- Email: [email protected]
Efficacy of pentosan polysulfate in the treatment of interstitial cystitis: A meta-analysis
ADULT
UROLOGY
ELSEVIER
EFFICACY OF PENTOSAN POLYSULFATE IN THE TREATMENT OF INTERSTITIAL CYSTITIS: A META-ANALYSIS PIEWAY
HWANG,
BARBARA
AUCLAIR, DANETTE BEECHINOR, THOMAS R. EINARSON
MICHELLE
DIMENT,
AND
ABSTRACT Objectives. To determine the efficacy of pentosan polysulfate (Elmiron) compared to placebo in the treatment of interstitial cystitis. Methods. The data sources used were MEDLINE, Excerpta Medica, and International Pharmaceutical Abstracts databases, and the manufacturer. Bibliographies of articles obtained were reviewed. The keywords used were pentosanpolysulfate, pentosanpolysulfate sodium, and pentosan. Inclusion criteria were blinded selection of English language, prospective, randomized, placebo-controlled comparative trials; ~8 weeks’ duration; ~-300 mg daily; adult humans with 2 1 symptoms including pain, urgency, frequency, and nocturia; symptoms for 2 12 months; normal urinalysis; negative findings for urine culture and cytology. Exclusion criteria were hemorrhagic cystitis; drug-, microbial-, or radiation-induced cystitis; carcinoma in situ; other influencing diseases. The outcome of success was defined as a 250% decrease in pain, urgency, frequency, and nocturia. The number of successes was extracted by blinded investigators, treating withdrawals as failures. The percentage difference in success rates of pentosan polysulfate and placebo, and the number needed to treat (NNT) were determined for each variable; P values and 95% confidence intervals (Cls) were determined for combined data. Homogeneity of effect was determined by calculating 0 (chi-squared). Article quality was assessed using the Chalmers scale to determine if quality affected outcome. Effective inter-rater reliability was determined using Rosenthal’s method. Significance was set at P ~0.05. Results. Four studies were included. Data were extracted from all four studies for pain (n = 398), three for urgency (n = 306), two for frequency (n = 160). and one study for nocturia (n = 106). The differences (95% confidence limits) were pain: 16.6% (95% Cl 8%, 25%), NNT = 7; urgency: 13.0% (1 .O%, 25%), NNT = 7.5: frequency: 16.7% (2.3%, 31.10/o), NNT = 6; nocturia: - 1% (- 19.8%, 2 1.8%). P values from homogeneity tests were not significant. Mean quality scores were 63.8%, 48.1%, 50.4%, and 65.6%, respectively, in the four studies; the effective inter-rater reliability was 0.96. Results did not differ when weighted by quality score. Conclusions. Pentosan polysulfate is more efficacious than placebo in the treatment of pain, urgency, and frequency associated with interstitial cystitis. Pentosan polysulfate is not significantly different from placebo UROLOGY 50: 39-43, 1997. 0 1997, Elsevier Sciin treating nocturia associated with interstitial cystitis. ence Inc. All rights reserved.
This research was funded by internal funds of the Doctor of Pharmacy Program at the University of Toronto, Course PHM 60.5. None of the material in this manuscript represents the policy or position of the Canadian Armed Forces or of the Government of Canada. From the Faculty of Pharmacy, University of Toronto, Toronto; Junssen-Ortho Canada Ltd.; Fuculty qf Pharmacy, University of Montreal, Montreal; Canadian Military Hospital, Langley, British Columbia; Wellesley Hospital, Toronto; and Department of Clinical Pharmacology, Hospital for Sick Children, Toronto, Canada Reprint requests: Dr. Thomas R. Einarson, Faculty of Pharmacy, University qfToronto, 19 Russell Street, Toronto, ONM5S 2S2, Canada Submitted: November 1, 1996, accepted (with revisions):January 8, 1997 0 1997, E~SEVIER SCIENCE ALL RIGHTS RESERVED
nterstitial cystitis Itreatment syndrome lacking moda1ity.l
(IC) is a chronic clinical a consistently successful It is largely a diagnosis of exclusion and is characterized by abdominal and perineal pain, severe urinary urgency, frequency, and nocturia. In some cases, the symptoms may be so severe as to be debilitating; many patients are unable to work or socialize, and suicidal ideations are not uncomm0n.l Patients suffer symptoms of IC for an average of 7 years before a diagnosis is made.l In the United States and Canada, 85,000 people have been diagnosed with IC, and although anyone may be affected,
INC. PII
0090-4295/97/$17,00 SOO90-4295(97)00110-6
39
90% of sufferers are women in their forties and fifties.2 Although numerous theories have been proposed, IC remains a condition of uncertain etiology. ‘a3 The bladder surface mucous coat contains glycosaminoglycans (GAGS), and appears to be the primary defensive layer lining the bladder urothelium, protecting it from bacteria and other harmful urinary constituents. One hypothesis is that there is a defect in this GAG layer, allowing diffusion of bacteria, toxins, and irritating components of urine through to the underlying bladder wall. This mechanism is thought to initiate the symptom complex.4-9 Pentosan polysulfate (PPS, Elmiron) is a semisynthetic sulfated polysaccharide that serves as a synthetic glycosaminoglycan. It is thought to adhere to the bladder surface, supplementing the defective natural GAG layer.” Its efficacy has been examined in a limited number of trials and results have varied. When individual studies are too small or too specific in design to draw conclusions, combining the findings utilizing meta-analytic techniques can strengthen the evidence of treatment efficacy. The purpose of this meta-analysis was to determine the clinical efficacy of PPS as compared to placebo in the treatment of interstitial cystitis. MATERIAL
AND
METHODS
DATA SOURCES A computerized literature search was conducted using the MEDLINE, Excerpta Medica, and International Pharmaceutical Abstracts (IPA) databases. The search was conducted on MEDLINE (January 1966 to June 1994) by using the Index Medicus medical subject heading pentosanpolysulfate. In Excerpta Medica (January 1988 to June 1994) and IPA (January 1973 to June 1994), the following keywords were used: pentosanpolysulfate, pentosan, and pentosanpolysulfate sodium. Bibliographies of original and review articles obtained from the literature search were examined to ensure all potential references had been identified. The manufacturer of PPS in Canada (Baker-Cummins, Kirkland, QC) was contacted for any additional references.
