International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Biology Contribution
Pudendal Nerve and Internal Pudendal Artery Damage May Contribute to Radiation-Induced Erectile Dysfunction Michael W. Nolan, DVM, PhD,*,y Angela J. Marolf, DVM,y E.J. Ehrhart, DVM, PhD,z Sangeeta Rao, BVSc, MVSc, PhD,x Susan L. Kraft, DVM, PhD,y Stephanie Engel, DVM, MS,x Hiroto Yoshikawa, DVM, PhD,y Anne E. Golden, AAS,y Todd H. Wasserman, MD,jj and Susan M. LaRue, DVM, PhDy *Department of Clinical Sciences, and Center for Comparative Medicine and Translational Research, North Carolina State University, Raleigh, North Carolina; Departments of yEnvironmental and Radiologic Health Sciences, zMicrobiology, Immunology and Pathology, and xClinical Sciences, Colorado State University, Fort Collins, Colorado; and jjDepartment of Radiation Oncology, Washington University, St. Louis, Missouri Received May 20, 2014, and in revised form Dec 8, 2014. Accepted for publication Dec 10, 2014.
Summary A canine model was developed to study radiationinduced erectile dysfunction (RI-ED). Stereotactic body radiation therapy caused slowing of motor nerve conduction velocities and axonal loss/degeneration in the pudendal nerve. Alterations in arterial tone were also noted in the internal pudendal artery and may explain why not all men with RI-ED respond to phosphodiesterase-5 inhibitors.
Purpose/Objectives: Erectile dysfunction is common after radiation therapy for prostate cancer; yet, the etiopathology of radiation-induced erectile dysfunction (RI-ED) remains poorly understood. A novel animal model was developed to study RI-ED, wherein stereotactic body radiation therapy (SBRT) was used to irradiate the prostate, neurovascular bundles (NVB), and penile bulb (PB) of dogs. The purpose was to describe vascular and neurogenic injuries after the irradiation of only the NVB or the PB, and after irradiation of all 3 sites (prostate, NVB, and PB) with varying doses of radiation. Methods and Materials: Dogs were treated with 50, 40, or 30 Gy to the prostate, NVB, and PB, or 50 Gy to either the NVB or the PB, by 5-fraction SBRT. Electrophysiologic studies of the pudendal nerve and bulbospongiosus muscles and ultrasound studies of pelvic perfusion were performed before and after SBRT. The results of these bioassays were correlated with histopathologic changes. Results: SBRT caused slowing of the systolic rise time, which corresponded to decreased arterial patency. Alterations in the response of the internal pudendal artery to vasoactive drugs were observed, wherein SBRT caused a paradoxical response to papaverine, slowing the systolic rise time after 40 and 50 Gy; these changes appeared
Reprint requests to: Dr Michael W. Nolan, DVM, PhD, 1052 William Moore Drive, Raleigh, NC 27613. Tel: (919) 513-6487; E-mail: [email protected] This work was partially supported by a sponsored research grant from Varian Medical Systems. Int J Radiation Oncol Biol Phys, Vol. 91, No. 4, pp. 796e806, 2015 0360-3016/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.12.025
Conflict of interest: none, except for the grant support provided by Varian Medical Systems. Supplementary material for this article can be found at www.redjournal.org.
Volume 91 Number 4 2015
Nerve/artery damage after prostatic SBRT
797
to have some dose dependency. The neurofilament content of penile nerves was also decreased at high doses and was more profound when the PB was irradiated than when the NVB was irradiated. These findings are coincident with slowing of motor nerve conduction velocities in the pudendal nerve after SBRT. Conclusions: This is the first report in which prostatic irradiation was shown to cause morphologic arterial damage that was coincident with altered internal pudendal arterial tone, and in which decreased motor function in the pudendal nerve was attributed to axonal degeneration and loss. Further investigation of the role played by damage to these structures in RI-ED is warranted. Ó 2015 Elsevier Inc. All rights reserved.
Introduction Erectile dysfunction (ED) is a common complication of treatment for localized prostate cancer. Although nervesparing prostatectomies have greatly reduced the risk of postoperative ED (1, 2), modern irradiation techniques such as intensity modulated radiation therapy (IMRT) and stereotactic body radiation therapy (SBRT) seem to have had little impact on the incidence of radiation-induced ED (RI-ED) (3, 4). Distilled to 4 necessities, normal penile erection requires functioning cavernous nerves, arterial inflow through the internal pudendal arteries, healthy erectile tissue that is capable of maintaining erection through venocclusion, and pudendal nerves capable of stimulating contraction of perineal musculature (which initiates the second phase of penile erection, and a final spike in intracorporeal pressures) (5). ED can occur if any of these components of penile erection fail. Erectile dysfunction is generally classified as psychogenic, neurogenic, endocrinologic, arteriogenic, or cavernosal (6). The causes and exact classification of RI-ED are unknown, although cavernosal (abnormal cavernosal distensibility), arteriogenic (low peak penile blood flow rates), and neurogenic (poor response to prostaglandin injections and histologic evidence of injury) dysfunction have all been associated with RI-ED (7-10). Irradiation of the bulbus penis, the posterolateral neurovascular bundles (which are composed of the venous plexus of Santorini, which drains the penis; branches of the internal pudendal artery that supply the prostate and penis; and the nerves originating from the pelvic plexus, which provide parasympathetic and sympathetic fibers to the prostate, urethra, seminal vesicles, and penile corpora cavernosa), the pudendal nerves, or some combination of these have been inconsistently implicated in RI-ED (11-14). Poor understanding of the mechanisms underlying the development of RI-ED is a major hindrance to the development of methods that may reduce the risk of its development. Historically, the only animal models available for studying RIED have involved delivery of single or multiple fractions of large pelvic radiation fields to rats, which were often associated with significant normal tissue toxicity (15-17). The rat model has recently been improved through the use of SBRT, which allows targeted irradiation of the prostate and has minimized apparent toxicosis. Despite improvements in the models available for investigation of the etiopathology of
RI-ED, several limitations still exist. For example, none of the rat irradiation protocols mimic clinical prostatic irradiation protocols. Also, apomorphine administration results in the activation of selective postsynaptic dopaminergic (D2) receptors, which in turn activate proerectile central neurologic pathways, involving nitric oxide (NO) signaling and ultimately inducing penile erection (18, 19). This system of inducing and measuring centrally mediated erections may not be as efficient at recapitulating the disturbances in erectile physiology that result in RI-ED in irradiated human prostate cancer patients as would a system involving measurement and study of locally induced penile erections. The purpose of this report is to describe the development of a complementary animal model. SBRT was used to irradiate the prostate, posterolateral prostatic neurovascular bundles (NVB), and penile bulb (PB) of dogs. Herein, we describe vascular and neurogenic injuries after the irradiation of only the NVB or PB and after irradiation of all 3 sites (prostate, NVB, and PB) with varying doses of radiation. This approach of geographic irradiations is rationalized by the fact that although the penile structure and function of dogs are similar to those of humans, there is conveniently a large separation between the prostate and penile bulb, facilitating irradiation of the prostate and NVB or PB without irradiating the other, thus allowing for evaluation of the role each of these tissues plays in producing RI-ED. Furthermore, the large size of dogs allows for the development of radiation fields similar in size and shape to those used for the treatment of prostate cancer in men, thereby enabling evaluation of radiation effects without the confounding volume effects that hinder comparison of rodent data with outcomes in human cancer patients. The rationale for studying effects at a variety of doses is that each functional and morphologic change induced by pelvic irradiation has a unique dose responsiveness, and better understanding of the magnitude of these changes induced within a range of doses clinically relevant to prostatic SBRT should help elucidate how each change contributes to clinical (or subclinical) erectile dysfunction.
