SEARCH

SEARCH BY CITATION

Keywords:

  • sling procedures;
  • retropubic haematoma;
  • magnetic resonance imaging;
  • postoperative complications

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

OBJECTIVE

To determine, using magnetic resonance imaging (MRI), the incidence of retropubic haematoma and any associated clinically significant effects after a xenograft (porcine dermis) sling (XS) or the tension-free vaginal tape (TVT) procedure.

PATIENTS AND METHODS

Between October 2003 and March 2004, 24 consecutive patients presenting with stress urinary incontinence (SUI) were enrolled in this prospective study; 12 each underwent an XS or TVT procedure. A vaginal balloon pack was used for only 3 h after XS and not after TVT. All patients had pelvic MRI 6–8 h after surgery. The primary outcome measure was the incidence and distribution of retropubic haematoma after each sling technique. Secondary outcome measures included the interval to the first three spontaneous voids, the bladder emptying efficiency of the first three voids, a visual analogue scale pain score at 24 h after surgery, and the short-term (6-month) cure rate for SUI.

RESULTS

Overall, six (25%) patients (four XS and two TVT) developed a retropubic haematoma. Most commonly, they spread along the right paravesico-urethral space between the right half of the levator ani and the bladder neck. Patients with large haematomas took significantly longer to void (median 14.5 vs 6.0 h, P = 0.048). There was no difference in pain score in patients with or with no haematoma. None of the patients had clinically detectable haematomas in the suprapubic wound. All six patients with haematomas were cured or improved at the 6-month follow-up.

CONCLUSIONS

MRI is a useful noninvasive method for detecting retropubic haematomas soon after surgery. There was a surprisingly high incidence of retropubic haematomas, especially after the XS procedure. Retropubic haematomas may influence postoperative voiding efficiency.


Abbreviations
SUI

stress urinary incontinence

XS

xenograft sling

TVT

tension-free vaginal tape

EE

(bladder) emptying efficiency

VAS

visual analogue scale.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Sling procedures have been used for over a century for treating female stress urinary incontinence (SUI). Modifications in sling technique have resulted in broader indications, reduced morbidity and a shorter hospital stay [1–3]. By using a rectus fascia substitute it is possible to avoid the Pfannenstiel incision and eliminate the morbidity associated with fascial harvesting. We have found that a suprapubic ‘needle down’ technique and finger-controlled tracking of the needle to the vaginal incision is a safe procedure when placing a xenograft (porcine dermis) sling (XS) [4]. After paraurethral dissection the suprapubic needle can be guided digitally, thereby minimizing the risk of visceral injury. However, this requires vaginal dissection and perforation of the endopelvic fascia from below to facilitate passage of the finger. This may result in bleeding which tracks up into the retropubic space. In a retrospective study, Deval et al.[5] analysed the records of 187 women who had a tension-free vaginal tape (TVT) procedure and found only one case of clinically documented suprapubic haematoma. There are no published prospective studies of the incidence of retropubic haematoma detected by imaging.

It is likely that retropubic haematoma could lead to increased morbidity after surgery, e.g. pain, fever, vaginal bleeding and suprapubic wound haematoma/infection. It is also possible that postoperative bladder emptying efficiency (EE) might be impaired and in turn the hospital stay prolonged by haematomas. Compared with other imaging methods the advantages of MRI include; (i) relative operator independence; (ii) direct multiplanar and multisequence image acquisition for better characterization of tissues; (iii) lack of ionizing radiation; and (iv) absence of bone artefact.

The aim of the present study was to determine, using MRI, the incidence of retropubic haematoma and any associated clinically significant effects after a XS or TVT procedure.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Between October 2003 and March 2004, 24 consecutive patients presenting with SUI were enrolled in a prospective study; 12 each had an XS or TVT procedure, assigned to either technique depending on the surgeon's preference. One surgeon (H.F.) did only the XS and the other (J.D.) only the TV. Ethical approval for the study was obtained from the local ethics committee. Patients were recruited once it was decided that surgery was a suitable treatment option for SUI.

