Treatment of fecal incontinence - review of observational studies (OS) and randomized controlled trials (RCT) related to injection of bulking agent into peri-anal tissue

Abstract

Purpose

Novel treatments are needed to augment medical therapy for fecal incontinence.

Methods

Medline and Google search (fecal incontinence and injection treatment), English publications.

Results

Twenty-two observational studies and 4 randomized controlled trials were identified. OS mostly with limited sample sizes reported promising results. Repeated injection was necessary in some patients. Effect on anal sphincter pressures was highly variable. Significant improvements in the length of anal high-pressure zone, asymmetry index and maximum tolerable rectal volume were suggested. Four randomized controlled trials (n=176) revealed: 1. Short-term benefits from injection of Bioplastique under ultrasound guidance compared with digital guidance; 2. Silicone biomaterial (PTQ) provided some advantages and was safer than carbon-coated beads (Durasphere); 3. PTQ did not demonstrate clinical benefit compared to control injection of saline; 4. There was significant improvement at 6 weeks post injection, but no difference between Bulkamid and Permacol. A 2010 Cochrane review, however, noted that these data were inconclusive due to limited number and methodological weaknesses.

Conclusion

Further studies are warranted to assess patient-centered outcomes (e.g. adequate relief) in addition to the attenuation of severity of incontinence symptoms in ambulatory patients. In nursing home residents, cost-effectiveness studies combining injection treatment and prompted voiding (to mitigate constraints of immobility and dementia) in preventing peri-anal skin complications deserves to be considered.

Introduction

Fecal incontinence affects ambulatory individuals and nursing home residents.^1–3 In the ambulatory care settings, only one-third of affected patients reported their problem to their care providers making it a silent affliction.³ Multiple factors underscore the pathophysiology of fecal incontinence. These include deficits of internal or external anal sphincter or pelvic floor muscle function; loss of endovascular cushions due to disruption of the hemorrhoidal plexus; impaired anorectal sensation associated with chronic constipation; poor rectal compliance and compromised accommodation from aging, inflammatory bowel disease, radiation enteritis, pelvic surgery; neuropathy affecting the pudendal, sacral, spinal or central nervous system; incomplete evacuation of stool, large stool volume liquid stool, and the irritant effect of bile salts in the rectum.⁴ In the nursing home setting additional risk factors are dementia and immobility (including patient restraints) which preclude residents from getting to the toilet in a timely manner.⁵ Fecal impaction and overflow fecal incontinence was related to drug-induced constipation is a myth⁶ because the high prevalence of constipation in nursing home residents was only partly due to side-effects of drugs⁷. Anorectal testing in a subgroup of nursing home subjects with fecal incontinence in one recent randomized controlled trial (RCT) comparing prompted voiding and usual treatment documented impaired sphincter function (risk for fecal incontinence), decreased rectal sensation and sphincter dyssynergia (risk for constipation and impaction).⁸

Usual management of fecal incontinence includes appraisal of predisposing factors, detailed drug history (constipating drugs or excessive laxatives) and a physical, neurological and rectal examination. Impacted stool requires manual disimpaction followed by a plan of controlled evacuation with suppositories or enemas at regular intervals. Specific treatments of underlying diarrhea or constipation and increased fiber are important; loperamide (Imodium) or diphenoxylate (Lomotil) dosed correctly increases stool firmness by slowing transit.⁴

Although most studies report positive results using biofeedback to treat fecal incontinence⁹; closer scrutiny of the elements of the intervention and controlled studies, however, have consistently failed to find benefit of the biofeedback element of the complex package of care; nor has any superiority been found for one modality over another^9–11. Nonetheless, a more recent RCT appeared to confirm efficacy of biofeedback treatment.¹² Fecal seepage a form of fecal incontinence associated with impaired rectal sensation, inappropriate elevation of anal sphincter pressure during defecation, “excessive” straining-induced inadvertent closure of the external anal sphincter; and biofeedback to improve rectal sensation and timing of sphincter relaxation has been reported to ameliorate this symptom.¹³

Fecal incontinence due to rectal prolapse, rectovaginal fistula, or neurological problems such as spinal cord injury may require operative treatment. The anatomical data of the integrity of external and internal anal sphincter muscles provided by anal ultrasound study facilitate more appropriate reconstruction. Pudendal nerve terminal motor latency which measures the neuromuscular integrity of the terminal portion of the pudendal nerve separates neuropathy (prolonged latency) from rectal wall disorders. Reconstructive surgery may not be successful in patients with pudendal neuropathy. Other surgical procedures include sphincter repair and artificial sphincter. Only colostomy, however, can guarantee total continence. Informing patients of this limitation can minimize disappointment.⁴ Sacral nerve stimulation for the treatment of fecal incontinence refractory to medical management continues to be used in Europe.¹⁴

