- Email: [email protected]
Immunoactive prophylaxis of recurrent urinary tract infections: a meta-analysis
International Journal of Antimicrobial Agents 33 (2009) 111–119
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Review
Immunoactive prophylaxis of recurrent urinary tract infections: a meta-analysis Kurt G. Naber a,∗ , Yong-Hyun Cho b , Tetsuro Matsumoto c , Anthony J. Schaeffer d a
Technical University Munich, Munich, Germany Department of Urology, St Marys Hospital, The Catholic University of Korea, Seoul, South Korea c Department of Urology, School of Medicine, University of Occupational and Environmental Health, Kitakyushu, Japan d Department of Urology, Northwestern University Medical School, Chicago, IL, USA b
a r t i c l e
i n f o
Keywords: Urinary tract infections Bacterial lysates Clinical review Meta-analysis OM-89 Uro-Vaxom SolcoUrovac
a b s t r a c t Recurrent urinary tract infections (UTIs) are very common, particularly among women in their reproductive years. Alternatives to antibacterial prophylaxis are needed, particularly measures to increase host defences. Various bacterial lysates have been proposed with this indication. The objective of this review and meta-analysis was to assess the efficacy and safety of bacterial lysates in the management of recurrent UTIs. Electronic databases identified 11 studies with the descriptors ‘urinary tract infection’, ‘immunotherapy’ or ‘vaccines’ and ‘double blind’. Seven of the studies dealt with an oral immunostimulant (OM-89), of which about five (1000 adult patients) were retained for analysis with an observation period of 6–12 months. The mean number of UTIs was significantly lower in OM-89-treated patients in all the trials analysed (mean 39%), as was the use of antibacterials. Four of the studies dealt with a vaginal vaccine, of which three small studies were retained for analysis (220 adult patients). The results suggest that this vaginal vaccine is effective when administered with a booster cycle (no recurrent UTI in 50% vs. 14% with placebo). No blind controlled studies could be identified with other bacterial lysates claiming the same indication. In conclusion, among the various immunotherapeutic products, studies were published only for one product (OM-89) that are in accordance with current standards. This product was shown to be effective under conditions of daily practice. The second product (vaginal vaccine) also appears to be effective but adequate phase III studies are necessary. © 2008 Published by Elsevier B.V. and the International Society of Chemotherapy.
1. Introduction Urinary tract infections (UTIs) are the most common medical complaint among women in their reproductive years. Women are 30 times more likely to have UTIs than men. Every year, 15% of sexually active women have at least one such infection and up to 60% of all women will develop a UTI at some time in their lives. At least one in four of these women will have a recurrence within a year [1]. UTIs are frequent in pre-menopausal, non-pregnant women who present with acute onset of dysuria, frequency or urgency, and suprapubic pain. These infections are confirmed by urine analysis documenting >10 white blood cells/mm3 and urine culture with >103 colony-forming units/mL of uropathogen in a mid-stream sample of urine. UTIs are defined as recurrent if there had been at least three episodes of uncomplicated infection documented by culture in the last 12 months. On average, each episode of acute UTI in pre-menopausal women was shown to be associated with
∗ Corresponding author. Present address: Karl-Bickleder-Str. 44c, D-94315 Straubing, Germany. Tel.: +49 9421 33369. E-mail address: [email protected] (K.G. Naber).
6.1 days of symptoms, 2.4 days of restricted activity, 1.2 days in which they were not able to attend classes or work and 0.4 days in bed [2]. Particularly in women there are several well-defined risk factors for recurrent UTI. In young women, the most important risk factors for acute cystitis are a history of previous episodes of cystitis and frequent or recent sexual activity; the use of spermicidal agents elevates the likelihood of infection. Women with frequent recurrences are more likely to have a maternal history of cystitis and to have had cystitis at an early age [3]. In post-menopausal women, UTI incidence is increased in diabetic patients (relative risk (RR) = 1.8) [4]; furthermore, it was shown that urinary incontinence (odds ratio (OR) = 5.79), a history of UTI before menopause (OR = 4.85), and non-secretor status (OR = 2.9) were most strongly associated with recurrent UTI [5,6]. Epithelial cells were collected from patients with recurrent UTI and compared with such cells obtained from controls; bacterial strains that caused cystitis adhered much more avidly to the epithelial cells from susceptible women. The presence or absence of blood group determinants on the surface of uroepithelial cells may influence an individual’s susceptibility to UTI. Susceptibility among women who do not secrete blood group antigens may be due to specific Escherichia coli-binding glycolipids
0924-8579/$ – see front matter © 2008 Published by Elsevier B.V. and the International Society of Chemotherapy. doi:10.1016/j.ijantimicag.2008.08.011
112
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
that are absent in women who secrete blood group antigens [7]. Some of these risk factors suggest possible therapeutic or prophylactic interventions. Additionally, there are diseases that seem to predispose to recurrent UTI, in particular lupus erythematosus [8] and rheumatoid arthritis [9]. 1.1. Proof of concept The idea of employing bacterial immune stimulants in order to reduce recurrent UTI was born some 40 years ago and they have been used, rather empirically, to prevent recurrent infections in the non-immunocompromised host. These early attempts were hampered at the time by a lack of knowledge of the potential mechanisms involved. Only in recent years has a better understanding of the innate immune system provided a solid rationale for such immune stimulants. The requirements for a bacterial extract to be effective are that it stimulates infection-combating, circulating elements within the lymphoid tissue and that the target organ envisioned is immunocompetent in order to translate adequately such a stimulus into an efficient response. 1.2. Bacterial immune stimulants Bacterial extracts can act as immune stimulants through the activation of monocyte-derived dendritic cells, as was shown for OM-89 [10]. Human monocyte-derived dendritic cells and circulating conventional dendritic cells co-express activating (CD32a) and inhibitory (CD32b) isoforms of immunoglobulin G (IgG) Fc gamma receptor (Fc␥R). The balance between these divergent receptors establishes a threshold of dendritic cell activation and enables immune complexes to mediate opposing effects on dendritic cell maturation and function. The activating and inhibitory Fc␥Rs on dendritic cells offer rational targets for immunotherapy based on the unique capacity of dendritic cells to play critical roles in both immunity and tolerance [11]. Some bacterial extracts are administered via the oral route with the goal of increasing the immune defences of organs with a mucous lining, although parenteral, intravaginal and intranasal applications have also been described. In animal models and in normal humans, gut-associated lymphoid tissue stimulation is able to induce a generalised response by the whole mucosa-associated lymphoid tissue (MALT), which is found in the digestive tract, the respiratory tract and the genitourinary tract. Migration and adherence of lamina propria lymphocytes are in part dictated by integrins and selectins. Lymphocyte circulation leads to the distribution of effector cells to particular parts of the body where homing of lymphocytes is controlled by the expression of various receptors on the cell surface and counter-receptors on the vascular endothelium [12]. A priori one would expect that a bacterial extract will act as a more or less specific immune stimulant for the bacteria extracted. However, this has not been the case and can be explained by the fact that different bacteria appear to share antigenic structures, to stimulate common Toll-like receptors (TLRs) and to share toxin secretion mechanisms. Innate immune recognition is based on socalled pattern recognition receptors (PRRs) or pathogen-associated molecular patterns [13] where various classes of pathogens can be recognised by the same PRRs; for example, TLR4, which recognises uropathogenic E. coli, is a critical component of the lipopolysaccharide receptor complex, which forms a detection system for Gram-negative bacteria [14]. One further example is the type VI toxin secretion pathway that is shared by Pseudomonas aeruginosa, Vibrio cholerae, Salmonella enterica, Yersinia pestis and E. coli [15].
1.3. Immunocompetence of target organs Animal studies have shown an anatomically well-organised lymphoid tissue in the urinary tract, resembling the MALT, which is associated with the epithelium of the renal pelvis and the ureter, respectively. Immunocytochemistry revealed S-100immunoreactive dendritic cells both in structurally organised and structurally non-organised lymphoid tissues [16]. Humans also have a well-organised lymphoid tissue in the urinary tract, particularly the urinary bladder. Host-specific factors associated with UTI include production of secretory IgA interfering with adhesion, presence of Tamm–Horsfall protein (THP) causing bacterial aggregation and washout, bactericidal properties of serum, and urodynamic factors, i.e. bacterial washout. THP, secreted by the renal tubules in the outer medulla, has specific receptors for several uropathogens and the bound bacteria are washed out in the urine [17,18]. Like the urinary bladder, the urethra may provide a favourable environment for colonisation by uropathogens but it is protected by powerful defence mechanisms that, besides trapping of bacteria by mucus and intermittent washout by urine, include the local production of immunoglobulins, cytokines and defensins in addition to the mobilisation of leukocytes [19]. The human penile urethra contains numerous IgA and J chain-positive plasma cells, and the epithelium expresses secretory component, the transport molecule for polymeric IgA, indicating that this region is an active site of secretory IgA-mediated immune defence. The mucosa contains intraepithelial dendritic cells, whilst T-lymphocytes are abundant and ubiquitous in all regions of the urethra. Both CD8+ and CD4+ subpopulations of T-lymphocytes are present in the lamina propria and epithelium, although CD8+ cells predominate. The majority of T-lymphocytes are positive for CD45RO (memory marker) and many are also positive for the ␣-E--7-integrin (mucosal-associated antigen). These data indicate that, in addition to the bladder, the urethra is a highly dynamic immunocompetent tissue possessing all the necessary elements for antigen presentation and both humoral and cellular mucosal immune responses. It is likely that this region plays a dominant role in protecting the urogenital tract against ascending infections [20]. 2. Methodology The primary objective of the review was to assess the efficacy and safety of immunostimulants based on bacterial lysates from a clinical point of view and taking into account clinically relevant endpoints. Meta-analysis may produce a stronger conclusion than can be provided by any individual study. The purposes of metaanalysis are to increase statistical power for primary endpoints, to increase general applicability (external validity) of findings, to resolve uncertainty when reports disagree, to provide quantitative estimates of effects, to call attention to strengths and weaknesses of a body of research in a particular area and to identify needs for new primary data collection. 2.1. Search strategy and material scrutinised Among the data sources consulted in the identification of trials were bibliographic databases (TOXLINE, PaperChase browsing the databases of the National Library of Medicine and the National Cancer Institute, i.e. MEDLINE, HealthSTAR, AIDSLINE and CANCERLIT, EMBASE, AMED, Cochrane Collaboration, PubMed (http://www.ncbi.nlm.nih.gov), TOXLINE Special, DART Special, HSDB, IRIS, ITER, GENETOX, ChemIDplus, Haz-Map), reference lists from pertinent review articles and books, and personal contacts with experts active in the area and manufacturers up to September 2008. All papers were screened and any dealing with prospec-
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
113
Fig. 1. Flowchart of double-blind studies identified and retained for analysis, by bacterial immunotherapy product.
tive clinical trials were retained for classification according to the selection criteria described below. In the case of double publications, those that were most recent and/or had appeared in a peer-reviewed journal were retained.
in the flowchart of studies identified and retained for analysis (Fig. 1).
2.1.1. Selection criteria and description of studies The search criteria were ‘urinary tract infection’ and ‘double blind’ or ‘single blind’ and ‘bacterial vaccines’ or ‘immunization’ or ‘immunotherapy’ or ‘vaccination’ or ‘vaccines’. To be retained, a study had to be a randomised, placebo-controlled, clinical trial with the primary aim of reducing the number of UTIs over a period of 6–12 months. All trials rendered eligible were classified by indication and summarised in a tabulated format by one reviewer. The standard table included a full reference, a quality rating, the type of pathology, demographic data and treatments, endpoints and adverse events. The trials were grouped by indication and by treatment. These tables were verified and discussed by the authors until consensus was reached. No formal validation process was employed. The electronic databases identified (i) urinary tract infection in 25 757 references and (ii) double or single blind in 88 121 + 10 220 references and (iii) bacterial vaccines or immunisation or immunotherapy or vaccination or vaccines in 155 209 references, but only 10 complied with the three descriptors. One additional clinical study was identified through other channels and was duly scrutinised. The search also yielded one meta-analysis [21] regarding OM-89 (Product ‘A’; Uro-Vaxom® ). As will be described in the results, three studies were eliminated because of duplication of data, particular patient selection or inadequate reporting (one study each). Furthermore, the electronic databases were searched for trials conducted with specific products identified (http://focosi. altervista.org/preventionprimaryimmunovaccine.html); that is, OM-89 capsules (Product ‘A’; five different serotypes of heat-killed uropathogenic E. coli); Product ‘B’ for intramuscular or vaginal administration (StroVac® and SolcoUrovac® , respectively; contains 10 heat-killed uropathogenic bacteria, including six different serotypes of uropathogenic E. coli, Proteus vulgaris, Klebsiella pneumoniae, Morganella morganii and Enterococcus faecalis); and Urostim® (‘Other products’; killed bacterial cells and their lysates of four microbial species: uropathogenic E. coli expressing type 1 pili, Rc mutant of E. coli, K. pneumoniae, Proteus mirabilis and E. faecalis) and Urvakol® (‘Other products’; uropathogenic E. coli, K. pneumoniae, P. mirabilis, P. aeruginosa and E. faecalis; adjuvant activity is provided by Propionibacterium acnes). However, this search did not identify additional trials complying with the predefined criteria. The resulting selection of trials is summarised
The guidelines provided by the Cochrane Collaboration Handbook for Reviews [22] and the corresponding software have been used in the analysis of the clinical data (Review Manager 4.2.7). Ordinal data were compared using the Mantel-HaenszelPeto method [23]. Significances were calculated using two-sided tests, with the threshold of significance being P ≤ 0.05 and for non-significance P > 0.1; values between P > 0.05 and P ≤ 0.1 were reported as trends. Mean and standard deviation correspond to those reported for the intention-to-treat analysis, where available, and N corresponds to the number of patients admitted to the trial and exposed to the treatment.