The inclusion criteria were that patients had urodynamically confirmed SUI, were scheduled for XS or TVT, and gave informed consent. Patients with a history of UTI in the last 6 weeks, neuropathic bladder, uterovaginal prolapse, detrusor instability or preoperative voiding dysfunction (maximum urinary flow rate, <15 mL/s, detrusor pressure at maximum urinary flow rate >40 cmH2O, or a postvoid residual volume of >50 mL) were excluded from the study.

Before surgery all patients were evaluated by a history, examination, urine analysis and urodynamic study. All patients had a pelvic examination to assess pelvic floor defects and bladder neck motion. Informed consent was obtained from all patients.

All procedures were scheduled early on the list to facilitate later MRI. All received thromboprophylaxis with enoxaparin 20 mg s.c. and antibiotic prophylaxis with ceftriaxone 1 g and gentamicin 240 mg i.v. before surgery. For all procedures the patients were under general anaesthesia. The XS was prepared using a 7 × 2 cm strip of porcine dermis (PelvicolTM, Bard, UK) secured at each end with a 0 polypropylene suture; the technique was described previously [4]. All slings were placed loosely under the bladder neck/proximal urethra. Patients returned to the ward without a urinary catheter, and a vaginal balloon pack was used for just 3 h after surgery. The operative duration and intraoperative blood loss were recorded, with any intraoperative complications. The surgical technique of the TVT procedure was as described by Ulmsten et al.[6], except that all patients were operated under general anaesthesia with no retropubic infiltration with bupivacaine.

After surgery, NSAIDS, e.g. parecoxib 40 mg i.v., paracetamol suppository 1 g and diclofenac suppository 100 mg, were used for pain control. Intravenous fluid was continued with solution-18 at 125 mL/h until the patient was able to drink fluid freely.

All patients had pelvic MRI 6–8 h after surgery, using a 1.5 T superconductive magnet (Magnetom Symphony; Siemens Corp, Erlangen, Germany), while supine and lying head first on the scanner. After an initial three-plane location sequence, the pelvic soft tissues were assessed with axial and sagittal T1-weighted spin-echo sequences (with a repetition time of 600 ms, echo time of 11 ms, and 4 mm slice thickness). T1-weighted fat-suppressed images were also acquired in the sagittal plane. Sagittal and axial T2-weighted turbo spin-echo sequences (with a repetition time of 4600 ms, echo time of 113 ms and 4 mm slice thickness) were also obtained. Images were obtained in the static phases. Because of the problem of breath-holding immediately after surgery the straining (dynamic) phase was not used, and nor was i.v. gadolinium used. The total image acquisition time was ≈ 20 min.

The investigators performing and analysing the MRI were unaware of both the patients’ clinical history and nature of the anti-incontinence procedure. All images were interpreted by a consultant radiologist (F.W.). Retropubic haematoma was defined as a haematoma of 2 × 2 × 2 cm or larger in its three largest diameters. The distribution of the haematoma was also noted; they were arbitrarily divided into small (2–3.9 cm), medium (4–5.9 cm) or large (≥6 cm).

After surgery patients were assessed for pain intensity using a visual analogue scale (VAS; 0 no pain, 10 worst pain), the interval to spontaneous voids, the EE [3,4] and any complications. All patients were followed for up to 6 months. The primary outcome measure was the incidence and distribution of retropubic haematoma after each sling technique. Secondary outcome measures included the interval to the first three spontaneous voids, EE of the first three voids, VAS pain score at 24-h after surgery, and the short-term (6 months) SUI cure rate. The surgical result was classified as ‘cured’ when the patient reported no leakage of urine under any circumstances, ‘improved’ when there was a ≥ 50% reduction in SUI, and as ‘failure’ with <50% reduction in SUI [7].

