Longitudinal Changes in the Pelvic Pain Only and Widespread Pain Phenotypes Over One Year in the MAPP-I Urologic Chronic Pelvic Pain Syndrome (UCPPS) Cohort

Objective: To examine how often urologic chronic pelvic pain syndrome (UCPPS) patients progressed from Pelvic Pain Only at baseline to Widespread Pain, or vice versa, during one-year longitudinal follow-up.

Men and women with UCPPS enrolled in the MAPP-I Epidemiology and Phenotyping Study completed a self-report body map to indicate their locations of pain every 2 months over 12 months. Patients were categorized at each assessment into one of three pain phenotypes: 1) Pelvic Pain Only, 2) an Intermediate group, 3) Widespread Pain. Only patients who completed 3 or more follow-ups were included in this longitudinal analysis. The primary outcome measure was pain classification at the majority (≥60%) of follow-up assessments. Longitudinal trends of somatic symptom burden were also assessed.

Among the 93 UCPPS participants with Pelvic Pain Only at baseline, only 2% (n=2) showed a Widespread Pain phenotype for the majority of assessments over 12 months. Among the 121 participants who had Widespread Pain at baseline, 6% (n=7) demonstrated Pelvic Pain Only for the majority of assessments over 12 months. Over half of participants (≥53%) stayed in their baseline phenotypic group. Somatic symptom burden remained stable over 12 months for each of the groups with high intra-class correlation coefficient (0.67 to 0.82).

It was uncommon for UCPPS patients to progress from Pelvic Pain Only to Widespread Pain, or vice versa, over 12 months. These data suggest that Pelvic Pain Only and Widespread Pain are distinct UCPPS phenotypes that are relatively stable over 12 months of follow up.

Interstitial cystitis, chronic prostatitis, chronic pelvic pain syndrome, phenotyping

Patients with urologic chronic pelvic pain syndrome (UCPPS, which includes interstitial cystitis [IC/BPS] and chronic prostatitis [CP/CPPS]) are heterogeneous. A variety of potential clinical phenotypes and diverse etiologies have been described. Previous work from the MAPP-I Research Network have identified several UCPPS pain phenotypes based on the distribution of pain across the body at the baseline visit: 1) Pelvic Pain Only (those with pain localized to the pelvic region), 2) an Intermediate group with one to two areas of pain outside of areas associated with urological or pelvic pain (e.g., back pain, shoulder pain, etc), and 3) Widespread Pain (those with greater than 2 locations of pain beyond the pelvic region).1 It has been hypothesized in the literature that UCPPS may progress from a pelvic condition to a systemic (i.e., widespread) condition over time.1–4 There are no longitudinal studies, however, to either support or refute this hypothesis. This is an important question to address. If progression from pelvic pain to systemic pain occurs often, clinicians need to identify the mechanisms behind the progression and mitigate the process since systemic UCPPS is difficult to manage. Conversely, if pain phenotypes rarely progress or change over time, this would indicate that Pelvic Pain Only and Widespread Pain are distinct UCPPS phenotypes that are stable over time. In MAPP-I we followed UCPPS participants longitudinally and collected pain spatial distribution and somatic symptom burden data every 2 months over a 12-month period. In this paper we examined longitudinal changes in pain phenotypes, focusing on how often UCPPS progressed from Pelvic Pain Only at baseline to Widespread Pain during the follow-up period; and conversely, how often UCPPS regressed from Widespread Pain to Pelvic Pain Only over time, as well as fluctuations to other phenotypes. In addition, UCPPS patients commonly report multiple somatic symptoms across multiple organ systems.3, 5 Using the CMSI (Complex Multiple Symptoms Inventory)6 that was administrated every 2 months, we examined longitudinal trends of somatic symptoms and the relationship between widespread pain and somatic symptom trajectories over 12 months of follow up.

