THE ANATOMICAL RELATIONSHIPS IN THE SPACE OF RETZIUS FOR PENILE IMPLANTS: AN MRI ANALYSIS

Abstract

Background:

The three-piece inflatable penile prosthesis (IPP) includes an easy-to-use pump and fluid filled reservoir which is placed in either the space of Retzius (SOR) or in an alternative ectopic location. Reservoir placement in the SOR is a blind procedure despite the SOR being surrounded by many critical structures. To date only a handful of cadaveric studies have described the relevant anatomy.

Aim:

To use magnetic resonance imaging (MRI) as an in-vivo model to study relevant retropubic anatomy critical for SOR reservoir placement.

Methods:

The study population included men with elevated prostate specific antigen or biopsy proven prostate cancer who (i) underwent pelvic MRI, (ii) without prior pelvic or inguinal surgery, and (iii) without pelvic radiation therapy. All MRIs were completed with a 3-Tesla scanner and endorectal coil. Both T1 and T2 weighted images were captured in both axial and sagittal planes. All images were reviewed by two independent reviewers under the supervision of a dedicated body MRI radiologist. Bladder volume was calculated using an ellipsoid formula.

Outcomes:

Relevant measurements included (i) the distance between the external inguinal ring (EIR) at the level of the pubic tubercle to the external iliac vein (EIV), (ii) the distance from the EIR at the pubic tubercle to the bladder (accounting for bladder volume) and (iii) the distance from the midline pubic symphysis to the bladder (accounting for bladder volume). Pearson correlation was used to determine correlated measurements.

Results:

A total of 24 patients were included. Median participant age was 63 years (interquartile range, 59–66). The mean EIR-EIV distance was 3.0 ± 0.4 cm, the mean EIR-bladder distance was 1.8 ± 1.0 cm and the mean distance from the superior pubic symphysis to bladder was 0.9 ± 0.3 cm. There was a weak correlation between bladder volume and distance between the EIR and bladder (r=−0.30, p=0.16).

Clinical Implications:

The use of MRI as an in-vivo model is a high-fidelity tool to study real time unaltered anatomy and allows for surgical preparation, diagnosis of anatomic variants and acts as a valuable teaching tool.

Strengths & Limitations:

This is the first in-vivo model to report relevant retropubic anatomy in penile implant surgery. Our study is limited by sample size and inclusion of participants with no history of prior pelvic intervention.

Conclusion:

We demonstrate the utility of MRI as an in-vivo model, as opposed to cadaveric models, for the understanding of relevant retropubic anatomy for implant surgeons.

INTRODUCTION

The definitive management of medication-refractory erectile dysfunction (ED) is a penile implant. These devices has evolved from purely rigid devices to malleable/semi-rigid and eventually inflatable devices.^1,2 Concerns regarding the concealability and poor flaccidity profile of malleable devices have elevated inflatable devices to first line surgical options at least in certain parts of the world.³ Early inflatable devices such as 2-piece devices allowed surgeons to place a reservoir into the scrotum under direct visualization, but were limited by the amount of fluid stored in the reservoir.^2,3

Current technology utilizes a 3-piece device with an easy to use pump in the scrotum and a fluid reservoir that historically has been placed in the space of Retzius (SOR) or more recently sub-muscularly or even ectopically in a high sub-rectus space.³ Utilizing the SOR as a reservoir location is a blind procedure and thus anatomical landmarks are important to appreciate when utilizing this approach.⁴ The SOR is surrounded by many vital structures both visceral and vascular, and inaccurate placement may result in severe injuries especially in a patient following radical pelvic surgery.⁵ To date, anatomic studies have been conducted only in cadaveric models, which have limitations in tissue anatomy secondary to the use of fixative agents.^6,7

Our objective was to use magnetic resonance imaging (MRI) as an in-vivo model, as opposed to cadaveric models, to assess relevant retropubic anatomy that is critical for SOR reservoir.

MATERIALS AND METHODS

Study Population:

A retrospective review was performed of select patients who (i) underwent pelvic MRI, (ii) without prior pelvic or inguinal surgery, and (iii) without pelvic radiation therapy. Study approval was obtained from the Institutional Review Board at our institution (IRB 16-459).