INCLUSIONAND EXCLUSION
CRITERIA The following criteria were used to select studies for inclusion: (1) study presentation: English language; (2) study design: original prospective, randomized, placebo-controlled comparative trial, minimum 8 weeks’ duration, a PPS minimum daily dose of 300 mg orally; (3) study subjects: human adults presenting with one or more symptoms of interstitial cystitis including pain, urgency, frequency, and nocturia, a symptom history of at least 12 months, normal urinalysis, and negative findings for urine culture and cytology. The following disease states constituted exclusion criteria: hemorrhagic cystitis; drug-, microbial-, or radiation-induced cystitis; carcinoma in situ; and other influencing diseases such as diabetic nephropathy. Blinded study selection was performed by two investigators using only the selection criteria to avoid selection bias. The methods section of each article was photocopied, all identi-
40
fying statements and titles (ie, names of investigators, journals, geographical locations, dates) were removed, and an identification number generated by a hand calculator was randomly assigned. If there was disagreement between the investigators, consensus was reached on the eligibility of a study for inclusion in the meta-analysis. DATA EXTRACTION Data from the results sections were extracted by two investigators in a blinded manner, similar to that of the study selection, using the results sections that had been photocopied and had identifiers removed. The outcome of success was determined for the symptom variables of pain, urgency, frequency, and nocturia. Success was defined as a decrease in a symptom by 50% or more. The number of subjects reporting success for each variable was independently extracted. Study withdrawals were included in the analysis and treated as failures. Disagreement between the two investigators on the values for outcomes was settled through consensus. All data for each outcome variable were entered into a 2 X 2 table format for statistical analysis. STATISTICAL ANALYSIS For each of the four variables, the percent difference between success rates of PPS and placebo and the corresponding number needed to treat (NNT) were determined. NNT is defined as the number of patients requiring treatment to produce a 50% improvement in a symptom in 1 patient. We calculated 95% confidence intervals for individual studies. Data were combined using the method described by DerSimonian and Laird” with homogeneity chi square (Q) being calculated to evaluate the combinability of results. The quality of all studies included in the meta-analysis was evaluated by four investigators using a modification of the scale of Chalmers et al.” The articles were evaluated based on the following criteria: basic descriptive material; the study protocol; blinding procedures; testing procedures; statistical analysis; and presentation of results. The maximum number of points was 100, and a quality score index was determined from the total number of points earned for each study. Interrater reliability was determined using Rosenthal’s method13 for effective reliability. A coefficient of 20.80 was considered adequate evidence of reliability of scores. Data were reanalyzed, weighted by quality to determine if outcomes were influenced.
RESULTS
Ten articles were identified in the original search. Four studies were eligible for meta-analysis based on the inclusion and exclusion criteria, for a total of 448 subjects.5-7~‘4 Subject demographics are summarized in Table I. Our inclusion criteria specified that selected subjects have a normal urinalysis; however, most authors did not mention urinalysis. These studies were accepted provided there was no evidence of abnormal urinalysis. One study I4 was suspected to be a continuation of an earlier study.2 It was confirmed by contacting the author (G. Benson, personal communication) that the studies were separate and involved no overlap of patients. Six studies were rejected, including one by Parsons” that duplicated the study of Mulholland et al.,’ four were not randomized, placeboUROLOGY
50 (11, I997
TABLE
1. Demographic data of patients studies used in meta-analysis
Demographic
Characteristics
No. women No. men No. not specified No. patients total Mean age (yr) No. withdrawals No. in pentosanpolysulfate No. in placebo group
in all
Totals
group
413 (93.2%) 30 5* 448 44.2 53 168’ 165+
* Reported gender only for those subjects who completed the study.‘O ’ The number of patients reporting outcomes di;ffered for each variable in one study, ” therefore it was not possible to detennine the totalnumber in pentosanpolysulfate and placebo groups; the value reported reflects the three remaining studies.
controlled studies,8,16-‘8 and one focused on hemorrhagic cystitis.lg Overall success rates for pentosan were 37% for pain, 28% for urgency, 54% for frequency, and 48% for nocturia. Results were significantly higher than placebo for pain, urgency, and frequency, but not for nocturia. Table II summarizes the results of the significance tests used in the meta-analysis. Homogeneity of effects was confirmed by the nonsignificant P values for studies evaluating pain, urgency, and frequency (Table III). The mean quality index scores were 63.8% (Parsons et u1.i4), 48.1% (Holm-Bentzen et uI.~>, 50.4% (Parsons and Mulholland6), and 65.6% (Mulholland et a1.7). The average inter-rater correlation of 0.867 produced an effective inter-rater reliability of 0.96 (P
The cause of interstitial cystitis remains unknown and the diagnosis is based on exclusion. It has been stated that a difference, to be a difference, must make a difference. The small but significant difference in the efficacy of PPS compared to placebo in treating pain, urgency, and frequency may have substantial clinical implications in those patients for whom it is effective. Because of the lack of successful alternative treatments, PPS is a useful addition. Several trials could not be included in this metaanalysis, two of which were case series studies. One of thesel’ reported a 58% improvement in pain (n = 60), whereas another8 reported a 92% improvement in the pain, urgency, and frequency associated with interstitial cystitis (n = 24). Both trials support our results. Nonsignificant results were found for nocturia, and this finding may be partly explained by study UROLOGY
50 (11, 1997
duration. Although pain and urgency may respond quickly to treatment, changes in voiding profiles may take many months. The studies in our sample evaluated patients after 3 months of treatment. A final consideration is that of statistical power for nocturia. Data could be extracted from one study having a sample size of 46 subjects, thus providing inadequate information to make a sound conclusion. A clinical trial is warranted to settle the issue. The recommended dose of PPS is 300 mg/day. The acquisition cost of PPS is $1.19 per lOO-mg capsule (Baker-Cummins, personal communication), thus the monthly cost would be $107.10 plus dispensing fees. This cost may be justified for those patients who improve with PPS when compared to the immeasurable costs of chronic pain, time off work, cost of other treatment, and multiple physician consultations, as well as psychologic and social implications. An in-depth economic analysis that incorporates the burden of illness is warranted and we recommend this issue be addressed in future studies. Meta-analysis is defined by DerSimonian and Lairdl’ as the statistical analysis of a collection of analytical results for the purpose of integrating the findings. Any meta-analysis will have some limitations involved in the process and this review was no exception. Publication bias is a common problem as publishers and authors may favor the publication of studies with positive results. This preference may prevent the inclusion of unpublished negative studies that could change results from significant to nonsignificant. Although the manufacturer of PPS was contacted, no other attempt was made to verify the existence of unpublished studies. Rosenthal” describes a method to calculate the number of unpublished studies that must exist to change results from significant to nonsignificant (P = 0.05). We determined that 12 studies on urgency and 81 studies on pain with zero difference would be required to change our results. Interstitial cystitis is not a common disease; all four studies used in this meta-analysis were multicentered trials requiring a great deal of cooperation and resources to obtain a sufficient sample size. Considerable time and effort has been invested in conducting these trials, making the existence of an unpublished study a remote possibility. A potential consideration in interpreting results of this meta-analysis is that three of the four studies meeting our inclusion criteria were performed by groups comprised of the same investigators.6l7,14 Data for urgency were extracted from these three studies only; however, data for pain were extracted from all four studies. One could argue that a systematic difference may be present as a result of the same investigators conducting similar studies, 41
TABLE
II.
Outcome Pain
Aggregated
Authors
Combined
Urgency
Pentosan Placebo Total
76 41 117
128 153 281
204 194 398
16.6
Pentosan Placebo Total Pentosan Placebo Total Pentosan Placebo Total
22 13 35 12 5 17 11 5 16
52 61 113 20 23 43 40 42 82
74 74 148 32 28 60 51 47 98
12.1
Pentosan Placebo Total
45 23 68
112 126 238
157 149 306
12.9
et CJ/.~ Pentosan
23 17 40 20 10 30
11 14 25 29 36 65
34 31 65 49 46 95
12.8
Placebo Total Pentosan Placebo Total Pentosan Placebo Total
43 27 70
40 50 90
83 77 160
16.7
Pentosan Placebo Total
25 26 51
27 27 54
52 53 106
-1 .o
total
et al.’
total
Parsons and Mulholland’j Combined
Nocturia
KEY: NNT = number
total
Holm-Bentzen
needed
et CI/.~
to weal, NIA = not applicable-
d$f erence
is negative
thus the potential for bias exists. The study by Holm-Bentzen et aL5 was nonsignificant, showing only 8.3% improvement in 105 patients. However, that study had the lowest quality score, which may be reflected in the results. On the other hand, the combined rate of the other three studies was 18.9% in 293 patients and statistically significant, while two of those three studies were nonsignificant, possibly due to sample size. 42
Percent Difference 20.2
et c11.~
Holm-Bentzen
Total 74 74 148 52 53 105 27 20 47 51 47 98
Parsons and Mulholland6
Frequency
Patients Failure
with pentosan
46 61 107 30 35 65 15 17 32 37 40 77
Parsons et a/. I4
Combined
treated
28 13 41 22 18 40 12 3 15 14 7 21
et a/,7
Mulholland
Success
outcomes
Pentosan Placebo Total Pentosan Placebo Total Pentosan Placebo Total Pentosan Placebo Total
Parsons and Mulhollancf’ Mulholland
for patient
Grout,
Parsons et al. I4
Holm-Bentzen
results
or placebo
95% Confidence Limit Lower - UuDer 4.8
NNT
35.8
5
8.3
-15.8
32.5
12
29.4
-2.5
61.4
4
12.5
-7.0
32.1
8
25.2
7
-3.0
27
9
19.6
-8.0
47
6
10.9
-7.0
29
10
25
8
44
8
40
6
31.1
6
21.8
N/A
19.1
8.0
1 .o
-19
-2.0
2.3
-19.8
CP = 0.887).
The possibility of subject selection bias exists. Two of the studies selected subjects that had moderate or severe disease,7z14 whereas the, remaining studies did not specify disease severity.‘,6 We recommend that future studies stratify results according to disease severity. This stratification will allow a more precise definition of the patient groups who will receive the most benefit from PPS. In addition, an extended folUROLOGY
50 01, 1997
TABLE Variable Pain Urgency Frequency Nocturia
I I I.
Homogeneity
of effects
cl*
Degrees of Freedom
P
2.33 0.45 0.17 N/A’
3 2 1 N/At
0.507 0.799 0.680 N/A’
* Chi-square Q calculated using DerSimonian and Laird’s ’ Not applicable; only one study was evaluated.
method.”
low-up would provide insight into the long-term effectiveness of the drug. Our inclusion criteria restricted studies to the English language, which may create some bias. This bias was minimized since our literature search did not identify any trials published in other languages. CONCLUSIONS Pentosan polysulfate is more efficacious than placebo in the treatment of pain, urgency, and frequency associated with interstitial cystitis, however, it is not significantly different from placebo in the treatment of nocturia. REFERENCES 1. Ratliff T, Klutke C, and McDougall E: The etiology of interstitial cystitis. Urol Clin North Am 21(l): 21-30, 1994. 2. Held PJ, Hanno PM, Wein AJ, Pauly MV, and Cann MA: Epidemiology of interstitial cystitis: 2, in Hanno PM, Staskin DR, Krane RJ, and Wein AJ (Eds): Interstitial Cystitis. New York, Springer-Verlag, 1990, chapt 4, pp 29-48. 3. Koziol J: Epidemiology of interstitial cystitis. Urol Clin North Am 21: 7-20, 1994. 4. Parsons CL: Prevention of urinary tract infection by the exogenous glycosaminoglycan sodium pentosanpolysulfate. J Urol 127: 167-169,1982. 5. Holm-Bentzen M, Jacobsen F, Nerstrom B, Lose G, Kristensen J, Pedersen R, Krarup T, Feggetter J, Bates P, Bar-
UROLOGY
50 (I), 1997
nard R, et al: A prospective double-blind clinically controlled multicenter trial of sodium pentosanpolysulfate in the treatment of interstitial cystitis and related painful bladder disease. J Urol 138: 503-507, 1987. 6. Parsons C, and Mulholland G: Successful therapy of interstitial cystitis with pentosanpolysulfate. J Urol 138: 513516, 1987. 7. Mulholland G, Hanno P, Parsons C, Sant G, and Staskin D: Pentosan polysulfate sodium for therapy of interstitial cystitis. A double-blind placebo-controlled clinical study. Urology 35: 553-559, 1990. 8. Fritjofsson A, Fall M, Juhlin R, Persson B, and Ruutu M: Treatment of ulcer and nonulcer interstitial cystitis with sodium pentosanpolysulfate: a multicenter trial. J Urol 138: 508-512,1987. 9. Parsons CL, and Koprowski PF: Interstitial cystitis: successful management by increasing urinary voiding intervals. Urology 37: 207-212, 1991. 10. Pentosan polysulfate (ElmironTM) product monograph. Kirkland, QC, Baker-Cummins, 1996. 11. DerSimonian R, and Laird N: Meta-analysis in clinical trials. Controlled Clin Trials 7: 177-188, 1986. 12. Chalmers TC, Smith H Jr, Blackbum B, Silverman B, Schroeder B, Reitman D, and Ambroz A: A method for assessing the quality of a randomized control trial. Controlled Clin Trials 2: 31-49, 1981. 13. Rosenthal R: Me&-Analytic Procedures for Social Research. Beverly Hills, Calif, Sage Publications, 1984. 14. Parsons C, Benson G, Childs S, Hanno P, Sant G, and Webster G: A quantitatively controlled method to study prospectively interstitial cystitis and demonstrate the efficacy of pentosanpolysulfate. J Urol 150: 845-848, 1993. 15. Parsons C: Sodium pentosanpolysulfate treatment of interstitial cystitis: an update. Urology 29(suppl 4): 14-16, 1987. 16. MacDermott G, Charpied G, Tesluk H, and Stone A: Can histological assessment predict the outcome in interstitial cystitis? Br J Urol 67: 44-47, 1991. 17. Hanno P, and Wein A: Conservative therapy of interstitial cystitis. Semin Urol 9: 143-147, 1991. 18. Parsons C, Schmidt J, and Pollen J: Successful treatment of interstitial cystitis with sodium pentosanpolysulfate. J Urol 130: 51-53, 1983. 19. Hampson S, and Woodhouse C: Sodium pentosanpolysulfate in the management of haemorrhagic cystitis: experience with 14 patients. Eur Uro125: 40-42, 1994. 20. Rosenthal R: The file drawer problem and tolerance for null results. Psychological Bull 86: 638-641, 1979.