Participants. The MAPP-I Epidemiology and Phenotyping (EP) Study enrolled 191 men and 233 women with UCPPS at six clinical sites across the US from 2009 to 2014. Adult men and women who met criteria of IC/BPS or CP/CPPS with pain ≥1 (0–10 scale) and associated urinary symptoms in the previous 6 months were eligible. The study design, inclusion and exclusion criteria, and measures have been described previously.7 The EP Study was a 12-month longitudinal study in which the body map1 and CMSI6 (see below), among other measures, were administered at baseline and every 2 months afterward. Participants provided written informed consent following IRB-approved protocols at each clinical site. Assessment of pain phenotypes. The body map, used to classify participants into the three pain phenotypes, has been previously described and is shown in Supplemental Figure 1.1 Participants were asked to select any sites on the body map to indicate where they experienced pain in the past week. If one or more sites within one of the seven body regions (shown in color) was checked, that region was considered positive for pain. Participants reporting pain in the pelvic region (sites 14, 15, 16) only were considered to have Pelvic Pain Only (PPO). Those reporting pain in 1 to 2 regions in addition to the pelvic region were considered to be in the Intermediate Pain group (IP). Those reporting pain in 3 or more of regions in addition to the pelvis were considered to have Widespread Pain (WP). At baseline and every 2 months for 12 months, participants completed the body map, and were classified into one of 3 pain phenotypes (PPO, IP, WP) based on their pain distribution at that time point. Assessment of somatic symptoms. The Complex Multiple Symptoms Inventory (CMSI)6 is a 41-item symptom checklist (39 items for men) of somatic symptoms including common pain symptoms (e.g., abdominal pain) in addition to non-painful common symptoms (e.g., ringing in the ears). At baseline, participants checked symptoms occurring for at least three months in the past year and every two months checked those present in the past two months. CMSI total scores can range from 0–41 for women and 0–39 for men. Higher CMSI scores represent higher somatic symptom burden. Statistical Analysis: Normality of the continuous data was examined with the Shapiro–Wilk test and quantile–quantile plots. Descriptive statistics of normally distributed continuous variables were presented as mean ± SD. Continuous variables not normally distributed were described by the median and the first (Q25) and third (Q75) quartiles. Frequencies and proportions were calculated for categorical variables In order to best describe an individual’s predominant pain status during follow-ups over the 12 months, our primary outcome was determined as their group classification for the majority of follow-up assessments. “Majority follow-up pain status” was defined as having ≥60% of observed follow-ups categorized into a single pain phenotype category. For example, if a participant fell into the Widespread Pain category at four follow-up visits and the Intermediate Pain group at the other two follow-up visits, the majority follow-up pain status for that participant was assigned as WP. Some participants fluctuated between different pain phenotypes from visit to visit and did not have a “majority follow-up pain status.” They were referred as the “Variable” category in Figure 1. A second analysis examined the extent of fluctuation of pain categories and CMSI scores within individual participants over the course of 12 months (see Supplemental Figure 2 below). Marginal distribution of pain phenotype (PPO, IP, WP) and month of assessment (2, 4, 6, 8, 10, and 12) was created within each of the groups based on the grouping at baseline. For example, for the 93 participants who reported PPO at baseline, a 6×3 table was created to indicate the visit number and the pain phenotype. A formal test of homogeneity was performed with Cochran-Mantel-Haenszel test,8, 9 where a p-value > 0.05 indicated no statistically significant heterogeneity from visit to visit. Intra-class correlation coefficient (ICC) was used to quantify the consistency of CMSI (somatic symptom burden) over time within each pain phenotype. ICCs range between 0 and 1; values closer to 1 indicate higher consistency of CMSI scores over time. We examined two potential moderators of longitudinal trends: (1) Sub-group with more recent onset of UCPPS: those with symptoms <2 years (n=36) versus ≥2 years (n=57).10 (2) Sex differences: Baseline pain group distribution and longitudinal trends were compared between men and women. Only participants (n=341) who provided data at baseline and at 3 or more of the 6 follow-up visits (i.e., ≥50% of follow-up assessments) were included in all analyses. Missing data was not imputed. Statistical analyses were performed using SAS 9.4 software (SAS Institute, USA). Plots were created using SAS 9.4 software and R version 3.6.0 (The R Foundation, https://www.R-project.org/).