2.2. Analysis of data
3. Results 3.1. Product ‘A’: OM-89 (Uro-Vaxom® ) From the original seven studies identified, only five were retained for further analysis. One study [24] was excluded because, although double-blind, it dealt with a special group of patients: in a 6-month, double-blind, placebo-controlled clinical trial with cross-over, the authors administered the bacterial extract OM-89 in 70 spinal cord injury patients with chronic lower UTIs. Compared with patients given placebo, in the treated patients there was a statistically significant decrease in the degree of bacteriuria, a considerably decreased incidence of infectious episodes and a lesser requirement for antibiotics. One further trial [25] with 64 outpatients suffering from an acute recurrent UTI was excluded because there was insufficient reporting of demographics and the publication did not account for drop-outs or missing data. In this trial, dysuria, bacteriuria, leukocyturia and consumption of antibacterials were reported to show a significant reduction with OM-89 in comparison with the placebo after 6 months. The five studies [26–30] retained for further analysis present a large amount of commonalities, albeit differing in some aspects. In three studies [27–30] also men were included, but always less than 20% of the total patients. Unfortunately, the results could not be stratified by gender. In the largest study [26], however, only female patients were included, thus the proportion of men decreased to less than 7% in the total population. A summary of treatment schedules, various definitions of UTIs, admission and exclusion criteria, and the patients’ flow is given in Table 1. In all these studies, patients were recruited suffering from acute UTI, had a history of recurrent
85/81 76/74
72/68 61/59
UTIs, were treated with an antibacterial until cure while at the same time starting treatment with OM-89, one capsule daily during 3 months. Thereafter, there was a follow-up period of an additional 3 months. In one study [26] the patients received a booster treatment in Months 6–9 (one capsule daily during 10 days per month), again with an additional follow-up of 3 months, thus completing a study period of 12 months. However, the study design does not allow for a distinction between effects due to the booster treatment and the possible ‘carry-over’ effects of the initial treatment cycle. In all the studies examined the patients were comparable at baseline with regard to demographic and clinical data. In addition to the planned visits, the patients could return to the consultation if they were symptomatic; these UTIs were assimilated to the following planned visit. Only two trials [26,27] reported the type of bacteria identified at baseline, which was E. coli in ca. 70% of cases in both trials. In the study by Bauer et al. (unpublished data on file, OM Pharma) it was found that the percentages of patients with a UTI during the study period were: GCP, good clinical practice; ITT, intent to treat analysis; PP, per protocol (analysis excluding protocol violators); N.D., not defined; S.D., standard deviation. a Midstream urine sample. b Rated sum score: fever, dysuria, pollakisuria, stranguria, burning, flank, suprapubic and perineal pain, fullness, haematuria, odorous and cloudy urine, other. c 105 bacteria/mL midstream urine or 104 bacteria/mL catheter urine. d Completed 12 months in parentheses.
17/95 17/133 45 (16.1) 50.4 (19.8) 45.3 (18.4) 51.2 (20.9) −−− −−− −−− −−− +++ +++ >105 c >105 c GCP (PP) Not stated (PP) Schulman et al. [29] Tammen [30]
−−− −−− +++ >105 Not stated (PP) Magasi et al. [28]
[++]b [++]b [++]b >105 a GCP (ITT/PP) Pisani et al. [27]
+++ GCP (ITT/PP) Bauer et al. [26]
≥103a
+++
+++
Pyelonephritis, catheter, neurogenic bladder, etc. Antibiotic-resistant UTI, pyelonephritis, catheter, obstruction, etc. Pyelonephritis, catheter, obstruction, etc. Anatomical abnormalities, etc. Post-coital UTI, pyelonephritis, catheter, obstruction, etc.
Bacteriuria N.D.
58/54 26/140 Range 16–82 Bacteriuria
63/59
66/71 9/128
0/453
46.9 (15.9)
81/79
195/191 (184/186)d 231/222 39.8 (15.1) 41.7 (15.3)
Bacteriuria + 2 symptoms Bacteriuria
Admitted
Age (years) (mean (S.D.))
Definition of UTI Excluded Pollakisuria, leukocyturia Micturition pain Dysuria
UTI variables rated
Bacteriuria (bacteria/mL)
Quality (reported as) Reference
Table 1 Summary of demographic data, definitions of urinary tract infection (UTI), admission and exclusion criteria, quality standards, reporting and patient flow.
Gender (M/F) Placebo OM-89
Completed
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
No. of patients OM-89/placebo
114
• E. coli, 47.3% for OM-89-treated patients vs. 59.1% for placebotreated patients; and • other bacteria combined (enterococci, streptococci, staphylococci, Klebsiella, Proteus, mixed infections, other), 32.8% of the OM-89-treated patients vs. 71.9% of the placebo-treated patients. 3.1.1. Mean number of urinary tract infections In all the studies examined, the authors reported the mean number of UTIs during the trial or described the actual number of UTIs, which could be transformed in the corresponding means. The data were calculated in two ways: ‘as reported’ in the studies (that is data after 6 months and 12 months); and employing a cut-off at 6 months (Table 2). For the mean number of UTIs, the weighted mean difference (WMD) (inverse of the variance) has been used since outcomes are measured in a standard way across studies. In view of the heterogeneity between trials, both the fixed- and the random-effect models have been calculated. Treatment yields significantly fewer UTIs than placebo in the fixed-effect model with a WMD of −0.36 (95% confidence interval (CI) −0.48 to −0.24) for the data after 6 months or 12 months (test for heterogeneity, P = 0.002; test for overall effect, P < 0.00001). The WMD for the data after 6 months was −0.26 (95% CI −0.36 to −0.16) (test for heterogeneity, P = 0.0004; test for overall effect, P < 0.00001). In the random-effect model, the WMDs were −0.516 (95% CI −0.80 to −0.22) for the data after 6 months or 12 months and −0.42 (95% CI −0.69 to −0.16) for the data after 6 months, with significances similar to those reported with the fixed-effect model. Even after allowing for differences in the rating method employed for diagnosing UTIs, the standardised mean difference (SMD) continues to show a highly significant difference in favour of the active treatment. Examining the mean number of UTIs with OM-89, as a function of number of UTIs with placebo, the studies with the largest number of UTIs with placebo were those showing the largest reduction with the active medication (Fig. 2). 3.1.2. Proportion of patients without recurrent urinary tract infection There were significantly more patients without any UTI among the OM-89-treated patients at the end of the studies, regardless of study duration. Regarding the percentages of patients with UTIs, shown as the difference between OM-89 (41.7% of patients) and placebo (62.4% of patients) in Fig. 3, the difference was −20.7% patients with UTIs with OM-89 (P < 0.001; OR = 0.43, 95% CI 0.34–0.55). Taking as the cut-off the end of the 6 months, available for all studies, the difference in patients with recurrence of UTI between OM-89 (35.1% of patients) and placebo (54.2% of patients)
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
115
Table 2 Mean number of recurrent urinary tract infections (UTIs), by trial and mean considered. Study
OM-89
Placebo
Difference as % of placebo
N
Mean (S.D.)
N
Mean (S.D.)
Bauer et al. [26] [12 months] Pisani et al. [27] Schulman et al. [29] Magasi et al. [28] Tammen [30]
195 [184] 66 85 58 76
0.47 (0.78) [0.84 (1.34)] 0.23 (0.37) 1.22 (1.72) 0.47 (0.79) 0.82 (0.75)
215 [186] 71 81 54 74
0.6 (0.93) [1.28 (1.68)] 0.41 (0.54) 1.96 (1.5) 1.35 (1.02) 1.26 (1.75)
−21.7 [−34.4] −43.9 −37.8 −65.2 −34.9
Mean
480
0.63 (1.03)
495
0.98 (1.29)
−35.7
Mean (including 12 months)
469
0.77 (1.23)
466
1.27 (1.54)
−39.4
S.D., standard deviation. Table 4 Incidence of dysuria at final visit. OM-89 (%) Placebo (%) Significance (P-value) RR (%) Bauer et al. [26] (6 months) Magasi et al. [28] Schulman et al. [29] Tammen [30]
10.8 3.3 9.8 6.6
18.8 16.7 16.7 18.9
Total
8.7
18.1
Difference
−9.4
<0.05 <0.05 N.S. <0.05
57.1 20.0 58.5 34.8
<0.001
48.1
<0.001
48.1
RR, relative risk; N.S., not significant.
Fig. 2. Mean number of episodes of urinary tract infections (N UTIs) (± standard error of the mean) with OM-89 as a function of the number of episodes with placebo in the studies examined.
Fig. 3. Percentage of patients without urinary tract infections: difference between OM-89 and placebo at the end of the studies.
was −19.1% (P < 0.001; OR = 0.46, 95% CI 0.36–0.59). Thus, both outcomes were significantly in favour of active treatment in all the trials, although there is significant heterogeneity between trials (P = 0.001). In conclusion, three out of five patients will have no further UTI in the 6 months following OM-89 treatment. Of the remaining OM89-treated patients still having acute episodes, 35% will have fewer UTIs than with placebo, on average. 3.1.3. Consumption of antibacterials for urinary tract infections This variable was examined using the SMD method (the difference between two means divided by an estimate of the within-group standard deviation), indicated when an outcome is measured in a variety of ways across studies (using different scales), as is the case here. One study [28] did not provide data regarding the use of antibacterials during the trial; the remaining four did (representing 88.4% of the population), but employed different ways of presenting their data. As shown in Table 3, the SMD shows a significant difference in favour of active treatment. In conclusion, although it is difficult to assign a specific value to this assertion owing to differences in the presentation of data, it may be concluded that with OM-89 there was a significant reduction in the use of antibacterials, in line with the reduction in UTIs described above.
Table 3 Consumption of antibacterials and standardised mean difference (SMD). Study
Treatment
Placebo
Weight
N
Mean (S.D.)
N
Mean (S.D.)
Bauer et al. [26] Pisani et al. [27] Schulman et al. [29] Tammen [30]
195 66 85 76
2.44 (1.75) 0.16 (0.36) 2.90 (18.00) 0.85 (1.65)
215 71 81 74
2.79 (2.03) 0.35 (0.48) 6.90 (19.00) 2.78 (5.40)
Total (95% CI)
422
S.D., standard deviation; CI, confidence interval. Test for heterogeneity, P = 0.32. Test for overall effect, P = 0.0001.
441
43.28 17.26 20.81 18.65 100
SMD (random) [95% CI]
Variable
−0.18 [−0.38, 0.01] −0.44 [−0.78,−0.10] −0.22 [−0.52, 0.09] −0.48 [−0.81,−0.16]
Prescriptions Patients with antibacterials Days with antibacterials/3 months Prescriptions
−0.29 [−0.44,−0.14]
116
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119 Table 6 Incidence of bacteriuria at final visit.
Table 5 Incidence of leukocyturia at final visit. OM-89 (%)
Placebo (%)
Significance (P-value)
Pisani et al. [27] Magasi et al. [28] Schulman et al. [29]
14.8 15.0 15.9
19.0 33.3 34.6
N.S. <0.05 <0.01
78.0 45.0 45.8
Total
15.2
28.6
<0.001
53.4
Difference
−13.3
<0.001
53.4
OM-89 (%)
Placebo (%)
2 (P-value)
19.5 3.7 3.3 11.0 –
20.9 11.4 21.7 17.9 –
N.S. (<0.1) <0.01 N.S.
