Urologic Chronic Pelvic Pain Syndrome Flares: A Comprehensive, Systematic Review and Meta-Analysis of the Peer-Reviewed Flare Literature.

Overall, 59 articles provided data related to UCPPS flares ( Table 1 ). Most included data from North America (75%), followed by Asia (12%), Northern Europe (including the United Kingdom; 10%), Russia (1.7%), and one international study (1.7%). Approximately half were descriptive in nature (including qualitative and cross-sectional articles [59%]), and recruited participants from only one or a few specialty clinics (e.g., Urology [49%]). Fifteen articles included participants from the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network (25%). Over half studied female IC/BPS participants only (or predominantly female [61%]), 22% studied male CP/CPPS participants only, and 19% studied both female and male participants. Most articles included participants with confirmed diagnoses (90%). Participants’ mean age ranged from 36–64 years (median: 32–51 years) and the majority self-identified as White (57–100%), followed by Asian in studies conducted in Asia. Participants’ mean condition duration ranged from >0.6–17 years (median: 0.1–12 years) and their typical pelvic pain intensity ranged from a mean of 2.9–9.1 out of 10 (median: 4–6). The most commonly used term for symptom exacerbations by article authors was “flare” or derivations thereof, followed by “exacerbation,” “relapse,” “aggravation,” and “attack” ( Table 1 ). Additional terms used by investigators were “increased,” “intense,” or “breakthrough” pain; and “worse,” “worst,” or “acute worsening” symptoms. Nine articles assessed participants’ own terminology for exacerbations or incorporated their exacerbation-related quotes into the manuscript (not including studies that used exacerbation-related language in their study instructions or moderator/interviewer guide, or that did not provide their guide). These studies included an older mixed-methods study of IC/BPS participants; 8 as well as several more recent qualitative and mixed-methods studies of mainly IC/BPS participants. 9 – 16 The most commonly used term by participants in the older study was “attack,” 8 whereas the most commonly used term in more recent studies was “flare” or derivations thereof. 9 – 16 Additional terms mentioned were “bad,” “bad bladder,” or “stay-in-bed” days; “episodes,” “pelvic floor episodes,” “incidents,” “exacerbation,” and “inflammation.” 9 , 11 – 13 Participants also sub-divided their flares into “minor” or “moderate” flares versus “major” flares and other similar terms. 9 , 13 , 15 Twenty-one articles provided information on flare manifestation, 8 , 9 , 12 , 13 , 17 – 33 although only six documented participants’ experiences prospectively over time. 21 – 23 , 25 , 28 , 29 Most participants reported at least one flare during their lifetime and/or symptom fluctuations (57–100%; Appendix Tables 1 – 2 ). However, typical or recent flare frequency varied considerably across studies, driven largely by the stringency of the flare definition and inclusion criteria of the study. The highest flare frequency was observed for flares of very short duration (<1 hour) defined as “urologic or pelvic pain symptoms that are much worse than usual” in a study of UCPPS participants not following any specific therapy (25%), 24 whereas the lowest frequency was observed for flares of longer duration defined as a 50% increase in “breakthrough pain” lasting ≥2 weeks in another study of IC/BPS participants on triple gabapentin, amitriptyline, and non-steroidal anti-inflammatory drug therapy (6.7% every 2–6 months or less). 22 With respect to flare duration, eight articles provided data, 8 , 9 , 12 , 13 , 24 , 25 , 28 , 33 including six from the MAPP Research Network 9 , 13 , 24 , 25 , 28 , 33 and only one that documented flare duration prospectively (limited to days-long flares). 25 Three studies assessed participants’ lifetime history of flares of any duration, all of which were performed in the MAPP Network ( Appendix Table 1 ). 9 , 13 , 24 In one of these studies, 37% of participants reported ever experiencing flares lasting only minutes or <1 hour; 61% reported flares lasting several hours but less than one day; and 76% reported flares lasting ≥1 day. 24 A small, but unknown, percentage of participants also reported flares with extreme durations (e.g., only during urination or up to months or years) in two qualitative studies. 9 , 13 These values were skewed towards longer flares in studies that assessed participants’ typical flare experiences: 16% reported flares lasting <1 day in one study 33 and 38–84% reported flares lasting ≥1 day (including 16% lasting ≥1 week 33 ) in two studies. 8 , 33 Variability in flare duration was also observed within the same participant in three studies (28–51% variable duration; 8 , 24 94% intra-individual variability 28 ). With respect to symptom intensity, fourteen articles provided data, 8 , 9 , 12 , 13 , 19 , 20 , 22 , 24 , 25 , 27 – 29 , 32 , 33 half of which were conducted in the MAPP Network 9 , 13 , 24 , 25 , 28 , 29 , 33 and only four of which documented participants’ symptoms prospectively. 