The demographics and characteristics of MAPP-I UCPPS participants stratified by their pain phenotypes at baseline are displayed in Supplemental Table 1. Detailed comparison of the 3 baseline groups (PPO, IP, WP) have been described previously.1 Of note, there were no differences in urologic pain intensity, urinary frequency, urgency, ICSI, or ICPI between the 3 baseline groups. There were also no differences in duration of UCPPS symptoms (p=0.28), or medication use between the 3 baseline groups at 0 month (p=0.52) and at 12 month (p=0.084). The WP group reported highest intensity of non-urologic pain at baseline, followed by IP and PPO groups. Longitudinal Trends of Pain Phenotypes. The primary outcome measure was the group classification for the majority of follow-up assessments. Figure 1 shows the frequency of “majority follow-up pain status” during the 12-month follow-up period. Among the 93 UCPPS participants with PPO at baseline, 53% (n=49) remained as PPO; 22% (n=20) were categorized in the IP group for majority of visits. Only 2% (n=2) of participants categorized as PPO at baseline progressed to WP as their “majority follow-up pain status” over 12 months. 24% (n=22) fluctuated between different groups. Among the 127 with IP at baseline, 17% (n=22) had PPO and 13% (n=17) showed WP as their “majority follow-up pain status” over 12 months. Among the 121 with WP at baseline, only 6% (n=7) had PPO as their “majority follow-up pain status” over 12 months. Supplemental Figure 2 shows the fluctuations by number of pain regions at bi-monthly follow-ups (see the supplement for details). To formally assess the homogeneity or stability of the three pain phenotypes over time, we performed a Cochran-Mantel-Haenszel test within each of the three groups. In the group who had PPO at baseline, after controlling for subject specific effects, group membership (PPO, IP, and WP) change from visit to visit was not statistically significant (p = 0.8874). Similarly, for patients with IP at baseline, belonging to each of the three pain phenotype groups did not change from visit to visit (p=0.0861). However, in the group who reported WP at baseline, change in phenotype to PPO, IP, or WP was statistically significant from visit to visit (p=0.0195), indicating decrease in pain distribution in the baseline WP group. As shown in Figure 1, the majority of follow-up status change from the participants who reported WP at baseline was to the IP group (24%, n=29) at follow-up; only 6% (n=7) had PPO as their “majority follow-up pain status” over 12 months. Sub-group with More Recent Onset of UCPPS. There were no statistically significant differences in majority follow-up pain status between participants in the PPO group at baseline with <2 years versus ≥ 2 years of symptoms (results not shown). There were also no differences in duration of UCPPS symptoms between the 3 baseline groups (PPO, IP, WP) (p=0.28, see Supplemental Table 1). Sex Comparison. There were non-significant differences between men and women in their baseline pain group distribution (PPO, IP, WP, p=0.12). Among the 46 women with PPO at baseline, 46.5% (n=20), 39.5% (n=17) and 13.9% (n=6) had PPO, IP and WP status at 12 months, respectively. Among the 47 men with PPO at baseline, 70.2% (n=33), 27.7% (n=13) and 2.1% (n=1) had PPO, IP and WP status at 12 months, respectively. The Breslow-Day test does not suggest evidence of homogeneity of odds ratios across sex for transitioning between PPO and Widespread pain categories from baseline to 48 weeks (p = 0.03). Women were more likely than men to transit between PPO and WP. Longitudinal Trends of Somatic Symptoms. There were significant differences between the CMSI somatic symptom burden of participants in the PPO, IP, and WP groups at baseline (respective means ± SD; 7.3 ± 4.5; 10.1 ± 6.1 and 15.1 ± 8.5, p<0.0001). The Spearman correlation coefficient between the number of non-pelvic body regions at baseline and the baseline CMSI was 0.501 (p<0.0001, see Supplemental Figure 3 for scatter plot). In addition, the Spearman correlation between CMSI scores and number of pain regions across all visits remained high (0.611, p<0.0001). Examination of the 12-month trajectories of CMSI scores for each of the body map-based phenotypes showed generally strong within-individual consistency across time, both in terms of mean values (see Supplemental Table 2) and ICCs across all time points. ICC for PPO=0.68; for IP=0.72; for WP=0.82. Overall, the somatic symptom burden results parallel those for the body map measures in terms of pain phenotype differences and stability of group differences over time.