Methods and Materials Animals Twenty-two sexually intact male-mixed breed hound dogs were used for the irradiation studies. All procedures
798
Nolan et al.
performed in these dogs were conducted in accordance with a protocol approved by the Institutional Animal Care and Use Committee. Additionally, tissues from 4 2-year-old to 4-year-old dogs of mixed breed were used as unirradiated controls for comparison of pathology data; these dogs were not subject to IACUC oversight for this project because they were killed as part of an unrelated study. Mixed-breed dogs were used rather than hounds because of their availability; this avoided the euthanizing of healthy unirradiated dogs for this study. Owing to similar ethical concerns in the use of large animals, unirradiated control animals were not used for the functional assays (eg ultrasound, electrophysiology); instead, preirradiation baseline data from each dog served as an internal control, allowing for comparison of temporal changes within individual animals. Control and treated dogs were of the same species and subspecies (Canis lupus familiaris), and it is unlikely that there were significant breed-associated differences in the histologic characteristics evaluated. Control dogs were matched to the ages and sizes of treated dogs as closely as possible. Dogs were subjected to erectile function evaluation training during the 2 to 6 weeks after a 2-week acclimation period. Training involved the use of rhythmic manual stimulation. Erection quality was quantified by size of erection (maximal transverse penile diameter at the level of the bulbus penis) and firmness of erection (subjectively graded: 0 Z lack of erection, 1 Z minimally firm erection with lack of turgidity, 2 Z moderate firmness, and 3 Z very firm and fully engorged). Training was complete upon the demonstration of repeatable induction of measurable erection with grade 3 firmness. Evaluations were repeated at least once monthly after the completion of radiation therapy. Dogs were deemed to have ED if they demonstrated a lack of change in diameter of the penis upon manual stimulation, accompanied by a firmness score 1; the findings had to be repeated twice, at least 1 week apart, in the absence of clinically identifiable pelvicoabdominal pain during physical examination, including digital rectal palpation of the prostate.
International Journal of Radiation Oncology Biology Physics
was represented by a 0.5-cm expansion of the prostatic and neurovascular bundle volumes, and a 0.7-cm expansion of the penile bulb volume; the rectal wall was excluded from the PTVs. The total prescribed radiation doses and treated tissue volumes varied (Table 1 and Fig. 1) but always involved delivery of 95% of the prescribed dose to the PTV in 5 equal fractions. Treatment groups were named (total dose in Gray/total treatment time in days/sites irradiated) to facilitate rapid recognition of the pertinent treatment characteristics of each irradiated dog. The doseevolume histogram from a representative subject (group 50/5d/all) in which all 3 PTVs were irradiated with 50 Gy is presented in Figure 1. Daily image-guided patient position verification was performed with on-board kilovoltage cone-beam computed tomography (20). After verification of position, intensity modulated SBRT was delivered using a Varian Trilogy linear accelerator (Varian Trilogy, Varian Medical Systems Inc, Palo Alto, CA).
Follow-up evaluations Endocrine evaluation, ultrasonography, and electrophysiologic evaluations were performed in most dogs before radiation therapy and repeated 4 months after completion of radiation therapy, and immediately before termination (for additional details on these experimental methods, see supplemental material available at www.redjournal.org). The dogs were monitored daily for toxicity. Toxicity was scored according to the Radiation Therapy Oncology Group urinary and rectal toxicity scale (21). The dogs were humanely killed within 3 weeks of reaching any of the following endpoints: (1) confirmation of erectile dysfunction; (2) development of clinically evident grade 3 or higher enterocolonic or genitourinary toxicity that was not responsive to aggressive medical therapy; or (3) 1 year after commencing radiation therapy. Postmortem examination was performed on all dogs, including gross necropsy, routine histopathology, histochemical evaluation of fibrosis, and immunohistochemical evaluation of changes in nerve and vessel morphology.
Radiation therapy Table 1
Radiation simulation and dose fractions were delivered with the dogs under general anesthesia; mean arterial blood pressure was maintained above 65 mm Hg. The dogs received a warm soapy enema 12 and 2 hours before anesthetic induction, and after induction, a sterile 8-Fr Foley urinary catheter was placed, as was a rectal balloon (Immobilizer Treatment Device [REF RB-100F]; RadiaDyne, Houston, TX) that was insufflated with 30 cc of air. The dogs were placed in dorsal recumbency in a foam trough. Computed tomography and magnetic resonance imaging were performed for radiation therapy planning. The prostate, bilateral prostatic neurovascular bundles (NVB) and penile bulb (PB) were delineated as the gross target volume (GTV). The planning target volume (PTV)
SBRT prescription information
Treated volume Total Total dose, Fractional treatment Treatment group n Gy dose, Gy time, days Prostate NVB PB 50/5d/all 50/11d/all 50/11d/ NVB 50/11d/PB 40/11d/all 30/11d/all Control
5 4 4
50 50 50
10 10 10
5 11 11
x x
x x x
x x
3 50 10 11 3 40 8 11 x x 3 30 6 11 x x 4 Unirradiated controls used for pathology/ histology studies
x x x
Abbreviations: NVB Z neurovascular bundles; PB Z penile bulb; SBRT Z stereotactic body radiation therapy.
Volume 91 Number 4 2015 Table 2
Nerve/artery damage after prostatic SBRT
Dose-volume limits for organs at risk
799
value of .05 was considered to determine statistical significance; a P value of .1 was also considered in interpretation of data to avoid committing type II errors. Pathology data were studied with the Kruskal-Wallis test; Dunn’s test was performed for multiple comparisons. Statistical software was used for all data analyses (SAS version 9.3, SAS Institute Inc, Cary, NC, and Prism 6 for Windows, GraphPad, La Jolla, CA).
Organ
Volume
Dose
Femoral heads Periprostatic anterior rectal wall Periprostatic lateral rectal wall
Less than 1 cc Maximum point dose
>30 Gy 105% of Rx
Maximum point dose Less than 0.3 cc cumulative (both sides) Maximum point dose
100% of Rx >90% of Rx
45% of Rx
Results
Maximum point dose Maximum point dose Less than 1 cc Maximum point dose Less than 0.8 cc Maximum point dose Maximum point dose Less than 1 cc
20 Gy 29 Gy >19.5 Gy 22 Gy >20 Gy 105% of Rx 105% of Rx >18.3 Gy
Not all dogs that were irradiated received complete endocrine, ultrasonographic, and electrophysiologic evaluations at each timepoint. Table E1 (available online at www. redjournal.org) summarizes which treatment group each dog belonged to, which data were successfully collected for each dog, and whether or not that dog experienced either ED or severe treatment-related complications.
