CLINICAL STUDY
Prostate Perfusion Mapped by Technetium-99m Macroaggregated Albumin after Selective Arterial Injection Jonathan T. Abele, MD, Ronald Moore, MD, Wayne Tymchak, MD, and Richard J. Owen, MD
ABSTRACT Purpose: To determine if perfusion of the prostate can be mapped using technetium-99m (99mTc) macroaggregated albumin (MAA) after selective prostate artery catheterization. Materials and Methods: Selective prostate artery injections of MAA were performed and analyzed in 14 patients; 9 patients received unilateral injection, and 5 patients received bilateral injections (37 MBq/1 mCi per injection). Fused single-photon emission computed tomography/computed tomography (SPECT/CT) images were subsequently acquired using a fiducial marker technique. Perfusion distribution was assessed, and relative intraprostatic versus extraprostatic activity was quantified and compared between groups. Results: The percentage of the prostate gland containing activity was significantly greater for the bilateral injection group compared with the unilateral injection group (76.6% vs 44.3%, P o .05). The percentage of relative intraprostatic versus extraprostatic activity was significantly lower for the bilateral injection group compared with the unilateral injection group (40.3% vs 75.9%, P o .05). Sites of visualized extraprostatic activity included the seminal vesicles (8 of 14 patients), internal iliac vessels (7 of 14 patients), bladder wall (5 of 14 patients), space of Retzius (3 of 14 patients), rectal wall (3 of 14 patients), and penis (1 of 14 patients). Conclusions: Perfusion mapping with 99mTc-MAA can be effectively performed with SPECT/CT after selective prostate artery catheterization. The relative percentage of intraprostatic versus extraprostatic activity can be quantified, and the distribution of activity within and outside the prostate gland can be determined.
ABBREVIATIONS MAA = macroaggregated albumin, PSA = prostate-specific antigen, ROI = region of interest, SPECT = single-photon emission computed tomography, 99mTc = technetium-99m
Interest in selective arterial catheterization of the prostate artery has increased owing to the potential clinical applications, in particular, prostate artery embolization, as a minimally invasive method to treat benign prostatic hyperplasia (1–5). Another potential clinical application From the Departments of Radiology and Diagnostic Imaging (J.T.A., R.J.O.), Surgery, Division of Urology (R.M.), and Cardiology (W.T.), University of Alberta, 2A2.41 WMC, 8440 112 Street, Edmonton, Alberta, Canada, T6G 2B7. Received May 21, 2014; final revision received November 5, 2014; accepted November 7, 2014. Address correspondence to J.T.A.; E-mail: [email protected] R.M. is financially supported by the Mr. Lube Chair in Uro-Oncology Research. None of the other authors have identified a conflict of interest. & SIR, 2015 J Vasc Interv Radiol 2015; 26:418–425 http://dx.doi.org/10.1016/j.jvir.2014.11.018
of interest is vascular photodynamic therapy for localized prostate carcinoma (6–8). This technique currently uses systemic administration of photosensitizers; however, more selective administration to the prostate gland may result in greater efficacy and reduced systemic side effects. The blood supply to the prostate is classically described to be via the inferior vesical artery, a branch of the anterior division of the internal iliac artery (9). More recent reports using computed tomography (CT) angiography and digital subtraction angiography demonstrated the blood supply to the prostate to be variable among individuals and between sides in the same individual (5,10). On either pelvic side in an individual, the prostate gland may be supplied by either one or two prostate arteries. These arteries are quite variable in their origins, arising from the internal
Volume 26
’
Number 3
’
March
’
2015
pudendal artery, superior vesical artery, anterior common gluteal-pudendal trunk, obturator artery, or a common trunk with rectal branches. In addition, anastomoses with other arteries are often demonstrated including the internal pudendal arteries, ipsilateral and contralateral prostatic branches, rectal arteries, vesical arteries, and lateral accessory pudendal arteries (5,10). Because of this variability, meticulous technique and planning is recommended in performing prostate artery embolization. Complicated arterial anatomy may require alterations to the procedure, such as an increase in embolization particle size or downstream vascular coiling to protect other vascular territories. Prostate artery embolization is not recommended in cases in which the arterial anatomy is unsuitable or there are extensive atherosclerotic changes (3). The ability to map the perfusion at a given catheter position accurately before embolization would have great clinical benefit because of the complex and variable anatomy involved; the high rate of vascular anastomoses; and concerns regarding potential secondary ischemia to regional structures such as the bladder, colon, rectum, and seminal vesicles. A similar practice has been described before selective hepatic artery yttrium-90 microsphere embolization using technetium-99m (99mTc) macroaggregated albumin (MAA) (11,12). To our knowledge, the use of 99mTc-MAA to map perfusion of the prostate gland during selective prostate artery injection in humans has not been previously described. The primary objective of this study was to determine if perfusion of the prostate gland could be mapped using 99m Tc-MAA after selective prostate artery catheterization. Secondary objectives included determining the specificity of arterial cannulation (intraprostatic vs extraprostatic activity) and using clinical criteria to determine the safety of the MAA injection procedure.
MATERIALS AND METHODS Patients Patients undergoing cardiac catheterization as an outpatient for documentation or treatment of coronary vascular disease between September 2009 and April 2012 were approached for inclusion in the study. There were 17 patients (mean age, 68.5 y; range, 52–81 y) who met the inclusion and exclusion criteria for enrollment (Table 1) and provided informed consent to participate in the study. Because this was a feasibility study, the patients were not randomly assigned to treatment groups, and the angiographer was not blinded. The initial 10 consecutive patients underwent unilateral injection with the remaining patients undergoing bilateral injection. The study protocol was approved by the institutional ethics review board. A history and physical examination were performed before enrollment. Urine dipstick and blood tests were performed before the procedure,
419
Table 1 . Exclusion and Inclusion Criteria Inclusion Criteria
Exclusion Criteria
1. Male sex
1. Age o 17 y
2. Undergoing
2. Emergency coronary catheterization
outpatient diagnostic or interventional coronary angiography 3. Myocardial infarction o 3 mo 4. Congenital heart disease 5. Stroke o 3 mo 6. Untreated cancer 7. Coagulopathy 8. Increased risk of contrast-induced kidney injury 9. GFR o 50 mL/min/1.73 m2 10. Diabetes mellitus 11. Solitary kidney 12. Prior organ transplant 13. Expected contrast load 4 3 mL/kg 14. Inability to sign consent 15. Lack of femoral artery access (ie, radial artery approach to coronary catheterization) GFR ¼ glomerular filtration rate.
including measurement of prostate-specific antigen (PSA), complete blood count and differential, blood clotting factors (prothrombin time and international normalized ratio), and serum creatinine.
Previous Angiographer Experience All selective prostate artery injections were performed by a single interventional radiologist (14 years of experience in diagnostic and interventional angiography). The interventional radiologist had limited previous experience with selective prostate artery angiography including limited animal and human research studies and clinical experience in the treatment of bleeding complications after transurethral prostate resection. The interventional radiologist had no previous experience with prostate artery embolization.