22 , 25 , 28 , 29 These articles indicated that pelvic pain and urologic symptom intensity were elevated during flares (mean=5.2–8.8 [median=8.0] for pain 8 , 20 , 22 , 24 , 25 , 27 – 29 , 33 and mean=3.5–5.6 for urologic symptoms out of 10 24 , 25 , 28 , 29 ), and greater than non-flare or baseline symptoms by a mean of 2.6–5.8 points [median=4.0] for pain 22 , 25 , 28 , 29 , 33 and a mean of 1.7–2.2 points for urologic symptoms ( Appendix Table 1 ). 25 , 28 , 29 Data from a small number of studies also indicated that: 1) symptom intensity could vary across flares from the same participant (intra-individual variability=38% for pain and 26% for urologic symptoms 28 ); 9 , 13 2) symptom intensity was significantly greater for flares of longer duration; 24 3) most flares involved increases in both pain and urologic symptoms (61%), and only a small percentage involved increases in pain (20%) or urologic symptoms only (5.8%); 28 and 4) flares could also be accompanied by non-UCPPS symptoms, such as diarrhea and extra-pelvic pain ( Appendix Table 3 ). 9 , 13 , 24 , 28 , 29 One small study attempted to identify symptoms independently associated with flare onset and found that only pelvic/genital pain (≥2 point increase) and urination-related pain (≥1 point increase) were independently predictive. 29 With respect to other flare characteristics, a small number of studies observed variability in flare onset (sudden [54% 17 ] to gradual [45% 17 ] 9 , 13 ), remission (rapid to gradual 9 , 13 ), and diurnal timing (no typical time of day or night [91%] 30 ). Thirty-nine articles investigated flare triggers, including 13 that examined potential triggers empirically and one that tested potential triggers experimentally. A wide variety of potential triggers were investigated and reported in these studies. The most commonly investigated trigger was diet and the most commonly reported dietary triggers were fruits and fruit juices (particularly citrus fruits and tomatoes), spicy food and chili/hot peppers, caffeinated (e.g., tea, coffee) and alcoholic beverages, and chocolate ( Table 2 , Appendix Figure 2 and Appendix Tables 4 – 6 ). Female participants tended to be more likely to report food triggers than male participants. Only one study investigated individual dietary triggers empirically, observing positive associations for greater recalled intake of tomatoes; yoghurt; spicy foods; caffeinated, carbonated beverages; and alcohol with flare onset, but only among participants who believed that diet had triggered their flare. 34 Other commonly reported triggers were stress; physical activity, sedentary behavior, and jostling forms of transportation; various infections and infection-like symptoms; menstrual cycle variation; cold weather and other cold experiences (e.g., sitting on a cold surface); and additional exposures, such as tight clothing ( Appendix Tables 7 – 10 ). In the small number of studies that investigated these factors empirically, some evidence of association was observed for stress, particularly in the one small cohort 35 and trial 36 on this topic; abdominal muscle exercises; 34 vaginal and urinary tract infection-like symptoms, 34 which likely reflects the similarity in symptoms between these infections and flares; spring, autumn, and winter season; 23 and increases in pollen. 37 Thirty articles investigated flare treatment and prevention strategies, including eight small studies that tested these empirically ( Table 3 and Appendix Tables 11 – 12 ). Common medical strategies included contacting health care providers for general medical care, hydrodistension, and bladder instillations; and taking various prescription, over-the-counter, and complementary medications and products. Only two of these strategies were tested empirically. Hormonal therapy was associated with symptom improvement in IC/BPS patients with menstrual cycle-related flares in one small retrospective cohort study, 38 and prednisolone improved symptoms in female IC/BPS patients when added to triple gabapentin, amitriptyline, and non-steroidal anti-inflammatory drug therapy in one small single arm trial. 22 Common non-medical or behavioral strategies for flare treatment and prevention included thermal therapy (heat and cold); increasing water or fluid intake; resting or relaxation techniques, such as meditation; light exercise; consuming baking soda; and dietary changes, such as avoiding perceived trigger foods and adopting a bland diet (e.g., rice, bread). This last strategy was tested in one small single arm trial 39 and two small randomized controlled trials with 40 and without 41 an anti-inflammatory diet; each of these studies observed symptom improvement. Combinations of medical and behavioral strategies were also tested in one case report and two pilot randomized controlled trials, observing symptom improvement for acupuncture, gabapentin, ibuprofen, and ranitidine combined; 42 and for over 10 behavioral strategies combined. 43 , 44 Finally, in studies that assessed participants’ preferences for flare treatment, many mentioned the importance of a structured flare plan, particularly one including self-management. 