Periprostatic posterior rectal wall Skin Small intestine Spinal cord Urethra Urinary bladder
Feasibility of a canine model of SBRT Statistical analysis A nonparametric approach was undertaken to perform comparisons between the categories of interest in the ultrasound variables. A Friedman test for nonparametric measures analysis of variance was used, which also accounted for repeated measurements on the same dog over time. Means were reported for description of the data. A P
Erectile function testing The initial prescription was based on the results of a recent phase I clinical trial, which demonstrated the short-term safety and efficacy of protocols delivering 50 Gy to the prostate in 5 fractions by linear acceleratorebased SBRT for low-risk to intermediate-risk prostate cancer in men (22). ED was observed in 2 of the 5 dogs treated in group
Fig. 1. Top, postcontrast computed tomographic images, with the various planning target volumes (PTVs) highlighted, with prostate (yellow), neurovascular bundle (NVB) (blue) and penile bulb (PB) (magenta). Left, parasagittal view of the caudal abdomen and pelvis with the red scout line referring to the image in the middle pane, and the green line corresponding to the right pane. Middle, transverse view of the caudal abdomen, at the level of the prostate gland. Right. transverse view of the perineum at the level of the PB. The regions included in the yellow, blue, and magenta PTVs were treated in groups 50/5d/all, 50/11d/all, 40/11d/all, and 30/11d/all; blue was treated in group 50/11d/NVB; magenta was treated in group 50/11d/PB. Bottom, A dose-volume histogram showing coverage of the various PTVs and low doses in surrounding organs for this dog (from group 50/5d/all).
800
Nolan et al.
50/5d/all; the time from the beginning of SBRT to the development of ED was 158 and 192 days. The 3 dogs not experiencing ED were killed 99, 104, and 151 days after the beginning of SBRT as a result of the development of grade IV colorectal toxicity. Because of the high rate of severe colorectal toxicity in this treatment group (50/5d/all), the fractionation schedule was altered for all remaining irradiations in such a manner that SBRT was delivered over 11 days, with a minimum interfraction interval of 48 hours. Two of 4 dogs treated in group 50/11d/all experienced ED, at 216 and 325 days, respectively. The remaining 2 dogs were killed at 125 and 396 days, with grade III and IV colorectal toxicity, respectively. Although increasing the duration of the interfraction interval decreased the incidence of severe colorectal toxicity, such events remained problematic. The next cohort of dogs (group 50/11d/NVB) received the same dosing scheme, but rather than irradiation of the entire prostate, the NVB, and the PB, only the NVB were treated. This reduced the volume of rectal tissue exposed to high-dose radiation, which was hypothesized to further reduce the risk of late colorectal toxicity. One of 4 dogs in this group (50/11d/NVB) experienced ED 105 days after SBRT. Two dogs in that group were killed on days 110 and 118 because of grade IV colorectal toxicity, and the fourth dog was killed on day 398 with grade III colorectal toxicity. The next cohort of dogs (group 50/11d/PB) was treated with 50 Gy (over 11 days), delivered only to the PB. None of the dogs in this group experienced ED; all were killed 1 year after SBRT without evidence of colorectal toxicity. Increasing the minimum duration of the interfraction interval from 24 to 48 hours improved complication rates, but the incidence of grade III and IV colorectal toxicity in dogs whose prostates, NVBs, or both were treated with 50 Gy was still too high to allow continued use of the canine model of RI-ED. The initial intent for the dose deescalation portion of the study was decrease dose increments of 10% (with regard to both fractional and total dose); however, concerns over the severity of colorectal effects at the initially prescribed dose necessitated deescalation in larger (20%) increments because of ethical concerns about animal welfare. No dogs in groups 40/11d/ all or 30/11d/all experienced measurable ED during the study period (1 year after completion of SBRT). No dog in either of these groups showed clinical evidence of colorectal toxicity. In summary, ED was observed in 5 of 22 (22.7%) irradiated dogs in this study. All 5 dogs with ED had 50 Gy delivered to either the prostate, the NVB, and the PB or only the NVB (groups 50/5d/all, 50/11d/all, and 50/11d/ NVB); of the dogs allocated to such treatment, the rate of ED was 38.5% (5 of 13). The median time to onset of RIED was 192 days (95% confidence interval, 119-265; range, 105-325); 4 of those 5 dogs experienced ED at least 5 months after irradiation. Conversely, the median survival time of dogs in the same treatment groups (groups 50/5d/
International Journal of Radiation Oncology Biology Physics
all, 50/11d/all, and 50/11d/NVB) that did not experience ED was 118 days (95% confidence interval 100-136 days; range, 99-398 days). Only 2 of those 7 dogs survived longer than 5 months; the remaining dogs were killed early because of severe colorectal toxicity. Endocrine testing No statistically significant alterations in serum concentrations of testosterone and luteinizing hormone were observed (Table E2, available online at www. redjournal.org).
Pudendal nerve function and internal pudendal arterial tone are altered by prostatic irradiation Electromyography of the bulbospongiosus muscle was performed before and after SBRT in the 2 dogs from group 50/5d/all that experienced clinically evident ED: 3 dogs in group 50/11d/all, 2 dogs in group 50/11d/NVB, and all 3 dogs in groups 50/11d/PB, 40/11d/all, and 30/11d/all. All of the dogs had a normal electromyogram before SBRT and showed spontaneous activity after SBRT. The majority of the abnormal activity was characterized by fibrillation potentials and positive sharp waves. There were no statistically detectable changes in motor nerve conduction velocity (MNCV) of the pudendal nerve with time within any of the treatment groups, nor were there differences between the groups at any point in time. The numeric values of MNCV at 12 months were, however, lower than baseline in both dogs from group 50/5d/all, 2 dogs in group 50/11d/all, and all 3 dogs in group 50/11d/PB. There was also a temporal decline in MNCV in all dogs in groups 40/11d/all and 30/ 11d/all, but the MNCV was higher than baseline in both dogs in group 50/11d/NVB, for whom serial measurements were made. The postirradiation compound muscle action potential (CMAP) amplitude of the pudendal nerve was lower than baseline in all cases from groups 50/5d/all and 50/11d/all. CMAP amplitudes were measured in 2 dogs from group 50/11d/NVB and were higher at the proximal and distal sites at the terminal examination than at baseline in 1 dog and lower in the other. Finally, there was a temporal decline in 2 of 3 dogs from group 40/11d/all and a temporal increase in 2 of 3 dogs from group 30/11d/all. Data describing the MNCV and CMAP amplitudes are summarized in Table E3 (available online at www. redjournal.org) and Figure 2, respectively. Sensory nerve action potentials were not elicited in any of the dogs. No cord dorsum potentials were successfully recorded in this set of dogs. Systolic rise times (SRT, briefly, the time measured from the start of systolic acceleration to the peak forward frequency on a Doppler waveform) that were measured in the internal pudendal artery before papaverine administration hastened after SBRT. Conversely, the rise times increased in response to papaverine after irradiation. Additional arterial blood velocity data collected by ultrasonography
Volume 91 Number 4 2015
Nerve/artery damage after prostatic SBRT
40
12 months post-SBRT
20
10
p A G ,p Gr rou rox o u p im p A, a l Gr B, p dist Gr ou rox al o u p B im p C , d al G , p ist Gr rou rox al o u p C im p , a D d l G , p ist Gr rou rox al o u p im p D, al E d Gr , p ist Gr oup rox al im ou p E, d al F i , Gr p sta ou ro l p xim F, a di l st al
0
Gr ou
There was lymphocytic cuffing of arteries, veins, and nerves in the prostate glands of several dogs from groups 50/5d/all and 50/11d/all. Medium-sized arteries were sclerotic, with prominent reorganizing thrombi, particularly in dogs killed more than 6 months after irradiation. Ganglia and nerves appeared to have increased interstitial collagen. Within the neurovascular plexuses in dogs from both groups, there was hyaline change in small arteries and arterioles, reactive perivascular fibrosis, widespread interstitial fibrosis, and atrophy of peripheral nerves, with moderate axon loss (Fig. 5). The glans penis was histologically normal in all dogs from groups 50/5d/all and 50/ 11d/all. There was marked axonal vacuolization in the Pre-Papaverine SRT as a Function of Time Post-SBRT Pre-SBRT
0.10 0.05
F up Gr o
p
E
D Gr ou
p
C Gr ou
p
p
B Gr ou
Gr ou
up
A
0.00 Gr o
0
Group C Group D Group E
-20
Group F
-40
5
10
15
Time (months)
Pathologic changes in nerves and vessels after SBRT
Systolic Rise Time (s)
Group B
0
are summarized in Figures 3 and 4 and also in Table E4 (available online at www.redjournal.org).