MRI Measurements:

All patients underwent magnetic resonance imaging with a 3-Tesla magnetic MRI scanner with an endorectal coil with T1 and T2 weighted imaging. Images were taken in the both the axial and sagittal planes. MRI images were reviewed by two independent reviewers. Landmark identification and measurement training were completed under the supervision of a dedicated body MRI radiologist. The accuracy of all measurements was assessed by the same radiologist. Measurements obtained included (i) distances (bilaterally, left and right sides) from the external inguinal ring (EIR) at the level of the pubic tubercle to the external iliac vein (EIV), (ii) the distances bilaterally from the EIR at the pubic tubercle to the bladder (accounting for bladder volume) and (iii) distance from the midline pubic symphysis to the bladder (accounting for bladder volume) (Figure 1).

Figure 1.

Magnetic Resonance Imaging Highlighting Anatomic Landmarks for Inflatable Penile Prosthesis. (A) External inguinal ring (red) to external iliac vein (blue) (axial T2 weighted image) (B) Superior aspect of pubic symphysis (PS) to bladder (midline-sagital T2 weighted image (C) External inguinal ring (red) to nearest bladder point (axial T2 weighted image).

Statistical Analysis:

Distances measured bilaterally were averaged. Bladder volume was calculated using an ellipsoid formula of length x width x height x π/6. Pearson correlation was sought to determine if bladder volume was correlated to the distance from the EIR. All analyses were completed using Stata v17 (StataCorp LLC, College Station, TX).

RESULTS

Study Population:

A total of 24 patients were included in the analysis. All men had an MRI for either an elevated prostate specific antigen level or with biopsy proven prostate cancer. Patient characteristics are displayed in Table 1. Overall median participant age was 63 years (interquartile range (IQR), 59–66), and median BMI was 28.0 kg/m² (IQR 25.5–31.5). The majority were never or former smokers (n=20, 83%).

Table 1.

Patient Demographics

Age (years, median, IQR)	63 (59–66)
Body mass index (kg/m2, median, IQR)	28 (26–32)
Smoking Status (n, %)
Never	13 (54)
Former	7 (29)
Current	4 (17)
Comorbidities (n, %)
Coronary Artery Disease	0 (0)
Stroke	1 (4)
Diabetes	6 (25)
Dyslipidemia	11 (46)
Hypertension	14 (58)

IQR: interquartile range

MRI Measurements:

The summary of measurements is displayed in Table 2. The mean EIR-EIV distance was 3.0 ± 0.4 cm with a range of 2.4 – 4.2 cm. The mean EIR-bladder distance was 1.8 ± 1.0 cm with a range of 0.0 – 3.6 cm. The mean distance from the superior pubic symphysis to bladder was 0.9 ± 0.3 cm with a range of 0.4 – 1.5 cm. The mean bladder volume was 99.4 ± 38.9 cm³ with a range of 35.6 – 180.0 cm³. There was only a weak correlation between bladder volume and distance between EIR and bladder (r=−0.30, p=0.16).

Table 2.

MRI Measurements

ID	External Inguinal Ring to External Iliac Vein (cm)	External Inguinal Ring to Bladder (cm)	Bladder Volume (cm3)	Superior Public Symphysis to Bladder (cm)
1	3.0	3.2	35.6	0.9
2	3.7	2.0	70.8	1.5
3	4.2	2.5	101.6	0.5
5	3.1	2.7	149.8	0.6
6	2.6	1.3	75.6	0.7
7	2.9	2.2	82.5	1.2
8	2.4	3.6	84.7	0.6
9	3.3	2.2	82.3	1.1
10	3.2	2.0	79.4	1.0
11	2.7	2.5	108.2	0.4
12	3.4	2.3	96.6	1.3
13	2.5	2.4	75.1	0.7
14	3.4	1.1	85.4	0.8
15	3.1	1.7	169.3	1.0
16	2.9	0.7	115.7	0.8
17	3.4	0.4	153.3	0.8
18	3.1	2.6	110.3	0.4
19	3.1	1.4	66.4	1.3
20	2.6	2.1	88.5	0.8
21	2.5	1.7	65.9	1.1
22	2.5	0.0	73.9	0.7
23	3.2	0.6	61.1	1.0
24	3.0	0.2	180.0	0.8
Mean (SD)	3.0 ± 0.4	1.8 ± 1.0	99.4 ± 38.9	0.9 ± 0.3
Median (IQR)	3.1 (2.7–3.2)	2.0 (1.1–2.5)	85.0 (74.5–113.0)	0.8 (0.7–1.0)
Range	2.4–4.2	0.0–3.6	35.6–180.0	0.4–1.5