Total
12.4
18.6
Difference
−6.2
RR (%)
RR, relative risk; N.S., not significant.
3.1.4. Findings at the end of the trials 3.1.4.1. Dysuria. Four of the five trials examined reported on this variable, representing 84.8% of the studied population (Table 4). Examining the outcome at 6 months, only a minority of patients reported dysuria, significantly less so among the OM-89-treated patients: −9.4% of the patients (P < 0.001; OR = 0.43, 95% CI 0.28–0.66). Examining the outcome as reported at the end of the studies (that is, the Bauer et al. trial [26] at 12 months) the results are similar, i.e. with OM-89 −6.6% of the patients had dysuria (P < 0.01; OR = 0.49, 95% CI 0.3–0.77). In conclusion, the incidence of dysuria at final visit was significantly lower with OM-89.
Bauer et al. [26] (6 months) Pisani et al. [27] Magasi et al. [28] Schulman et al. [29] Tammen [30]
RR (%) 93.1 32.5 15.4 61.1
<0.05
66.8
<0.05
66.8
RR, relative risk; N.S., not significant.
Bauer et al. trial [26] at 12 months) the results are similar, i.e. with OM-89 −5.4% patients had bacteriuria (P < 0.05; OR = 0.65, 95% CI 0.44–0.96). In conclusion, the incidence of bacteriuria at final visit was lower with OM-89 but the findings are inconsistent between trials. However, the incidence of bacteriuria among the OM-89-treated patients is not much higher than that reported in the general female population, which has been reported to lie between 2.8% (women aged 50–59 years) and 17% (women aged >75 years) [31]. 3.1.5. Safety of OM-89 in placebo-controlled clinical trials The differences of incidence of adverse events in comparative trials showed that they were slightly more frequent during OM-89 therapy than with placebo (+0.8%) and somewhat more patients withdrew from the trials due to side effects (+0.6%). No serious adverse events were attributed to OM-89 in these clinical studies; no disease- or age-related mortality was recorded in the studied population. The most frequent adverse events reported are detailed in Table 7.
3.1.4.2. Leukocyturia. Only three of the five trials examined reported on this variable, representing 42.6% of the population studied (Table 5). Examining the outcome at 6 months, only a minority of patients reported leukocyturia >5/field, significantly less so among the OM-89-treated patients: −13.3% of the patients (P < 0.001; OR = 0.45, 95% CI 0.28–0.72). In conclusion, the incidence of leukocyturia at final visit was significantly lower with OM-89 but the findings were derived from only three out of five trials, representing less than one-half the population studied.
4. Product ‘B’
3.1.4.3. Bacteriuria. Four of the five trials examined reported on this variable, representing 85.7% of the population studied (Table 6). Examining the outcome at 6 months, only a minority of patients reported bacteriuria, significantly less so among the OM-89treated patients: −6.2% of the patients (P < 0.05; OR = 0.62, 95% CI 0.42–0.91). However, this outcome was driven mainly by one study [28] as shown in the sensitivity analysis. There was significant heterogeneity between studies. In one trial [29] the data were provided for both 105 bacteria/mL and 104 bacteria/mL as the cut-off; taking this lower level, the differences become significant in this trial too (OM-89: 19.5% of patients; placebo 35.9% of patients). Examining the outcome as reported at the end of the studies (that is, the
There were no blind trials identified for the parenteral formulation of this product. This formulation causes significant undesirable side effects at the injection site [32] as well as systemic reactions [33]. The published studies of recent years dealt with the intravaginal formulation only, which is discussed herein. Because of the similarity of the data, design and outcomes, it was concluded that the study by Uehling et al. [34] published in 2001 was a preliminary presentation of the data published 2 years later and was therefore excluded from analysis. In the first study with this product [35], a total of 91 women with recurrent urinary infections were entered into a double-blind randomised study. Subjects received three vaginal suppositories at
Table 7 Adverse events reported in controlled clinical trials by ≥0.7% of patients. Common adverse events
OM-89
Placebo
Fisher’s test
Respiratory, thoracic and mediastinal disorders
‘Flu’, bronchitis, pharyngitis
4.3%
3.8%
N.S.
Nervous system disorders
Headache Somnolence and sleep disorders
2.2% 0.1%
2.8% 0.7%
N.S. N.S.
Gastrointestinal disorders
Gastric intolerance/dyspepsia Diarrhoea Nausea/vomiting
1.9% 1.3% 0.7%
2.0% 1.2% 1.0%
N.S. N.S. N.S.
Reproductive system and breast disorders Skin and subcutaneous tissue disorders
Vaginitis, leukorrhoea Allergic reaction/rash
1.0% 1.0%
2.6% 1.3%
P = 0.056 N.S.
Placebo > OM-89
Renal and urinary disorders
Renal pain, nephrolithiasis Pollakisuria and ejaculatory troubles, leukocyturia
0.7% 0.0%
0.0% 0.7%
P = 0.064 P = 0.064
OM-89 > placebo Placebo > OM-89
Musculoskeletal, connective tissue and bone disorders
Back pain
0.4%
1.7%
P = 0.048
Placebo > OM-89
N.S., not significant.
Comment
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
weekly intervals. Depending on the treatment group, each suppository contained one or two vaccine doses or suppository material only. Each patient was followed for 5 months to record infection episodes and to obtain urine, vaginal irrigates and serum in order to measure immunological responses. The verum-treated women who were off antibiotic prophylaxis (prevention discontinued in approximately one-half the patients) had a significant delay in the interval to re-infection during the first 8 weeks, and the mean interval until re-infection was delayed from 8.7 weeks for placebo-treated women to 13 weeks for vaccine-treated women (no significances provided). However, the overall incidence of urinary infections was similar, regardless of the treatment received (on average, 1.4 UTIs per patient). In the second and third study by the same group [36,37], a total of 54 and 75 women, respectively, were entered into doubleblind, placebo-controlled trials using the vaginal suppositories. Patients were withdrawn from prophylactic antibiotics and randomly assigned to one of three treatment groups, namely placebo only, primary immunisation (given on Weeks 0, 1 and 2) or primary plus booster immunisations (booster given on Weeks 6, 10 and 14). All women were followed for 6 months to determine the time until first recurrence and the number of infections. In the second study, the women receiving ‘primary + booster’ remained free of infections for a significantly longer period than those receiving placebo or ‘primary’ immunisation only. The patients receiving ‘primary + booster’ immunisation had significantly fewer infections (average 1.1 UTIs/patient) than the placebo-treated women (average 1.5 UTIs/patient) or those receiving ‘primary’ only (average 1.7 UTIs/patient). The interest of this study is limited by the small number of patients treated and the fact that placebo-treated patients were on average 13 years older than the verum-treated patients. Five of the 36 verum-treated patients experienced vaginal irritation within 1 day of immunisation or had transient diarrhoea. In the third study, the results were similar; that is, placebo was equivalent to ‘primary’ immunisation, and both were inferior to ‘primary + booster’ vaccine (Kaplan–Meier; P < 0.1). Thirty percent of the patients in the placebo group, 57% of the ‘primary’ only group and 72.5% of the patients in the ‘primary + booster’ group remained free of E. coli infection. Furthermore, the results suggested that the vaccine may provide the most benefit to women in the 20–50-yearold age group. The interest of this study is again limited by the small number of patients treated and the fact that placebo-treated patients were on average 9 years older than the verum-treated patients. Twelve percent of the women reported episodes of a burning sensation shortly after treatment both with the active product and with placebo. There were single occurrences of mild to moderate adverse events (four with the active product and four with placebo). A quantitative analysis of anti-E. coli IgA and IgG antibodies in the urine and vaginal fluid samples collected during the course of these two studies did not reveal any significant differences between the three groups. The findings regarding the difference between verum and placebo, as a percentage of patients with UTIs, are illustrated in Fig. 4. In conclusion, with product ‘B’, whilst the shorter ‘primary’ immunisation has not been effective, the longer ‘primary + booster’ treatment was probably effective. Confirmation by larger phase III studies by independent investigators is absolutely necessary.
5. Discussion and conclusions Management of recurrent UTI is complex due to the various factors predisposing to such bacterial invasions. It is further com-
117
Fig. 4. Percentage of patients without urinary tract infections: difference between verum and placebo. Product ‘B’ vaginal suppositories ‘primary’ on Weeks 0, 1 and 2 and ‘primary + booster’ additionally on Weeks 6, 10 and 14.
plicated by the increasing prevalence of antibiotic-resistant strains of E. coli. These facts stress the need for therapies directed at diminishing susceptibility and enhancing the host’s defences against UTI rather than administering antibacterials on a broad base. The urinary tract possesses the immunological and cellular machinery for an efficient antibacterial defence and it seems a sensible approach to activate it. Among the various immunotherapeutic products offered on the market, only two have published studies that are in accordance with current standards. Of these, product ‘A’ (OM-89) is fairly well documented and has been shown to be more effective than placebo in several randomised trials. Product ‘B’ has been shown to be effective only when administered with a booster cycle in three small phase II trials, and larger studies by independent investigators are awaited with interest. Concentrating on OM-89, the data show that the product reduces the number of recurrences of UTI by ca. 36%, and one out of five patients will be entirely free of UTIs for at least 6 months, with a corresponding reduction of antibacterial treatments. At the end of the 6-month observation period, fewer OM-89-treated patients had dysuria (RR = 48%) and possibly leukocyturia (RR = 53%) or bacteriuria (RR = 67%). A further aspect worthwhile examining is the potential reduction of vulvovaginal infections with OM-89 (OR = 0.42, 95% CI 0.18–0.98) versus placebo in women with recurrent UTIs. However, the studies examined do not provide any information on how to optimise treatment. Since only approximately one-third of the UTIs are avoided with this treatment, it would be important to know whether gender, type of infection, secretor status, hormonal status, concentration of THP or other variables are predictors of response to the product. The studies with OM-89 appear to have good internal and external validity; that is, the investigations were conducted with outpatients with relatively broad admission or exclusion criteria and thus can be generalised to the circumstances of daily practice. It seems quite probable that a particular subpopulation might benefit more than others. Unfortunately, there are no studies comparing directly antimicrobial prophylaxis with immunoactive prophylaxis; head-to-head studies are planned. From the reported results in the literature, it seems that the efficacy of antimicrobial prophylaxis is superior to immunisation regimens presently available [2]. However, since 1990 there has been a steadily increasing rate of resistance to trimethoprim/sulfamethoxazole, reaching >30% in some areas; in the USA
118
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
the average resistance rate for E. coli is 18%. Moreover, resistance rates to quinolones have been increasing worldwide at alarming rates, reaching >20% in Spain [38–41]. Another potential downside of antibacterial prophylaxis is the side effects such as rash, nausea, diarrhoea, vaginitis and oral candidiasis [42,43]. Treatments with antibacterials have to be monitored closely since they can also cause major adverse reactions, e.g. phototoxicity and rare but serious hepatotoxicity has been reported with fluoroquinolones [44], whilst nitrofurantoin is known infrequently to cause both acute and chronic pulmonary reactions (interstitial lung disease and pulmonary fibrosis may develop with long-term use) [45], as well as rare cases of blood dyscrasia, erythema exsudativum multiforme (Stevens–Johnson syndrome) and epidermolysis acuta toxica (Lyell syndrome) with trimethoprim/sulfamethoxazole or with nitrofurantoin, to mention a few [46]. Additionally, oral administration of antibiotics for treatment of UTI can cause ecological disturbances in the normal urethral [47] and intestinal microflora [48]. To increase the mucosal defences through immunoactive prophylaxis looks like a compelling alternative that merits further studies and improvements. The future may bring new and possibly more effective immunostimulants, such as the vaccine derived from purified E. coli FimH adhesion [49], P. mirabilis fimbriae (PMF and purified recombinant structural fimbrial proteins of these fimbriae) [50,51] and others, currently tested in animals.