Injection Technique After routine cardiac catheterization (performed by a cardiologist who was not one of the study investigators) with femoral arterial access in place, the patients were transferred to the angiography suite. The existing femoral vascular sheath was accessed, and a C2 RIM (Cook, Inc, Bloomington, Indiana) or SIM 2 (Cordis Corp, Bridgewater, New Jersey) catheter was used for selective catheterization and imaging of the contralateral or ipsilateral internal iliac artery. CT angiography was not performed before the procedure in these patients. Angiographic runs were carried out in multiple projections to
420
’
Prostate Perfusion Mapped by
99m
Tc-MAA
identify the prostatic arteries. Typically, ipsilateral anterior oblique projections (25–30 degrees) were used to identify the vessel origin from the anterior division of the internal iliac artery. When the prostate artery was identified, it was selectively catheterized using a 2.3-F RAPIDTRANSIT microcatheter (Codman Neuro, Raynham, Massachusetts) and a 0.014-inch Transend platinum guide wire (Boston Scientific, Marlborough, Massachusetts). Angiographic runs were carried out to confirm the catheter position and to confirm that the microcatheter was thought to lie within the prostate artery. Contrast material within the urinary bladder related to the previous coronary angiogram helped to localize the prostate gland. The catheter tip was positioned outside the expected position of the prostatic capsule and before the vessel divided within the gland. Where the prostate artery arose from another vessel, typically the obturator or inferior vesical artery, the catheter was advanced into the prostate artery origin. When a satisfactory selective prostate artery injection site had been obtained, 37 MBq (1 mCi) of 99mTc-MAA in 0.5–1 mL of saline was injected slowly by hand. In patients 1–10 (first phase), unilateral injections were attempted. In patients 11–17 (second phase), bilateral injections were attempted. One of the bilateral attempts was converted to a unilateral injection because of occlusion of one of the internal iliac arteries. Conebeam CT was not used in this study. After MAA injection, the diagnostic catheters were removed, the femoral sheath was removed with manual pressure, and hemostasis was achieved. The patients were sent for CT imaging of the pelvis and single-photon emission computed tomography (SPECT) imaging of the pelvis. When the imaging was completed, the patient received routine clinical care as provided after angiography.
Macroaggregated Albumin Each injection comprised 37 MBq (1 mCi) of 99mTcMAA in 0.5–1.0 mL of normal saline. The exact particle count injected varied depending on the day and time of injection. The MAA particles were calibrated at 1.2–1.4 million particles/GBq at 7:30 AM. Based on this calibration, the estimated range of total number of injected particles is calculated as 44,000–103,000 particles per injection, depending on the time of day. Patients with bilateral injections potentially received twice the number of particles as patients with a unilateral injection.
CT Imaging Next, the patient was immediately transferred to the CT imaging suite. Three radiopaque fiducial markers were placed on the skin of the patient overlying the right anterior superior iliac spine, the left anterior superior iliac spine, and the left greater trochanter. These were fixed to the skin and labeled with a small quantity (18 MBq) of 99mTc-pertechnetate or 99mTc-MAA.
Abele et al
’
JVIR
Non–contrast-enhanced CT imaging through the pelvis was performed including from a few centimeters above the most superior marker to a few centimeters below the most inferior marker. The scans were acquired on either a 64-slice or a 128-slice CT scanner (Siemens Healthcare, Erlangen, Germany) with a 512 512 matrix and 1-mm slice thickness at 1-mm slice intervals. The images were reconstructed in transverse, sagittal, and coronal planes using routine soft tissue and bone reconstruction kernels.
SPECT Imaging Immediately after CT imaging, the patient was transferred to the nuclear medicine department for SPECT imaging. SPECT images were acquired at a single bed position located at the pelvis with complete inclusion of the fiducial markers. SPECT imaging was performed using the following parameters: low-energy high-resolution collimator, dual-head gamma camera (Philips Healthcare, Best, The Netherlands), 128 angles (64 per head), 20second acquisition per angle, 128 128 matrix, zoom 1.0. All CT and SPECT images were acquired within 2 hours of the MAA injection for all patients.
Image Reconstruction and Evaluation All SPECT studies were reconstructed using OASIS software (Segami Corp, Columbia, MD). Reconstruction parameters included ordered subset expectation maximization three-dimensional iterative reconstruction with four iterations and eight subsets. No additional corrections were applied. All SPECT images were manually aligned with the CT images using fiducial marker alignment. A minimum threshold of 200 counts per voxel was used to define the region of interest (ROI) for the SPECT images. For intraprostatic activity, the ROI was confined to the prostate gland using the CT scan for definition. The total activity was calculated by widening the ROI to include the prostate plus all regional extraprostatic structures. The amount and volume of extraprostatic activity was determined by subtracting the two (total minus intraprostatic). Prostate volumes were determined by defining a three-dimensional ROI using the CT images. The reviewer was not blinded to the type of injection (unilateral vs bilateral); however, the same methodology was used for all patients.
Clinical Follow-up and Safety Monitoring Adverse events were recorded and categorized according to the following criteria: intensity or severity, expectedness, relatedness, outcome, treatment, or action taken. These were described as either minor or major complications using standard definitions (13). A data safety monitoring board was convened after the first five patients and again after the subsequent five patients before proceeding on to patients for bilateral prostate artery embolization (phase 2). The data safety
Volume 26
’
Number 3
’
March
’
2015
421
monitoring board was composed of three physicians (one urologist, one radiologist, one nuclear medicine physician) who were independent of the study. Followup PSA, urinalysis, and clinical examination were performed on all patients.
Statistics The continuous variables for each group expressed as the mean and SD were compared with a two-tailed Student t test (www.studentsttest.com). P values o .05 were considered significant.
RESULTS MAA Injection Angiography was performed in 17 patients on an intention-to-inject basis. In one patient, the prostate artery could not be identified on either side because of severe atherosclerosis. MAA was not injected in this patient, and this patient was excluded from analysis. In the remaining 16 patients, 5 bilateral injections and 11 unilateral injections were performed. Total dose of intravenous contrast material including the cardiac angiogram was 1.21–2.83 mL/kg (mean, 1.79 mL/kg; SD, 0.39) (OMNIPAQUE 240; GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom). Baseline blood work and urinalysis were normal in all study patients.
Perfusion Mapping Of the remaining 16 patients, 14 patients were analyzed. One patient was excluded because the fiducial markers
were inadvertently moved between the CT and SPECT studies and reliable fusion could not be performed. One patient was excluded because of data corruption of the SPECT raw data with resultant inability to process the SPECT study reliably. In one patient (patient no. 15) for whom a bilateral injection was attempted, only one prostate artery was cannulated. The contralateral prostate artery could not be identified, and this patient was converted to a unilateral injection. There were 9 unilateral injections and 5 bilateral injections (n ¼ 14). In two patients with unilateral injections (patient no. 7 and patient no. 11), all activity was localized outside of the prostate. Individual patient data are summarized in Table 2. For analysis purposes, the patients were evaluated in three groups: (a) unilateral injection group including all unilateral injections; (b) bilateral injection group including all bilateral injections; and (c) unilateral optimal injection subgroup, which excluded patient no. 7 and patient no. 11 in whom no prostatic activity was demonstrated (obvious missed injections) (Fig 1). The unilateral optimal subgroup analysis was performed to help account for the limited previous experience of the angiographer in selective prostate artery injection (an angiographer with more experience in selective prostate artery injection may not have any missed injections). The unilateral group had an average age of 65.7 years (SD, 9.0), the bilateral group had an average age of 61.6 years (SD, 7.8), and the unilateral optimal subgroup had an average age of 67.3 years (SD, 8.6). Statistical comparison of the unilateral and unilateral optimal groups with the bilateral group yielded P values 4 .05. The mean prostatic volumes based on the CT data measured
Table 2 . Patient Injection Data Unilateral or
Prostate with
Injected Activity
Sites of Extraprostatic
Age (y)
Bilateral Injection
Activity (%)
within Prostate (%)
Activity
3 4
58 81
U U
51.7 59.8
81.7 62.8
SV, IV SV, IV, BW
5
64
U
31.0
67.7
SV
6 7
61 52
U U
42.4 0
60.3 0
SV SR, BW
8
61
U
45.0
100.0
—
9 11
77 68
U U
41.3 0
58.7 0
BW IV, RW
15
68
U
38.7
100.0
—
12 13
72 67
B B
78.3 99.5
67.4 43.8
SV, IV IV, BW, RW, P
14
55
B
73.0
34.1
BW, SR, IV, SV
16 17
60 54
B B
48.5 83.7
37.3 19.0
SV, RW SV, IV, SR
Patient No.