Despite the usual treatments as outlined above, some ambulatory patients with fecal incontinence remain symptomatic and are in need of innovative therapeutic approaches. In the elderly fecal incontinence is a real risk factor for nursing home placement. Approaches in ambulatory patients are applicable to nursing home residents. In addition, the use of a rectal stimulant and weekly enemas to achieve complete rectal emptying reduced the frequency of fecal incontinence by 35% and the incidence of soiling by 42%.¹⁵ Fecal impaction may require digital manual disimpaction, followed by tap water enemas two or three times each week, and possible use of rectal suppositories.¹⁶ The practice of routine prophylactic use of stool softeners, stimulant laxatives, and osmotic products against constipation, however, deserve to be re-examined. In the presence of impaired sphincter function and rectal sensation, the fluidity of the stool induced by such treatments predisposes nursing home residents to fecal incontinence. Dementia and immobility limit the effectiveness of biofeedback^2,17 in nursing home residents. Assisted toileting was studied in a recent RCT. The multi-component intervention significantly changed multiple risk factors associated with fecal incontinence and increased bowel movements but did not decrease frequency of fecal incontinence.⁸ Other novel approaches to augment assisted toileting may be necessary.

Injection of bulking agents to augment anal sphincter function has been reported with increasing frequency in observational studies (OS). Promising results of observational studies (OS) supported the conduct of RCT. The objective of this report is to review the efficacy of injection of bulking agents into the peri-anal locations for the treatment of fecal incontinence.

Methods

A Google and a Medline search (12/15/2010) were conducted using the criteria of “fecal incontinence” and “injections. The Google search identified one article.¹⁸ The Medline search (limited to publications in English and focused on humans) produced 70 publications which were then screened for data on non autologus bulking agents.^19–42 Conference abstracts were excluded.

Results

For the observational studies (Table 1), targeted study patients mostly had dysfunction of the internal^{20,21,25,27,30–36} or external^20,29,32 anal sphincter. In some reports, patients with unspecified^19,21 or varied etiologies (iatrogenic, obstetric injury, neurogenic, idiopathic)^{20–24,27–28,30–32,34–38} were included. The extent of fecal incontinence ranged from partial³⁸, mild and moderate^18,19,21 to persistent leakage and soiling³³ or severe³⁴. Almost all of these patient had failed prior conservative management such as pelvic floor physiotherapy or exercises^18–20,24, sphincter exercises²², electric stimulation³⁸, biofeedback^{18–20,27–28,34}; diet modifications^{18–20,24,27,28,33}; medications^19,20,28 such as anti-diarrheal drugs^23,33–35, loperamide^22,27, diphenoxylate plus atropine²⁷, stool bulking agents^21–23,33; sacral nerve stimulation²⁸; sphincteroplasty²⁸, artificial bowel sphincter²⁸, or prior PTQ macroplastique bulking agent²⁸. The site of injection is highly variable. The bulking agents were administered by peri-anal^21,27,30,38, intra-anal³², submucosal (just above the dentate line)^{18,22,28,29,33,36,38}, trans-sphincteric^18,21,34,35 or inter-sphincteric^{19,20,22–25} injections.

Table 1.

Observational studies suggesting injection therapy may have a role in treating fecal incontinence.