Acknowledgment The authors are thankful to Dr Reto G. Brignoli for assistance with the statistical analysis and preparation of the manuscript. Funding: This project was partially funded by OM Pharma, Geneva, Switzerland. The sponsors had no involvement in the study design, analysis and interpretation of the data, writing of the manuscript, or in the decision to submit the manuscript for publication. Competing interests: KGN: Bayer (investigator, speaker at scientific meetings), Bionorica (investigator, consultant), Eumedica Pharmaceuticals (consultant), MerLion Pharmaceuticals (investigator, consultant), Daiichi (speaker at scientific meetings, scientific publication), MUCOS Pharma (consultant, speaker at scientific meetings), NanoVibronix (consultant), Ocean Spray Cranberries (investigator), OM Pharma (investigator, consultant), Peninsula/Johnson & Johnson/Janssen-Cilag (investigator, speaker at scientific meetings), Pharmatoka (investigator, consultant), Polyphor (consultant), Protez (consultant), Rosen Pharma (consultant), sanofi-aventis (investigator, speaker at scientific meetings, scientific publication), UroVision (investigator) and Zambon (investigator, speaker at scientific meetings). Y-HC: Asia Pharm (consultant, speaker at scientific meetings), Handok Pharmaceuticals (consultant, speaker at scientific meetings) and Pacific Pharma (consultant, speaker at scientific meetings). TM: Daiichi Sankyo (consultant, investigator, speaker at scientific meetings), Asteras (consultant, investigator), Kyorin (investigator, speaker at scientific meetings), Meiji (consultant, investigator), Nisshin Science (consultant, investigator), Takeda (speaker at scientific meetings), Taisho-Toyama (investigator, speaker at scientific meetings) and Combinature/MerLion (consultant). AJS: American Urological Association (editorial fees and faculty honorarium), IMS Health (consultant), Alita Pharmaceuticals, Inc. (consultant), NovaBay Pharmaceuticals, Inc. (consultant), cme2 (speaker at scientific meeting), Blackwell Publishing (book royalty), Advanstar Communications Inc. (consultant), Elsevier Science USA (book royalty) and Regeneron Pharmaceuticals, Inc. (consultant). Ethical approval: Not required.
References [1] Lindsay EN. Managing recurrent urinary tract infections in women. Womens Health 2005;1:39–50. [2] Naber KG (Chairman), Bishop MC, Bjerklund-Johansen TE, Botto H, C¸ek M, Grabe M, et al. Guidelines on the management of urinary and male genital tract infections. European Association of Urology; 2006. p. 7–8, Appendices 12.1 [Criteria for the diagnosis of a UTI], p. 109. [3] Fihn SD. Acute uncomplicated urinary tract infection in women. N Engl J Med 2003;349:259–66. [4] Boyko EJ, Fihn SD, Scholes D, Abraham L, Monsey B. Risk of urinary tract infection and asymptomatic bacteriuria among diabetic and nondiabetic postmenopausal women. Am J Epidemiol 2005;161:557–64. [5] Raz R, Gennesin Y, Wasser J, Stoler Z, Rosenfeld S, Rottensterich E, et al. Recurrent urinary tract infections in postmenopausal women. Clin Infect Dis 2000;30:152–6. [6] Ishitoya S, Yamamoto S, Mitsumori K, Ogawa O, Terai A. Non-secretor status is associated with female acute uncomplicated pyelonephritis. BJU Int 2002;89:851–4. [7] Schaeffer AJ, Rajan N, Cao Q, Anderson BE, Pruden DL, Sensibar J, et al. Host pathogenesis in urinary tract infections. Int J Antimicrob Agents 2001;17:245–51. [8] Hidalgo-Tenorio C, Jimenez-Alonso J, de Dios Luna J, Tallada M, Martinez-Brocal A, Sabio JM. Urinary tract infections and lupus erythematosus. Ann Rheum Dis 2004;63:431–7. [9] Hughes LB, Criswell LA, Beasley TM, Edberg JC, Kimberly RP, Moreland LW, et al. Genetic risk factors for infection in patients with early rheumatoid arthritis. Genes Immun 2004;5:641–7. [10] Schmidhammer S, Ramoner R, Holtl L, Bartsch G, Thurnher M, Zelle-Rieser C. An Escherichia coli-based oral vaccine against urinary tract infections potently activates human dendritic cells. Urology 2002;60:521–6. [11] Boruchov AM, Heller G, Veri MC, Bonvini E, Ravetch JV, Young JW. Activating and inhibitory IgG Fc receptors on human DCs mediate opposing functions. J Clin Invest 2005;115:2914–23. [12] David J, Terhorst C. Organs and cells of the immune system. In: Scientific American Medicine; Ch. 6. Immunology. WebMD Corp.; 1999. [13] Jerome KR, Corey L. The danger within. N Engl J Med 2004;350:411–2. [14] Anders HJ, Patole PS. Toll-like receptors recognize uropathogenic Escherichia coli and trigger inflammation in the urinary tract. Nephrol Dial Transplant 2005;20:1529–32. [15] Yahr TL. A critical new pathway for toxin secretion? N Engl J Med 2006;355:1171–2. [16] Kerschbaum HH, Singh SK, Hermann A. Lymphoid tissue in the kidney of the musk shrew, Suncus murinus. Tissue Cell 1995;27:421–4. [17] Thilagarajah R, Witherow RO, Walker MM. Quantitative histopathology can aid diagnosis in painful bladder syndrome. J Clin Pathol 1998;51:211–4. [18] Franz M, Hörl WH. Common errors in diagnosis and management of urinary tract infection. I. Pathophysiology and diagnostic techniques. Nephrol Dial Transplant 1999;14:2746–53. [19] Kunin CM, Evans C, Bartholomew D, Bates DG. The antimicrobial defense mechanism of the female urethra: a reassessment. J Urol 2002;168:413–9. [20] Pudney J, Anderson DJ. Immunobiology of the human penile urethra. Am J Pathol 1995;147:155–65. [21] Bauer HW, Rahlfs VW, Lauener PA, Blessmann GS. Prevention of recurrent urinary tract infections with immuno-active E. coli fractions: a meta-analysis of five placebo-controlled double-blind studies. Int J Antimicrob Agents 2002;19:451–6. [22] Alderson P, Green S, Higgins JPT, editors. Cochrane Handbook for Systematic Reviews of Interventions [version 4.2.2, updated March 2004]. In: The Cochrane Library, Issue 1. Chichester, UK: John Wiley & Sons, Ltd.; 2004. http://www.cochrane.org/resources/handbook/index.htm [accessed 8 September 2008]. [23] Bax L, Yu LM, Ikeda N, Tsuruta H, Moons KGM. Development and validation of MIX: comprehensive free software for meta-analysis of causal research data. BMC Med Res Methodol 2006;6:50. [24] Hachen HJ. Oral immunotherapy in paraplegic patients with chronic urinary tract infections: a double-blind, placebo-controlled trial. J Urol 1990;143:759–62. [25] Frey C, Obolensky W, Wyss H. Treatment of recurrent urinary tract infections: efficacy of an orally administered biological response modifier. Urol Int 1986;41:444–6. [26] Bauer HW, Alloussi S, Egger G, Blumlein HM, Cozma G, Schulman CC; Multicenter UTI Study Group. A long-term, multicenter, double-blind study of an Escherichia coli extract (OM-89) in female patients with recurrent urinary tract infections. Eur Urol 2005;47:542–8. [27] Pisani E, Palla R, Bono AV. Double-blind randomised clinical study of OM-8930 vs placebo in patients suffering from recurrent urinary tract infections. Geneva, Switzerland: OM Pharma; 1992. (Data on file). Quoted in: Chiavaroli C, Moore A. An hypothesis to link the opposing immunological effects induced by the bacterial lysate OM-89 in urinary tract infection and rheumatoid arthritis. BioDrugs 2006;20:141–9. [28] Magasi P, Panovics J, Illes A, Nagy M. Uro-Vaxom and the management of recurrent urinary tract infection in adults: a randomized multicenter double-blind trial. Eur Urol 1994;26:137–40.
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119 [29] Schulman CC, Corbusier A, Michiels H, Taenzer HJ. Oral immunotherapy of recurrent urinary tract infections: a double-blind placebo-controlled multicenter study. J Urol 1993;150:917–21. [30] Tammen H; The German Urinary Tract Infection Study Group. Immunobiotherapy with Uro-Vaxom in recurrent urinary tract infection. Br J Urol 1990;65:6–9. [31] Scottish Intercollegiate Guidelines Network. Management of suspected bacterial urinary tract infection in adults. A national clinical guideline. July 2006. [32] Pam D. Vaginal vaccine for recurrent urinary-tract infections. Lancet Infect Dis 2002;2:68. [33] Überempfindlichkeitsreaktionen auf ‘Impfstoff’ STROVAC [Hypersensitivity reactions to ‘vaccine’ STROVAC]. Arznei-Telegram 2007; 38(2):31. [34] Uehling DT, Hopkins WJ, Beierle LM, Kryger JV, Heisey DM. Vaginal mucosal immunization for recurrent urinary tract infection: extended phase II clinical trial. J Infect Dis 2001;183(Suppl. 1):S81–3. [35] Uehling DT, Hopkins WJ, Balish E, Xing Y, Heisey DM. Vaginal mucosal immunization for recurrent urinary tract infection: phase II clinical trial. J Urol 1997;157:2049–52. [36] Uehling DT, Hopkins WJ, Elkahwaji JE, Schmidt DM, Leverson GE. Phase 2 clinical trial of a vaginal mucosal vaccine for urinary tract infections. J Urol 2003;170:867–9. [37] Hopkins WJ, Elkahwaji J, Beierle LM, Leverson GE, Uehling DT. Vaginal mucosal vaccine for recurrent urinary tract infections in women: results of a phase 2 clinical trial. J Urol 2007;177:1349–53. [38] Hooton TM, Stamm WE. Diagnosis and treatment of uncomplicated urinary tract infection. Infect Dis Clin North Am 1997;11:551–81. [39] Kahlmeter G. The ECO.SENS Project: a prospective, multinational, multicentre epidemiological survey of the prevalence and antimicrobial susceptibility of urinary tract pathogens—interim report. J Antimicrob Chemother 2000;46(Suppl. 1):15–22. [40] Gupta K, Sahm DF, Mayfield D, Stamm WE. Antimicrobial resistance among uropathogens that cause community-acquired urinary tract infections in women: a nationwide analysis. Clin Infect Dis 2001;33:89–94. [41] Manges AR, Johnson JR, Foxman B, O’Bryan TT, Fullerton KE, Riley LW. Widespread distribution of urinary tract infections caused by a multidrug-
[42]
[43]
[44] [45]
[46] [47]
[48]
[49]
[50]
[51]
119
resistant Escherichia coli clonal group. N Engl J Med 2001;345:1007– 13. Albert X, Huertas I, Pereiró I, Sanfélix J, Gosalbes V, Perrota C. Antibiotics for preventing recurrent urinary tract infection in non-pregnant women. Cochrane Database Syst Rev 2004;(3). CD001209. Gradwohl SE (Team Leader); University of Michigan Health System (Agency for Healthcare Research and Quality (AHRQ)). Urinary tract infection. Ann Arbor, MI: University of Michigan Health System; 2005. http://www.guidelines. gov/summary/summary.aspx?view id=1&doc id=7407 [accessed 3 September 2008]. Ball P, Mandell L, Niki Y, Tillotson G. Comparative tolerability of the newer fluoroquinolone antibacterials. Drug Saf 1999;21:407–21. Sheehan RE, Wells AU, Milne DG, Hansell DM. Nitrofurantoin-induced lung disease: two cases demonstrating resolution of apparently irreversible CT abnormalities. J Comput Assist Tomogr 2000;24:259–61. Howe RA, Spencer RC. Cotrimoxazole: rationale for re-examining its indications for use. Drug Saf 1996;14:213–8. Norinder BS, Norrby R, Palmgren AC, Hollenberg S, Eriksson U, Nord CE. Microflora changes with norfloxacin and pivmecillinam in women with recurrent urinary tract infection. Antimicrob Agents Chemother 2006;50: 1528–30. Edlund C, Nord CE. Effect on the human normal microflora of oral antibiotics for treatment of urinary tract infections. J Antimicrob Chemother 2000;46(Suppl. 1):41–8. Langermann S, Mollby R, Burlein JE, Palaszynski SR, Auguste CG, DeFusco A, et al. Vaccination with FimH adhesin protects cynomolgus monkeys from colonization and infection by uropathogenic Escherichia coli. J Infect Dis 2000;181:774–8. Pellegrino R, Galvalisi U, Scavone P, Sosa V, Zunino P. Evaluation of Proteus mirabilis structural fimbrial proteins as antigens against urinary tract infections. FEMS Immunol Med Microbiol 2003;36:103–10. Li X, Lockatell CV, Johnson DE, Lane MC, Warren JW, Mobley HL. Development of an intranasal vaccine to prevent urinary tract infection by Proteus mirabilis. Infect Immun 2004;72:66–75.
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Review
Immunoactive prophylaxis of recurrent urinary tract infections: a meta-analysis Kurt G. Naber a,∗ , Yong-Hyun Cho b , Tetsuro Matsumoto c , Anthony J. Schaeffer d a
Technical University Munich, Munich, Germany Department of Urology, St Marys Hospital, The Catholic University of Korea, Seoul, South Korea c Department of Urology, School of Medicine, University of Occupational and Environmental Health, Kitakyushu, Japan d Department of Urology, Northwestern University Medical School, Chicago, IL, USA b
a r t i c l e
i n f o
Keywords: Urinary tract infections Bacterial lysates Clinical review Meta-analysis OM-89 Uro-Vaxom SolcoUrovac
a b s t r a c t Recurrent urinary tract infections (UTIs) are very common, particularly among women in their reproductive years. Alternatives to antibacterial prophylaxis are needed, particularly measures to increase host defences. Various bacterial lysates have been proposed with this indication. The objective of this review and meta-analysis was to assess the efficacy and safety of bacterial lysates in the management of recurrent UTIs. Electronic databases identified 11 studies with the descriptors ‘urinary tract infection’, ‘immunotherapy’ or ‘vaccines’ and ‘double blind’. Seven of the studies dealt with an oral immunostimulant (OM-89), of which about five (1000 adult patients) were retained for analysis with an observation period of 6–12 months. The mean number of UTIs was significantly lower in OM-89-treated patients in all the trials analysed (mean 39%), as was the use of antibacterials. Four of the studies dealt with a vaginal vaccine, of which three small studies were retained for analysis (220 adult patients). The results suggest that this vaginal vaccine is effective when administered with a booster cycle (no recurrent UTI in 50% vs. 14% with placebo). No blind controlled studies could be identified with other bacterial lysates claiming the same indication. In conclusion, among the various immunotherapeutic products, studies were published only for one product (OM-89) that are in accordance with current standards. This product was shown to be effective under conditions of daily practice. The second product (vaginal vaccine) also appears to be effective but adequate phase III studies are necessary. © 2008 Published by Elsevier B.V. and the International Society of Chemotherapy.
1. Introduction Urinary tract infections (UTIs) are the most common medical complaint among women in their reproductive years. Women are 30 times more likely to have UTIs than men. Every year, 15% of sexually active women have at least one such infection and up to 60% of all women will develop a UTI at some time in their lives. At least one in four of these women will have a recurrence within a year [1]. UTIs are frequent in pre-menopausal, non-pregnant women who present with acute onset of dysuria, frequency or urgency, and suprapubic pain. These infections are confirmed by urine analysis documenting >10 white blood cells/mm3 and urine culture with >103 colony-forming units/mL of uropathogen in a mid-stream sample of urine. UTIs are defined as recurrent if there had been at least three episodes of uncomplicated infection documented by culture in the last 12 months. On average, each episode of acute UTI in pre-menopausal women was shown to be associated with
∗ Corresponding author. Present address: Karl-Bickleder-Str. 44c, D-94315 Straubing, Germany. Tel.: +49 9421 33369. E-mail address: [email protected] (K.G. Naber).
6.1 days of symptoms, 2.4 days of restricted activity, 1.2 days in which they were not able to attend classes or work and 0.4 days in bed [2]. Particularly in women there are several well-defined risk factors for recurrent UTI. In young women, the most important risk factors for acute cystitis are a history of previous episodes of cystitis and frequent or recent sexual activity; the use of spermicidal agents elevates the likelihood of infection. Women with frequent recurrences are more likely to have a maternal history of cystitis and to have had cystitis at an early age [3]. In post-menopausal women, UTI incidence is increased in diabetic patients (relative risk (RR) = 1.8) [4]; furthermore, it was shown that urinary incontinence (odds ratio (OR) = 5.79), a history of UTI before menopause (OR = 4.85), and non-secretor status (OR = 2.9) were most strongly associated with recurrent UTI [5,6]. Epithelial cells were collected from patients with recurrent UTI and compared with such cells obtained from controls; bacterial strains that caused cystitis adhered much more avidly to the epithelial cells from susceptible women. The presence or absence of blood group determinants on the surface of uroepithelial cells may influence an individual’s susceptibility to UTI. Susceptibility among women who do not secrete blood group antigens may be due to specific Escherichia coli-binding glycolipids
0924-8579/$ – see front matter © 2008 Published by Elsevier B.V. and the International Society of Chemotherapy. doi:10.1016/j.ijantimicag.2008.08.011
112
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
that are absent in women who secrete blood group antigens [7]. Some of these risk factors suggest possible therapeutic or prophylactic interventions. Additionally, there are diseases that seem to predispose to recurrent UTI, in particular lupus erythematosus [8] and rheumatoid arthritis [9]. 1.1. Proof of concept The idea of employing bacterial immune stimulants in order to reduce recurrent UTI was born some 40 years ago and they have been used, rather empirically, to prevent recurrent infections in the non-immunocompromised host. These early attempts were hampered at the time by a lack of knowledge of the potential mechanisms involved. Only in recent years has a better understanding of the innate immune system provided a solid rationale for such immune stimulants. The requirements for a bacterial extract to be effective are that it stimulates infection-combating, circulating elements within the lymphoid tissue and that the target organ envisioned is immunocompetent in order to translate adequately such a stimulus into an efficient response. 1.2. Bacterial immune stimulants Bacterial extracts can act as immune stimulants through the activation of monocyte-derived dendritic cells, as was shown for OM-89 [10]. Human monocyte-derived dendritic cells and circulating conventional dendritic cells co-express activating (CD32a) and inhibitory (CD32b) isoforms of immunoglobulin G (IgG) Fc gamma receptor (Fc␥R). The balance between these divergent receptors establishes a threshold of dendritic cell activation and enables immune complexes to mediate opposing effects on dendritic cell maturation and function. The activating and inhibitory Fc␥Rs on dendritic cells offer rational targets for immunotherapy based on the unique capacity of dendritic cells to play critical roles in both immunity and tolerance [11]. Some bacterial extracts are administered via the oral route with the goal of increasing the immune defences of organs with a mucous lining, although parenteral, intravaginal and intranasal applications have also been described. In animal models and in normal humans, gut-associated lymphoid tissue stimulation is able to induce a generalised response by the whole mucosa-associated lymphoid tissue (MALT), which is found in the digestive tract, the respiratory tract and the genitourinary tract. Migration and adherence of lamina propria lymphocytes are in part dictated by integrins and selectins. Lymphocyte circulation leads to the distribution of effector cells to particular parts of the body where homing of lymphocytes is controlled by the expression of various receptors on the cell surface and counter-receptors on the vascular endothelium [12]. A priori one would expect that a bacterial extract will act as a more or less specific immune stimulant for the bacteria extracted. However, this has not been the case and can be explained by the fact that different bacteria appear to share antigenic structures, to stimulate common Toll-like receptors (TLRs) and to share toxin secretion mechanisms. Innate immune recognition is based on socalled pattern recognition receptors (PRRs) or pathogen-associated molecular patterns [13] where various classes of pathogens can be recognised by the same PRRs; for example, TLR4, which recognises uropathogenic E. coli, is a critical component of the lipopolysaccharide receptor complex, which forms a detection system for Gram-negative bacteria [14]. One further example is the type VI toxin secretion pathway that is shared by Pseudomonas aeruginosa, Vibrio cholerae, Salmonella enterica, Yersinia pestis and E. coli [15].
1.3. Immunocompetence of target organs Animal studies have shown an anatomically well-organised lymphoid tissue in the urinary tract, resembling the MALT, which is associated with the epithelium of the renal pelvis and the ureter, respectively. Immunocytochemistry revealed S-100immunoreactive dendritic cells both in structurally organised and structurally non-organised lymphoid tissues [16]. Humans also have a well-organised lymphoid tissue in the urinary tract, particularly the urinary bladder. Host-specific factors associated with UTI include production of secretory IgA interfering with adhesion, presence of Tamm–Horsfall protein (THP) causing bacterial aggregation and washout, bactericidal properties of serum, and urodynamic factors, i.e. bacterial washout. THP, secreted by the renal tubules in the outer medulla, has specific receptors for several uropathogens and the bound bacteria are washed out in the urine [17,18]. Like the urinary bladder, the urethra may provide a favourable environment for colonisation by uropathogens but it is protected by powerful defence mechanisms that, besides trapping of bacteria by mucus and intermittent washout by urine, include the local production of immunoglobulins, cytokines and defensins in addition to the mobilisation of leukocytes [19]. The human penile urethra contains numerous IgA and J chain-positive plasma cells, and the epithelium expresses secretory component, the transport molecule for polymeric IgA, indicating that this region is an active site of secretory IgA-mediated immune defence. The mucosa contains intraepithelial dendritic cells, whilst T-lymphocytes are abundant and ubiquitous in all regions of the urethra. Both CD8+ and CD4+ subpopulations of T-lymphocytes are present in the lamina propria and epithelium, although CD8+ cells predominate. The majority of T-lymphocytes are positive for CD45RO (memory marker) and many are also positive for the ␣-E--7-integrin (mucosal-associated antigen). These data indicate that, in addition to the bladder, the urethra is a highly dynamic immunocompetent tissue possessing all the necessary elements for antigen presentation and both humoral and cellular mucosal immune responses. It is likely that this region plays a dominant role in protecting the urogenital tract against ascending infections [20]. 2. Methodology The primary objective of the review was to assess the efficacy and safety of immunostimulants based on bacterial lysates from a clinical point of view and taking into account clinically relevant endpoints. Meta-analysis may produce a stronger conclusion than can be provided by any individual study. The purposes of metaanalysis are to increase statistical power for primary endpoints, to increase general applicability (external validity) of findings, to resolve uncertainty when reports disagree, to provide quantitative estimates of effects, to call attention to strengths and weaknesses of a body of research in a particular area and to identify needs for new primary data collection. 2.1. Search strategy and material scrutinised Among the data sources consulted in the identification of trials were bibliographic databases (TOXLINE, PaperChase browsing the databases of the National Library of Medicine and the National Cancer Institute, i.e. MEDLINE, HealthSTAR, AIDSLINE and CANCERLIT, EMBASE, AMED, Cochrane Collaboration, PubMed (http://www.ncbi.nlm.nih.gov), TOXLINE Special, DART Special, HSDB, IRIS, ITER, GENETOX, ChemIDplus, Haz-Map), reference lists from pertinent review articles and books, and personal contacts with experts active in the area and manufacturers up to September 2008. All papers were screened and any dealing with prospec-
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
113
Fig. 1. Flowchart of double-blind studies identified and retained for analysis, by bacterial immunotherapy product.
tive clinical trials were retained for classification according to the selection criteria described below. In the case of double publications, those that were most recent and/or had appeared in a peer-reviewed journal were retained.
in the flowchart of studies identified and retained for analysis (Fig. 1).
2.1.1. Selection criteria and description of studies The search criteria were ‘urinary tract infection’ and ‘double blind’ or ‘single blind’ and ‘bacterial vaccines’ or ‘immunization’ or ‘immunotherapy’ or ‘vaccination’ or ‘vaccines’. To be retained, a study had to be a randomised, placebo-controlled, clinical trial with the primary aim of reducing the number of UTIs over a period of 6–12 months. All trials rendered eligible were classified by indication and summarised in a tabulated format by one reviewer. The standard table included a full reference, a quality rating, the type of pathology, demographic data and treatments, endpoints and adverse events. The trials were grouped by indication and by treatment. These tables were verified and discussed by the authors until consensus was reached. No formal validation process was employed. The electronic databases identified (i) urinary tract infection in 25 757 references and (ii) double or single blind in 88 121 + 10 220 references and (iii) bacterial vaccines or immunisation or immunotherapy or vaccination or vaccines in 155 209 references, but only 10 complied with the three descriptors. One additional clinical study was identified through other channels and was duly scrutinised. The search also yielded one meta-analysis [21] regarding OM-89 (Product ‘A’; Uro-Vaxom® ). As will be described in the results, three studies were eliminated because of duplication of data, particular patient selection or inadequate reporting (one study each). Furthermore, the electronic databases were searched for trials conducted with specific products identified (http://focosi. altervista.org/preventionprimaryimmunovaccine.html); that is, OM-89 capsules (Product ‘A’; five different serotypes of heat-killed uropathogenic E. coli); Product ‘B’ for intramuscular or vaginal administration (StroVac® and SolcoUrovac® , respectively; contains 10 heat-killed uropathogenic bacteria, including six different serotypes of uropathogenic E. coli, Proteus vulgaris, Klebsiella pneumoniae, Morganella morganii and Enterococcus faecalis); and Urostim® (‘Other products’; killed bacterial cells and their lysates of four microbial species: uropathogenic E. coli expressing type 1 pili, Rc mutant of E. coli, K. pneumoniae, Proteus mirabilis and E. faecalis) and Urvakol® (‘Other products’; uropathogenic E. coli, K. pneumoniae, P. mirabilis, P. aeruginosa and E. faecalis; adjuvant activity is provided by Propionibacterium acnes). However, this search did not identify additional trials complying with the predefined criteria. The resulting selection of trials is summarised
The guidelines provided by the Cochrane Collaboration Handbook for Reviews [22] and the corresponding software have been used in the analysis of the clinical data (Review Manager 4.2.7). Ordinal data were compared using the Mantel-HaenszelPeto method [23]. Significances were calculated using two-sided tests, with the threshold of significance being P ≤ 0.05 and for non-significance P > 0.1; values between P > 0.05 and P ≤ 0.1 were reported as trends. Mean and standard deviation correspond to those reported for the intention-to-treat analysis, where available, and N corresponds to the number of patients admitted to the trial and exposed to the treatment.