Data obtained from the case report forms were transferred to a computer spreadsheet and entries then checked for errors. All data were tested where appropriate for normality. The statistical significance was calculated using Fisher's exact test or the Mann–Whitney U-test where appropriate, with P < 0.05 considered to indicate statistical significance.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Table 1 shows the baseline characteristics of the patients undergoing the XS and TVT procedures. The MRI was well tolerated by all patients; none experienced discomfort during the procedure. The quality of all MR images was optimal; the sling material was not visible in either treatment group. Haematomas appeared as low to medium signal-intensity areas in T1-weighted sequences (Fig. 1) and as medium to high intensity areas in T2-weighted sequences (Fig. 2).

Table 1.  Baseline characteristics of patients treated with an XS or TVT procedure for SUI (12 patients each), and the incidence of retropubic haematoma after surgery
Median (range) or n, variableXSTVT
  • *

    P= 0.01, Mann–Whitney U-test.

Age, years49 (36–65)51 (37–67)
Parity 4 (1–8) 3 (1–6)
Postmenopausal, n 6 6
Hysterectomy, n 3 3
Previous surgery for SUI, n 4 3
Preoperative pad use/day 3 (1–6) 4 (1–8)
Operative duration, min*25 (20–45)35 (25–55)
VAS pain score at 24 h 3.5 (0–10) 3 (0–10)
Length of stay, days 2 (1–4) 1.5 (1–3)
Incidence of haematoma, n
Small (2–3.9 cm) 1 2
Medium (4–5.9 cm) 1 0
Large ≥ 6 cm 2 0
Total 4 2
image

Figure 1. Retropubic haematoma (H, small) after a TVT procedure. The haematoma (outlined) in the axial T1-weighted turbo spin-echo sequence appears as a low to medium signal-intensity area. B, urinary bladder; V, vagina; A, anal canal.

Download figure to PowerPoint

image

Figure 2. Retropubic haematoma (same patient as in Fig. 1) after a TVT procedure. Haematoma in the axial T2-weighted turbo spin-echo sequence appears as a medium to high signal-intensity area. There is marked backward projection of the symphyseal cartilage (SC). H, haematoma; B, bladder; V, vagina; A, anal canal.

Download figure to PowerPoint

Of the 24 patients assessed, six (25%) had a retropubic haematoma of 2–8.5 cm in diameter; the occurrence of haematoma with each type of surgery is also shown Table 1. Retropubic haematomas were typically in the space of Retzius above the levator ani muscles. Most commonly (three of four small or medium haematomas), they spread along the right paravesico-urethral space between the right half of the levator ani and bladder neck (Fig. 3). Two patients after XS developed a large haematoma of 3.8 × 4.2 × 6.5 cm and 6.6 × 4.5 × 8.5 cm, respectively. In each case the haematoma extended from the retropubic space to both right and left paravesical space above the levator ani (Fig. 4), and along the preperitoneal space behind the rectus muscle.

image

Figure 3. Retropubic haematoma (medium) after a XS procedure. In the axial T1-weighted image the haematoma appears as a low to medium signal-intensity area (outlined). In the axial view the haematoma is seen tracking along the right paravesical space from the retropubic space. H, haematoma; B, bladder; UT, uterus; R, rectum.

Download figure to PowerPoint

image

Figure 4. A retropubic haematoma (large) after an XS procedure. In this axial T1-weighted image the haematoma is extending from the retropubic space to the both right and left paravesical space (outlined). H, haematoma; B, bladder; UT, uterus; R, rectum.

Download figure to PowerPoint

The two patients with large haematomas took longer for their first void than those with no large haematoma (median 14.5 vs 6.0 h, P = 0.048). Although patients with no haematoma voided more efficiently during their first three voids, the differences were not statistically significant (Table 2).

Table 2.  EE differences in patients with and without haematoma. P-values are given for Mann–Whitney U-test
VoidMedian (interquartile range) EE, %P
with haematomaNo haematoma
N 618 
First40 (25–64)58.5 (38–78)0.081
Second60 (35–75)76 (57–94)0.119
Third68 (40–80)80.5 (60–94)0.187

There was no significant difference in the VAS pain score at 24 h in patients with or without haematomas (median VAS 3.5 vs 3.0). While the patient with the largest detected haematoma had a pain score of 5 and the one with second largest had a pain score of 3, two patients who scored 10 had no haematoma.