Agent	No.	Manometry	Effect of Treatment	Ref
PTQ™b	16	Median resting and maximal squeeze pressures improved at 6 and 12 months	FI scores significantly improved at 3 and 24 months.	18
Enteryx	21	Mean or maximum anal canal resting or squeeze pressures unchanged at 12 months	FI scores significantly improved at 6 weeks, 3 months, 6 months and 12 months.	19
PTQ	74	Mean resting anal canal and squeeze pressure significant increased	FI scores improved at a median follow-up of 28 months.	20
Silicone	35	Clinical improvement was not associated with increase in resting or squeeze pressures	Length of high-pressure zone increased from 1 to 1.7 cm. Asymmetry index showed a significant change.	21
NASHA Dx gel	34		Fifteen patients (44%) were responders (50% decrease in incontinence episodes) at 6 months, compared with 19 (56%) at 12 months.	22
Silicone	33	Not affected except for the maximum tolerable rectal volume, which was significantly reduced	The Wexner Continence Score was significantly reduced short term from 12.7 to 11.0 (P = 0.03) and long term to 10.4 (P = 0.02). The long-term effect on liquid stool incontinence continued to improve significantly (P < 0.01). Six patients (18%) reported major improvement in Wexner Continence Score at the time of final follow-up.	23
Carbon beads	11	Trend toward increase in measured pressures	Mean incontinence score was 12.3 (0.9) at baseline, 6.8 (1.6) at three-month, 6.7 (1.5) at six-month, 5.9 (0.9) at one-year, and 4.9 (0.9) at two-year follow-up (P = 0.003).	24
PTQ	15		FI scores significantly improved at 1 week, 1 month, 3 months, 6 months, 1 year, and 2 years.	25
PTQ	20	Unchanged mean resting or squeeze pressures	FI scores significantly improved at 1 month and 1 year (P < 0.005 and P = 0.02, respectively) but not at 2 years.	26
Coaptite	10	Significant improvement from baseline in the mean resting anal canal pressure	Eight patients (80%) had a marked improvement in continence, with a significant reduction in Fecal Incontinence Scoring System from 85.6 (9.4) to 28.0 (9.0) (p=0.008) at 12 months.	27
Durasphere	33	Significantly improved (resting pressure 34 to 42 mmHg; squeeze pressure 66 to 79 mmHg)	After a median follow-up of 20.8 (range, 10 – 22) months, the median Cleveland Clinic continence score decreased significantly from 12 to 8 (P < 0.001) and the median American Medical System score from 89 to 73 (P = 0.0074)	28
PTQ	24	Unchanged Mean resting and squeeze pressures	FI scores significantly improved 3 and 12 months.	29
Silicone	6		At 61-month one patient received a colostomy. In the remaining 5 the incontinence score was little changed: 11 (8 – 20) vs. 13 (9 – 19). Subjectively, 3 patients were improved; 1 had additional injections and one improved after a course of biofeedback.	30
PTQ	7	Maximum resting anal canal pressure significant improved at 6 months	FI scores significantly improved at 3 and 12 months.	31
Collagen	73		At a median follow-up of 12 months, 63% of patients had an improved incontinence score and 73% reported an overall improvement in symptoms. Logistic regression showed that older age and idiopathic FI were predictors of good outcome.	32
Durasphere	18	Maximum tolerable rectal volume at 12 months showed significant improvement	The mean follow-up is 28.5 months. Changes from baseline were not statistically significant up to 6 months.	33
Bioplastique	6	Median resting anal and squeeze pressure significant increased	FI scores Significantly improved at a median follow-up of 18 months.	34
Bioplastique	10	Unchanged maximum resting pressure		35
GAX	17		11 patients showed marked symptomatic improvement. One patient reported symptomatic improvement but remained in clinical grade 3, and two reported minimal improvement. There was no improvement in three patients, but one of these had a repeat injection and then showed significant improvement.	36
Buttock fat	1	Resting anal pressure increased	FI improved at 8 months.	37
Autologous fat (60 ml from abdominal wall)	14		All patients were continent during the first 2 to 3 post-injection months. At the 6th month, patients were divided into 3 scores. Score 1 (complete continence) comprised 3 patients who were continent for 9, 11, and 14 months post-injection, with normalization of their rectal neck pressure. Seven patients with Score 2 were incontinent to flatus and were re-injected; they were continent (Score 1) for a mean of 13.8 months and had normal rectal neck pressure. Four patients had Score 3 (no improvement), of whom 2 became continent after the 2nd injection and 2 after the 3rd. They were continent (Score 1) 6 to 16 months post-injection.	38
Teflon or Polytef (5 ml)	11		They were categorized into 3 scores: 1, cured; 2, improved, and 3, no change. Long-term cure (score 1) occurred in 45% after the 1st injection and in 64% after the 2nd injection. 36% showed partial improvement (score 2).	39

The majority of the observational studies reported improvement in fecal incontinence scores^{18–20,21–34} or significantly fewer incontinence episodes^22,29 over time. Some reports showed improvement in resting^{18,20,27,28,31,34} or squeeze^{18,20,24,28,31} anal sphincter pressures. Others showed no improvement in resting^{19,21,23,24,26,29,32,35} or squeeze^{19,21,23,26,29} anal sphincter pressures.

Quality of life scores showed significant^{19–21,24,27,33,34} or substantial^24,30,31 improvements. Improvements were recorded in physical function^30,34, social function score^30,34, lifestyle^18,21,27,31, coping/behavior^{18,19,21,24,27,31,33}, depression/self perception^18,21,22,33, embarrassment^{18,19,21,24,27,31,33} domains. Others reported limited¹⁸ or no^23,28,29 improvement. Quality of life scores were not recorded in many of the reports^{25,26,32,35–39}. Satisfaction improved significantly^29,33; was reported to be high after treatment^20,30, not only for patients cured but for some with post implant fecal incontinence, and did not change significantly over time²⁰. In the majority of studies, however, satisfaction score was not reported.^{18,19,21–28,31,32,34–39}

Specific report of no adverse events was described in some of the studies^{18,22,24,34,36} while others made no mention of whether adverse events occurred or not^{23,25,27,29,31,32}. Adverse event included sepsis²⁰, injection site abscess^19,20,21,26 or hematoma²¹, need for anti-diarrheal medication²⁰, post implant constipation²⁰, anal irritation/discomfort^{20,21,26,28,33,35}, mucosal ulceration^20,35, leakage/passage of injected material^28,33, and distal migration of the injected agent along the dentate line²⁸.