43
UROLOGY
ELSEVIER
EFFICACY OF PENTOSAN POLYSULFATE IN THE TREATMENT OF INTERSTITIAL CYSTITIS: A META-ANALYSIS PIEWAY
HWANG,
BARBARA
AUCLAIR, DANETTE BEECHINOR, THOMAS R. EINARSON
MICHELLE
DIMENT,
AND
ABSTRACT Objectives. To determine the efficacy of pentosan polysulfate (Elmiron) compared to placebo in the treatment of interstitial cystitis. Methods. The data sources used were MEDLINE, Excerpta Medica, and International Pharmaceutical Abstracts databases, and the manufacturer. Bibliographies of articles obtained were reviewed. The keywords used were pentosanpolysulfate, pentosanpolysulfate sodium, and pentosan. Inclusion criteria were blinded selection of English language, prospective, randomized, placebo-controlled comparative trials; ~8 weeks’ duration; ~-300 mg daily; adult humans with 2 1 symptoms including pain, urgency, frequency, and nocturia; symptoms for 2 12 months; normal urinalysis; negative findings for urine culture and cytology. Exclusion criteria were hemorrhagic cystitis; drug-, microbial-, or radiation-induced cystitis; carcinoma in situ; other influencing diseases. The outcome of success was defined as a 250% decrease in pain, urgency, frequency, and nocturia. The number of successes was extracted by blinded investigators, treating withdrawals as failures. The percentage difference in success rates of pentosan polysulfate and placebo, and the number needed to treat (NNT) were determined for each variable; P values and 95% confidence intervals (Cls) were determined for combined data. Homogeneity of effect was determined by calculating 0 (chi-squared). Article quality was assessed using the Chalmers scale to determine if quality affected outcome. Effective inter-rater reliability was determined using Rosenthal’s method. Significance was set at P ~0.05. Results. Four studies were included. Data were extracted from all four studies for pain (n = 398), three for urgency (n = 306), two for frequency (n = 160). and one study for nocturia (n = 106). The differences (95% confidence limits) were pain: 16.6% (95% Cl 8%, 25%), NNT = 7; urgency: 13.0% (1 .O%, 25%), NNT = 7.5: frequency: 16.7% (2.3%, 31.10/o), NNT = 6; nocturia: - 1% (- 19.8%, 2 1.8%). P values from homogeneity tests were not significant. Mean quality scores were 63.8%, 48.1%, 50.4%, and 65.6%, respectively, in the four studies; the effective inter-rater reliability was 0.96. Results did not differ when weighted by quality score. Conclusions. Pentosan polysulfate is more efficacious than placebo in the treatment of pain, urgency, and frequency associated with interstitial cystitis. Pentosan polysulfate is not significantly different from placebo UROLOGY 50: 39-43, 1997. 0 1997, Elsevier Sciin treating nocturia associated with interstitial cystitis. ence Inc. All rights reserved.
This research was funded by internal funds of the Doctor of Pharmacy Program at the University of Toronto, Course PHM 60.5. None of the material in this manuscript represents the policy or position of the Canadian Armed Forces or of the Government of Canada. From the Faculty of Pharmacy, University of Toronto, Toronto; Junssen-Ortho Canada Ltd.; Fuculty qf Pharmacy, University of Montreal, Montreal; Canadian Military Hospital, Langley, British Columbia; Wellesley Hospital, Toronto; and Department of Clinical Pharmacology, Hospital for Sick Children, Toronto, Canada Reprint requests: Dr. Thomas R. Einarson, Faculty of Pharmacy, University qfToronto, 19 Russell Street, Toronto, ONM5S 2S2, Canada Submitted: November 1, 1996, accepted (with revisions):January 8, 1997 0 1997, E~SEVIER SCIENCE ALL RIGHTS RESERVED
nterstitial cystitis Itreatment syndrome lacking moda1ity.l
(IC) is a chronic clinical a consistently successful It is largely a diagnosis of exclusion and is characterized by abdominal and perineal pain, severe urinary urgency, frequency, and nocturia. In some cases, the symptoms may be so severe as to be debilitating; many patients are unable to work or socialize, and suicidal ideations are not uncomm0n.l Patients suffer symptoms of IC for an average of 7 years before a diagnosis is made.l In the United States and Canada, 85,000 people have been diagnosed with IC, and although anyone may be affected,
INC. PII
0090-4295/97/$17,00 SOO90-4295(97)00110-6
39
90% of sufferers are women in their forties and fifties.2 Although numerous theories have been proposed, IC remains a condition of uncertain etiology. ‘a3 The bladder surface mucous coat contains glycosaminoglycans (GAGS), and appears to be the primary defensive layer lining the bladder urothelium, protecting it from bacteria and other harmful urinary constituents. One hypothesis is that there is a defect in this GAG layer, allowing diffusion of bacteria, toxins, and irritating components of urine through to the underlying bladder wall. This mechanism is thought to initiate the symptom complex.4-9 Pentosan polysulfate (PPS, Elmiron) is a semisynthetic sulfated polysaccharide that serves as a synthetic glycosaminoglycan. It is thought to adhere to the bladder surface, supplementing the defective natural GAG layer.” Its efficacy has been examined in a limited number of trials and results have varied. When individual studies are too small or too specific in design to draw conclusions, combining the findings utilizing meta-analytic techniques can strengthen the evidence of treatment efficacy. The purpose of this meta-analysis was to determine the clinical efficacy of PPS as compared to placebo in the treatment of interstitial cystitis. MATERIAL
AND
METHODS
DATA SOURCES A computerized literature search was conducted using the MEDLINE, Excerpta Medica, and International Pharmaceutical Abstracts (IPA) databases. The search was conducted on MEDLINE (January 1966 to June 1994) by using the Index Medicus medical subject heading pentosanpolysulfate. In Excerpta Medica (January 1988 to June 1994) and IPA (January 1973 to June 1994), the following keywords were used: pentosanpolysulfate, pentosan, and pentosanpolysulfate sodium. Bibliographies of original and review articles obtained from the literature search were examined to ensure all potential references had been identified. The manufacturer of PPS in Canada (Baker-Cummins, Kirkland, QC) was contacted for any additional references.