Recent studies have identified UCPPS pain phenotypes based on the distribution of pain across the body (e.g., Pelvic Pain Only versus Widespread Pain).1 It has been hypothesized in the literature that UCPPS may progress from a pelvic pain condition to a systemic pain condition over time.1–4 However there are no longitudinal studies to either support or refute this hypothesis. Clemens et al attempted to address the question by comparing the self-reported age of onset of bladder symptoms versus age of onset of other non-bladder symptoms (e.g., irritable bowel syndrome, fibromyalgia symptoms).11 Results from that study showed that the onset of bladder symptoms was not consistently earlier or later than the onset of non-bladder symptoms. However, the study by Clemens et al is very limited due to the reliance of retrospective recalls of self-reported age of onset of symptoms which can be very unreliable. The current study is the first prospective longitudinal study to examine this hypothesis. Data from this MAPP-I longitudinal study showed that it was uncommon for UCPPS to progress from Pelvic Pain Only to Widespread Pain (2%) over the 12-month follow-up period (see Figure 1). Very few participants (6%) regressed from Widespread Pain to Pelvic Pain Only over 12 months. Over half of patients (53–55%) stayed in their baseline pain category the majority of time. Approximately 20% of the patients fluctuated between categories without consistent status and thus were classified as having variable pain in Figure 1. In patients who had Pelvic Pain Only at baseline, we found no evidence that group membership changed significantly between visits (negative Cochran-Mantel-Haenszel test). Restricting the analyses to the sub-group with more recent onset of UCPPS (symptom duration <2 years) supported similar conclusions. There were also no differences in duration of UCPPS symptoms between the 3 baseline groups (PPO, IP, WP). Somatic symptom burden results paralleled body map analyses, with regard to higher symptom burden with more widespread pain and stability of these group differences over time. Overall, we found no evidence in our data to support the hypothesis that UCPPS generally progresses from a pelvic pain condition to a systemic pain condition over a 12-month period. Pelvic Pain Only and Widespread Pain are UCPPS phenotypes that are relatively stable over one year of follow up. We did find women were more likely than men to transit between Pelvic Pain Only and Widespread Pain, although the reasons behind this sex difference is unclear and warrants further study. About 20% of individuals moved from PPO or WP to IP and 15–20% of individuals in these groups showed no clear pattern of pain locations over the year. Some fluctuations of pain categories over time are expected, since pain in any body location can wax and wane, and this is very likely to occur in the most common pain locations such as back pain and headache. So movement from PPO to IP could occur with any new onset or flare up of one or more of these common complaints. The clinical significance of the IP category, or migration from PPO to IP, is therefore unclear, and suggests the most significant impact of these results is the lack of transition between PPO and WP – the two most interpretable categories. From a phenotyping perspective, it is important in future studies to identify the underlying mechanisms for the various pain phenotypes. For example, systemic pathophysiology (e.g., central sensitization, systemic neuro-inflammation, and “top-down” mechanisms) may play a more significant role in the Widespread Pain group, whereas local factors (e.g., pelvic floor dysfunction, loco-regional bladder-bowel-pelvic floor cross-talk and cross-sensitization) may be important in the Pelvic Pain Only group. From a therapeutic perspective, the body map can be used to categorize the heterogeneous UCPPS population into distinct clinical phenotypes to inform more rational management decisions. For example, one may consider bladder-centric or pelvic floor-centric treatments in patients presenting with Pelvic Pain Only phenotype. On the other hand, one may consider systemic therapies (e.g., tricyclics, gabapentinoids, SNRI, cognitive behavioral therapy or multi-disciplinary pain management) in those presenting with Widespread Pain or psychosocial difficulties. Using a “one size fits all” algorithm12 to treat IC/BPS or CP/CPPS patients without regards for their distinct clinical phenotypes (e.g., localized vs. systemic pain) and diverse pathogenesis may lead to treatment failures. This may explain, in part, why previous randomized controlled trials failed to demonstrate significant benefits in the entire heterogeneous UCPPS population, burying potential signals in patient sub-group(s) or phenotypes. A personalized approach to treatment based on clinical phenotyping of UCPPS should be a priority of future translational research of UCPPS due to its potential high impact to revolutionize the treatment paradigm. This study has several strengths, including the availability of longitudinal phenotyping data every two months to examine the temporal relationships of the phenotypes, and inclusion of both men and women with IC/BPS and CP/CPPS. Potential weaknesses include the relatively brief period of follow-up (12 months) compared to the often long natural history of these disorders. It is possible that the transition from PPO to WP may take many years and our follow-up duration failed to capture the transition. In the next MAPP-II Study, longer follow-up data over 3 years will be available. Our cohort had long duration of symptoms, which may make them less likely to change. To address this concern, we demonstrated no significant progression even in the subgroup with symptoms <2 years. However, one may argue that even the cutoff of 2 years was too long, and some patients may have already made a rapid transition from PPO to WP within the first few months of symptom onset. Unfortunately to be eligible for this study, participants had to have symptoms ≥6 months, and we did not have enough participants (n=7) with symptoms ≥6 months and <1 year to perform this comparison.

