0 and TP1 was also seen i.e. average reduction of 2.007 mm at TP1 (6 months) with p values 0.014 (t test) and 0.027 (Wilcoxon signed-ranked test) respectively.
Table 3 Number (percentage) of tumors of different tumor types after GKRS at different time points with either volume decrease or size control TP1
TP2
TP3
TP4
TP5
Number of tumors with volume decrease (%) NF
2 (33.33)
3 (42.86)
2 (28.57)
4 (50.00)
3 (33.33)
ACTH GH
4 (66.67) 1 (33.33)
4 (57.14) 3 (33.33)
4 (57.14) 6 (66.67)
5 (62.50) 6 (66.67)
5 (62.50) 6 (66.67)
Number of tumors with size control (%) NF
5 (83.33)
6 (85.71)
6 (85.71)
7 (87.50)
7 (77.78)
ACTH
5 (83.33)
6 (85.71)
6 (85.71)
6 (75.00)
6 (75.00)
GH
2 (66.67)
8 (88.89)
7 (77.78)
7 (77.78)
8 (88.89)
123
Pituitary
0.001* 0.007* 1.384
0.004* 0.024*
\0.001* \0.001*
1.560
0.737
0.050* 0.025*
0.002* \0.001*
1.204
1.012
t test
Standard Error (mm)
4.918 16 0.003* 0.009* 0.926 2.936 28 0.023*
4.565
3.806
18
22
0.034* 0.056
\0.001* \0.001* 0.592
0.912 2.176
2.601 33
36 0.063
0.005*
3.501
3.737 30
18 0.027*
0.005* 0.008*
0.014* 0.680
0.768 2.390
2.007 36 0.098
44
Wilcoxon signed-rank test t test
N Wilcoxon signed-rank test
0.018*
Mean difference (mm) Mean difference (mm)
Standard Error (mm)
p value
#
21 tumors with size control
123
* Significant at 5 % level One-sided test with Bonferroni correction #
N number of entries of dimensional indices
0.020* 0.875 31 TP5
2.471
0.075
0.015* 0.623
0.855
39 TP4
1.804
39 TP3
1.930
0.013*
t test
0.155 0.673
0.695 2.024 50 TP2
1.290 45 TP1
N
Mean difference (mm)
Standard Error (mm)
p value
#
All tumors
Table 4 Paired comparisons of tumor dimensional indices relative to the baseline time point (TP0)
Comparison of the dimensional indices relative to the base time point (TP0) showed remarkable reduction as early as at TP1 (6 months). The average reduction in dimensional indices ranged from 3.501 mm (TP1) to 4.918 mm (TP5). Comparison of tumor change using pituitary tumor volumes relative to the base time point (TP0) did not show any appreciable significant difference (results not shown).
Discussion
N
14 tumors with shrinkage
p value#
Wilcoxon signed-rank test
Using 14 tumors with shrinkage only
The study demonstrates a tumor control rate of 80.8 % was found at the end of a 4-year follow-up in our cohort of 21 patients. The effectiveness of GKRS on the control of the size of pituitary tumors has been well documented in different series with most having a tumor control rate of [90 % [4, 13–18, 23, 30]. Subgroup analysis according to tumor types revealed that rates of control of tumor size were 89, 75 and 78 % for growth hormone-secreting, adrenocorticotrophic hormone-secreting and non-functioning adenomas, respectively (Table 3). The slightly lower tumor control rates in this series is likely related to sampling bias, since in larger cohorts treated in the same centre, tumor control rates of 89, 92 and 93 %, respectively, were previously reported [18]. A previous study [5] reported at least a 20 % reduction of tumor volume in 42.3 % of their patients with nonfunctioning pituitary adenomas. Attanasio et al. [23] reported that tumor reduction of [25 % of basal volume occurred in 58 % of their cohort of acromegalic patients 24 months after GKRS, while Kong et al. [20], found an objective response rate (complete being [25 % and partial being [10 % tumor shrinkage) of 39.5 % at 2 years and 81.8 % at 4 years in a large group of patients (including both functional and non-functional adenomas). At the end of a 48-month follow-up period in our cohort, tumor shrinkage (volume decrease [25 %) was achieved in 52.4, 33.3, 62.5 and 66.7 % of all, NF, ACTH, and GH adenomas respectively (Table 3); which was comparable to these prior studies. Contrary to our previous study [6], there was no significant association between tumour control outcome and margin dose. This was probably due to the small sample size. In view of no observed effect of margin dose, baseline tumor volume or tumor type on tumor control outcome in our cohort, tumors of different types were grouped together in current study to increase the statistical power for the analysis of effective timeframe of GKRS. In this study dimensional indices were assessed as an indicator of tumor size change at different time points. These measurements are commonly used in clinical practice
Pituitary
for the evaluation of the efficacy of GKRS [20, 24, 31]. The paired time-point comparisons used in the current study are more accurate than cumulative rate in detection of tumor size change because such a comparison was made within the same subject in each pair of time points and thus variations due to subject differences can be eliminated. In our study, whole group (all subjects) analyses revealed that dimensional reductions of the tumors occur between the baseline and all subsequent time points. Pamir et al. [26] reported in their study that earliest volume decrease after GKRS occurred 6–9 months after the procedure, albeit with higher marginal doses ([27 Gy). Significant reduction in tumor dimensional indices tended to occur 1 year after GKRS in our study. Subsequent reduction in later time points appeared to be not that apparent. In the subgroup analyses of the 21 tumors with size control and 