Note.–Data from 14 patients included in analysis. Patient no. 1 was excluded from analysis because of an acquisition error. Patient no. 2 was excluded from analysis because of data corruption. Patient no. 10 was excluded from analysis because of inability to cannulate the prostatic arteries. Patient no. 15 was converted from a bilateral to a unilateral injection. B ¼ bilateral; BW ¼ bladder wall; IV ¼ internal iliac vessels; P ¼ penis; RW ¼ rectal wall; SR ¼ space of Retzius; SV ¼ seminal vesicle; U ¼ unilateral.
422
’
Prostate Perfusion Mapped by
99m
Tc-MAA
Abele et al
’
JVIR
(SD, 18.1) for the unilateral optimal injection group. In statistical comparison, there was a significant difference between the bilateral and unilateral optimal injection groups only (P o .05). The data and statistical analyses are summarized in Table 3.
Safety Data No minor or major complications were reported relating to the prostate artery MAA injection. One patient described difficulty voiding immediately after the angiogram, which he had experienced after a previous coronary angiogram; this was not thought to be related to the prostatic MAA injection procedure and required no therapy. The mean PSA before the procedure (1.66 [SD, 1.35]) was not significantly different from the mean PSA after the procedure (2.26 [SD, 1.70]; P 4 .05). The mean number of days between the PSA measurements before and after the procedure was 175.5 days (SD, 254.6).
DISCUSSION
Figure 1. Definition of unilateral group (U), bilateral group (B), and unilateral optimal subgroup (UO).
60.1 mL (SD, 24.3) for the unilateral group, 61.2 mL (SD, 19.6) for the bilateral group, and 65.5 mL (SD, 25.4) for the unilateral optimal group. Statistical comparison of the unilateral and unilateral optimal groups with the bilateral group yielded P values 4 .05. The percentage of the prostate gland containing activity (4 200 counts per voxel) was significantly greater for the bilateral injection group than the unilateral groups (P o .05). Mean percentages measured 76.6% for the bilateral injection group compared with 34.4% for the unilateral injection group and 44.3% for the unilateral optimal injection group. Only 2 of 14 patients demonstrated all injected activity within the prostate gland (Fig 2a–c). Both of these patients received unilateral injections. In the other patients, the sites of extraprostatic activity visualized included the seminal vesicles (8 of 14 patients), internal iliac vessels (7 of 14 patients), bladder wall (5 of 14 patients), space of Retzius (3 of 14 patients), rectal wall (3 of 14 patients), and penis (1 of 14 patients) (Figs 3a–c, 4a, b). The percentage of total injected activity localized within the prostate gland compared with extraprostatic sites was also compared. The unilateral injection group demonstrated 59.0% (SD, 36.7) of total injected activity within the prostate gland compared with 40.3% (SD, 17.7) for the bilateral injection group and 75.9%
This phase I feasibility study demonstrates that 99mTcMAA when selectively injected into the prostate artery can be used to map perfusion. By using SPECT/CT imaging to map the distribution of MAA, the ratio of MAA injected into the prostate gland compared with extraprostatic structures can be directly assessed. This assessment has potential clinical implications when considering prostatic embolization for benign prostatic hyperplasia or selective photodynamic therapy for prostate carcinoma. In particular, for patients with complex conditions, such as patients with atherosclerosis, this technique could be used to assess the distribution at a particular injection point before performing an actual therapeutic embolization. 99m Tc-MAA is currently widely used in a similar fashion before yttrium-90 microsphere hepatic artery embolization therapy for liver tumors. For these patients, the initial 99mTc-MAA perfusion map allows assessment of both intrahepatic arterial shunting and the distribution of extrahepatic activity at a particular injection point (12,14). Our study demonstrates that a similar technique is feasible in planning prostate therapy before embolization. As expected, our study demonstrates that a bilateral injection technique results in delivery of activity to a significantly larger percentage of the prostate gland than a unilateral injection technique. The bilateral injection technique resulted in activity within 76.6% of the prostate gland on average compared with 44.3% for the optimal unilateral injection group. This finding suggests that a bilateral injection may provide maximal effect when considering embolization therapy of the prostate gland.
Volume 26
’
Number 3
’
March
’
2015
423
Figure 2. Fused SPECT/CT images from patient no. 15 after a unilateral selective prostate artery injection of 99mTc-MAA (37 MBq/1 mCi). (a) Transverse image demonstrating activity within the prostate without evidence of extraprostatic activity (thick arrow). Focal activity is also seen at the skin surface related to a fiducial marker (thin arrow). (b) Coronal image demonstrating activity within the prostate without evidence of extraprostatic activity (thick arrow). A fiducial marker is seen at the skin surface (thin arrow). The urinary bladder contains iodinated contrast material from recently performed angiography. (c) Sagittal image demonstrating activity within the prostate gland (thick arrow), a fiducial marker at the skin surface (thin arrow), and iodinated contrast material within the urinary bladder.
Figure 3. Fused SPECT/CT images from patient no. 13 after a bilateral selective prostate artery injection of 99mTc-MAA (37 MBq/1 mCi each side). (a) Transverse image demonstrating activity within the prostate (thick arrow) and extraprostatic activity involving the rectal wall (arrow A). (b) Coronal image demonstrating activity within the prostate (thick arrow). (c) Sagittal image demonstrating activity within the prostate (thick arrow) and extraprostatic activity involving the rectal wall (arrow A) and bladder wall (arrow B).
Figure 4. Fused SPECT/CT images from patient no. 13 (same patient as Fig 3). (a) Transverse image demonstrating activity within the right seminal vesicle (arrow A). Some fiducial marker activity is seen at the anterior skin surface (arrow C). (b) Coronal image demonstrating activity within the right seminal vesicle (arrow A) and rectal wall (arrow B).
424
’
Prostate Perfusion Mapped by
99m
Tc-MAA
Abele et al
Table 3 . Statistical Comparisons between Groups U
B
UO
U vs B
UO vs B
65.7 61.6 67.3
P ¼ .40
P ¼ .26
P ¼ .96
P ¼ .75
34.4 76.6 44.3 P o .01
P ¼ .01
Age (y) Mean SD Prostate volume (mL) Mean
9.0
7.8
8.6
60.1 61.2 65.5
SD 24.3 19.6 25.4 Prostate with activity (%) Mean SD Activity in prostate (%)
21.1 18.6
9.3
Mean
59.0 40.3 75.9
SD
36.7 17.7 18.1
P ¼ .23
P o .01
B ¼ bilateral injection; U ¼ unilateral injection; UO ¼ unilateral optimal injection.