INCLUSIONAND EXCLUSION
CRITERIA The following criteria were used to select studies for inclusion: (1) study presentation: English language; (2) study design: original prospective, randomized, placebo-controlled comparative trial, minimum 8 weeks’ duration, a PPS minimum daily dose of 300 mg orally; (3) study subjects: human adults presenting with one or more symptoms of interstitial cystitis including pain, urgency, frequency, and nocturia, a symptom history of at least 12 months, normal urinalysis, and negative findings for urine culture and cytology. The following disease states constituted exclusion criteria: hemorrhagic cystitis; drug-, microbial-, or radiation-induced cystitis; carcinoma in situ; and other influencing diseases such as diabetic nephropathy. Blinded study selection was performed by two investigators using only the selection criteria to avoid selection bias. The methods section of each article was photocopied, all identi-
40
fying statements and titles (ie, names of investigators, journals, geographical locations, dates) were removed, and an identification number generated by a hand calculator was randomly assigned. If there was disagreement between the investigators, consensus was reached on the eligibility of a study for inclusion in the meta-analysis. DATA EXTRACTION Data from the results sections were extracted by two investigators in a blinded manner, similar to that of the study selection, using the results sections that had been photocopied and had identifiers removed. The outcome of success was determined for the symptom variables of pain, urgency, frequency, and nocturia. Success was defined as a decrease in a symptom by 50% or more. The number of subjects reporting success for each variable was independently extracted. Study withdrawals were included in the analysis and treated as failures. Disagreement between the two investigators on the values for outcomes was settled through consensus. All data for each outcome variable were entered into a 2 X 2 table format for statistical analysis. STATISTICAL ANALYSIS For each of the four variables, the percent difference between success rates of PPS and placebo and the corresponding number needed to treat (NNT) were determined. NNT is defined as the number of patients requiring treatment to produce a 50% improvement in a symptom in 1 patient. We calculated 95% confidence intervals for individual studies. Data were combined using the method described by DerSimonian and Laird” with homogeneity chi square (Q) being calculated to evaluate the combinability of results. The quality of all studies included in the meta-analysis was evaluated by four investigators using a modification of the scale of Chalmers et al.” The articles were evaluated based on the following criteria: basic descriptive material; the study protocol; blinding procedures; testing procedures; statistical analysis; and presentation of results. The maximum number of points was 100, and a quality score index was determined from the total number of points earned for each study. Interrater reliability was determined using Rosenthal’s method13 for effective reliability. A coefficient of 20.80 was considered adequate evidence of reliability of scores. Data were reanalyzed, weighted by quality to determine if outcomes were influenced.
RESULTS
Ten articles were identified in the original search. Four studies were eligible for meta-analysis based on the inclusion and exclusion criteria, for a total of 448 subjects.5-7~‘4 Subject demographics are summarized in Table I. Our inclusion criteria specified that selected subjects have a normal urinalysis; however, most authors did not mention urinalysis. These studies were accepted provided there was no evidence of abnormal urinalysis. One study I4 was suspected to be a continuation of an earlier study.2 It was confirmed by contacting the author (G. Benson, personal communication) that the studies were separate and involved no overlap of patients. Six studies were rejected, including one by Parsons” that duplicated the study of Mulholland et al.,’ four were not randomized, placeboUROLOGY
50 (11, I997
TABLE
1. Demographic data of patients studies used in meta-analysis
Demographic
Characteristics
No. women No. men No. not specified No. patients total Mean age (yr) No. withdrawals No. in pentosanpolysulfate No. in placebo group
in all
Totals
group
413 (93.2%) 30 5* 448 44.2 53 168’ 165+
* Reported gender only for those subjects who completed the study.‘O ’ The number of patients reporting outcomes di;ffered for each variable in one study, ” therefore it was not possible to detennine the totalnumber in pentosanpolysulfate and placebo groups; the value reported reflects the three remaining studies.