2.2. Analysis of data
3. Results 3.1. Product ‘A’: OM-89 (Uro-Vaxom® ) From the original seven studies identified, only five were retained for further analysis. One study [24] was excluded because, although double-blind, it dealt with a special group of patients: in a 6-month, double-blind, placebo-controlled clinical trial with cross-over, the authors administered the bacterial extract OM-89 in 70 spinal cord injury patients with chronic lower UTIs. Compared with patients given placebo, in the treated patients there was a statistically significant decrease in the degree of bacteriuria, a considerably decreased incidence of infectious episodes and a lesser requirement for antibiotics. One further trial [25] with 64 outpatients suffering from an acute recurrent UTI was excluded because there was insufficient reporting of demographics and the publication did not account for drop-outs or missing data. In this trial, dysuria, bacteriuria, leukocyturia and consumption of antibacterials were reported to show a significant reduction with OM-89 in comparison with the placebo after 6 months. The five studies [26–30] retained for further analysis present a large amount of commonalities, albeit differing in some aspects. In three studies [27–30] also men were included, but always less than 20% of the total patients. Unfortunately, the results could not be stratified by gender. In the largest study [26], however, only female patients were included, thus the proportion of men decreased to less than 7% in the total population. A summary of treatment schedules, various definitions of UTIs, admission and exclusion criteria, and the patients’ flow is given in Table 1. In all these studies, patients were recruited suffering from acute UTI, had a history of recurrent
85/81 76/74
72/68 61/59
UTIs, were treated with an antibacterial until cure while at the same time starting treatment with OM-89, one capsule daily during 3 months. Thereafter, there was a follow-up period of an additional 3 months. In one study [26] the patients received a booster treatment in Months 6–9 (one capsule daily during 10 days per month), again with an additional follow-up of 3 months, thus completing a study period of 12 months. However, the study design does not allow for a distinction between effects due to the booster treatment and the possible ‘carry-over’ effects of the initial treatment cycle. In all the studies examined the patients were comparable at baseline with regard to demographic and clinical data. In addition to the planned visits, the patients could return to the consultation if they were symptomatic; these UTIs were assimilated to the following planned visit. Only two trials [26,27] reported the type of bacteria identified at baseline, which was E. coli in ca. 70% of cases in both trials. In the study by Bauer et al. (unpublished data on file, OM Pharma) it was found that the percentages of patients with a UTI during the study period were: GCP, good clinical practice; ITT, intent to treat analysis; PP, per protocol (analysis excluding protocol violators); N.D., not defined; S.D., standard deviation. a Midstream urine sample. b Rated sum score: fever, dysuria, pollakisuria, stranguria, burning, flank, suprapubic and perineal pain, fullness, haematuria, odorous and cloudy urine, other. c 105 bacteria/mL midstream urine or 104 bacteria/mL catheter urine. d Completed 12 months in parentheses.
17/95 17/133 45 (16.1) 50.4 (19.8) 45.3 (18.4) 51.2 (20.9) −−− −−− −−− −−− +++ +++ >105 c >105 c GCP (PP) Not stated (PP) Schulman et al. [29] Tammen [30]
−−− −−− +++ >105 Not stated (PP) Magasi et al. [28]
[++]b [++]b [++]b >105 a GCP (ITT/PP) Pisani et al. [27]
+++ GCP (ITT/PP) Bauer et al. [26]
≥103a
+++
+++
Pyelonephritis, catheter, neurogenic bladder, etc. Antibiotic-resistant UTI, pyelonephritis, catheter, obstruction, etc. Pyelonephritis, catheter, obstruction, etc. Anatomical abnormalities, etc. Post-coital UTI, pyelonephritis, catheter, obstruction, etc.
Bacteriuria N.D.
58/54 26/140 Range 16–82 Bacteriuria
63/59
66/71 9/128
0/453
46.9 (15.9)
81/79
195/191 (184/186)d 231/222 39.8 (15.1) 41.7 (15.3)
Bacteriuria + 2 symptoms Bacteriuria
Admitted
Age (years) (mean (S.D.))
Definition of UTI Excluded Pollakisuria, leukocyturia Micturition pain Dysuria
UTI variables rated
Bacteriuria (bacteria/mL)
Quality (reported as) Reference
Table 1 Summary of demographic data, definitions of urinary tract infection (UTI), admission and exclusion criteria, quality standards, reporting and patient flow.
Gender (M/F) Placebo OM-89
Completed
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
No. of patients OM-89/placebo
114
• E. coli, 47.3% for OM-89-treated patients vs. 59.1% for placebotreated patients; and • other bacteria combined (enterococci, streptococci, staphylococci, Klebsiella, Proteus, mixed infections, other), 32.8% of the OM-89-treated patients vs. 71.9% of the placebo-treated patients. 3.1.1. Mean number of urinary tract infections In all the studies examined, the authors reported the mean number of UTIs during the trial or described the actual number of UTIs, which could be transformed in the corresponding means. The data were calculated in two ways: ‘as reported’ in the studies (that is data after 6 months and 12 months); and employing a cut-off at 6 months (Table 2). For the mean number of UTIs, the weighted mean difference (WMD) (inverse of the variance) has been used since outcomes are measured in a standard way across studies. In view of the heterogeneity between trials, both the fixed- and the random-effect models have been calculated. Treatment yields significantly fewer UTIs than placebo in the fixed-effect model with a WMD of −0.36 (95% confidence interval (CI) −0.48 to −0.24) for the data after 6 months or 12 months (test for heterogeneity, P = 0.002; test for overall effect, P < 0.00001). The WMD for the data after 6 months was −0.26 (95% CI −0.36 to −0.16) (test for heterogeneity, P = 0.0004; test for overall effect, P < 0.00001). In the random-effect model, the WMDs were −0.516 (95% CI −0.80 to −0.22) for the data after 6 months or 12 months and −0.42 (95% CI −0.69 to −0.16) for the data after 6 months, with significances similar to those reported with the fixed-effect model. Even after allowing for differences in the rating method employed for diagnosing UTIs, the standardised mean difference (SMD) continues to show a highly significant difference in favour of the active treatment. Examining the mean number of UTIs with OM-89, as a function of number of UTIs with placebo, the studies with the largest number of UTIs with placebo were those showing the largest reduction with the active medication (Fig. 2). 3.1.2. Proportion of patients without recurrent urinary tract infection There were significantly more patients without any UTI among the OM-89-treated patients at the end of the studies, regardless of study duration. Regarding the percentages of patients with UTIs, shown as the difference between OM-89 (41.7% of patients) and placebo (62.4% of patients) in Fig. 3, the difference was −20.7% patients with UTIs with OM-89 (P < 0.001; OR = 0.43, 95% CI 0.34–0.55). Taking as the cut-off the end of the 6 months, available for all studies, the difference in patients with recurrence of UTI between OM-89 (35.1% of patients) and placebo (54.2% of patients)
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
115
Table 2 Mean number of recurrent urinary tract infections (UTIs), by trial and mean considered. Study
OM-89
Placebo
Difference as % of placebo
N
Mean (S.D.)
N
Mean (S.D.)
Bauer et al. [26] [12 months] Pisani et al. [27] Schulman et al. [29] Magasi et al. [28] Tammen [30]
195 [184] 66 85 58 76
0.47 (0.78) [0.84 (1.34)] 0.23 (0.37) 1.22 (1.72) 0.47 (0.79) 0.82 (0.75)
215 [186] 71 81 54 74
0.6 (0.93) [1.28 (1.68)] 0.41 (0.54) 1.96 (1.5) 1.35 (1.02) 1.26 (1.75)
−21.7 [−34.4] −43.9 −37.8 −65.2 −34.9
Mean
480
0.63 (1.03)
495
0.98 (1.29)
−35.7
Mean (including 12 months)
469
0.77 (1.23)
466
1.27 (1.54)
−39.4
S.D., standard deviation. Table 4 Incidence of dysuria at final visit. OM-89 (%) Placebo (%) Significance (P-value) RR (%) Bauer et al. [26] (6 months) Magasi et al. [28] Schulman et al. [29] Tammen [30]
10.8 3.3 9.8 6.6
18.8 16.7 16.7 18.9
Total
8.7
18.1
Difference
−9.4
<0.05 <0.05 N.S. <0.05
57.1 20.0 58.5 34.8
<0.001
48.1
<0.001
48.1
RR, relative risk; N.S., not significant.
Fig. 2. Mean number of episodes of urinary tract infections (N UTIs) (± standard error of the mean) with OM-89 as a function of the number of episodes with placebo in the studies examined.
Fig. 3. Percentage of patients without urinary tract infections: difference between OM-89 and placebo at the end of the studies.
was −19.1% (P < 0.001; OR = 0.46, 95% CI 0.36–0.59). Thus, both outcomes were significantly in favour of active treatment in all the trials, although there is significant heterogeneity between trials (P = 0.001). In conclusion, three out of five patients will have no further UTI in the 6 months following OM-89 treatment. Of the remaining OM89-treated patients still having acute episodes, 35% will have fewer UTIs than with placebo, on average. 3.1.3. Consumption of antibacterials for urinary tract infections This variable was examined using the SMD method (the difference between two means divided by an estimate of the within-group standard deviation), indicated when an outcome is measured in a variety of ways across studies (using different scales), as is the case here. One study [28] did not provide data regarding the use of antibacterials during the trial; the remaining four did (representing 88.4% of the population), but employed different ways of presenting their data. As shown in Table 3, the SMD shows a significant difference in favour of active treatment. In conclusion, although it is difficult to assign a specific value to this assertion owing to differences in the presentation of data, it may be concluded that with OM-89 there was a significant reduction in the use of antibacterials, in line with the reduction in UTIs described above.
Table 3 Consumption of antibacterials and standardised mean difference (SMD). Study
Treatment
Placebo
Weight
N
Mean (S.D.)
N
Mean (S.D.)
Bauer et al. [26] Pisani et al. [27] Schulman et al. [29] Tammen [30]
195 66 85 76
2.44 (1.75) 0.16 (0.36) 2.90 (18.00) 0.85 (1.65)
215 71 81 74
2.79 (2.03) 0.35 (0.48) 6.90 (19.00) 2.78 (5.40)
Total (95% CI)
422
S.D., standard deviation; CI, confidence interval. Test for heterogeneity, P = 0.32. Test for overall effect, P = 0.0001.
441
43.28 17.26 20.81 18.65 100
SMD (random) [95% CI]
Variable
−0.18 [−0.38, 0.01] −0.44 [−0.78,−0.10] −0.22 [−0.52, 0.09] −0.48 [−0.81,−0.16]
Prescriptions Patients with antibacterials Days with antibacterials/3 months Prescriptions
−0.29 [−0.44,−0.14]
116
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119 Table 6 Incidence of bacteriuria at final visit.
Table 5 Incidence of leukocyturia at final visit. OM-89 (%)
Placebo (%)
Significance (P-value)
Pisani et al. [27] Magasi et al. [28] Schulman et al. [29]
14.8 15.0 15.9
19.0 33.3 34.6
N.S. <0.05 <0.01
78.0 45.0 45.8
Total
15.2
28.6
<0.001
53.4
Difference
−13.3
<0.001
53.4
OM-89 (%)
Placebo (%)
2 (P-value)
19.5 3.7 3.3 11.0 –
20.9 11.4 21.7 17.9 –
N.S. (<0.1) <0.01 N.S.
Total
12.4
18.6
Difference
−6.2
RR (%)
RR, relative risk; N.S., not significant.
3.1.4. Findings at the end of the trials 3.1.4.1. Dysuria. Four of the five trials examined reported on this variable, representing 84.8% of the studied population (Table 4). Examining the outcome at 6 months, only a minority of patients reported dysuria, significantly less so among the OM-89-treated patients: −9.4% of the patients (P < 0.001; OR = 0.43, 95% CI 0.28–0.66). Examining the outcome as reported at the end of the studies (that is, the Bauer et al. trial [26] at 12 months) the results are similar, i.e. with OM-89 −6.6% of the patients had dysuria (P < 0.01; OR = 0.49, 95% CI 0.3–0.77). In conclusion, the incidence of dysuria at final visit was significantly lower with OM-89.
Bauer et al. [26] (6 months) Pisani et al. [27] Magasi et al. [28] Schulman et al. [29] Tammen [30]
RR (%) 93.1 32.5 15.4 61.1
<0.05
66.8
<0.05
66.8
RR, relative risk; N.S., not significant.
Bauer et al. trial [26] at 12 months) the results are similar, i.e. with OM-89 −5.4% patients had bacteriuria (P < 0.05; OR = 0.65, 95% CI 0.44–0.96). In conclusion, the incidence of bacteriuria at final visit was lower with OM-89 but the findings are inconsistent between trials. However, the incidence of bacteriuria among the OM-89-treated patients is not much higher than that reported in the general female population, which has been reported to lie between 2.8% (women aged 50–59 years) and 17% (women aged >75 years) [31]. 3.1.5. Safety of OM-89 in placebo-controlled clinical trials The differences of incidence of adverse events in comparative trials showed that they were slightly more frequent during OM-89 therapy than with placebo (+0.8%) and somewhat more patients withdrew from the trials due to side effects (+0.6%). No serious adverse events were attributed to OM-89 in these clinical studies; no disease- or age-related mortality was recorded in the studied population. The most frequent adverse events reported are detailed in Table 7.
3.1.4.2. Leukocyturia. Only three of the five trials examined reported on this variable, representing 42.6% of the population studied (Table 5). Examining the outcome at 6 months, only a minority of patients reported leukocyturia >5/field, significantly less so among the OM-89-treated patients: −13.3% of the patients (P < 0.001; OR = 0.45, 95% CI 0.28–0.72). In conclusion, the incidence of leukocyturia at final visit was significantly lower with OM-89 but the findings were derived from only three out of five trials, representing less than one-half the population studied.
4. Product ‘B’
3.1.4.3. Bacteriuria. Four of the five trials examined reported on this variable, representing 85.7% of the population studied (Table 6). Examining the outcome at 6 months, only a minority of patients reported bacteriuria, significantly less so among the OM-89treated patients: −6.2% of the patients (P < 0.05; OR = 0.62, 95% CI 0.42–0.91). However, this outcome was driven mainly by one study [28] as shown in the sensitivity analysis. There was significant heterogeneity between studies. In one trial [29] the data were provided for both 105 bacteria/mL and 104 bacteria/mL as the cut-off; taking this lower level, the differences become significant in this trial too (OM-89: 19.5% of patients; placebo 35.9% of patients). Examining the outcome as reported at the end of the studies (that is, the
There were no blind trials identified for the parenteral formulation of this product. This formulation causes significant undesirable side effects at the injection site [32] as well as systemic reactions [33]. The published studies of recent years dealt with the intravaginal formulation only, which is discussed herein. Because of the similarity of the data, design and outcomes, it was concluded that the study by Uehling et al. [34] published in 2001 was a preliminary presentation of the data published 2 years later and was therefore excluded from analysis. In the first study with this product [35], a total of 91 women with recurrent urinary infections were entered into a double-blind randomised study. Subjects received three vaginal suppositories at
Table 7 Adverse events reported in controlled clinical trials by ≥0.7% of patients. Common adverse events
OM-89
Placebo
Fisher’s test
Respiratory, thoracic and mediastinal disorders
‘Flu’, bronchitis, pharyngitis
4.3%
3.8%
N.S.
Nervous system disorders
Headache Somnolence and sleep disorders
2.2% 0.1%
2.8% 0.7%
N.S. N.S.
Gastrointestinal disorders
Gastric intolerance/dyspepsia Diarrhoea Nausea/vomiting
1.9% 1.3% 0.7%
2.0% 1.2% 1.0%
N.S. N.S. N.S.
Reproductive system and breast disorders Skin and subcutaneous tissue disorders
Vaginitis, leukorrhoea Allergic reaction/rash
1.0% 1.0%
2.6% 1.3%
P = 0.056 N.S.
Placebo > OM-89
Renal and urinary disorders
Renal pain, nephrolithiasis Pollakisuria and ejaculatory troubles, leukocyturia
0.7% 0.0%
0.0% 0.7%
P = 0.064 P = 0.064
OM-89 > placebo Placebo > OM-89
Musculoskeletal, connective tissue and bone disorders
Back pain
0.4%
1.7%
P = 0.048
Placebo > OM-89
N.S., not significant.
Comment
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
weekly intervals. Depending on the treatment group, each suppository contained one or two vaccine doses or suppository material only. Each patient was followed for 5 months to record infection episodes and to obtain urine, vaginal irrigates and serum in order to measure immunological responses. The verum-treated women who were off antibiotic prophylaxis (prevention discontinued in approximately one-half the patients) had a significant delay in the interval to re-infection during the first 8 weeks, and the mean interval until re-infection was delayed from 8.7 weeks for placebo-treated women to 13 weeks for vaccine-treated women (no significances provided). However, the overall incidence of urinary infections was similar, regardless of the treatment received (on average, 1.4 UTIs per patient). In the second and third study by the same group [36,37], a total of 54 and 75 women, respectively, were entered into doubleblind, placebo-controlled trials using the vaginal suppositories. Patients were withdrawn from prophylactic antibiotics and randomly assigned to one of three treatment groups, namely placebo only, primary immunisation (given on Weeks 0, 1 and 2) or primary plus booster immunisations (booster given on Weeks 6, 10 and 14). All women were followed for 6 months to determine the time until first recurrence and the number of infections. In the second study, the women receiving ‘primary + booster’ remained free of infections for a significantly longer period than those receiving placebo or ‘primary’ immunisation only. The patients receiving ‘primary + booster’ immunisation had significantly fewer infections (average 1.1 UTIs/patient) than the placebo-treated women (average 1.5 UTIs/patient) or those receiving ‘primary’ only (average 1.7 UTIs/patient). The interest of this study is limited by the small number of patients treated and the fact that placebo-treated patients were on average 13 years older than the verum-treated patients. Five of the 36 verum-treated patients experienced vaginal irritation within 1 day of immunisation or had transient diarrhoea. In the third study, the results were similar; that is, placebo was equivalent to ‘primary’ immunisation, and both were inferior to ‘primary + booster’ vaccine (Kaplan–Meier; P < 0.1). Thirty percent of the patients in the placebo group, 57% of the ‘primary’ only group and 72.5% of the patients in the ‘primary + booster’ group remained free of E. coli infection. Furthermore, the results suggested that the vaccine may provide the most benefit to women in the 20–50-yearold age group. The interest of this study is again limited by the small number of patients treated and the fact that placebo-treated patients were on average 9 years older than the verum-treated patients. Twelve percent of the women reported episodes of a burning sensation shortly after treatment both with the active product and with placebo. There were single occurrences of mild to moderate adverse events (four with the active product and four with placebo). A quantitative analysis of anti-E. coli IgA and IgG antibodies in the urine and vaginal fluid samples collected during the course of these two studies did not reveal any significant differences between the three groups. The findings regarding the difference between verum and placebo, as a percentage of patients with UTIs, are illustrated in Fig. 4. In conclusion, with product ‘B’, whilst the shorter ‘primary’ immunisation has not been effective, the longer ‘primary + booster’ treatment was probably effective. Confirmation by larger phase III studies by independent investigators is absolutely necessary.
5. Discussion and conclusions Management of recurrent UTI is complex due to the various factors predisposing to such bacterial invasions. It is further com-
117
Fig. 4. Percentage of patients without urinary tract infections: difference between verum and placebo. Product ‘B’ vaginal suppositories ‘primary’ on Weeks 0, 1 and 2 and ‘primary + booster’ additionally on Weeks 6, 10 and 14.
plicated by the increasing prevalence of antibiotic-resistant strains of E. coli. These facts stress the need for therapies directed at diminishing susceptibility and enhancing the host’s defences against UTI rather than administering antibacterials on a broad base. The urinary tract possesses the immunological and cellular machinery for an efficient antibacterial defence and it seems a sensible approach to activate it. Among the various immunotherapeutic products offered on the market, only two have published studies that are in accordance with current standards. Of these, product ‘A’ (OM-89) is fairly well documented and has been shown to be more effective than placebo in several randomised trials. Product ‘B’ has been shown to be effective only when administered with a booster cycle in three small phase II trials, and larger studies by independent investigators are awaited with interest. Concentrating on OM-89, the data show that the product reduces the number of recurrences of UTI by ca. 36%, and one out of five patients will be entirely free of UTIs for at least 6 months, with a corresponding reduction of antibacterial treatments. At the end of the 6-month observation period, fewer OM-89-treated patients had dysuria (RR = 48%) and possibly leukocyturia (RR = 53%) or bacteriuria (RR = 67%). A further aspect worthwhile examining is the potential reduction of vulvovaginal infections with OM-89 (OR = 0.42, 95% CI 0.18–0.98) versus placebo in women with recurrent UTIs. However, the studies examined do not provide any information on how to optimise treatment. Since only approximately one-third of the UTIs are avoided with this treatment, it would be important to know whether gender, type of infection, secretor status, hormonal status, concentration of THP or other variables are predictors of response to the product. The studies with OM-89 appear to have good internal and external validity; that is, the investigations were conducted with outpatients with relatively broad admission or exclusion criteria and thus can be generalised to the circumstances of daily practice. It seems quite probable that a particular subpopulation might benefit more than others. Unfortunately, there are no studies comparing directly antimicrobial prophylaxis with immunoactive prophylaxis; head-to-head studies are planned. From the reported results in the literature, it seems that the efficacy of antimicrobial prophylaxis is superior to immunisation regimens presently available [2]. However, since 1990 there has been a steadily increasing rate of resistance to trimethoprim/sulfamethoxazole, reaching >30% in some areas; in the USA
118
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119
the average resistance rate for E. coli is 18%. Moreover, resistance rates to quinolones have been increasing worldwide at alarming rates, reaching >20% in Spain [38–41]. Another potential downside of antibacterial prophylaxis is the side effects such as rash, nausea, diarrhoea, vaginitis and oral candidiasis [42,43]. Treatments with antibacterials have to be monitored closely since they can also cause major adverse reactions, e.g. phototoxicity and rare but serious hepatotoxicity has been reported with fluoroquinolones [44], whilst nitrofurantoin is known infrequently to cause both acute and chronic pulmonary reactions (interstitial lung disease and pulmonary fibrosis may develop with long-term use) [45], as well as rare cases of blood dyscrasia, erythema exsudativum multiforme (Stevens–Johnson syndrome) and epidermolysis acuta toxica (Lyell syndrome) with trimethoprim/sulfamethoxazole or with nitrofurantoin, to mention a few [46]. Additionally, oral administration of antibiotics for treatment of UTI can cause ecological disturbances in the normal urethral [47] and intestinal microflora [48]. To increase the mucosal defences through immunoactive prophylaxis looks like a compelling alternative that merits further studies and improvements. The future may bring new and possibly more effective immunostimulants, such as the vaccine derived from purified E. coli FimH adhesion [49], P. mirabilis fimbriae (PMF and purified recombinant structural fimbrial proteins of these fimbriae) [50,51] and others, currently tested in animals.