None of the patients had clinically detectable haematoma in the suprapubic wound. There was no pyrexia after surgery in any of the patients with haematoma. The patient with the largest haematoma had had transient hypotension 4 h after surgery and responded to i.v. fluids. The same patient also complained of minor vaginal bleeding for one week after discharge, which settled spontaneously. No patients required peri-operative blood transfusion. The median hospital stay was similar in patients with or without haematoma (2 days). The outcome of the sling surgery is shown in Table 3; all six patients with haematoma were cured or improved at the 6-month follow-up. Of the two patients with a large haematoma, one was cured and another improved.

Table 3.  The outcome of sling surgery for SUI at 6 months follow-up
ContinenceHaematoma (six)No haematoma (18)
XSTVTXSTVT
N42810
Cured215 6
Improved212 3
Failed001 1

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

This study showed an overall 25% incidence of retropubic haematoma detected by MRI after an XS or TVT sling procedure. Early retropubic haematoma often appears as an area of low to medium signal-intensity on the T1-weighted sequence, due to residual oxyhaemoglobin and deoxyhaemoglobin present in intact red blood cells. As we did not use retropubic infiltration this could not be a confounding factor in interpreting the image.

Although a 3-h vaginal pack was used after XS but not after TVT, retropubic haematomas were more common after XS (Table 1), probably because of the greater paraurethral vaginal dissection required to facilitate retropubic digital guidance of the needle in the XS group. Recently, Muir et al.[8] showed that the major vessels in the retropubic space and anterior abdominal wall lie 0.9–6.7 cm lateral to the TVT needles. If the TVT needle is laterally aimed or rotated, major vascular injury can occur. Interestingly, in both the present cases of retropubic haematoma after TVT, there was a marked backward projection of symphyseal cartilage (Fig. 2). This might have deflected the trocar, resulting in laceration of the perivesical venous plexus and thus haematoma formation. We also found that retropubic haematomas most commonly extended along the right paravesical space (Fig. 3). Unterweger et al.[9] recently showed that the right levator muscle was significantly thinner overall than the left in nulliparous, parous and stress incontinent women. This may be a possible explanation of less potential space and greater resistance to spread of haematoma in the left paravesical space. This may also account for why bladder perforation is more common on the left side during blind trocar insertion. However, a more likely explanation is probably greater effort by a right-handed operator in the right paraurethral dissection than the left.

Ultrasonography [10] can be used to detect anatomical changes after sling surgery; the main disadvantages of this method for assessing retropubic haematoma are loss of contrast resolution with increasing depth, thus precluding visualization of the deep retropubic space, and discomfort from the transvaginal or transabdominal probe, especially soon after surgery.

In summary, this preliminary study shows that MRI is useful for detecting early retropubic haematoma formation after sling procedures. There was a surprisingly high incidence of retropubic haematoma, some of which were large. That there were more haematomas after an XS procedure suggests that blunt perforation (as opposed to by trocar) of the endopelvic fascia may indeed increase the incidence of retropubic haematoma. However, because there were few patients in the present study, differences were not significant (P = 0.346, Fisher's exact test). Interestingly, haematomas may affect bladder emptying ability after surgery, as there was a significant delay in the time to the first void in those with large haematomas, and there was also a trend towards a poorer EE for the first void in these patients.

The present study was primarily descriptive; the findings support a larger study in patients who have anti-incontinence procedures which involve blunt perforation of the endopelvic fascia to determine: (i) the effect of haematoma formation on bladder function and factors of general recovery; and (ii) the effectiveness of vaginal packing for preventing haematoma formation.

In conclusion, retropubic haematomas seem to be common after sling procedures, and MRI can reliably detect and define them. Large haematomas may have previously unrecognized effects on bladder emptying after surgery.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

This work was supported by educational grant from the National Institute of Health Sciences (NIHS), Limerick, Ireland and Pfizer Sales Ireland.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES