In April 1995 a 49 year old woman presented with a 24-hour history of persistent macroscopic haematuria. There was no associated frequency, urgency, incontinence, or pain. She had passed small blood clots in her urine, and on wiping her perineum had found a metal clip on the toilet paper, which she brought to the consultation. This clip was readily identifiable as a closed Filshie clip (Fig. 1).
She had undergone laparoscopic tubal ligation using Filshie clips seven years previously. Review of the laparoscopist's notes and contemporary medical records did not suggest the presence of any pelvic pathology, operative difficulty, or early postoperative complication. There had been no pregnancy since tubal ligation, no history of other abdomino-pelvic pathology or surgical procedure, no prior episodes of haematuria, and no history of inflammatory bowel disease, recurrent urinary tract infection, voiding difficulty, or dysuria. Her suggested menopause had occurred six months before presentation.
On examination she was well, afebrile and normotensive. Her abdomen was soft and non-tender. Vaginal examination revealed a normal-sized ante-verted uterus, closed cervical os and no defect in the vaginal epithelium. The urethral meatus appeared normal. Macroscopic haematuria was confirmed on gross inspection of the urine. Subsequent culture of a midstream sample showed it to be sterile.
A plain abdomino-pelvic X-ray showed a solitary Filshie clip on the left side of the pelvis. Cystoscopy demonstrated a deep bed of chronic mucosal ulceration on the right side of the bladder dome (Fig. 2). Laparoscopic examination of the pelvis revealed a mobile left fallopian tube transected approximately 1–1.5 cm from the left uterine cornu, with a Filshie clip loosely attached to peritoneum at the site of tubal separation (Fig. 3). This clip was removed through the 5 mm side-port. The right fallopian tube was disrupted approximately 2 cm from its uterine cornu, at which point it was firmly adherent to the serosa of the bladder dome (Fig. 3). No attempt was made to separate the right fallopian tube segments from the bladder wall. The only other abnormality noted within the pelvis was the presence of a thin band-adhesion between the anterior uterine wall and the serosa of the bladder dome (Fig. 3). Both ovaries and the appendix appeared normal. Combined chromo-cystodistension and laparoscopy did not demonstrate any connection between the lumen of the bladder and the peritoneal cavity, or between the lumina of the bladder and right fallopian tube segments. Subsequent transcervical chromotubation confirmed occlusion of both fallopian tubes and again failed to demonstrate any connection between the lumen of the proximal right fallopian tube and the bladder cavity.
The woman made an uneventful recovery and was prescribed combined hormone replacement therapy one week after the investigative surgery. There has been no reported recurrence of her haematuria to date.
Filshie clips are 12.7 mm long and 4 mm wide with jaws of titanium, lined with silicone rubber (Fig. 1). They effect sterilisation by being locked across the breadth of each fallopian tube to occlude the rubal lumina. Avascular necrosis occurs at the site of clip application with each tube eventually dividing to leave two healed and occluded stumps. In the majority of patients the Filshie clip remains attached at the site of tube separation, eventually becoming covered by a thin layer of peritoneum. However, occasionally one or both clips will dislodge and settle elsewhere within the peritoneal cavity; most often the Pouch of Douglas or paracolic gutters. Nevertheless, whether properly sited or otherwise, it is rare for the continued presence of Filshie clips within the peritoneal cavity to result in significant morbidity, despite a number of women attributing a variety of abdominal and pelvic symptoms to their clips. A hydrosalpinx or a pyo-salpinx may occur if the clips are applied to damaged or infected fallopian tubes, but such complications are a function of occlusion of the tubal lumina, rather than of the Filshie clip itself. Indeed, only three reports appear in the medical literature of morbidity related to the continued presence of a Filshie clip per se, and in all three, morbidity was a function of the Filshie clip acting as a foreign-body nidus for chronic infection1–3.
The migration of a Filshie clip across tissue planes, with subsequent expulsion via the urethra, vagina or rectum is rare. In a study population of 5326 women the Food and Drug Administration in the United States of America noted only three such occurrences; suggesting an incidence of approximately 0.6 per thousand women4 (Mr M. Filshie, personal communication). In one woman, the clip found its way across the rectum 18 months after sterilisation. In another, a Filshie clip was expelled via the urethra 10 months following its application to the fallopian tube. In the third patient the clip migrated across the vaginal wall, 34 months after sterilisation, but hysterectomy had been performed in the intervening period. It is noteworthy that in these instances of clip migration and loss, including the case presented here, transperitoneal migration and subsequent loss of a Filshie clip from a body cavity was a late complication of sterilisation, and was not associated with any significant harm or long term adverse sequelae.
On the basis of the operative findings in this report, it is interesting to speculate on the mechanism underlying clip migration. It is possible that there was misapplication of the Filshie clip at the time of sterilisation. However, this is unlikely as there was no documented operative difficulty or adhesion formation distorting the pelvic anatomy at the time of sterilisation. In the healthy pelvis we have found it difficult to incorporate the full thickness of a fallopian tube, override the round ligament, and include part of the bladder wall between the jaws of a Filshie clip, as would need to have occurred if this had been a misapplication.
It is perhaps more likely that a low grade inflammatory process was responsible for the migration of the Filshie clip into the urinary bladder. If one looks at the situation of other foreign bodies that have found their way from the peritoneal cavity into the bladder lumen—such as surgical needles and gauze pads inadvertently left behind at laparotomy—the mechanism is one of surrounding peritoneal inflammation, local adhesion formation to the bladder wall with subsequent chronic inflammation, and eventual erosion into the bladder cavity5,6. The process operating in the case reported here was clearly chronic in nature, since seven years had passed between the sterilisation and the passing of the Filshie clip in the urine. Furthermore, the dense adhesion formation between the right fallopian tube segments and the bladder dome, and the adhesions between the anterior uterine wall and the bladder wall serosa noted at the time of laparoscopy, strongly support the activity of a chronic inflammatory process. Controlled studies undertaken in primates have shown that Filshie clips are capable of inducing a foreign body inflammatory response7, and this may be seen clinically, although uncommonly, in the situation of adhesion formation around dislodged Filshie clips. That there is tissue reaction following the application of the Filshie clip is also seen clinically in cases of ectopic pregnancy following sterilisation, which are thought to be due to formation of either a direct tubo-peritoneal fistula or tubo-tubal fistula; both chronic inflammatory sequelae.
Laparoscopic occlusion of the fallopian tubes using Filshie clips has become one of the most prevalent methods of contraception in the developed world. It is safe, technically straightforward, and is associated with only a 0.2–0.5% incidence of failure. This report adds to the small body of literature concerning late complications of female sterilisation; a matter of significant medico-legal importance. In the limited experience so far, it is notable that women who have passed a Filshie clip via the rectum, vagina or urethra have not been placed at significant risk of harm or long term adverse sequelae. Given this, operative investigation of a woman who has expelled a clip from a body cavity is unlikely to be clinically useful, and perhaps should be accepted as unnecessary if she is otherwise asymptomatic. Where the possibility of fertility is of concern, hysterosalpingo-graphy is sufficient in assessing occlusion of the fallopian tubes, and will also provide information about any remaining Filshie clips within the pelvis.
The authors would like to thank Mr M. Filshie for his comments, and Ms L. Connolly for proofreading this report.