However, almost all patients demonstrated at least some activity outside of the prostate gland. The percentage of total activity outside of the prostate was significantly higher for the bilateral injection group (59.7%) compared with the optimal unilateral injection group (24.1%). This finding is understandable given that a unilateral injection is likely performed on the side with easier access, whereas a bilateral injection also requires an injection on the side with more difficult access. The clinical implications are uncertain but with respect to complications related to iatrogenic extraprostatic embolization, a unilateral injection may be safer than a bilateral one. Ultimately, performing an angiographic assessment with 99mTc-MAA injection and SPECT/CT imaging before performing therapeutic prostate embolization, similar to current practice before hepatic artery radioembolization procedures, may provide planning for maximal efficacy and safety. A limitation of the present study is a relative lack of experience of the angiographer in selective prostate artery catheterization. CT angiography and cone-beam CT were not performed before the procedure, and both of these techniques have been described to improve prostate artery selection (3,10,15). These limitations combined with relatively complex patients already being clinically evaluated for atherosclerotic disease may have resulted in suboptimal injections, especially compared with other more experienced centers. However, another potential benefit of the technique is that relatively inexperienced angiographers can use the 99mTc-MAA distribution to map the success of selective prostate artery catheterization protocols and techniques to gain experience. This benefit is of particular importance in patients with complex anatomy or marked atherosclerotic disease. Another limitation of this study is that the SPECT and CT images were acquired on separate scanners and fused manually using a fiducial marker technique. Although this technique has been used successfully for this type of
’
JVIR
fusion in the past, the more modern technique is to acquire SPECT/CT images using a hybrid SPECT/CT camera system. This limitation may have resulted in slight misregistration for some of the data in our study, although qualitatively the registration was optimized. Studies with obvious misregistration (ie, markers moved between the CT and SPECT acquisitions) were excluded from analysis. From this respect, the results remain valid; however, a modern hybrid SPECT/CT system may provide an easier imaging protocol and more robust registration. The overall study size analyzed was small (n = 14) with an uneven distribution (unilateral, n = 9; bilateral, n = 5). The original study design for this phase I feasibility study compared 10 patients with unilateral injections with 10 patients with bilateral injections. The overall enrollment rate was low, related particularly to increased use of radial artery access for coronary angiography during the study period at our institution. For practical reasons, the study was shortened to 5 patients in the bilateral group. Even with this limitation, the data fulfilled our primary study objective. The overall small patient numbers and asymmetric distribution make comparison between the groups less robust, and further evaluation in this regard is indicated. The decision to use a dose of 37 MBq (1 mCi) per injection was based on past experience in obtaining adequate SPECT images in a reasonable acquisition time using 99mTc for other applications (eg, sentinel lymph node scintigraphy). Because this study was a feasibility study, there was no specific evaluation regarding the optimal injection dose or MAA particle count for this application. The patients with bilateral injections potentially received twice the number of particles as the patients with a unilateral injection (depending on the time of day). It is assumed that there are double the number of capillaries available with a bilateral injection (overall similar particle-to-capillary ratio as the unilateral injection). This assumption is supported by previous animal (canine) research but has not been proven in humans, which is a limitation of our study (16). Further evaluation of prostate perfusion with different particle numbers would be valuable to determine how this might affect overall distribution. Finally, the distribution of activity within the prostate gland was not correlated with the vascular anatomy specifically, which is a limitation of the present study. For example, one of the patients was converted from a planned bilateral injection to a unilateral injection because on initial angiographic assessment one of the internal iliac arteries was occluded. This patient met the study criteria for unilateral injection and was included in this group. It is possible that chronic unilateral iliac occlusion may affect the prostate vascular distribution. All patients included in this study had a clinical suspicion for coronary atherosclerotic disease, and the presence or absence of atherosclerosis involving the
Volume 26
’
Number 3
’
March
’
2015
internal iliac arteries was not directly evaluated in our study. In conclusion, this study demonstrates the use of 99m Tc-MAA to map perfusion during selective prostate artery catheterization, analogous in technique to current practice in hepatic artery radioembolization planning. The analysis demonstrates that bilateral injections result in perfusion of a larger percentage of the prostate gland than unilateral injections but also result in a higher relative percentage of extraprostatic activity. This technique has the potential to improve the efficacy and safety of embolization therapy and other intravascular therapies (ie, photodynamic therapy), particularly in patients with complex vascular anatomy or advanced atherosclerosis. It also may be useful for angiographers with little experience in selective prostate artery catheterization to improve technique and training. Further evaluation of this technique with respect to benefit of integrating with actual clinical therapies would be valuable.
REFERENCES 1. Sun F, Sánchez FM, Crisótomo V, et al. Benign prostatic hyperplasia: transcatheter arterial embolization as potential treatment—preliminary study in pigs. Radiology 2008; 246:783–789. 2. Mauro MA. Science to practice: can hyperplastic prostate follow uterine fibroids and be managed with transcatheter arterial embolization? Radiology 2008; 246:657–658. 3. Pisco JM, Pereira J, Tinto HR, Fernandes L, Bilhim T. How to perform prostatic arterial embolization. Tech Vasc Interv Radiol 2012; 15:286–289.
425
4. Fernandes L, Tinto HR, Pereira J, Duarte M, Bilhim T, Pisco JM. Prostatic arterial embolization: post-procedure follow-up. Tech Vasc Interv Radiol 2012; 15:294–299. 5. Bilhim T, Pisco JM, Tinto HR, et al. Prostatic arterial supply: anatomic and imaging findings relevant for selective arterial embolization. J Vasc Interv Radiol 2012; 23:1403–1415. 6. Bozzini G, Colin P, Betrouni N, et al. Photodynamic therapy in urology: what can we do now and where are we heading? Photodiagnosis Photodyn Ther 2012; 9:261–273. 7. Arumainayagam N, Moore CM, Ahmed HU, Emberton M. Photodynamic therapy for focal ablation of the prostate. World J Surg 2010; 28: 571–576. 8. Nomura T, Mimata H. Focal therapy in the management of prostate cancer. Adv Urol 2012; 2012:391437. 9. Priest R, Garzotto M, Kaufman J. Benign prostatic hyperplasia: a brief overview of pathogenesis, diagnosis, and therapy. Tech Vasc Interv Radiol 2012; 15:261–264. 10. Bilhim T, Tinto HR, Fernandes L, Pisco JM. Radiological anatomy of prostatic arteries. Tech Vasc Interv Radiol 2012; 15:276–285. 11. Jiang M, Nowakowski FS, Wang J, et al. Characterization of extrahepatic distribution of Tc-99m macroaggregated albumin in hepatic perfusion imaging studies prior to yttrium-90 microsphere therapy. Cancer Biother Radiopharm 2011; 26:511–518. 12. Ahmadzadehfar H, Sabet A, Biermann K, et al. The significance of 99m Tc-MAA SPECT/CT liver perfusion imaging in treatment planning for 90 Y-microsphere selective internal radiation treatment. J Nucl Med 2010; 51:1206–1212. 13. Sacks D, McClenny TE, Cardella JF, Lewis CA. Society of Interventional Radiology clinical practice guidelines. J Vasc Interv Radiol 2003; 14: S199–S202. 14. Salem R, Lewandowski RJ, Gates VL, et al. Research reporting standards for radioembolization of hepatic malignancies. J Vasc Interv Radiol 2011; 22:265–278. 15. Bagla S, Rholl KS, Sterling KM, et al. Utility of cone-beam CT imaging in prostatic artery embolization. J Vasc Interv Radiol 2013; 24:1603–1607. 16. Moore RB, Xiao Z, Owen RJ, et al. Photodynamic therapy of the canine prostate: intra-arterial drug delivery. Cardiovasc Intervent Radiol 2008; 31: 164–176.