0.15
Group A 20
-60
Fig. 2. Compound muscle action potentials (CMAP) before and after stereotactic body radiation therapy (SBRT) at proximal and distal recording sites. Error bars represent the standard error of the mean.
0.20
40 Change in SRT after papaverine (%)
CMAP (mV)
Papaverine-induced change in SRT as a function of time post-SBRT
Baseline
30
801
Fig. 3. Ultrasonographic evaluation of prepapaverine systolic rise times in the internal pudendal artery. Error bars represent the standard error of the mean. SBRT Z stereotactic body radiation therapy.
Fig. 4. Ultrasonographic evaluation of the change in systolic rise times (SRT) in the internal pudendal artery after intracavernosal papaverine administration. SBRT Z stereotactic body radiation therapy. prostate of 1 dog from group 50/11d/NVB. The neurovascular bundles from dogs in group 50/11d/NVB displayed the same histologic abnormalities as those in groups 50/5d/all and 50/11d/all. No histologic abnormalities were noted in the penis of dogs from group 50/11d/ NVB. Nerves and vessels within the prostate and the NVB appeared normal in dogs from Group 50/11d/PB. Within the prostate of dogs from group 40/11d/all, there was moderate multifocal lymphocytic inflammation, with moderate perineural and perivascular lymphocytic cuffing. There was severe adventitial thickening, and intimal hypertrophy with subendothelial hemorrhage and occasional medial necrosis in small- and medium-sized arteries. Many of these vessels were also thrombosed, often with remodeling around the thrombi. The peripheral nerve bundles had increased amounts of interstitial collagen. The neurovascular bundles from dogs in group 40/5d/all were characterized by moderate interstitial fibrosis of the fibrofatty connective tissue, moderate increases in interstitial collagen within peripheral nerves and ganglia, and moderate vascular changes, including severe intimal proliferation and mild to moderate perivascular fibrosis of medium-sized arteries, multifocal lymphocytic cuffing of arteries and arterioles, and thrombosis and reorganization of several medium to large arteries. No histologic abnormalities were noted in the penis in group 40/11d/all. In the prostate of dogs from group 30/11d/all, nerves and vessels appeared normal. The neurovascular bundles and penis from dogs in this group were histologically unremarkable. A summary of outcomes and statistics from semiquantitative pathologic evaluations is presented in Table 3 Figure 5, and also in Figure E1 (available online at www.redjournal.org). Further statistical analysis of the percent collagen within nerves was performed for all treatment groups, to correct for differences in the time after SBRT at which dogs were killed. A multivariable linear regression model of ranks for mean values was used to evaluate the effects
802
Nolan et al.
International Journal of Radiation Oncology Biology Physics
Fig. 5. Top row, Masson’s trichrome-stained, paraffin-embedded section of prostatic nerves ( 200 magnification; each nerve is denoted by a central white asterisk) from a normal dog (left, untreated control) showing normal collagen content and axonal structure, and from a dog treated with 50 Gy (right, from group 50/11d/all) showing severe axonal vacuolization and degeneration. Bottom row. Photomicrographs ( 200 magnification) of representative paraffin-embedded penile nerves, prepared with neurofilament immunohistochemistry from an unirradiated dog (left, from dog in control group) and a dog treated with 50 Gy to the prostate, neurovascular bundle, and penile bulb (group 50/11d/all). Neurofilament appears brown (with nuclei staining blue); this example shows lower expression of neurofilament in the dog from group 50/11d/all, compared with that from the unirradiated dog. of treatment group and survival time. The results are presented in Tables E5 and E6 (available online at www.redjournal.org).
Discussion This report describes the development and feasibility of a canine model for studying the causes of RI-ED. SBRT was chosen rather than more conventional fractionated irradiation techniques because of the convenience of a severely hypofractionated protocol; SBRT has been shown to result in a similar incidence of RI-ED in comparison with conventional protocols (3). Pelvic irradiation may alter endocrine status and adversely affect erectile function. There were, however, no significant changes in circulating serum concentrations of either testosterone or luteinizing hormone in these dogs. Dogs in which only the PB was irradiated, and dogs that received a lower total dose (and dose per fraction) of radiation, experienced neither clinically appreciable colorectal toxicity nor measurable ED. This does not, however, imply that dogs in groups 50/11d/PB, 40/11d/all, and 30/ 11d/all had no evidence of subclinical ED; in fact, neurophysiologic and perfusion studies evidenced neurogenic
injuries, vascular injuries, or both in these dogs, which could contribute to RI-ED. Observed electromyographic changes in the bulbospongiosus muscles of all irradiated dogs in this study indicate hypersensitive, denervated myofibers, which can be observed with either primary muscle or nerve disease. To further investigate potential nerve damage, nerve conduction studies were performed. Motor nerve conduction studies of the pudendal nerve were performed before radiation treatment and repeated at the examinations 4 months and 1 year after treatment. Inasmuch as there are no established normal values for MNCV of the pudendal nerve in dogs, it was intended that the pretreatment values were to be used as comparison for the posttreatment values in each dog. However, CMAPs were not obtained at all timepoints in all of the dogs. It is likely that the reason for inadequate data collection was inappropriate technique, stemming from: (1) the stimulating electrode not being in close enough proximity to the pudendal nerve because it travels between the coccygeus and superficial gluteal muscles; and (2) placement of the recording electrode too far distal to the bulbospongiosus muscle. Sensory nerve conduction results and cord dorsum potentials were not successfully recorded (see supplementary materials, available online at www.redjournal.org).