14 tumors with shrinkage [25 %, the results were even more convincing. There are significant differences between baseline and all the subsequent time points, as early as 6 months after GKRS. A further reduction trend was observed in later time points, although in a less drastic manner. Similarly, a review of 82 patients with nonsecretory tumors by Jagannathan et al. [28] revealed that the median time to tumor shrinkage on MR imaging was 9 months (range 6–48 months). Interestingly, no significant difference in tumor volumes across various TPs was found, as compared to dimensional indices. The finding was not surprisingly because the estimated volumes were derived from the product of 3 dimensional index random variables. The observed significant reduction on dimensional indices but along with an insignificant reduction on the estimated ellipsoid volume suggested a non-uniform reduction in the 3 dimensions. Hypothetically, some of the tumor shrinkage only occurred in any 1 or 2 dimensions, thence the volume would not have the same degree of reduction effect size as the individual dimension did, which would give rise to such a phenomenon. Nevertheless, we did not have sufficient data to verify. This may warrant further study in future. Nevertheless, previous studies found that endocrine remission in ‘radioresponsive’ secretory adenomas occurs usually within the first 2 years [4]. However, no direct correlation between tumor volume response and endocrine response following GKRS has been reported so far [4, 7, 32]. Small basal tumor volume was a determinant of volume change [26] after GKRS, and was associated with endocrine remission [6, 13, 25], though not found in one study [24]. In current study, the confounding effect of the antisecretory medication in patients with acromegaly after GKRS was also evaluated. 4 patients were on continual antisecretory medication (Pegvisomant) after GKRS; 2 showed tumor shrinkage, 1 remained static and 1 showed
increment at latest follow-up scan. For the other three patients with medication being stopped after IGF-1 normalization (2 on Sandostatin and 1 on Pegvisomant)), 2 showed tumor shrinkage and 1 remained static at latest follow-up scan. Pegvisomant lowers IGF-1 level but has no proven anti-proliferative effect [33], therefore, co-administration of this drug should not affect the imaging findings of these 5 patients. For the two remaining patients on Sandostatin, since the medication was stopped between 9 months to 1 year, the drug effect should be minimal. In patients with secreting tumors, it would be tempting to have a comparison between anti-tumor and anti-secretory efficacy of GKRS. All patients with ACTH-adenomas had normal laboratory tests except patient 18. All the acromegalic patients (except patient 12) eventually had IGF-1 level normalized (with or without medications after GKRS), 4 with tumor shrinkage, 1 static and 1 with increment. The antitumor and antisecretory efficacies of GKRS could not be compared in those patients with concurrent use of anti-secretory medication. Patient 18 had 3 tumors under control after GKRS (2 with shrinkage and 1 static) but persistently elevated ACTH level (latest— 446 pg/ml). Patient 12 had persistently elevated IGF-1 level, but static tumor size. There is effective tumor control but failure of anti-secretory efficacy of GKRS in these two subjects. Five large tumors with tumor size [8 cm3 have been treated with GKRS in this study. It is generally accepted that GKRS compared with conventional fractionated radiation therapy, provides a decreased time interval between the treatment and the achievement of remission in functioning pituitary adenomas. Furthermore, given the sharp radiation dose fall-off of the GKRS, it is safe and effective to treat large tumors as described above as long as the adenoma is not too close to the optic apparatus. Furthermore, for the five subjects with tumor size increment, no pituitary carcinoma was detected. 2 patients (1 ACTH staining adenoma and 1 non-functioning adenoma) were uneventful and had regular follow-up; 1 patient had non-functioning adenoma but was lost in follow-up; 1 patient had functioning, invasive ACTH adenoma requiring 1 surgical resection and 1 CyberKnife radiosurgery after GKRS, and 1 acromegalic patient with invasive tumor to cavernous sinus was placed on antisecretory medication. Limitations This was a retrospective study to examine the time frame at which pituitary tumor size reduction occurred after the administration of GKRS. Our study is limited by a relatively small sample size and pooled data of different tumor types. Hence, it is exploratory in nature. Prospective
123
Pituitary
studies with a larger sample size and longer follow-up period are needed to confirm our results. Further validation of our results by measuring actual tumor volumes obtained from volume scans is recommended [27, 29, 34]. The time point at which failure of GKRS occurs i.e. pituitary tumors begin to grow, should also be investigated in a larger-scale study. In addition, comprehensive head-to-head comparison with published studies in the literature was difficult due to heterogeneous methodology and study cohorts [23, 25] in analyzing treatment outcome [10].