Data related to four RCT^40–43 were reviewed.⁴⁴ There were short-term benefits from injection of Bioplastique delivered under ultrasound guidance compared with digital guidance.⁴³ A silicone biomaterial (PTQ) was shown to provide some advantages and was safer than carbon-coated beads (Durasphere) in the short-term.⁴⁰ PTQ did not demonstrate obvious clinical benefit compared to control injection of normal saline.⁴² A small study revealed significant improvement at 6 weeks post injection but no difference between Bulkamid and Permacol.⁴¹

Discussions

In managing patients with recent onset fecal incontinence, including in the nursing home setting, the traditional measures to minimize fecal incontinence due to treatable causes such as fecal impaction or infectious diarrhea are applicable.⁴⁵ For new onset fecal incontinence, if a rectal examination has not been performed in the recent past few days, a digital rectal examination should be performed to exclude fecal impaction and overflow incontinence. If impaction is absent, basal anal sphincter tone is low and the patient is receiving stool softener or laxative, these medications should be discontinued as they may contribute to diarrhea and fecal incontinence. In the course of managing a patient with an infection (presumably with antibiotics), the empirical steps to manage new onset fecal incontinence in addition to the above is as follows. Since the patient will likely have diarrhea, a stool sample needs to be sent for C. difficile toxin assay. If the patient is on enteral nutritional supplementation, osmotic diarrhea-induced fecal incontinence should be considered if the stool studies are negative. If a patient is receiving stool softener or laxative as prophylactic treatment, discontinuing the medication may improve incontinence. If a patient does not have fecal impaction, anal sphincter tone appears to be normal, is off stool softener or laxative, and not on enteral nutritional supplementation; or if C. difficile is positive, there is blood and pus (leukocyte) in the stool, bloody diarrhea, rectal bleeding, significant abdominal pain, referral for lower gastrointestinal endoscopic examination may be indicated to further work up the etiology of fecal incontinence⁴⁷. After these treatable conditions are excluded, some individual may persist with fecal incontinence and are in need of adjunct treatments.

Observational studies mostly with limited sample sizes reported promising results^18–39. In some patients, repeated injections were necessary^30,36,39. The mechanism of the impact of injection treatment was not fully understood. Effect on anal sphincter pressures was highly variable. Significant improvements in the length of the high-pressure zone²¹, asymmetry index²¹ and maximum tolerable rectal volume³³ were also suggested.

In a recent Cochrane report⁴⁴ all randomized or quasirandomized controlled trials comparing use of injectable bulking agents with any alternative treatments or placebo were reviewed. Four eligible RCT were identified with a total of 134 patients.^40–43 All trials except one were at an uncertain or high risk of bias. Most trials reported short-term benefits regardless of the material used as outcome measures improved over time. The report noted that a definitive conclusion could not be drawn regarding the efficacy of peri-anal injection of bulking agents for fecal incontinence due to the limited number of identified randomized controlled trials and associated methodological weaknesses. Within the available data, there was no reliable evidence for effectiveness of one treatment over another in improving fecal incontinence. The report concluded that larger well-designed trials with adequate numbers of subjects using reliable validated outcome measures would be needed to allow definitive assessment of the treatment for both safety and effectiveness.

Adequate relief⁴⁶ has not been used as an outcome measure in injection treatment studies. The approach has limitations. Adequate relief has been criticized as subjective and a relatively insensitive outcome measure.¹² As an endpoint it is confounded with initial symptom severity as measured by baseline reporting of adequate relief. The confounding effects, however, can be eliminated by excluding those who report adequate relief at screening from study participation.⁴⁷ Despite these limitations adequate relief remains a specific patient-centered outcome measure.

None of the OS and RCT reported to date has focused attention on nursing home residents. The consideration to add injection treatment in nursing home resident is based on combining it with prompted voiding⁸ to mitigate the constraints of dementia and mobility. Cost-effectiveness analysis will be needed to determine prevention of peri-anal skin complications if achievable will offset the cost of injection treatment.

Conclusion

Further studies are warranted to assess patient-centered outcomes (e.g. adequate relief) in addition to the attenuation of severity of incontinence symptoms in ambulatory patients. In nursing home residents, cost-effectiveness studies combining injection treatment and prompted voiding (to mitigate constraints of immobility and dementia) in preventing peri-anal skin complications deserves to be considered.

Abbreviations

GAX

glutaraldehyde cross-linked

NASHA Dx gel

stabilized nonanimal hyaluronic acid with dextranomer

PTQ

polydimethylsiloxane elastomer particle paste silicone biomaterial

QOL

quality of life

OS

observational study

RCT

randomized controlled trail