controlled studies,8,16-‘8 and one focused on hemorrhagic cystitis.lg Overall success rates for pentosan were 37% for pain, 28% for urgency, 54% for frequency, and 48% for nocturia. Results were significantly higher than placebo for pain, urgency, and frequency, but not for nocturia. Table II summarizes the results of the significance tests used in the meta-analysis. Homogeneity of effects was confirmed by the nonsignificant P values for studies evaluating pain, urgency, and frequency (Table III). The mean quality index scores were 63.8% (Parsons et u1.i4), 48.1% (Holm-Bentzen et uI.~>, 50.4% (Parsons and Mulholland6), and 65.6% (Mulholland et a1.7). The average inter-rater correlation of 0.867 produced an effective inter-rater reliability of 0.96 (P
The cause of interstitial cystitis remains unknown and the diagnosis is based on exclusion. It has been stated that a difference, to be a difference, must make a difference. The small but significant difference in the efficacy of PPS compared to placebo in treating pain, urgency, and frequency may have substantial clinical implications in those patients for whom it is effective. Because of the lack of successful alternative treatments, PPS is a useful addition. Several trials could not be included in this metaanalysis, two of which were case series studies. One of thesel’ reported a 58% improvement in pain (n = 60), whereas another8 reported a 92% improvement in the pain, urgency, and frequency associated with interstitial cystitis (n = 24). Both trials support our results. Nonsignificant results were found for nocturia, and this finding may be partly explained by study UROLOGY
50 (11, 1997
duration. Although pain and urgency may respond quickly to treatment, changes in voiding profiles may take many months. The studies in our sample evaluated patients after 3 months of treatment. A final consideration is that of statistical power for nocturia. Data could be extracted from one study having a sample size of 46 subjects, thus providing inadequate information to make a sound conclusion. A clinical trial is warranted to settle the issue. The recommended dose of PPS is 300 mg/day. The acquisition cost of PPS is $1.19 per lOO-mg capsule (Baker-Cummins, personal communication), thus the monthly cost would be $107.10 plus dispensing fees. This cost may be justified for those patients who improve with PPS when compared to the immeasurable costs of chronic pain, time off work, cost of other treatment, and multiple physician consultations, as well as psychologic and social implications. An in-depth economic analysis that incorporates the burden of illness is warranted and we recommend this issue be addressed in future studies. Meta-analysis is defined by DerSimonian and Lairdl’ as the statistical analysis of a collection of analytical results for the purpose of integrating the findings. Any meta-analysis will have some limitations involved in the process and this review was no exception. Publication bias is a common problem as publishers and authors may favor the publication of studies with positive results. This preference may prevent the inclusion of unpublished negative studies that could change results from significant to nonsignificant. Although the manufacturer of PPS was contacted, no other attempt was made to verify the existence of unpublished studies. Rosenthal” describes a method to calculate the number of unpublished studies that must exist to change results from significant to nonsignificant (P = 0.05). We determined that 12 studies on urgency and 81 studies on pain with zero difference would be required to change our results. Interstitial cystitis is not a common disease; all four studies used in this meta-analysis were multicentered trials requiring a great deal of cooperation and resources to obtain a sufficient sample size. Considerable time and effort has been invested in conducting these trials, making the existence of an unpublished study a remote possibility. A potential consideration in interpreting results of this meta-analysis is that three of the four studies meeting our inclusion criteria were performed by groups comprised of the same investigators.6l7,14 Data for urgency were extracted from these three studies only; however, data for pain were extracted from all four studies. One could argue that a systematic difference may be present as a result of the same investigators conducting similar studies, 41
TABLE
II.
Outcome Pain
Aggregated
Authors
Combined
Urgency
Pentosan Placebo Total
76 41 117
128 153 281
204 194 398
16.6
Pentosan Placebo Total Pentosan Placebo Total Pentosan Placebo Total
22 13 35 12 5 17 11 5 16
52 61 113 20 23 43 40 42 82
74 74 148 32 28 60 51 47 98
12.1
Pentosan Placebo Total
45 23 68
112 126 238
157 149 306
12.9
et CJ/.~ Pentosan
23 17 40 20 10 30
11 14 25 29 36 65
34 31 65 49 46 95
12.8
Placebo Total Pentosan Placebo Total Pentosan Placebo Total
43 27 70
40 50 90
83 77 160
16.7
Pentosan Placebo Total
25 26 51
27 27 54
52 53 106
-1 .o
total
et al.’
total
Parsons and Mulholland’j Combined
Nocturia
KEY: NNT = number
total
Holm-Bentzen
needed
et CI/.~
to weal, NIA = not applicable-
d$f erence
is negative
thus the potential for bias exists. The study by Holm-Bentzen et aL5 was nonsignificant, showing only 8.3% improvement in 105 patients. However, that study had the lowest quality score, which may be reflected in the results. On the other hand, the combined rate of the other three studies was 18.9% in 293 patients and statistically significant, while two of those three studies were nonsignificant, possibly due to sample size. 42
Percent Difference 20.2
et c11.~
Holm-Bentzen
Total 74 74 148 52 53 105 27 20 47 51 47 98
Parsons and Mulholland6
Frequency
Patients Failure
with pentosan
46 61 107 30 35 65 15 17 32 37 40 77
Parsons et a/. I4
Combined
treated
28 13 41 22 18 40 12 3 15 14 7 21
et a/,7
Mulholland
Success
outcomes
Pentosan Placebo Total Pentosan Placebo Total Pentosan Placebo Total Pentosan Placebo Total
Parsons and Mulhollancf’ Mulholland
for patient
Grout,
Parsons et al. I4
Holm-Bentzen
results
or placebo
95% Confidence Limit Lower - UuDer 4.8
NNT
35.8
5
8.3
-15.8
32.5
12
29.4
-2.5
61.4
4
12.5
-7.0
32.1
8
25.2
7
-3.0
27
9
19.6
-8.0
47
6
10.9
-7.0
29
10
25
8
44
8
40
6
31.1
6
21.8
N/A
19.1
8.0
1 .o
-19
-2.0
2.3
-19.8
CP = 0.887).