Conclusion Consistent with the findings in prior studies, GKRS is efficacious in tumor size control. Effective tumor size reduction tended to occur 1 year after the administration of GKRS and further tumor reduction continued up to the end of our follow-up period (4 years post GKRS) especially for those with size control/with shrinkage, although to a lesser extent. If GKRS fails to achieve any significant change during such a period of time, alternative treatments such as further GKRS or microsurgery might be prescribed in a timely manner if clinically indicated. Conflict of interest The authors reported no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Ethical standard We declared that the experiments comply with the current laws of our country.
References 1. Laws ER Jr, Vance ML (1999) Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am 10:327–336 2. Radhakrishnan K, Mokri B, Parisi JE, O’Fallon WM, Sunku J, Kurland LT (1995) The trends in incidence of primary brain tumors in the population of Rochester, Minnesota. Ann Neurol 37:67–73 3. Mortini P, Losa M, Barzaghi R, Boari N, Giovanelli M (2005) Results of transsphenoidal surgery in a large series of patients with pituitary adenoma. Neurosurgery 56:1222–1233; discussion 1233 4. Laws ER, Sheehan JP, Sheehan JM, Jagnathan J, Jane JA Jr, Oskouian R (2004) Stereotactic radiosurgery for pituitary adenomas: a review of the literature. J Neurooncol 69:257–272 5. Losa M, Valle M, Mortini P, Franzin A, da Passano CF, Cenzato M, Bianchi S, Picozzi P, Giovanelli M (2004) Gamma knife surgery for treatment of residual nonfunctioning pituitary adenomas after surgical debulking. J Neurosurg 100:438–444 6. Sheehan JP, Pouratian N, Steiner L, Laws ER, Vance ML (2011) Gamma knife surgery for pituitary adenomas: factors related to radiological and endocrine outcomes. J Neurosurg 114:303–309 7. Sheehan JP, Niranjan A, Sheehan JM, Jane JA Jr, Laws ER, Kondziolka D, Flickinger J, Landolt AM, Loeffler JS, Lunsford LD (2005) Stereotactic radiosurgery for pituitary adenomas: an intermediate review of its safety, efficacy, and role in the neurosurgical treatment armamentarium. J Neurosurg 102:678–691
123
8. Marko NF, LaSota E, Hamrahian AH, Weil RJ (2012) Comparative effectiveness review of treatment options for pituitary microadenomas in acromegaly. J Neurosurg 117:522–538 9. Barkan AL, Halasz I, Dornfeld KJ, Jaffe CA, Friberg RD, Chandler WF, Sandler HM (1997) Pituitary irradiation is ineffective in normalizing plasma insulin-like growth factor I in patients with acromegaly. J Clin Endocrinol Metab 82:3187–3191 10. Brada M, Ajithkumar TV, Minniti G (2004) Radiosurgery for pituitary adenomas. Clin Endocrinol (Oxf) 61:531–543 11. Prasad D (2006) Clinical results of conformal radiotherapy and radiosurgery for pituitary adenoma. Neurosurgery Clin N Am 17:129–141, vi 12. Rim CH, Yang DS, Park YJ, Yoon WS, Lee JA, Kim CY (2011) Radiotherapy for pituitary adenomas: long-term outcome and complications. Radiat Oncol J 29:156–163 13. Castinetti F, Nagai M, Morange I, Dufour H, Caron P, Chanson P, Cortet-Rudelli C, Kuhn JM, Conte-Devolx B, Regis J, Brue T (2009) Long-term results of stereotactic radiosurgery in secretory pituitary adenomas. J Clin Endocrinol Metab 94:3400–3407 14. Feigl GC, Bonelli CM, Berghold A, Mokry M (2002) Effects of gamma knife radiosurgery of pituitary adenomas on pituitary function. J Neurosurg 97:415–421 15. Franzin A, Spatola G, Losa M, Picozzi P, Mortini P (2012) Results of gamma knife radiosurgery in acromegaly. Int J Endocrinol 2012:342034 16. Kim M, Paeng S, Pyo S, Jeong Y, Lee S, Jung Y (2006) Gamma knife surgery for invasive pituitary macroadenoma. J Neurosurg 105(Suppl):26–30 17. Pollock BE, Cochran J, Natt N, Brown