Volume 91 Number 4 2015 Table 3
Nerve/artery damage after prostatic SBRT
803
Summary statistics for semiquantitative pathology data Median values, %
Factor Nerve density, prostatic capsule Neurofilament content, prostatic nerves Collagen content, prostatic nerves Neurofilament content, penile nerves Collagen content, penile nerves
Arterial patency, NVB
Statistics
50/5d/ All (A)
50/11d/ All (B)
50/11d/ NVB (C)
50/11d/ PB (D)
40/11d/ All (E)
30/11d/ All (F)
Control (0 Gy)
3.04
4.54
2.02
1.76
4.09
5.91
3.52
61.73
64.01
68.43
73.43
68.36
68.20
76.75
21.73
24.74
41.50
43.82
49.06
52.40
48.16
41.95
43.78
61.66
56.71
71.87
60.98
75.51
31.15
27.09
64.12
62.11
72.20
74.27
67.70
23.30
21.00
24.90
28.00
23.00
27.30
32.40
Statistically significant comparisons (P
Radiation Oncology biology
physics
www.redjournal.org
Biology Contribution
Pudendal Nerve and Internal Pudendal Artery Damage May Contribute to Radiation-Induced Erectile Dysfunction Michael W. Nolan, DVM, PhD,*,y Angela J. Marolf, DVM,y E.J. Ehrhart, DVM, PhD,z Sangeeta Rao, BVSc, MVSc, PhD,x Susan L. Kraft, DVM, PhD,y Stephanie Engel, DVM, MS,x Hiroto Yoshikawa, DVM, PhD,y Anne E. Golden, AAS,y Todd H. Wasserman, MD,jj and Susan M. LaRue, DVM, PhDy *Department of Clinical Sciences, and Center for Comparative Medicine and Translational Research, North Carolina State University, Raleigh, North Carolina; Departments of yEnvironmental and Radiologic Health Sciences, zMicrobiology, Immunology and Pathology, and xClinical Sciences, Colorado State University, Fort Collins, Colorado; and jjDepartment of Radiation Oncology, Washington University, St. Louis, Missouri Received May 20, 2014, and in revised form Dec 8, 2014. Accepted for publication Dec 10, 2014.
Summary A canine model was developed to study radiationinduced erectile dysfunction (RI-ED). Stereotactic body radiation therapy caused slowing of motor nerve conduction velocities and axonal loss/degeneration in the pudendal nerve. Alterations in arterial tone were also noted in the internal pudendal artery and may explain why not all men with RI-ED respond to phosphodiesterase-5 inhibitors.
Purpose/Objectives: Erectile dysfunction is common after radiation therapy for prostate cancer; yet, the etiopathology of radiation-induced erectile dysfunction (RI-ED) remains poorly understood. A novel animal model was developed to study RI-ED, wherein stereotactic body radiation therapy (SBRT) was used to irradiate the prostate, neurovascular bundles (NVB), and penile bulb (PB) of dogs. The purpose was to describe vascular and neurogenic injuries after the irradiation of only the NVB or the PB, and after irradiation of all 3 sites (prostate, NVB, and PB) with varying doses of radiation. Methods and Materials: Dogs were treated with 50, 40, or 30 Gy to the prostate, NVB, and PB, or 50 Gy to either the NVB or the PB, by 5-fraction SBRT. Electrophysiologic studies of the pudendal nerve and bulbospongiosus muscles and ultrasound studies of pelvic perfusion were performed before and after SBRT. The results of these bioassays were correlated with histopathologic changes. Results: SBRT caused slowing of the systolic rise time, which corresponded to decreased arterial patency. Alterations in the response of the internal pudendal artery to vasoactive drugs were observed, wherein SBRT caused a paradoxical response to papaverine, slowing the systolic rise time after 40 and 50 Gy; these changes appeared
Reprint requests to: Dr Michael W. Nolan, DVM, PhD, 1052 William Moore Drive, Raleigh, NC 27613. Tel: (919) 513-6487; E-mail: [email protected] This work was partially supported by a sponsored research grant from Varian Medical Systems. Int J Radiation Oncol Biol Phys, Vol. 91, No. 4, pp. 796e806, 2015 0360-3016/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.12.025
Conflict of interest: none, except for the grant support provided by Varian Medical Systems. Supplementary material for this article can be found at www.redjournal.org.
Volume 91 Number 4 2015
Nerve/artery damage after prostatic SBRT
797
to have some dose dependency. The neurofilament content of penile nerves was also decreased at high doses and was more profound when the PB was irradiated than when the NVB was irradiated. These findings are coincident with slowing of motor nerve conduction velocities in the pudendal nerve after SBRT. Conclusions: This is the first report in which prostatic irradiation was shown to cause morphologic arterial damage that was coincident with altered internal pudendal arterial tone, and in which decreased motor function in the pudendal nerve was attributed to axonal degeneration and loss. Further investigation of the role played by damage to these structures in RI-ED is warranted. Ó 2015 Elsevier Inc. All rights reserved.
Introduction Erectile dysfunction (ED) is a common complication of treatment for localized prostate cancer. Although nervesparing prostatectomies have greatly reduced the risk of postoperative ED (1, 2), modern irradiation techniques such as intensity modulated radiation therapy (IMRT) and stereotactic body radiation therapy (SBRT) seem to have had little impact on the incidence of radiation-induced ED (RI-ED) (3, 4). Distilled to 4 necessities, normal penile erection requires functioning cavernous nerves, arterial inflow through the internal pudendal arteries, healthy erectile tissue that is capable of maintaining erection through venocclusion, and pudendal nerves capable of stimulating contraction of perineal musculature (which initiates the second phase of penile erection, and a final spike in intracorporeal pressures) (5). ED can occur if any of these components of penile erection fail. Erectile dysfunction is generally classified as psychogenic, neurogenic, endocrinologic, arteriogenic, or cavernosal (6). The causes and exact classification of RI-ED are unknown, although cavernosal (abnormal cavernosal distensibility), arteriogenic (low peak penile blood flow rates), and neurogenic (poor response to prostaglandin injections and histologic evidence of injury) dysfunction have all been associated with RI-ED (7-10). Irradiation of the bulbus penis, the posterolateral neurovascular bundles (which are composed of the venous plexus of Santorini, which drains the penis; branches of the internal pudendal artery that supply the prostate and penis; and the nerves originating from the pelvic plexus, which provide parasympathetic and sympathetic fibers to the prostate, urethra, seminal vesicles, and penile corpora cavernosa), the pudendal nerves, or some combination of these have been inconsistently implicated in RI-ED (11-14). Poor understanding of the mechanisms underlying the development of RI-ED is a major hindrance to the development of methods that may reduce the risk of its development. Historically, the only animal models available for studying RIED have involved delivery of single or multiple fractions of large pelvic radiation fields to rats, which were often associated with significant normal tissue toxicity (15-17). The rat model has recently been improved through the use of SBRT, which allows targeted irradiation of the prostate and has minimized apparent toxicosis. Despite improvements in the models available for investigation of the etiopathology of
RI-ED, several limitations still exist. For example, none of the rat irradiation protocols mimic clinical prostatic irradiation protocols. Also, apomorphine administration results in the activation of selective postsynaptic dopaminergic (D2) receptors, which in turn activate proerectile central neurologic pathways, involving nitric oxide (NO) signaling and ultimately inducing penile erection (18, 19). This system of inducing and measuring centrally mediated erections may not be as efficient at recapitulating the disturbances in erectile physiology that result in RI-ED in irradiated human prostate cancer patients as would a system involving measurement and study of locally induced penile erections. The purpose of this report is to describe the development of a complementary animal model. SBRT was used to irradiate the prostate, posterolateral prostatic neurovascular bundles (NVB), and penile bulb (PB) of dogs. Herein, we describe vascular and neurogenic injuries after the irradiation of only the NVB or PB and after irradiation of all 3 sites (prostate, NVB, and PB) with varying doses of radiation. This approach of geographic irradiations is rationalized by the fact that although the penile structure and function of dogs are similar to those of humans, there is conveniently a large separation between the prostate and penile bulb, facilitating irradiation of the prostate and NVB or PB without irradiating the other, thus allowing for evaluation of the role each of these tissues plays in producing RI-ED. Furthermore, the large size of dogs allows for the development of radiation fields similar in size and shape to those used for the treatment of prostate cancer in men, thereby enabling evaluation of radiation effects without the confounding volume effects that hinder comparison of rodent data with outcomes in human cancer patients. The rationale for studying effects at a variety of doses is that each functional and morphologic change induced by pelvic irradiation has a unique dose responsiveness, and better understanding of the magnitude of these changes induced within a range of doses clinically relevant to prostatic SBRT should help elucidate how each change contributes to clinical (or subclinical) erectile dysfunction.