Prostate Perfusion Mapped by Technetium-99m Macroaggregated Albumin after Selective Arterial Injection Jonathan T. Abele, MD, Ronald Moore, MD, Wayne Tymchak, MD, and Richard J. Owen, MD
ABSTRACT Purpose: To determine if perfusion of the prostate can be mapped using technetium-99m (99mTc) macroaggregated albumin (MAA) after selective prostate artery catheterization. Materials and Methods: Selective prostate artery injections of MAA were performed and analyzed in 14 patients; 9 patients received unilateral injection, and 5 patients received bilateral injections (37 MBq/1 mCi per injection). Fused single-photon emission computed tomography/computed tomography (SPECT/CT) images were subsequently acquired using a fiducial marker technique. Perfusion distribution was assessed, and relative intraprostatic versus extraprostatic activity was quantified and compared between groups. Results: The percentage of the prostate gland containing activity was significantly greater for the bilateral injection group compared with the unilateral injection group (76.6% vs 44.3%, P o .05). The percentage of relative intraprostatic versus extraprostatic activity was significantly lower for the bilateral injection group compared with the unilateral injection group (40.3% vs 75.9%, P o .05). Sites of visualized extraprostatic activity included the seminal vesicles (8 of 14 patients), internal iliac vessels (7 of 14 patients), bladder wall (5 of 14 patients), space of Retzius (3 of 14 patients), rectal wall (3 of 14 patients), and penis (1 of 14 patients). Conclusions: Perfusion mapping with 99mTc-MAA can be effectively performed with SPECT/CT after selective prostate artery catheterization. The relative percentage of intraprostatic versus extraprostatic activity can be quantified, and the distribution of activity within and outside the prostate gland can be determined.
ABBREVIATIONS MAA = macroaggregated albumin, PSA = prostate-specific antigen, ROI = region of interest, SPECT = single-photon emission computed tomography, 99mTc = technetium-99m
Interest in selective arterial catheterization of the prostate artery has increased owing to the potential clinical applications, in particular, prostate artery embolization, as a minimally invasive method to treat benign prostatic hyperplasia (1–5). Another potential clinical application From the Departments of Radiology and Diagnostic Imaging (J.T.A., R.J.O.), Surgery, Division of Urology (R.M.), and Cardiology (W.T.), University of Alberta, 2A2.41 WMC, 8440 112 Street, Edmonton, Alberta, Canada, T6G 2B7. Received May 21, 2014; final revision received November 5, 2014; accepted November 7, 2014. Address correspondence to J.T.A.; E-mail: [email protected] R.M. is financially supported by the Mr. Lube Chair in Uro-Oncology Research. None of the other authors have identified a conflict of interest. & SIR, 2015 J Vasc Interv Radiol 2015; 26:418–425 http://dx.doi.org/10.1016/j.jvir.2014.11.018
of interest is vascular photodynamic therapy for localized prostate carcinoma (6–8). This technique currently uses systemic administration of photosensitizers; however, more selective administration to the prostate gland may result in greater efficacy and reduced systemic side effects. The blood supply to the prostate is classically described to be via the inferior vesical artery, a branch of the anterior division of the internal iliac artery (9). More recent reports using computed tomography (CT) angiography and digital subtraction angiography demonstrated the blood supply to the prostate to be variable among individuals and between sides in the same individual (5,10). On either pelvic side in an individual, the prostate gland may be supplied by either one or two prostate arteries. These arteries are quite variable in their origins, arising from the internal
Volume 26
’
Number 3
’
March
’
2015
pudendal artery, superior vesical artery, anterior common gluteal-pudendal trunk, obturator artery, or a common trunk with rectal branches. In addition, anastomoses with other arteries are often demonstrated including the internal pudendal arteries, ipsilateral and contralateral prostatic branches, rectal arteries, vesical arteries, and lateral accessory pudendal arteries (5,10). Because of this variability, meticulous technique and planning is recommended in performing prostate artery embolization. Complicated arterial anatomy may require alterations to the procedure, such as an increase in embolization particle size or downstream vascular coiling to protect other vascular territories. Prostate artery embolization is not recommended in cases in which the arterial anatomy is unsuitable or there are extensive atherosclerotic changes (3). The ability to map the perfusion at a given catheter position accurately before embolization would have great clinical benefit because of the complex and variable anatomy involved; the high rate of vascular anastomoses; and concerns regarding potential secondary ischemia to regional structures such as the bladder, colon, rectum, and seminal vesicles. A similar practice has been described before selective hepatic artery yttrium-90 microsphere embolization using technetium-99m (99mTc) macroaggregated albumin (MAA) (11,12). To our knowledge, the use of 99mTc-MAA to map perfusion of the prostate gland during selective prostate artery injection in humans has not been previously described. The primary objective of this study was to determine if perfusion of the prostate gland could be mapped using 99m Tc-MAA after selective prostate artery catheterization. Secondary objectives included determining the specificity of arterial cannulation (intraprostatic vs extraprostatic activity) and using clinical criteria to determine the safety of the MAA injection procedure.
MATERIALS AND METHODS Patients Patients undergoing cardiac catheterization as an outpatient for documentation or treatment of coronary vascular disease between September 2009 and April 2012 were approached for inclusion in the study. There were 17 patients (mean age, 68.5 y; range, 52–81 y) who met the inclusion and exclusion criteria for enrollment (Table 1) and provided informed consent to participate in the study. Because this was a feasibility study, the patients were not randomly assigned to treatment groups, and the angiographer was not blinded. The initial 10 consecutive patients underwent unilateral injection with the remaining patients undergoing bilateral injection. The study protocol was approved by the institutional ethics review board. A history and physical examination were performed before enrollment. Urine dipstick and blood tests were performed before the procedure,
419
Table 1 . Exclusion and Inclusion Criteria Inclusion Criteria
Exclusion Criteria
1. Male sex
1. Age o 17 y
2. Undergoing
2. Emergency coronary catheterization
outpatient diagnostic or interventional coronary angiography 3. Myocardial infarction o 3 mo 4. Congenital heart disease 5. Stroke o 3 mo 6. Untreated cancer 7. Coagulopathy 8. Increased risk of contrast-induced kidney injury 9. GFR o 50 mL/min/1.73 m2 10. Diabetes mellitus 11. Solitary kidney 12. Prior organ transplant 13. Expected contrast load 4 3 mL/kg 14. Inability to sign consent 15. Lack of femoral artery access (ie, radial artery approach to coronary catheterization) GFR ¼ glomerular filtration rate.
including measurement of prostate-specific antigen (PSA), complete blood count and differential, blood clotting factors (prothrombin time and international normalized ratio), and serum creatinine.
Previous Angiographer Experience All selective prostate artery injections were performed by a single interventional radiologist (14 years of experience in diagnostic and interventional angiography). The interventional radiologist had limited previous experience with selective prostate artery angiography including limited animal and human research studies and clinical experience in the treatment of bleeding complications after transurethral prostate resection. The interventional radiologist had no previous experience with prostate artery embolization.