PD, Erickson D, Link MJ, Garces YI, Foote RL, Stafford SL, Schomberg PJ (2008) Gamma knife radiosurgery for patients with nonfunctioning pituitary adenomas: results from a 15-year experience. Int J Radiat Oncol Biol Phys 70:1325–1329 18. Sheehan JP, Jagannathan J, Pouratian N, Steiner L (2006) Stereotactic radiosurgery for pituitary adenomas: a review of the literature and our experience. Front Horm Res 34:185–205 19. Kobayashi T (2009) Long-term results of stereotactic gamma knife radiosurgery for pituitary adenomas. Specific strategies for different types of adenoma. Prog Neurol Surg 22:77–95 20. Kong DS, Lee JI, Lim DH, Kim KW, Shin HJ, Nam DH, Park K, Kim JH (2007) The efficacy of fractionated radiotherapy and stereotactic radiosurgery for pituitary adenomas: long-term results of 125 consecutive patients treated in a single institution. Cancer 110:854–860 21. Pollock BE, Brown PD, Nippoldt TB, Young WF Jr (2008) Pituitary tumor type affects the chance of biochemical remission after radiosurgery of hormone-secreting pituitary adenomas. Neurosurgery 62:1271–1276; discussion 1276–1278 22. Vik-Mo EO, Oksnes M, Pedersen PH, Wentzel-Larsen T, Rodahl E, Thorsen F, Schreiner T, Aanderud S, Lund-Johansen M (2007) Gamma knife stereotactic radiosurgery for acromegaly. Eur J Endocrinol 157:255–263 23. Attanasio R, Epaminonda P, Motti E, Giugni E, Ventrella L, Cozzi R, Farabola M, Loli P, Beck-Peccoz P, Arosio M (2003) Gamma-knife radiosurgery in acromegaly: a 4-year follow-up study. J Clin Endocrinol Metab 88:3105–3112 24. Jezkova J, Marek J, Hana V, Krsek M, Weiss V, Vladyka V, Lisak R, Vymazal J, Pecen L (2006) Gamma knife radiosurgery for acromegaly–long-term experience. Clin Endocrinol (Oxf) 64:588–595 25. Castro DG, Cecilio SA, Canteras MM (2010) Radiosurgery for pituitary adenomas: evaluation of its efficacy and safety. Radiat Oncol 5:109 26. Pamir MN, Kilic T, Belirgen M, Abacioglu U, Karabekiroglu N (2007) Pituitary adenomas treated with gamma knife
Pituitary
27.
28.
29.
30.
radiosurgery: volumetric analysis of 100 cases with minimum 3 year follow-up. Neurosurgery 61:270–280; discussion 280 Starke RM, Williams BJ, Jane JA Jr, Sheehan JP (2012) Gamma knife surgery for patients with nonfunctioning pituitary macroadenomas: predictors of tumor control, neurological deficits, and hypopituitarism. J Neurosurg 117:129–135 Jagannathan J, Yen CP, Pouratian N, Laws ER, Sheehan JP (2009) Stereotactic radiosurgery for pituitary adenomas: a comprehensive review of indications, techniques and long-term results using the gamma knife. J Neurooncol 92:345–356 Wowra B, Stummer W (2002) Efficacy of gamma knife radiosurgery for nonfunctioning pituitary adenomas: a quantitative follow up with magnetic resonance imaging-based volumetric analysis. J Neurosurg 97:429–432 Mingione V, Yen CP, Vance ML, Steiner M, Sheehan J, Laws ER, Steiner L (2006) Gamma surgery in the treatment of nonsecretory pituitary macroadenoma. J Neurosurg 104:876–883
31. Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD (2002) Radiosurgery for residual or recurrent nonfunctioning pituitary adenoma. J Neurosurg 97:408–414 32. Losa M, Gioia L, Picozzi P, Franzin A, Valle M, Giovanelli M, Mortini P (2008) The role of stereotactic radiotherapy in patients with growth hormone-secreting pituitary adenoma. J Clin Endocrinol Metab 93:2546–2552 33. Higham CE, Chung TT, Lawrence J, Drake WM, Trainer PJ (2009) Long-term experience of pegvisomant therapy as a treatment for acromegaly. Clin Endocrinol (Oxf) 71:86–91 34. Jagannathan J, Sheehan JP, Pouratian N, Laws ER Jr, Steiner L, Vance ML (2008) Gamma knife radiosurgery for acromegaly: outcomes after failed transsphenoidal surgery. Neurosurgery 62:1262–1269; discussion 1269–1270
123