Methods and Materials Animals Twenty-two sexually intact male-mixed breed hound dogs were used for the irradiation studies. All procedures
798
Nolan et al.
performed in these dogs were conducted in accordance with a protocol approved by the Institutional Animal Care and Use Committee. Additionally, tissues from 4 2-year-old to 4-year-old dogs of mixed breed were used as unirradiated controls for comparison of pathology data; these dogs were not subject to IACUC oversight for this project because they were killed as part of an unrelated study. Mixed-breed dogs were used rather than hounds because of their availability; this avoided the euthanizing of healthy unirradiated dogs for this study. Owing to similar ethical concerns in the use of large animals, unirradiated control animals were not used for the functional assays (eg ultrasound, electrophysiology); instead, preirradiation baseline data from each dog served as an internal control, allowing for comparison of temporal changes within individual animals. Control and treated dogs were of the same species and subspecies (Canis lupus familiaris), and it is unlikely that there were significant breed-associated differences in the histologic characteristics evaluated. Control dogs were matched to the ages and sizes of treated dogs as closely as possible. Dogs were subjected to erectile function evaluation training during the 2 to 6 weeks after a 2-week acclimation period. Training involved the use of rhythmic manual stimulation. Erection quality was quantified by size of erection (maximal transverse penile diameter at the level of the bulbus penis) and firmness of erection (subjectively graded: 0 Z lack of erection, 1 Z minimally firm erection with lack of turgidity, 2 Z moderate firmness, and 3 Z very firm and fully engorged). Training was complete upon the demonstration of repeatable induction of measurable erection with grade 3 firmness. Evaluations were repeated at least once monthly after the completion of radiation therapy. Dogs were deemed to have ED if they demonstrated a lack of change in diameter of the penis upon manual stimulation, accompanied by a firmness score 1; the findings had to be repeated twice, at least 1 week apart, in the absence of clinically identifiable pelvicoabdominal pain during physical examination, including digital rectal palpation of the prostate.
International Journal of Radiation Oncology Biology Physics
was represented by a 0.5-cm expansion of the prostatic and neurovascular bundle volumes, and a 0.7-cm expansion of the penile bulb volume; the rectal wall was excluded from the PTVs. The total prescribed radiation doses and treated tissue volumes varied (Table 1 and Fig. 1) but always involved delivery of 95% of the prescribed dose to the PTV in 5 equal fractions. Treatment groups were named (total dose in Gray/total treatment time in days/sites irradiated) to facilitate rapid recognition of the pertinent treatment characteristics of each irradiated dog. The doseevolume histogram from a representative subject (group 50/5d/all) in which all 3 PTVs were irradiated with 50 Gy is presented in Figure 1. Daily image-guided patient position verification was performed with on-board kilovoltage cone-beam computed tomography (20). After verification of position, intensity modulated SBRT was delivered using a Varian Trilogy linear accelerator (Varian Trilogy, Varian Medical Systems Inc, Palo Alto, CA).
Follow-up evaluations Endocrine evaluation, ultrasonography, and electrophysiologic evaluations were performed in most dogs before radiation therapy and repeated 4 months after completion of radiation therapy, and immediately before termination (for additional details on these experimental methods, see supplemental material available at www.redjournal.org). The dogs were monitored daily for toxicity. Toxicity was scored according to the Radiation Therapy Oncology Group urinary and rectal toxicity scale (21). The dogs were humanely killed within 3 weeks of reaching any of the following endpoints: (1) confirmation of erectile dysfunction; (2) development of clinically evident grade 3 or higher enterocolonic or genitourinary toxicity that was not responsive to aggressive medical therapy; or (3) 1 year after commencing radiation therapy. Postmortem examination was performed on all dogs, including gross necropsy, routine histopathology, histochemical evaluation of fibrosis, and immunohistochemical evaluation of changes in nerve and vessel morphology.
Radiation therapy Table 1
Radiation simulation and dose fractions were delivered with the dogs under general anesthesia; mean arterial blood pressure was maintained above 65 mm Hg. The dogs received a warm soapy enema 12 and 2 hours before anesthetic induction, and after induction, a sterile 8-Fr Foley urinary catheter was placed, as was a rectal balloon (Immobilizer Treatment Device [REF RB-100F]; RadiaDyne, Houston, TX) that was insufflated with 30 cc of air. The dogs were placed in dorsal recumbency in a foam trough. Computed tomography and magnetic resonance imaging were performed for radiation therapy planning. The prostate, bilateral prostatic neurovascular bundles (NVB) and penile bulb (PB) were delineated as the gross target volume (GTV). The planning target volume (PTV)
SBRT prescription information
Treated volume Total Total dose, Fractional treatment Treatment group n Gy dose, Gy time, days Prostate NVB PB 50/5d/all 50/11d/all 50/11d/ NVB 50/11d/PB 40/11d/all 30/11d/all Control
5 4 4
50 50 50
10 10 10
5 11 11
x x
x x x
x x
3 50 10 11 3 40 8 11 x x 3 30 6 11 x x 4 Unirradiated controls used for pathology/ histology studies
x x x
Abbreviations: NVB Z neurovascular bundles; PB Z penile bulb; SBRT Z stereotactic body radiation therapy.
Volume 91 Number 4 2015 Table 2
Nerve/artery damage after prostatic SBRT
Dose-volume limits for organs at risk
799
value of .05 was considered to determine statistical significance; a P value of .1 was also considered in interpretation of data to avoid committing type II errors. Pathology data were studied with the Kruskal-Wallis test; Dunn’s test was performed for multiple comparisons. Statistical software was used for all data analyses (SAS version 9.3, SAS Institute Inc, Cary, NC, and Prism 6 for Windows, GraphPad, La Jolla, CA).
Organ
Volume
Dose
Femoral heads Periprostatic anterior rectal wall Periprostatic lateral rectal wall
Less than 1 cc Maximum point dose
>30 Gy 105% of Rx
Maximum point dose Less than 0.3 cc cumulative (both sides) Maximum point dose
100% of Rx >90% of Rx
45% of Rx
Results
Maximum point dose Maximum point dose Less than 1 cc Maximum point dose Less than 0.8 cc Maximum point dose Maximum point dose Less than 1 cc
20 Gy 29 Gy >19.5 Gy 22 Gy >20 Gy 105% of Rx 105% of Rx >18.3 Gy
Not all dogs that were irradiated received complete endocrine, ultrasonographic, and electrophysiologic evaluations at each timepoint. Table E1 (available online at www. redjournal.org) summarizes which treatment group each dog belonged to, which data were successfully collected for each dog, and whether or not that dog experienced either ED or severe treatment-related complications.