SD: standard deviation; IQR: interquartile range; MRI: magnetic resonance imaging

DISCUSSION

Penile implant surgery remains the standard treatment option for medication-refractory ED. The use of a 3-piece device mandates placement of a fluid-filled reservoir which for concealment purposes has historically been placed in the SOR. This is a blind procedure conducted with a finger through the EIR with attendant risks to the bladder deep and external iliac veins laterally. We herein report the first in-vivo study, using MRI, to explore relevant retropubic anatomy for IPP placement.

The SOR is an extraperitoneal anatomical space also known as the pre-vesical space. It consists of loose fibrous tissue and fat, and its borders consist of the umbilicus superiorly, pubovesical ligament inferiorly, the transversalis fascia anteriorly and the umbilicovesical fascia posteriorly.^8,9 This space is also importantly adjacent to the bladder and iliac vasculature. While various techniques have been described for SOR reservoir placement, they all consist of identification of the pubic tubercle to identify the EIR, followed by medial dissection over the pubic bone through the transversalis fascia into the SOR.¹⁰ In circumstances where SOR placement is challenging (ie. previous pelvic surgery or known scarring) a counter incision may be considered or the reservoir may be placed ectopically.

An understanding of appropriate anatomical landmarks is critical due to some of the reported complications following device placement. This includes direct hernia development, bladder injury, device infection or migration, bowel injury and possible vascular injury.^5,11–13 While rare, the gravity of these possible complications is substantial, and therefore precise anatomical knowledge remains paramount for blind reservoir placement. As a consequence, steps to minimize these risks include bladder decompression and avoidance of deep dissection lateral to the inguinal ring have been suggested, but do not obviate the importance of an understanding of each individual patients anatomy.⁷

Previous anatomical analyses have focused on cadaveric models.^6,7,14 While these provide a great educational tool for learning anatomy, cadaveric models have limitations, the most important being tissue contraction and fixation.¹⁵ In the cadaveric study by Henry et al. the authors found that the mean distance from EIR-EIV was 3.2 cm whereas we found this mean distance to be smaller at 3.0 cm.⁷ They also found the mean distance from EIR to bladder was 2.6 cm for a filled bladder and 6.5 cm for a decompressed bladder.⁷ While we did not specify any criteria for bladder filling in our study, we found this distance to be much smaller, at about 1.8 cm on average. Limited other cadaveric studies exist for SOR reservoir placement, but some have focused on alternative placement methods such as ectopic reservoir insertion.⁶

Other educational methods such as visual dissection or imaging may be more ideal than classic cadaveric models.¹⁵ MRI for example, provides a non-invasive and high-fidelity tool to review real time anatomy rather than the possibly altered post-mortem state. Not only are such models incredibly useful for teaching purposes, the clinical implications of this are evident as it reveals relevant anatomical landmarks, potential unexpected anatomic variations, and assists in surgical planning to avoid inadvertent and possibly catastrophic injuries.

This study is limited in its use of a small number of patients and using patients with no prior pelvic intervention or radiation, which would both be important groups to examine in future studies. However, it is the first report of its kind to utilize measurements from living patients with cross-sectional imaging as opposed to fixed cadaveric tissue for IPP placement, which may still have some differences from the true surgical setting as with any radiographic imaging. We also did not obtain measurements with and without bladder filling which may provide more insight with regards to possible correlations between bladder volume and measurements in the SOR.

CONCLUSION

We describe here the utility of MRI for the understanding of the relevant pelvic anatomy prior to penile implant surgery. Furthermore, the anatomical distances from external inguinal ring, bladder and external iliac veins in this in-vivo model are remarkably close. Future collaborative and multicenter studies would improve the reliability and clinical usefulness of MRI in this setting.