9 , 11 , 32 , 45 Participants in one study also mentioned being more open to taking UCPPS medications during flares than at other times. 32 Thirteen articles presented data on the impact of flares on participants’ QOL, including 10 qualitative/mixed-methods and three cross-sectional analyses. 9 – 16 , 24 , 27 , 32 , 33 , 45 During flares, participants reported a wide range of impact – from difficulty concentrating, irritability, anxiety, drowsiness from pain medications, difficulty sleeping, and cancelling social plans to leaving/staying home from work or school, staying in bed or close to a toilet, going to the emergency room, and feeling depressed and occasionally suicidal. 9 – 15 , 27 , 45 The degree of impact tended to be related to the intensity, duration, frequency, and unpredictability of symptoms, particularly pain. 9 , 13 , 24 , 33 Findings for flare frequency persisted in one study after adjustment for average pain intensity, suggesting that flares negatively impact participants’ lives over and above typical pain intensity. Other flare characteristics were not tested in this analysis. 33 In addition to an immediate impact, participants also reported a longer-term impact of flares. This included small adjustments to participants’ lives, such as changes to their diet, clothing, and level of travel, spontaneity, and sexual activity to much larger changes, such as social isolation from avoiding social events with food, educational and career changes, employment loss, disability leave, marital strain, and depression. Participants also reported a constant vigilance for flare triggers, attempt to control these factors, and worry about future flares. 9 – 11 , 13 , 16 , 32 Only three, small cohort analyses investigated additional aspects of flares that might inform flare pathophysiology, all of which were conducted in the MAPP Research Network. Schrepf and colleagues 46 observed no association for toll-like receptor-2 inflammation, toll-like receptor-4 inflammation, and diurnal cortisol slope with flare frequency in IC/BPS participants. Xu and colleagues 29 observed that participants with chronic overlapping pain conditions reported increased extra-pelvic pain intensity and widespread-ness during flares than at non-flare times, possibly suggesting increased central sensitization or global hypersensitivity during flares; and Harte and colleagues 47 observed that participants had greater pressure pain sensitivity during flares, again possibly suggesting increased global hypersensitivity during flares.

We searched the literature for records including the concepts of UCPPS and exacerbation using a combination of keywords and controlled vocabulary in Embase.com 1947-, Ovid Medline 1946-, Scopus 1823-, The Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, and Clinicaltrials.gov 1997- (PROSPERO ID: 288603). Initial searches were performed on November 3, 2021 and again on February 28, 2023, with no added limits. In April 2023, we discovered that several qualitative articles that included flare-related quotes were not captured with the initial search terms. Therefore, we created a supplemental search to capture the concepts of UCPPS and qualitative research (performed on April 14, 2023; see Appendix for fully reproducible search strategies). We also hand-searched the references of all identified articles and contacted select study authors for additional data. We included English-language, full-text articles that addressed IC/BPS and/or CP/CPPS in humans. Articles were excluded if they: 1) focused exclusively on chronic bacterial prostatitis; lupus, radiation-induced, or other forms of cystitis; endometriosis; gynecological chronic pelvic pain; or sequelae of other conditions (e.g., Crohn’s disease); or 2) did not include separate analyses of UCPPS participants in studies of multiple conditions. UCPPS-specific articles were further excluded if they: 1) did not address flares; 2) focused exclusively on iatrogenic flares following organ-specific procedures (e.g., prostate biopsy, bladder instillation); or 3) focused exclusively on pain with or following sexual activity, ejaculation, or urination; or pain relieved by urination, as these are well-known characteristics of UCPPS that have already been incorporated into patient-reported outcome measures. 