Periprostatic posterior rectal wall Skin Small intestine Spinal cord Urethra Urinary bladder
Feasibility of a canine model of SBRT Statistical analysis A nonparametric approach was undertaken to perform comparisons between the categories of interest in the ultrasound variables. A Friedman test for nonparametric measures analysis of variance was used, which also accounted for repeated measurements on the same dog over time. Means were reported for description of the data. A P
Erectile function testing The initial prescription was based on the results of a recent phase I clinical trial, which demonstrated the short-term safety and efficacy of protocols delivering 50 Gy to the prostate in 5 fractions by linear acceleratorebased SBRT for low-risk to intermediate-risk prostate cancer in men (22). ED was observed in 2 of the 5 dogs treated in group
Fig. 1. Top, postcontrast computed tomographic images, with the various planning target volumes (PTVs) highlighted, with prostate (yellow), neurovascular bundle (NVB) (blue) and penile bulb (PB) (magenta). Left, parasagittal view of the caudal abdomen and pelvis with the red scout line referring to the image in the middle pane, and the green line corresponding to the right pane. Middle, transverse view of the caudal abdomen, at the level of the prostate gland. Right. transverse view of the perineum at the level of the PB. The regions included in the yellow, blue, and magenta PTVs were treated in groups 50/5d/all, 50/11d/all, 40/11d/all, and 30/11d/all; blue was treated in group 50/11d/NVB; magenta was treated in group 50/11d/PB. Bottom, A dose-volume histogram showing coverage of the various PTVs and low doses in surrounding organs for this dog (from group 50/5d/all).
800
Nolan et al.
50/5d/all; the time from the beginning of SBRT to the development of ED was 158 and 192 days. The 3 dogs not experiencing ED were killed 99, 104, and 151 days after the beginning of SBRT as a result of the development of grade IV colorectal toxicity. Because of the high rate of severe colorectal toxicity in this treatment group (50/5d/all), the fractionation schedule was altered for all remaining irradiations in such a manner that SBRT was delivered over 11 days, with a minimum interfraction interval of 48 hours. Two of 4 dogs treated in group 50/11d/all experienced ED, at 216 and 325 days, respectively. The remaining 2 dogs were killed at 125 and 396 days, with grade III and IV colorectal toxicity, respectively. Although increasing the duration of the interfraction interval decreased the incidence of severe colorectal toxicity, such events remained problematic. The next cohort of dogs (group 50/11d/NVB) received the same dosing scheme, but rather than irradiation of the entire prostate, the NVB, and the PB, only the NVB were treated. This reduced the volume of rectal tissue exposed to high-dose radiation, which was hypothesized to further reduce the risk of late colorectal toxicity. One of 4 dogs in this group (50/11d/NVB) experienced ED 105 days after SBRT. Two dogs in that group were killed on days 110 and 118 because of grade IV colorectal toxicity, and the fourth dog was killed on day 398 with grade III colorectal toxicity. The next cohort of dogs (group 50/11d/PB) was treated with 50 Gy (over 11 days), delivered only to the PB. None of the dogs in this group experienced ED; all were killed 1 year after SBRT without evidence of colorectal toxicity. Increasing the minimum duration of the interfraction interval from 24 to 48 hours improved complication rates, but the incidence of grade III and IV colorectal toxicity in dogs whose prostates, NVBs, or both were treated with 50 Gy was still too high to allow continued use of the canine model of RI-ED. The initial intent for the dose deescalation portion of the study was decrease dose increments of 10% (with regard to both fractional and total dose); however, concerns over the severity of colorectal effects at the initially prescribed dose necessitated deescalation in larger (20%) increments because of ethical concerns about animal welfare. No dogs in groups 40/11d/ all or 30/11d/all experienced measurable ED during the study period (1 year after completion of SBRT). No dog in either of these groups showed clinical evidence of colorectal toxicity. In summary, ED was observed in 5 of 22 (22.7%) irradiated dogs in this study. All 5 dogs with ED had 50 Gy delivered to either the prostate, the NVB, and the PB or only the NVB (groups 50/5d/all, 50/11d/all, and 50/11d/ NVB); of the dogs allocated to such treatment, the rate of ED was 38.5% (5 of 13). The median time to onset of RIED was 192 days (95% confidence interval, 119-265; range, 105-325); 4 of those 5 dogs experienced ED at least 5 months after irradiation. Conversely, the median survival time of dogs in the same treatment groups (groups 50/5d/
International Journal of Radiation Oncology Biology Physics
all, 50/11d/all, and 50/11d/NVB) that did not experience ED was 118 days (95% confidence interval 100-136 days; range, 99-398 days). Only 2 of those 7 dogs survived longer than 5 months; the remaining dogs were killed early because of severe colorectal toxicity. Endocrine testing No statistically significant alterations in serum concentrations of testosterone and luteinizing hormone were observed (Table E2, available online at www. redjournal.org).
Pudendal nerve function and internal pudendal arterial tone are altered by prostatic irradiation Electromyography of the bulbospongiosus muscle was performed before and after SBRT in the 2 dogs from group 50/5d/all that experienced clinically evident ED: 3 dogs in group 50/11d/all, 2 dogs in group 50/11d/NVB, and all 3 dogs in groups 50/11d/PB, 40/11d/all, and 30/11d/all. All of the dogs had a normal electromyogram before SBRT and showed spontaneous activity after SBRT. The majority of the abnormal activity was characterized by fibrillation potentials and positive sharp waves. There were no statistically detectable changes in motor nerve conduction velocity (MNCV) of the pudendal nerve with time within any of the treatment groups, nor were there differences between the groups at any point in time. The numeric values of MNCV at 12 months were, however, lower than baseline in both dogs from group 50/5d/all, 2 dogs in group 50/11d/all, and all 3 dogs in group 50/11d/PB. There was also a temporal decline in MNCV in all dogs in groups 40/11d/all and 30/ 11d/all, but the MNCV was higher than baseline in both dogs in group 50/11d/NVB, for whom serial measurements were made. The postirradiation compound muscle action potential (CMAP) amplitude of the pudendal nerve was lower than baseline in all cases from groups 50/5d/all and 50/11d/all. CMAP amplitudes were measured in 2 dogs from group 50/11d/NVB and were higher at the proximal and distal sites at the terminal examination than at baseline in 1 dog and lower in the other. Finally, there was a temporal decline in 2 of 3 dogs from group 40/11d/all and a temporal increase in 2 of 3 dogs from group 30/11d/all. Data describing the MNCV and CMAP amplitudes are summarized in Table E3 (available online at www. redjournal.org) and Figure 2, respectively. Sensory nerve action potentials were not elicited in any of the dogs. No cord dorsum potentials were successfully recorded in this set of dogs. Systolic rise times (SRT, briefly, the time measured from the start of systolic acceleration to the peak forward frequency on a Doppler waveform) that were measured in the internal pudendal artery before papaverine administration hastened after SBRT. Conversely, the rise times increased in response to papaverine after irradiation. Additional arterial blood velocity data collected by ultrasonography
Volume 91 Number 4 2015
Nerve/artery damage after prostatic SBRT
40
12 months post-SBRT
20
10
p A G ,p Gr rou rox o u p im p A, a l Gr B, p dist Gr ou rox al o u p B im p C , d al G , p ist Gr rou rox al o u p C im p , a D d l G , p ist Gr rou rox al o u p im p D, al E d Gr , p ist Gr oup rox al im ou p E, d al F i , Gr p sta ou ro l p xim F, a di l st al
0
Gr ou
There was lymphocytic cuffing of arteries, veins, and nerves in the prostate glands of several dogs from groups 50/5d/all and 50/11d/all. Medium-sized arteries were sclerotic, with prominent reorganizing thrombi, particularly in dogs killed more than 6 months after irradiation. Ganglia and nerves appeared to have increased interstitial collagen. Within the neurovascular plexuses in dogs from both groups, there was hyaline change in small arteries and arterioles, reactive perivascular fibrosis, widespread interstitial fibrosis, and atrophy of peripheral nerves, with moderate axon loss (Fig. 5). The glans penis was histologically normal in all dogs from groups 50/5d/all and 50/ 11d/all. There was marked axonal vacuolization in the Pre-Papaverine SRT as a Function of Time Post-SBRT Pre-SBRT
0.10 0.05
F up Gr o
p
E
D Gr ou
p
C Gr ou
p
p
B Gr ou
Gr ou
up
A
0.00 Gr o
0
Group C Group D Group E
-20
Group F
-40
5
10
15
Time (months)
Pathologic changes in nerves and vessels after SBRT
Systolic Rise Time (s)
Group B
0
are summarized in Figures 3 and 4 and also in Table E4 (available online at www.redjournal.org).