Injection Technique After routine cardiac catheterization (performed by a cardiologist who was not one of the study investigators) with femoral arterial access in place, the patients were transferred to the angiography suite. The existing femoral vascular sheath was accessed, and a C2 RIM (Cook, Inc, Bloomington, Indiana) or SIM 2 (Cordis Corp, Bridgewater, New Jersey) catheter was used for selective catheterization and imaging of the contralateral or ipsilateral internal iliac artery. CT angiography was not performed before the procedure in these patients. Angiographic runs were carried out in multiple projections to
420
’
Prostate Perfusion Mapped by
99m
Tc-MAA
identify the prostatic arteries. Typically, ipsilateral anterior oblique projections (25–30 degrees) were used to identify the vessel origin from the anterior division of the internal iliac artery. When the prostate artery was identified, it was selectively catheterized using a 2.3-F RAPIDTRANSIT microcatheter (Codman Neuro, Raynham, Massachusetts) and a 0.014-inch Transend platinum guide wire (Boston Scientific, Marlborough, Massachusetts). Angiographic runs were carried out to confirm the catheter position and to confirm that the microcatheter was thought to lie within the prostate artery. Contrast material within the urinary bladder related to the previous coronary angiogram helped to localize the prostate gland. The catheter tip was positioned outside the expected position of the prostatic capsule and before the vessel divided within the gland. Where the prostate artery arose from another vessel, typically the obturator or inferior vesical artery, the catheter was advanced into the prostate artery origin. When a satisfactory selective prostate artery injection site had been obtained, 37 MBq (1 mCi) of 99mTc-MAA in 0.5–1 mL of saline was injected slowly by hand. In patients 1–10 (first phase), unilateral injections were attempted. In patients 11–17 (second phase), bilateral injections were attempted. One of the bilateral attempts was converted to a unilateral injection because of occlusion of one of the internal iliac arteries. Conebeam CT was not used in this study. After MAA injection, the diagnostic catheters were removed, the femoral sheath was removed with manual pressure, and hemostasis was achieved. The patients were sent for CT imaging of the pelvis and single-photon emission computed tomography (SPECT) imaging of the pelvis. When the imaging was completed, the patient received routine clinical care as provided after angiography.
Macroaggregated Albumin Each injection comprised 37 MBq (1 mCi) of 99mTcMAA in 0.5–1.0 mL of normal saline. The exact particle count injected varied depending on the day and time of injection. The MAA particles were calibrated at 1.2–1.4 million particles/GBq at 7:30 AM. Based on this calibration, the estimated range of total number of injected particles is calculated as 44,000–103,000 particles per injection, depending on the time of day. Patients with bilateral injections potentially received twice the number of particles as patients with a unilateral injection.
CT Imaging Next, the patient was immediately transferred to the CT imaging suite. Three radiopaque fiducial markers were placed on the skin of the patient overlying the right anterior superior iliac spine, the left anterior superior iliac spine, and the left greater trochanter. These were fixed to the skin and labeled with a small quantity (18 MBq) of 99mTc-pertechnetate or 99mTc-MAA.
Abele et al
’
JVIR
Non–contrast-enhanced CT imaging through the pelvis was performed including from a few centimeters above the most superior marker to a few centimeters below the most inferior marker. The scans were acquired on either a 64-slice or a 128-slice CT scanner (Siemens Healthcare, Erlangen, Germany) with a 512 512 matrix and 1-mm slice thickness at 1-mm slice intervals. The images were reconstructed in transverse, sagittal, and coronal planes using routine soft tissue and bone reconstruction kernels.
SPECT Imaging Immediately after CT imaging, the patient was transferred to the nuclear medicine department for SPECT imaging. SPECT images were acquired at a single bed position located at the pelvis with complete inclusion of the fiducial markers. SPECT imaging was performed using the following parameters: low-energy high-resolution collimator, dual-head gamma camera (Philips Healthcare, Best, The Netherlands), 128 angles (64 per head), 20second acquisition per angle, 128 128 matrix, zoom 1.0. All CT and SPECT images were acquired within 2 hours of the MAA injection for all patients.
Image Reconstruction and Evaluation All SPECT studies were reconstructed using OASIS software (Segami Corp, Columbia, MD). Reconstruction parameters included ordered subset expectation maximization three-dimensional iterative reconstruction with four iterations and eight subsets. No additional corrections were applied. All SPECT images were manually aligned with the CT images using fiducial marker alignment. A minimum threshold of 200 counts per voxel was used to define the region of interest (ROI) for the SPECT images. For intraprostatic activity, the ROI was confined to the prostate gland using the CT scan for definition. The total activity was calculated by widening the ROI to include the prostate plus all regional extraprostatic structures. The amount and volume of extraprostatic activity was determined by subtracting the two (total minus intraprostatic). Prostate volumes were determined by defining a three-dimensional ROI using the CT images. The reviewer was not blinded to the type of injection (unilateral vs bilateral); however, the same methodology was used for all patients.
Clinical Follow-up and Safety Monitoring Adverse events were recorded and categorized according to the following criteria: intensity or severity, expectedness, relatedness, outcome, treatment, or action taken. These were described as either minor or major complications using standard definitions (13). A data safety monitoring board was convened after the first five patients and again after the subsequent five patients before proceeding on to patients for bilateral prostate artery embolization (phase 2). The data safety
Volume 26
’
Number 3
’
March
’
2015
421
monitoring board was composed of three physicians (one urologist, one radiologist, one nuclear medicine physician) who were independent of the study. Followup PSA, urinalysis, and clinical examination were performed on all patients.
Statistics The continuous variables for each group expressed as the mean and SD were compared with a two-tailed Student t test (www.studentsttest.com). P values o .05 were considered significant.
RESULTS MAA Injection Angiography was performed in 17 patients on an intention-to-inject basis. In one patient, the prostate artery could not be identified on either side because of severe atherosclerosis. MAA was not injected in this patient, and this patient was excluded from analysis. In the remaining 16 patients, 5 bilateral injections and 11 unilateral injections were performed. Total dose of intravenous contrast material including the cardiac angiogram was 1.21–2.83 mL/kg (mean, 1.79 mL/kg; SD, 0.39) (OMNIPAQUE 240; GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom). Baseline blood work and urinalysis were normal in all study patients.
Perfusion Mapping Of the remaining 16 patients, 14 patients were analyzed. One patient was excluded because the fiducial markers
were inadvertently moved between the CT and SPECT studies and reliable fusion could not be performed. One patient was excluded because of data corruption of the SPECT raw data with resultant inability to process the SPECT study reliably. In one patient (patient no. 15) for whom a bilateral injection was attempted, only one prostate artery was cannulated. The contralateral prostate artery could not be identified, and this patient was converted to a unilateral injection. There were 9 unilateral injections and 5 bilateral injections (n ¼ 14). In two patients with unilateral injections (patient no. 7 and patient no. 11), all activity was localized outside of the prostate. Individual patient data are summarized in Table 2. For analysis purposes, the patients were evaluated in three groups: (a) unilateral injection group including all unilateral injections; (b) bilateral injection group including all bilateral injections; and (c) unilateral optimal injection subgroup, which excluded patient no. 7 and patient no. 11 in whom no prostatic activity was demonstrated (obvious missed injections) (Fig 1). The unilateral optimal subgroup analysis was performed to help account for the limited previous experience of the angiographer in selective prostate artery injection (an angiographer with more experience in selective prostate artery injection may not have any missed injections). The unilateral group had an average age of 65.7 years (SD, 9.0), the bilateral group had an average age of 61.6 years (SD, 7.8), and the unilateral optimal subgroup had an average age of 67.3 years (SD, 8.6). Statistical comparison of the unilateral and unilateral optimal groups with the bilateral group yielded P values 4 .05. The mean prostatic volumes based on the CT data measured
Table 2 . Patient Injection Data Unilateral or
Prostate with
Injected Activity
Sites of Extraprostatic
Age (y)
Bilateral Injection
Activity (%)
within Prostate (%)
Activity
3 4
58 81
U U
51.7 59.8
81.7 62.8
SV, IV SV, IV, BW
5
64
U
31.0
67.7
SV
6 7
61 52
U U
42.4 0
60.3 0
SV SR, BW
8
61
U
45.0
100.0
—
9 11
77 68
U U
41.3 0
58.7 0
BW IV, RW
15
68
U
38.7
100.0
—
12 13
72 67
B B
78.3 99.5
67.4 43.8
SV, IV IV, BW, RW, P
14
55
B
73.0
34.1
BW, SR, IV, SV
16 17
60 54
B B
48.5 83.7
37.3 19.0
SV, RW SV, IV, SR
Patient No.