6 As no well-accepted definition of flares exists, we included all studies that reported any form of symptom exacerbation that was transient in nature (i.e., observed to resolve) or that participants described as a symptom exacerbation ( Appendix Figure 1A ). We did not include studies that described sustained symptom worsening (i.e., worsening that was not observed or described to resolve; 1B), or symptom recurrence or worsening following a transiently successful therapy (e.g., “wearing off” of the effects of a therapy; 1C). In addition, given the general sparsity of the flare literature, we included studies of any design, methodology, or size, as long as they addressed at least one of the following aspects of flares: patient terminology, manifestation, triggers, treatments and prevention strategies (including studies that described/tested interventions to reduce flare triggers even if they did not evaluate flares as their outcome), QOL impact, or pathophysiologic mechanisms. We identified 1,486 unique abstracts: 1,149 from the initial, 87 from the updated, and 250 from the supplemental search ( Figure 1 ). These were reviewed for relevance by two authors each. 1,088 citations did not meet eligibility criteria and were removed, leaving 398 for full-text review (303 from the original, 18 from the updated, and 77 from the supplemental search). Full-text articles of all original UCPPS research studies were reviewed, even if they did not mention flares in the abstract, to ensure identification of flare-related data found only in the body of the manuscript. Full-text articles were reviewed by two authors each and discrepancies were resolved by team consensus. 352 full-text articles were excluded at this stage, leaving 47 eligible articles (32 from the original, 3 from the updated, and 12 from the supplemental search). Hand-searches identified an additional 12 articles, bringing the total number of eligible articles to 59. Data extraction was performed by at least two authors. When articles included overlapping information, data from the largest analysis and/or the one dedicated to addressing flare-related questions were selected. Given the wide range of study designs/methodologies included, data were summarized in several different ways: as 1) the number of studies that mentioned a concept for qualitative studies; 2) a summary prevalence estimate for cross-sectional and descriptive case-control studies; 3) individual results for cohort-based studies, case reports, and single arm trials because of the small number of studies with similar information; and 4) a summary difference in mean change across arms in pilot randomized controlled trials. Summary estimates were calculated by generalized linear mixed (“random effects”) models. Models for prevalence estimates included a logit link and robust variance estimation 7 to account for inclusion of multiple estimates per study (e.g., multiple individual citrus fruit values in the citrus fruit category). Results were presented for all participants combined, as well as separately by sex to approximate clinical diagnosis (IC/BPS or CP/CPPS). With respect to methodologic quality, we did not use a formal instrument, as few are designed to evaluate data from the wide range of study designs and methodologies included in our review (i.e., from qualitative to experimental studies). Instead, we presented aspects of study design and conduct that we considered important for interpreting study findings (e.g., clinical confirmation of diagnoses, sampling frame, study design, etc.). Finally, we did not present data for known aspects of UCPPS, including procedure-based iatrogenic flares; symptom worsening with/after bladder filling, urination, ejaculation, and sexual activity; and symptom improvement following urination.

Overall, our review identified a large number of flare-related articles (N=59), although most were limited to one aspect of flares and/or participant self-report (i.e., qualitative or cross-sectional studies). Only a small number of studies documented flare manifestation prospectively or tested proposed flare triggers and treatments empirically. In addition, almost one quarter were derived from one main study population, the US-based, multi-site MAPP Epidemiology and Phenotyping Study. Nevertheless, accumulated articles still provide useful, foundational data for future flare research. The term “flare” or derivations thereof were used commonly and well-understood by participants, although additional terms, such as exacerbation or relapse, were also used. Flares tended to involve significant increases in pelvic pain, but less data were available on flare frequency and duration. Painful, frequent, long, and/or unpredictable flares were highly impactful, even over and above participants’ non-flare symptoms. A large number of triggers were proposed to cause flares and many therapeutic strategies were reported by participants, providing numerous hypotheses for empirical research; however, few studies investigated flare triggers or treatments empirically, or attempted to understand the biologic mechanisms underlying flare occurrence. Despite their frequent mention in the lay literature and growing number of studies in the peer-reviewed literature, no standard definition of flares exists. In our review, patient-reported flare experiences tended to be painful and transient, but were otherwise variable in frequency, duration, presence and intensity of additional symptoms, and predictability. Reported flares occurred infrequently (e.g., less than once/year) to multiple times per day; lasted a few seconds (only during urination) or minutes to multiple years; involved small increases in pelvic pain, as well as other UCPPS and extra-pelvic symptoms, to large increases in symptoms; and occurred predictably for some participants and unpredictably for others. This wide variability makes it difficult to define and study flares. It also raises the possibility of varying pathophysiology and triggers across flare experiences. For instance, patient-reported flares that occur only during urination could potentially reflect the bladder hypersensitivity feature of UCPPS itself or possibly acute pelvic floor muscle spasms. Minutes-long painful flares could also potentially reflect these spasms. In contrast, longer hours- to days-long painful flares may be more likely to reflect a pathophysiologic process that includes interactions between a trigger (e.g., diet) and proposed biologic flare mediators (e.g., bladder mast cell/histamine-mediated mechanisms; increased central sensitization; or global hypersensitivity). 