0.15
Group A 20
-60
Fig. 2. Compound muscle action potentials (CMAP) before and after stereotactic body radiation therapy (SBRT) at proximal and distal recording sites. Error bars represent the standard error of the mean.
0.20
40 Change in SRT after papaverine (%)
CMAP (mV)
Papaverine-induced change in SRT as a function of time post-SBRT
Baseline
30
801
Fig. 3. Ultrasonographic evaluation of prepapaverine systolic rise times in the internal pudendal artery. Error bars represent the standard error of the mean. SBRT Z stereotactic body radiation therapy.
Fig. 4. Ultrasonographic evaluation of the change in systolic rise times (SRT) in the internal pudendal artery after intracavernosal papaverine administration. SBRT Z stereotactic body radiation therapy. prostate of 1 dog from group 50/11d/NVB. The neurovascular bundles from dogs in group 50/11d/NVB displayed the same histologic abnormalities as those in groups 50/5d/all and 50/11d/all. No histologic abnormalities were noted in the penis of dogs from group 50/11d/ NVB. Nerves and vessels within the prostate and the NVB appeared normal in dogs from Group 50/11d/PB. Within the prostate of dogs from group 40/11d/all, there was moderate multifocal lymphocytic inflammation, with moderate perineural and perivascular lymphocytic cuffing. There was severe adventitial thickening, and intimal hypertrophy with subendothelial hemorrhage and occasional medial necrosis in small- and medium-sized arteries. Many of these vessels were also thrombosed, often with remodeling around the thrombi. The peripheral nerve bundles had increased amounts of interstitial collagen. The neurovascular bundles from dogs in group 40/5d/all were characterized by moderate interstitial fibrosis of the fibrofatty connective tissue, moderate increases in interstitial collagen within peripheral nerves and ganglia, and moderate vascular changes, including severe intimal proliferation and mild to moderate perivascular fibrosis of medium-sized arteries, multifocal lymphocytic cuffing of arteries and arterioles, and thrombosis and reorganization of several medium to large arteries. No histologic abnormalities were noted in the penis in group 40/11d/all. In the prostate of dogs from group 30/11d/all, nerves and vessels appeared normal. The neurovascular bundles and penis from dogs in this group were histologically unremarkable. A summary of outcomes and statistics from semiquantitative pathologic evaluations is presented in Table 3 Figure 5, and also in Figure E1 (available online at www.redjournal.org). Further statistical analysis of the percent collagen within nerves was performed for all treatment groups, to correct for differences in the time after SBRT at which dogs were killed. A multivariable linear regression model of ranks for mean values was used to evaluate the effects
802
Nolan et al.
International Journal of Radiation Oncology Biology Physics
Fig. 5. Top row, Masson’s trichrome-stained, paraffin-embedded section of prostatic nerves ( 200 magnification; each nerve is denoted by a central white asterisk) from a normal dog (left, untreated control) showing normal collagen content and axonal structure, and from a dog treated with 50 Gy (right, from group 50/11d/all) showing severe axonal vacuolization and degeneration. Bottom row. Photomicrographs ( 200 magnification) of representative paraffin-embedded penile nerves, prepared with neurofilament immunohistochemistry from an unirradiated dog (left, from dog in control group) and a dog treated with 50 Gy to the prostate, neurovascular bundle, and penile bulb (group 50/11d/all). Neurofilament appears brown (with nuclei staining blue); this example shows lower expression of neurofilament in the dog from group 50/11d/all, compared with that from the unirradiated dog. of treatment group and survival time. The results are presented in Tables E5 and E6 (available online at www.redjournal.org).
Discussion This report describes the development and feasibility of a canine model for studying the causes of RI-ED. SBRT was chosen rather than more conventional fractionated irradiation techniques because of the convenience of a severely hypofractionated protocol; SBRT has been shown to result in a similar incidence of RI-ED in comparison with conventional protocols (3). Pelvic irradiation may alter endocrine status and adversely affect erectile function. There were, however, no significant changes in circulating serum concentrations of either testosterone or luteinizing hormone in these dogs. Dogs in which only the PB was irradiated, and dogs that received a lower total dose (and dose per fraction) of radiation, experienced neither clinically appreciable colorectal toxicity nor measurable ED. This does not, however, imply that dogs in groups 50/11d/PB, 40/11d/all, and 30/ 11d/all had no evidence of subclinical ED; in fact, neurophysiologic and perfusion studies evidenced neurogenic
injuries, vascular injuries, or both in these dogs, which could contribute to RI-ED. Observed electromyographic changes in the bulbospongiosus muscles of all irradiated dogs in this study indicate hypersensitive, denervated myofibers, which can be observed with either primary muscle or nerve disease. To further investigate potential nerve damage, nerve conduction studies were performed. Motor nerve conduction studies of the pudendal nerve were performed before radiation treatment and repeated at the examinations 4 months and 1 year after treatment. Inasmuch as there are no established normal values for MNCV of the pudendal nerve in dogs, it was intended that the pretreatment values were to be used as comparison for the posttreatment values in each dog. However, CMAPs were not obtained at all timepoints in all of the dogs. It is likely that the reason for inadequate data collection was inappropriate technique, stemming from: (1) the stimulating electrode not being in close enough proximity to the pudendal nerve because it travels between the coccygeus and superficial gluteal muscles; and (2) placement of the recording electrode too far distal to the bulbospongiosus muscle. Sensory nerve conduction results and cord dorsum potentials were not successfully recorded (see supplementary materials, available online at www.redjournal.org).
Volume 91 Number 4 2015 Table 3
Nerve/artery damage after prostatic SBRT
803
Summary statistics for semiquantitative pathology data Median values, %
Factor Nerve density, prostatic capsule Neurofilament content, prostatic nerves Collagen content, prostatic nerves Neurofilament content, penile nerves Collagen content, penile nerves
Arterial patency, NVB
Statistics
50/5d/ All (A)
50/11d/ All (B)
50/11d/ NVB (C)
50/11d/ PB (D)
40/11d/ All (E)
30/11d/ All (F)
Control (0 Gy)
3.04
4.54
2.02
1.76
4.09
5.91
3.52
61.73
64.01
68.43
73.43
68.36
68.20
76.75
21.73
24.74
41.50
43.82
49.06
52.40
48.16
41.95
43.78
61.66
56.71
71.87
60.98
75.51
31.15
27.09
64.12
62.11
72.20
74.27
67.70
23.30
21.00
24.90
28.00
23.00
27.30
32.40
Statistically significant comparisons (P