Note.–Data from 14 patients included in analysis. Patient no. 1 was excluded from analysis because of an acquisition error. Patient no. 2 was excluded from analysis because of data corruption. Patient no. 10 was excluded from analysis because of inability to cannulate the prostatic arteries. Patient no. 15 was converted from a bilateral to a unilateral injection. B ¼ bilateral; BW ¼ bladder wall; IV ¼ internal iliac vessels; P ¼ penis; RW ¼ rectal wall; SR ¼ space of Retzius; SV ¼ seminal vesicle; U ¼ unilateral.
422
’
Prostate Perfusion Mapped by
99m
Tc-MAA
Abele et al
’
JVIR
(SD, 18.1) for the unilateral optimal injection group. In statistical comparison, there was a significant difference between the bilateral and unilateral optimal injection groups only (P o .05). The data and statistical analyses are summarized in Table 3.
Safety Data No minor or major complications were reported relating to the prostate artery MAA injection. One patient described difficulty voiding immediately after the angiogram, which he had experienced after a previous coronary angiogram; this was not thought to be related to the prostatic MAA injection procedure and required no therapy. The mean PSA before the procedure (1.66 [SD, 1.35]) was not significantly different from the mean PSA after the procedure (2.26 [SD, 1.70]; P 4 .05). The mean number of days between the PSA measurements before and after the procedure was 175.5 days (SD, 254.6).
DISCUSSION
Figure 1. Definition of unilateral group (U), bilateral group (B), and unilateral optimal subgroup (UO).
60.1 mL (SD, 24.3) for the unilateral group, 61.2 mL (SD, 19.6) for the bilateral group, and 65.5 mL (SD, 25.4) for the unilateral optimal group. Statistical comparison of the unilateral and unilateral optimal groups with the bilateral group yielded P values 4 .05. The percentage of the prostate gland containing activity (4 200 counts per voxel) was significantly greater for the bilateral injection group than the unilateral groups (P o .05). Mean percentages measured 76.6% for the bilateral injection group compared with 34.4% for the unilateral injection group and 44.3% for the unilateral optimal injection group. Only 2 of 14 patients demonstrated all injected activity within the prostate gland (Fig 2a–c). Both of these patients received unilateral injections. In the other patients, the sites of extraprostatic activity visualized included the seminal vesicles (8 of 14 patients), internal iliac vessels (7 of 14 patients), bladder wall (5 of 14 patients), space of Retzius (3 of 14 patients), rectal wall (3 of 14 patients), and penis (1 of 14 patients) (Figs 3a–c, 4a, b). The percentage of total injected activity localized within the prostate gland compared with extraprostatic sites was also compared. The unilateral injection group demonstrated 59.0% (SD, 36.7) of total injected activity within the prostate gland compared with 40.3% (SD, 17.7) for the bilateral injection group and 75.9%
This phase I feasibility study demonstrates that 99mTcMAA when selectively injected into the prostate artery can be used to map perfusion. By using SPECT/CT imaging to map the distribution of MAA, the ratio of MAA injected into the prostate gland compared with extraprostatic structures can be directly assessed. This assessment has potential clinical implications when considering prostatic embolization for benign prostatic hyperplasia or selective photodynamic therapy for prostate carcinoma. In particular, for patients with complex conditions, such as patients with atherosclerosis, this technique could be used to assess the distribution at a particular injection point before performing an actual therapeutic embolization. 99m Tc-MAA is currently widely used in a similar fashion before yttrium-90 microsphere hepatic artery embolization therapy for liver tumors. For these patients, the initial 99mTc-MAA perfusion map allows assessment of both intrahepatic arterial shunting and the distribution of extrahepatic activity at a particular injection point (12,14). Our study demonstrates that a similar technique is feasible in planning prostate therapy before embolization. As expected, our study demonstrates that a bilateral injection technique results in delivery of activity to a significantly larger percentage of the prostate gland than a unilateral injection technique. The bilateral injection technique resulted in activity within 76.6% of the prostate gland on average compared with 44.3% for the optimal unilateral injection group. This finding suggests that a bilateral injection may provide maximal effect when considering embolization therapy of the prostate gland.
Volume 26
’
Number 3
’
March
’
2015
423
Figure 2. Fused SPECT/CT images from patient no. 15 after a unilateral selective prostate artery injection of 99mTc-MAA (37 MBq/1 mCi). (a) Transverse image demonstrating activity within the prostate without evidence of extraprostatic activity (thick arrow). Focal activity is also seen at the skin surface related to a fiducial marker (thin arrow). (b) Coronal image demonstrating activity within the prostate without evidence of extraprostatic activity (thick arrow). A fiducial marker is seen at the skin surface (thin arrow). The urinary bladder contains iodinated contrast material from recently performed angiography. (c) Sagittal image demonstrating activity within the prostate gland (thick arrow), a fiducial marker at the skin surface (thin arrow), and iodinated contrast material within the urinary bladder.
Figure 3. Fused SPECT/CT images from patient no. 13 after a bilateral selective prostate artery injection of 99mTc-MAA (37 MBq/1 mCi each side). (a) Transverse image demonstrating activity within the prostate (thick arrow) and extraprostatic activity involving the rectal wall (arrow A). (b) Coronal image demonstrating activity within the prostate (thick arrow). (c) Sagittal image demonstrating activity within the prostate (thick arrow) and extraprostatic activity involving the rectal wall (arrow A) and bladder wall (arrow B).
Figure 4. Fused SPECT/CT images from patient no. 13 (same patient as Fig 3). (a) Transverse image demonstrating activity within the right seminal vesicle (arrow A). Some fiducial marker activity is seen at the anterior skin surface (arrow C). (b) Coronal image demonstrating activity within the right seminal vesicle (arrow A) and rectal wall (arrow B).
424
’
Prostate Perfusion Mapped by
99m
Tc-MAA
Abele et al
Table 3 . Statistical Comparisons between Groups U
B
UO
U vs B
UO vs B
65.7 61.6 67.3
P ¼ .40
P ¼ .26
P ¼ .96
P ¼ .75
34.4 76.6 44.3 P o .01
P ¼ .01
Age (y) Mean SD Prostate volume (mL) Mean
9.0
7.8
8.6
60.1 61.2 65.5
SD 24.3 19.6 25.4 Prostate with activity (%) Mean SD Activity in prostate (%)
21.1 18.6
9.3
Mean
59.0 40.3 75.9
SD
36.7 17.7 18.1
P ¼ .23
P o .01
B ¼ bilateral injection; U ¼ unilateral injection; UO ¼ unilateral optimal injection.