29 , 47 – 49 Finally, flares that extend months to years may potentially reflect an overall worsening of patients’ conditions rather than a transient experience. Although complex, this issue is not unlike the hypothesized heterogeneity of UCPPS itself. UCPPS is often classified and studied as one or two conditions (i.e., IC/BPS and CP/CPPS), but may potentially represent several more pathophysiologically-distinct conditions. Therefore, future research should explore the potential for pathophysiologically-distinct flares to facilitate preventive and therapeutic flare research. In the meantime, however, strategies such as collecting information to characterize flares by frequency, duration, intensity, and other characteristics; and performing analyses stratified by or limited to certain flare manifestations should facilitate future flare research. The development of a validated flare outcome measure would also aid future research. Additional flare research is needed because, despite data to suggest that flares impact patients’ QOL and health care-seeking behavior, little scientific data exist on flare triggers and treatments. Although many triggers have been proposed and many therapies have been used by patients, few studies have investigated these triggers and treatments empirically to inform the efficacy of prevention and treatment strategies. Data from the small body of empirical research provide preliminary support for diet, stress, abdominal muscle exercises, and pollen as potential flare triggers; and for steroids, hormones, dietary modification, and various other medical and behavioral strategies combined as potential prevention and treatment strategies, but additional empirical research that tests these strategies is still clearly needed. Although this type of research would be facilitated by a better understanding of flare pathophysiology and potential flare heterogeneity, much could still be learned even in the absence of such data simply by using an empirical approach. Strengths of our review include its systematic nature, rigorous execution (including double data extraction and review), and comprehensive evaluation of multiple aspects of flares. Similar to most reviews, limitations include those of the existing literature, most notably the dearth of empirical studies. In summary, our comprehensive review demonstrated that flares are painful and impactful, but otherwise poorly understood in terms of manifestation, triggers, treatment, and underlying pathophysiology. Development of a validated flare outcome measure and collection of detailed flare manifestation data would facilitate much-needed, new prospective and empirical research on UCPPS flares to reduce their high patient burden.

Urologic chronic pelvic pain syndrome (UCPPS) is an umbrella term that encompasses interstitial cystitis/bladder pain syndrome (IC/BPS) and chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS). Both of these syndromes are characterized by chronic pelvic, bladder, or prostate pain; and urinary symptoms, such as urinary urgency and frequency. The etiology and pathophysiology of UCPPS are poorly-understood, and its symptoms are difficult to diagnose and treat. 1 UCPPS affects an estimated 1.8–6.5% 2 , 3 of the US population and contributes to considerable healthcare expenditures, 4 as well as significant reductions in physical and mental quality of life (QOL). 2 , 5 Similar to many other chronic pain conditions, UCPPS symptoms are widely known to fluctuate, including periods of symptom exacerbation (or “flares”) and remission. Although flares are well-recognized by patients and providers, and have been discussed extensively in the lay literature (e.g., self-help books and patient websites), they are much less well-represented in the scientific literature. Most of the extant literature is limited to descriptions of patients’ beliefs about flare triggers and their treatment and management strategies. Therefore, to support future empirical research, we systematically reviewed and summarized the existing, peer-reviewed literature on numerous aspects of flares to serve as a foundation for future flare-related research. Aspects included in our comprehensive, systematic review and meta-analysis were terminology; manifestation; perceived and tested triggers, therapies, prevention strategies; impact on QOL; and insight into pathophysiologic mechanisms.