It was uncommon for UCPPS patients to progress from Pelvic Pain Only to Widespread Pain, or vice versa, over 12 months. These data suggest that Pelvic Pain Only and Widespread Pain are distinct UCPPS phenotypes that are relatively stable over 12 months of follow up. Supplementary Material Funding: The MAPP Research Network acknowledges support through NIH grants: U01 DK823 Funding for the MAPP Research Network was obtained under a cooperative agreement from National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health (NIH) [DK082315 (Andriole, G; Lai, H), DK082316 (Landis, J), DK082325 (Buchwald, D), DK082333 (Lucia, M), DK082342 (Klumpp, D; Schaeffer A), DK082344 (Kreder, K), DK082345 (Clauw, D; Clemens, JQ), DK082370 (Mayer, E; Rodriguez L), DK103227 (Moses, M), DK103260 (Anger, J; Freeman, M), DK103271 (Nickel, J)]. Abbreviations: - CMSI Complex Multiple Symptoms Inventory - CP/CPPS chronic prostatitis/chronic pelvic pain syndrome - IC/BPS interstitial cystitis/bladder pain syndrome - ICC Intra-class coefficient - IP intermediate pain - MAPP Multidisciplinary Approach to the Study of Chronic Pelvic Pain - PPO pelvic pain only - UCPPS urologic chronic pelvic pain syndrome - WP widespread pain Appendix: MAPP-I Research Network Study Group Footnotes Clinical Trial: ClinicalTrials.gov Identifier: NCT01098279

- 1.Lai HH, Jemielita T, Sutcliffe S, et al. Characterization of Whole Body Pain in Urological Chronic Pelvic Pain Syndrome at Baseline: A MAPP Research Network Study. J Urol. 2017;198:622–631. [DOI] [PMC free article] [PubMed] [Google Scholar] - 2.Nickel JC, Tripp DA, Pontari M, et al. Interstitial cystitis/painful bladder syndrome and associated medical conditions with an emphasis on irritable bowel syndrome, fibromyalgia and chronic fatigue syndrome. J Urol. 2010;184:1358–1363. [DOI] [PubMed] [Google Scholar] - 3.Lai HH, North CS, Andriole GL, Sayuk GS, Hong BA. Polysymptomatic, polysyndromic presentation of patients with urological chronic pelvic pain syndrome. J Urol. 2012;187:2106–2112. [DOI] [PMC free article] [PubMed] [Google Scholar] - 4.Warren JW, van de Merwe JP, Nickel JC. Interstitial Cystitis/Bladder Pain Syndrome and Nonbladder Syndromes: Facts and Hypotheses. Urology. 2011. [DOI] [PubMed] [Google Scholar] - 5.Lai HH, North CS, Andriole GL, et al. Urological symptoms in a subset of patients with urological chronic pelvic pain syndrome and a polysymptomatic, polysyndromic pattern of presentation. J Urol. 2014;191:1802–1807. [DOI] [PMC free article] [PubMed] [Google Scholar] - 6.Williams DA, Schilling S. Advances in the assessment of fibromyalgia. Rheum Dis Clin North Am. 2009;35:339–357. [DOI] [PMC free article] [PubMed] [Google Scholar] - 7.Landis JR, Williams DA, Lucia MS, et al. The MAPP research network: design, patient characterization and operations. BMC Urol. 2014;14:58. [DOI] [PMC free article] [PubMed] [Google Scholar] - 8.Cochran WG. Some Methods for Strengthening the Common Chi-Square Tests. Biometrics. 1954;10:417–451. [Google Scholar] - 9.Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22:719–748. [PubMed] [Google Scholar] - 10.Rodriguez LV, Stephens AJ, Clemens JQ, et al. Symptom Duration in Patients With Urologic Chronic Pelvic Pain Syndrome is not Associated With Pain Severity, Nonurologic Syndromes and Mental Health Symptoms: A Multidisciplinary Approach to the Study of Chronic Pelvic Pain Network Study. Urology. 2019;124:14–22. [DOI] [PMC free article] [PubMed] [Google Scholar] - 11.Clemens JQ, Elliott MN, Suttorp M, Berry SH. Temporal ordering of interstitial cystitis/bladder pain syndrome and non-bladder conditions. Urology. 2012;80:1227–1231. [DOI] [PMC free article] [PubMed] [Google Scholar] - 12.Hanno PM, Erickson D, Moldwin R, Faraday MM. Diagnosis and treatment of interstitial cystitis/bladder pain syndrome: AUA guideline amendment. J Urol. 2015;193:1545–1553. [DOI] [PubMed] [Google Scholar] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article.