However, almost all patients demonstrated at least some activity outside of the prostate gland. The percentage of total activity outside of the prostate was significantly higher for the bilateral injection group (59.7%) compared with the optimal unilateral injection group (24.1%). This finding is understandable given that a unilateral injection is likely performed on the side with easier access, whereas a bilateral injection also requires an injection on the side with more difficult access. The clinical implications are uncertain but with respect to complications related to iatrogenic extraprostatic embolization, a unilateral injection may be safer than a bilateral one. Ultimately, performing an angiographic assessment with 99mTc-MAA injection and SPECT/CT imaging before performing therapeutic prostate embolization, similar to current practice before hepatic artery radioembolization procedures, may provide planning for maximal efficacy and safety. A limitation of the present study is a relative lack of experience of the angiographer in selective prostate artery catheterization. CT angiography and cone-beam CT were not performed before the procedure, and both of these techniques have been described to improve prostate artery selection (3,10,15). These limitations combined with relatively complex patients already being clinically evaluated for atherosclerotic disease may have resulted in suboptimal injections, especially compared with other more experienced centers. However, another potential benefit of the technique is that relatively inexperienced angiographers can use the 99mTc-MAA distribution to map the success of selective prostate artery catheterization protocols and techniques to gain experience. This benefit is of particular importance in patients with complex anatomy or marked atherosclerotic disease. Another limitation of this study is that the SPECT and CT images were acquired on separate scanners and fused manually using a fiducial marker technique. Although this technique has been used successfully for this type of
’
JVIR
fusion in the past, the more modern technique is to acquire SPECT/CT images using a hybrid SPECT/CT camera system. This limitation may have resulted in slight misregistration for some of the data in our study, although qualitatively the registration was optimized. Studies with obvious misregistration (ie, markers moved between the CT and SPECT acquisitions) were excluded from analysis. From this respect, the results remain valid; however, a modern hybrid SPECT/CT system may provide an easier imaging protocol and more robust registration. The overall study size analyzed was small (n = 14) with an uneven distribution (unilateral, n = 9; bilateral, n = 5). The original study design for this phase I feasibility study compared 10 patients with unilateral injections with 10 patients with bilateral injections. The overall enrollment rate was low, related particularly to increased use of radial artery access for coronary angiography during the study period at our institution. For practical reasons, the study was shortened to 5 patients in the bilateral group. Even with this limitation, the data fulfilled our primary study objective. The overall small patient numbers and asymmetric distribution make comparison between the groups less robust, and further evaluation in this regard is indicated. The decision to use a dose of 37 MBq (1 mCi) per injection was based on past experience in obtaining adequate SPECT images in a reasonable acquisition time using 99mTc for other applications (eg, sentinel lymph node scintigraphy). Because this study was a feasibility study, there was no specific evaluation regarding the optimal injection dose or MAA particle count for this application. The patients with bilateral injections potentially received twice the number of particles as the patients with a unilateral injection (depending on the time of day). It is assumed that there are double the number of capillaries available with a bilateral injection (overall similar particle-to-capillary ratio as the unilateral injection). This assumption is supported by previous animal (canine) research but has not been proven in humans, which is a limitation of our study (16). Further evaluation of prostate perfusion with different particle numbers would be valuable to determine how this might affect overall distribution. Finally, the distribution of activity within the prostate gland was not correlated with the vascular anatomy specifically, which is a limitation of the present study. For example, one of the patients was converted from a planned bilateral injection to a unilateral injection because on initial angiographic assessment one of the internal iliac arteries was occluded. This patient met the study criteria for unilateral injection and was included in this group. It is possible that chronic unilateral iliac occlusion may affect the prostate vascular distribution. All patients included in this study had a clinical suspicion for coronary atherosclerotic disease, and the presence or absence of atherosclerosis involving the
Volume 26
’
Number 3
’
March
’
2015
internal iliac arteries was not directly evaluated in our study. In conclusion, this study demonstrates the use of 99m Tc-MAA to map perfusion during selective prostate artery catheterization, analogous in technique to current practice in hepatic artery radioembolization planning. The analysis demonstrates that bilateral injections result in perfusion of a larger percentage of the prostate gland than unilateral injections but also result in a higher relative percentage of extraprostatic activity. This technique has the potential to improve the efficacy and safety of embolization therapy and other intravascular therapies (ie, photodynamic therapy), particularly in patients with complex vascular anatomy or advanced atherosclerosis. It also may be useful for angiographers with little experience in selective prostate artery catheterization to improve technique and training. Further evaluation of this technique with respect to benefit of integrating with actual clinical therapies would be valuable.
REFERENCES 1. Sun F, Sánchez FM, Crisótomo V, et al. Benign prostatic hyperplasia: transcatheter arterial embolization as potential treatment—preliminary study in pigs. Radiology 2008; 246:783–789. 2. Mauro MA. Science to practice: can hyperplastic prostate follow uterine fibroids and be managed with transcatheter arterial embolization? Radiology 2008; 246:657–658. 3. Pisco JM, Pereira J, Tinto HR, Fernandes L, Bilhim T. How to perform prostatic arterial embolization. Tech Vasc Interv Radiol 2012; 15:286–289.
425
4. Fernandes L, Tinto HR, Pereira J, Duarte M, Bilhim T, Pisco JM. Prostatic arterial embolization: post-procedure follow-up. Tech Vasc Interv Radiol 2012; 15:294–299. 5. Bilhim T, Pisco JM, Tinto HR, et al. Prostatic arterial supply: anatomic and imaging findings relevant for selective arterial embolization. J Vasc Interv Radiol 2012; 23:1403–1415. 6. Bozzini G, Colin P, Betrouni N, et al. Photodynamic therapy in urology: what can we do now and where are we heading? Photodiagnosis Photodyn Ther 2012; 9:261–273. 7. Arumainayagam N, Moore CM, Ahmed HU, Emberton M. Photodynamic therapy for focal ablation of the prostate. World J Surg 2010; 28: 571–576. 8. Nomura T, Mimata H. Focal therapy in the management of prostate cancer. Adv Urol 2012; 2012:391437. 9. Priest R, Garzotto M, Kaufman J. Benign prostatic hyperplasia: a brief overview of pathogenesis, diagnosis, and therapy. Tech Vasc Interv Radiol 2012; 15:261–264. 10. Bilhim T, Tinto HR, Fernandes L, Pisco JM. Radiological anatomy of prostatic arteries. Tech Vasc Interv Radiol 2012; 15:276–285. 11. Jiang M, Nowakowski FS, Wang J, et al. Characterization of extrahepatic distribution of Tc-99m macroaggregated albumin in hepatic perfusion imaging studies prior to yttrium-90 microsphere therapy. Cancer Biother Radiopharm 2011; 26:511–518. 12. Ahmadzadehfar H, Sabet A, Biermann K, et al. The significance of 99m Tc-MAA SPECT/CT liver perfusion imaging in treatment planning for 90 Y-microsphere selective internal radiation treatment. J Nucl Med 2010; 51:1206–1212. 13. Sacks D, McClenny TE, Cardella JF, Lewis CA. Society of Interventional Radiology clinical practice guidelines. J Vasc Interv Radiol 2003; 14: S199–S202. 14. Salem R, Lewandowski RJ, Gates VL, et al. Research reporting standards for radioembolization of hepatic malignancies. J Vasc Interv Radiol 2011; 22:265–278. 15. Bagla S, Rholl KS, Sterling KM, et al. Utility of cone-beam CT imaging in prostatic artery embolization. J Vasc Interv Radiol 2013; 24:1603–1607. 16. Moore RB, Xiao Z, Owen RJ, et al. Photodynamic therapy of the canine prostate: intra-arterial drug delivery. Cardiovasc Intervent Radiol 2008; 31: 164–176.