Acknowledgment The authors are thankful to Dr Reto G. Brignoli for assistance with the statistical analysis and preparation of the manuscript. Funding: This project was partially funded by OM Pharma, Geneva, Switzerland. The sponsors had no involvement in the study design, analysis and interpretation of the data, writing of the manuscript, or in the decision to submit the manuscript for publication. Competing interests: KGN: Bayer (investigator, speaker at scientific meetings), Bionorica (investigator, consultant), Eumedica Pharmaceuticals (consultant), MerLion Pharmaceuticals (investigator, consultant), Daiichi (speaker at scientific meetings, scientific publication), MUCOS Pharma (consultant, speaker at scientific meetings), NanoVibronix (consultant), Ocean Spray Cranberries (investigator), OM Pharma (investigator, consultant), Peninsula/Johnson & Johnson/Janssen-Cilag (investigator, speaker at scientific meetings), Pharmatoka (investigator, consultant), Polyphor (consultant), Protez (consultant), Rosen Pharma (consultant), sanofi-aventis (investigator, speaker at scientific meetings, scientific publication), UroVision (investigator) and Zambon (investigator, speaker at scientific meetings). Y-HC: Asia Pharm (consultant, speaker at scientific meetings), Handok Pharmaceuticals (consultant, speaker at scientific meetings) and Pacific Pharma (consultant, speaker at scientific meetings). TM: Daiichi Sankyo (consultant, investigator, speaker at scientific meetings), Asteras (consultant, investigator), Kyorin (investigator, speaker at scientific meetings), Meiji (consultant, investigator), Nisshin Science (consultant, investigator), Takeda (speaker at scientific meetings), Taisho-Toyama (investigator, speaker at scientific meetings) and Combinature/MerLion (consultant). AJS: American Urological Association (editorial fees and faculty honorarium), IMS Health (consultant), Alita Pharmaceuticals, Inc. (consultant), NovaBay Pharmaceuticals, Inc. (consultant), cme2 (speaker at scientific meeting), Blackwell Publishing (book royalty), Advanstar Communications Inc. (consultant), Elsevier Science USA (book royalty) and Regeneron Pharmaceuticals, Inc. (consultant). Ethical approval: Not required.
References [1] Lindsay EN. Managing recurrent urinary tract infections in women. Womens Health 2005;1:39–50. [2] Naber KG (Chairman), Bishop MC, Bjerklund-Johansen TE, Botto H, C¸ek M, Grabe M, et al. Guidelines on the management of urinary and male genital tract infections. European Association of Urology; 2006. p. 7–8, Appendices 12.1 [Criteria for the diagnosis of a UTI], p. 109. [3] Fihn SD. Acute uncomplicated urinary tract infection in women. N Engl J Med 2003;349:259–66. [4] Boyko EJ, Fihn SD, Scholes D, Abraham L, Monsey B. Risk of urinary tract infection and asymptomatic bacteriuria among diabetic and nondiabetic postmenopausal women. Am J Epidemiol 2005;161:557–64. [5] Raz R, Gennesin Y, Wasser J, Stoler Z, Rosenfeld S, Rottensterich E, et al. Recurrent urinary tract infections in postmenopausal women. Clin Infect Dis 2000;30:152–6. [6] Ishitoya S, Yamamoto S, Mitsumori K, Ogawa O, Terai A. Non-secretor status is associated with female acute uncomplicated pyelonephritis. BJU Int 2002;89:851–4. [7] Schaeffer AJ, Rajan N, Cao Q, Anderson BE, Pruden DL, Sensibar J, et al. Host pathogenesis in urinary tract infections. Int J Antimicrob Agents 2001;17:245–51. [8] Hidalgo-Tenorio C, Jimenez-Alonso J, de Dios Luna J, Tallada M, Martinez-Brocal A, Sabio JM. Urinary tract infections and lupus erythematosus. Ann Rheum Dis 2004;63:431–7. [9] Hughes LB, Criswell LA, Beasley TM, Edberg JC, Kimberly RP, Moreland LW, et al. Genetic risk factors for infection in patients with early rheumatoid arthritis. Genes Immun 2004;5:641–7. [10] Schmidhammer S, Ramoner R, Holtl L, Bartsch G, Thurnher M, Zelle-Rieser C. An Escherichia coli-based oral vaccine against urinary tract infections potently activates human dendritic cells. Urology 2002;60:521–6. [11] Boruchov AM, Heller G, Veri MC, Bonvini E, Ravetch JV, Young JW. Activating and inhibitory IgG Fc receptors on human DCs mediate opposing functions. J Clin Invest 2005;115:2914–23. [12] David J, Terhorst C. Organs and cells of the immune system. In: Scientific American Medicine; Ch. 6. Immunology. WebMD Corp.; 1999. [13] Jerome KR, Corey L. The danger within. N Engl J Med 2004;350:411–2. [14] Anders HJ, Patole PS. Toll-like receptors recognize uropathogenic Escherichia coli and trigger inflammation in the urinary tract. Nephrol Dial Transplant 2005;20:1529–32. [15] Yahr TL. A critical new pathway for toxin secretion? N Engl J Med 2006;355:1171–2. [16] Kerschbaum HH, Singh SK, Hermann A. Lymphoid tissue in the kidney of the musk shrew, Suncus murinus. Tissue Cell 1995;27:421–4. [17] Thilagarajah R, Witherow RO, Walker MM. Quantitative histopathology can aid diagnosis in painful bladder syndrome. J Clin Pathol 1998;51:211–4. [18] Franz M, Hörl WH. Common errors in diagnosis and management of urinary tract infection. I. Pathophysiology and diagnostic techniques. Nephrol Dial Transplant 1999;14:2746–53. [19] Kunin CM, Evans C, Bartholomew D, Bates DG. The antimicrobial defense mechanism of the female urethra: a reassessment. J Urol 2002;168:413–9. [20] Pudney J, Anderson DJ. Immunobiology of the human penile urethra. Am J Pathol 1995;147:155–65. [21] Bauer HW, Rahlfs VW, Lauener PA, Blessmann GS. Prevention of recurrent urinary tract infections with immuno-active E. coli fractions: a meta-analysis of five placebo-controlled double-blind studies. Int J Antimicrob Agents 2002;19:451–6. [22] Alderson P, Green S, Higgins JPT, editors. Cochrane Handbook for Systematic Reviews of Interventions [version 4.2.2, updated March 2004]. In: The Cochrane Library, Issue 1. Chichester, UK: John Wiley & Sons, Ltd.; 2004. http://www.cochrane.org/resources/handbook/index.htm [accessed 8 September 2008]. [23] Bax L, Yu LM, Ikeda N, Tsuruta H, Moons KGM. Development and validation of MIX: comprehensive free software for meta-analysis of causal research data. BMC Med Res Methodol 2006;6:50. [24] Hachen HJ. Oral immunotherapy in paraplegic patients with chronic urinary tract infections: a double-blind, placebo-controlled trial. J Urol 1990;143:759–62. [25] Frey C, Obolensky W, Wyss H. Treatment of recurrent urinary tract infections: efficacy of an orally administered biological response modifier. Urol Int 1986;41:444–6. [26] Bauer HW, Alloussi S, Egger G, Blumlein HM, Cozma G, Schulman CC; Multicenter UTI Study Group. A long-term, multicenter, double-blind study of an Escherichia coli extract (OM-89) in female patients with recurrent urinary tract infections. Eur Urol 2005;47:542–8. [27] Pisani E, Palla R, Bono AV. Double-blind randomised clinical study of OM-8930 vs placebo in patients suffering from recurrent urinary tract infections. Geneva, Switzerland: OM Pharma; 1992. (Data on file). Quoted in: Chiavaroli C, Moore A. An hypothesis to link the opposing immunological effects induced by the bacterial lysate OM-89 in urinary tract infection and rheumatoid arthritis. BioDrugs 2006;20:141–9. [28] Magasi P, Panovics J, Illes A, Nagy M. Uro-Vaxom and the management of recurrent urinary tract infection in adults: a randomized multicenter double-blind trial. Eur Urol 1994;26:137–40.
K.G. Naber et al. / International Journal of Antimicrobial Agents 33 (2009) 111–119 [29] Schulman CC, Corbusier A, Michiels H, Taenzer HJ. Oral immunotherapy of recurrent urinary tract infections: a double-blind placebo-controlled multicenter study. J Urol 1993;150:917–21. [30] Tammen H; The German Urinary Tract Infection Study Group. Immunobiotherapy with Uro-Vaxom in recurrent urinary tract infection. Br J Urol 1990;65:6–9. [31] Scottish Intercollegiate Guidelines Network. Management of suspected bacterial urinary tract infection in adults. A national clinical guideline. July 2006. [32] Pam D. Vaginal vaccine for recurrent urinary-tract infections. Lancet Infect Dis 2002;2:68. [33] Überempfindlichkeitsreaktionen auf ‘Impfstoff’ STROVAC [Hypersensitivity reactions to ‘vaccine’ STROVAC]. Arznei-Telegram 2007; 38(2):31. [34] Uehling DT, Hopkins WJ, Beierle LM, Kryger JV, Heisey DM. Vaginal mucosal immunization for recurrent urinary tract infection: extended phase II clinical trial. J Infect Dis 2001;183(Suppl. 1):S81–3. [35] Uehling DT, Hopkins WJ, Balish E, Xing Y, Heisey DM. Vaginal mucosal immunization for recurrent urinary tract infection: phase II clinical trial. J Urol 1997;157:2049–52. [36] Uehling DT, Hopkins WJ, Elkahwaji JE, Schmidt DM, Leverson GE. Phase 2 clinical trial of a vaginal mucosal vaccine for urinary tract infections. J Urol 2003;170:867–9. [37] Hopkins WJ, Elkahwaji J, Beierle LM, Leverson GE, Uehling DT. Vaginal mucosal vaccine for recurrent urinary tract infections in women: results of a phase 2 clinical trial. J Urol 2007;177:1349–53. [38] Hooton TM, Stamm WE. Diagnosis and treatment of uncomplicated urinary tract infection. Infect Dis Clin North Am 1997;11:551–81. [39] Kahlmeter G. The ECO.SENS Project: a prospective, multinational, multicentre epidemiological survey of the prevalence and antimicrobial susceptibility of urinary tract pathogens—interim report. J Antimicrob Chemother 2000;46(Suppl. 1):15–22. [40] Gupta K, Sahm DF, Mayfield D, Stamm WE. Antimicrobial resistance among uropathogens that cause community-acquired urinary tract infections in women: a nationwide analysis. Clin Infect Dis 2001;33:89–94. [41] Manges AR, Johnson JR, Foxman B, O’Bryan TT, Fullerton KE, Riley LW. Widespread distribution of urinary tract infections caused by a multidrug-
[42]
[43]
[44] [45]
[46] [47]
[48]
[49]
[50]
[51]
119
resistant Escherichia coli clonal group. N Engl J Med 2001;345:1007– 13. Albert X, Huertas I, Pereiró I, Sanfélix J, Gosalbes V, Perrota C. Antibiotics for preventing recurrent urinary tract infection in non-pregnant women. Cochrane Database Syst Rev 2004;(3). CD001209. Gradwohl SE (Team Leader); University of Michigan Health System (Agency for Healthcare Research and Quality (AHRQ)). Urinary tract infection. Ann Arbor, MI: University of Michigan Health System; 2005. http://www.guidelines. gov/summary/summary.aspx?view id=1&doc id=7407 [accessed 3 September 2008]. Ball P, Mandell L, Niki Y, Tillotson G. Comparative tolerability of the newer fluoroquinolone antibacterials. Drug Saf 1999;21:407–21. Sheehan RE, Wells AU, Milne DG, Hansell DM. Nitrofurantoin-induced lung disease: two cases demonstrating resolution of apparently irreversible CT abnormalities. J Comput Assist Tomogr 2000;24:259–61. Howe RA, Spencer RC. Cotrimoxazole: rationale for re-examining its indications for use. Drug Saf 1996;14:213–8. Norinder BS, Norrby R, Palmgren AC, Hollenberg S, Eriksson U, Nord CE. Microflora changes with norfloxacin and pivmecillinam in women with recurrent urinary tract infection. Antimicrob Agents Chemother 2006;50: 1528–30. Edlund C, Nord CE. Effect on the human normal microflora of oral antibiotics for treatment of urinary tract infections. J Antimicrob Chemother 2000;46(Suppl. 1):41–8. Langermann S, Mollby R, Burlein JE, Palaszynski SR, Auguste CG, DeFusco A, et al. Vaccination with FimH adhesin protects cynomolgus monkeys from colonization and infection by uropathogenic Escherichia coli. J Infect Dis 2000;181:774–8. Pellegrino R, Galvalisi U, Scavone P, Sosa V, Zunino P. Evaluation of Proteus mirabilis structural fimbrial proteins as antigens against urinary tract infections. FEMS Immunol Med Microbiol 2003;36:103–10. Li X, Lockatell CV, Johnson DE, Lane MC, Warren JW, Mobley HL. Development of an intranasal vaccine to prevent urinary tract infection by Proteus mirabilis. Infect Immun 2004;72:66–75.