The possibility of subject selection bias exists. Two of the studies selected subjects that had moderate or severe disease,7z14 whereas the, remaining studies did not specify disease severity.‘,6 We recommend that future studies stratify results according to disease severity. This stratification will allow a more precise definition of the patient groups who will receive the most benefit from PPS. In addition, an extended folUROLOGY
50 01, 1997
TABLE Variable Pain Urgency Frequency Nocturia
I I I.
Homogeneity
of effects
cl*
Degrees of Freedom
P
2.33 0.45 0.17 N/A’
3 2 1 N/At
0.507 0.799 0.680 N/A’
* Chi-square Q calculated using DerSimonian and Laird’s ’ Not applicable; only one study was evaluated.
method.”
low-up would provide insight into the long-term effectiveness of the drug. Our inclusion criteria restricted studies to the English language, which may create some bias. This bias was minimized since our literature search did not identify any trials published in other languages. CONCLUSIONS Pentosan polysulfate is more efficacious than placebo in the treatment of pain, urgency, and frequency associated with interstitial cystitis, however, it is not significantly different from placebo in the treatment of nocturia. REFERENCES 1. Ratliff T, Klutke C, and McDougall E: The etiology of interstitial cystitis. Urol Clin North Am 21(l): 21-30, 1994. 2. Held PJ, Hanno PM, Wein AJ, Pauly MV, and Cann MA: Epidemiology of interstitial cystitis: 2, in Hanno PM, Staskin DR, Krane RJ, and Wein AJ (Eds): Interstitial Cystitis. New York, Springer-Verlag, 1990, chapt 4, pp 29-48. 3. Koziol J: Epidemiology of interstitial cystitis. Urol Clin North Am 21: 7-20, 1994. 4. Parsons CL: Prevention of urinary tract infection by the exogenous glycosaminoglycan sodium pentosanpolysulfate. J Urol 127: 167-169,1982. 5. Holm-Bentzen M, Jacobsen F, Nerstrom B, Lose G, Kristensen J, Pedersen R, Krarup T, Feggetter J, Bates P, Bar-
UROLOGY
50 (I), 1997
nard R, et al: A prospective double-blind clinically controlled multicenter trial of sodium pentosanpolysulfate in the treatment of interstitial cystitis and related painful bladder disease. J Urol 138: 503-507, 1987. 6. Parsons C, and Mulholland G: Successful therapy of interstitial cystitis with pentosanpolysulfate. J Urol 138: 513516, 1987. 7. Mulholland G, Hanno P, Parsons C, Sant G, and Staskin D: Pentosan polysulfate sodium for therapy of interstitial cystitis. A double-blind placebo-controlled clinical study. Urology 35: 553-559, 1990. 8. Fritjofsson A, Fall M, Juhlin R, Persson B, and Ruutu M: Treatment of ulcer and nonulcer interstitial cystitis with sodium pentosanpolysulfate: a multicenter trial. J Urol 138: 508-512,1987. 9. Parsons CL, and Koprowski PF: Interstitial cystitis: successful management by increasing urinary voiding intervals. Urology 37: 207-212, 1991. 10. Pentosan polysulfate (ElmironTM) product monograph. Kirkland, QC, Baker-Cummins, 1996. 11. DerSimonian R, and Laird N: Meta-analysis in clinical trials. Controlled Clin Trials 7: 177-188, 1986. 12. Chalmers TC, Smith H Jr, Blackbum B, Silverman B, Schroeder B, Reitman D, and Ambroz A: A method for assessing the quality of a randomized control trial. Controlled Clin Trials 2: 31-49, 1981. 13. Rosenthal R: Me&-Analytic Procedures for Social Research. Beverly Hills, Calif, Sage Publications, 1984. 14. Parsons C, Benson G, Childs S, Hanno P, Sant G, and Webster G: A quantitatively controlled method to study prospectively interstitial cystitis and demonstrate the efficacy of pentosanpolysulfate. J Urol 150: 845-848, 1993. 15. Parsons C: Sodium pentosanpolysulfate treatment of interstitial cystitis: an update. Urology 29(suppl 4): 14-16, 1987. 16. MacDermott G, Charpied G, Tesluk H, and Stone A: Can histological assessment predict the outcome in interstitial cystitis? Br J Urol 67: 44-47, 1991. 17. Hanno P, and Wein A: Conservative therapy of interstitial cystitis. Semin Urol 9: 143-147, 1991. 18. Parsons C, Schmidt J, and Pollen J: Successful treatment of interstitial cystitis with sodium pentosanpolysulfate. J Urol 130: 51-53, 1983. 19. Hampson S, and Woodhouse C: Sodium pentosanpolysulfate in the management of haemorrhagic cystitis: experience with 14 patients. Eur Uro125: 40-42, 1994. 20. Rosenthal R: The file drawer problem and tolerance for null results. Psychological Bull 86: 638-641, 1979.
43