Urinary nerve growth factor but not prostaglandin E2 increases in patients with interstitial cystitis/bladder pain syndrome and detrusor overactivity


  • Hsin-Tzu Liu,

    1. Department of Urology, Buddhist Tzu Chi General Hospital,
    2. Department of Urology, Tzu Chi University, Hualien, Taiwan, and
    Search for more papers by this author
  • Pradeep Tyagi,

    1. William Beaumont Hospital, Royal Oak, MI, USA
    Search for more papers by this author
  • Michael B. Chancellor,

    1. William Beaumont Hospital, Royal Oak, MI, USA
    Search for more papers by this author
  • Hann-Chorng Kuo

    Corresponding author
    1. Department of Urology, Buddhist Tzu Chi General Hospital,
    2. Department of Urology, Tzu Chi University, Hualien, Taiwan, and
      Hann-Chorng Kuo, Department of Urology, Buddhist Tzu Chi General Hospital, 707, Section 3, Chung Yang Road, Hualien, Taiwan.
      e-mail: hck@tzuchi.com.tw, htl1@ms43.hinet.net
    Search for more papers by this author

Hann-Chorng Kuo, Department of Urology, Buddhist Tzu Chi General Hospital, 707, Section 3, Chung Yang Road, Hualien, Taiwan.
e-mail: hck@tzuchi.com.tw, htl1@ms43.hinet.net


Study Type – Aetiology (case series)
Level of Evidence 4


To compare urinary nerve growth factor (NGF) and prostaglandin E2 (PGE2) levels among patients with detrusor overactivity (DO), increased bladder sensation (ISB), interstitial cystitis/bladder pain syndrome(IC/BPS) and controls.


Urine samples were collected from 40 women with IC/BPS, 54 with overactive bladder (OAB) and 27 normal women as controls, all with a full bladder. Patients with OAB were further classified into subgroups of DO or IBS by urodynamic results. Urinary NGF and PGE2 levels were measured using an enzyme-linked immunosorbent assay. Urinary NGF and PGE2 levels were normalized by urinary creatinine (Cr) levels and compared among all subgroups.


Urinary NGF levels were increased in the 40 women with IC/BPS and 23 with DO but not in 31 with IBS and the 27 controls. Mean (sd) urinary NGF/Cr levels were not significantly different between patients with IC/BPS, at 1.35 (0.36), and DO, at 1.93 (0.77). Urinary NGF/Cr levels were significantly higher in women with IC/BPS than in women with IBS, at 0.25 (0.10) (P = 0.01). Using receiver operating characteristic curves for assessing urinary NGF/Cr levels in patients with IC/BPS and IBS, IC/BPS was diagnosed with a sensitivity and specificity of 75% and 65.5%, respectively, based on a urinary NGF/Cr threshold of 0.015. However, urinary PGE2/Cr levels were not significantly different among all subgroups.


Urinary NGF/Cr levels are elevated in women with IC/BPS or DO, but not in those with IBS. The differential diagnosis of women with IC/BPS from those with frequency-urgency syndrome is possible based on urinary NGF/Cr levels but not urinary PGE2/Cr level.


interstitial cystitis/bladder pain syndrome


overactive bladder


detrusor overactivity


nerve growth factor


creatinine level




increased bladder sensation.


Interstitial cystitis/bladder pain syndrome (IC/BPS) is a chronic sterile inflammatory disease of the bladder of unknown aetiology characterized by irritative voiding symptoms including urgency and pain [1]. Although there are many theories, the cause of this condition remains obscure. The chronic pain symptoms in IC/BPS might be due to an abnormality or activation of the afferent sensory system in the urinary bladder or CNS sensitization [2]. The diagnosis of IC/BPS is made by clinical and cystoscopic hydrodistension and exclusion of other bladder disorders [3].

Overactive bladder (OAB) and IC/BPS are usually diagnosed by subjective symptoms. Patients with OAB present with urgency with or without urgency incontinence, while patients with IC/BPS present with bladder pain and frequency. Although a urodynamic study is frequently used to detect detrusor overactivity (DO) in patients with OAB, ≈45% of patients might not have such a urodynamic finding [4]. Cystoscopic hydrodistension has been the standard procedure to diagnose IC [1], but not all patients with BPS have such a cystoscopic finding. Therefore, the search for a biomarker to predict the presence of DO or IC/BPS has been enthusiastically undertaken by urologists. However, there are no diagnostic tools for an objective classification between patients with IC/BPS or OAB unless thewy are assessed by cystoscopic hydrodistension.

In the urinary tract nerve growth factor (NGF) is produced by bladder smooth muscle and urothelium [5]. Increased levels of NGF have also been reported in the bladder tissue and urine of patients with painful inflammatory conditions of the lower urinary tract, such as sensory urgency, IC/BPS or bladder cancer [6,7]. Patients with OAB, DO and BOO have been found to have increased urinary NGF level [8–10], which can decrease when the OAB symptoms improve [10]. Upper urinary tract diseases such as ureteric stone, urothelial cancer and bacterial infection that cause inflammation of the urinary tract can also be associated with an elevated urinary NGF level.

In the urinary bladder, prostaglandin E2 (PGE2) is a cytoprotective eicosanoid which inhibits apoptosis of epithelial cells [11]. Intravesical instillation with PGE2 induces detrusor contraction, while topical application of PGE2 to the urethra causes urethral relaxation in rats [12]. An increased PGE2 level in the bladder is likely to be associated with long-standing storage dysfunction [13]. Urinary PGE2 levels were elevated in patients with UTI, and successful treatment of UTI lowered urinary PGE2 levels [14]. In patients with OAB, the urinary PGE2 level has been found to be significantly increased and negatively correlated with the maximum cystometric capacity [15].

As OAB symptoms overlap with symptoms of IC/BPS, a more detailed analysis of urinary biomarkers is needed to establish their role as a noninvasive diagnostic tool for the differential diagnosis of IC/BPS and OAB. The present study was designed to compare the urinary NGF and PGE2 levels in patients with IC/BPS and OAB. The difference in urinary levels of these proteins might provide a diagnostic basis for IC/BPS.


Forty women with IC/BPS, 54 women with OAB and 27 women as normal controls were enrolled in this study. This study was approved by the institution review board of the hospital. Informed consent was obtained from all participants before collecting urine samples for measurement of NGF and PGE2 before any treatment.

Patients with IC/BPS had characteristic symptoms (suprapubic pain, severe frequency and urgency) and cystoscopic findings (glomerulation, petechiae, or mucosal fissure) after hydrodistension under anaesthesia [1]. All patients were investigated thoroughly and were excluded if they did not meet the criteria of the National Institute of Diabetes and Digestive and Kidney Diseases [3]. Women with urgency-frequency with or without urgency incontinence confirmed by a 3-day voiding diary were enrolled as having OAB. Control subjects included those who were free of LUTS or any urogenital disease. Patients with previous bladder or urethral surgery, active UTI, neurogenic lesion, or a postvoid residual urine volume of >50 mL were excluded.

All patients with OAB had a video-urodynamic study; a final diagnosis of normal bladder, increased bladder sensation (IBS) or DO was made based on the video-urodynamic study results. Patients with a normal bladder and evidence of BOO or detrusor underactivity were excluded from the study. IBS was defined as the patient feeling an early and persistent strong desire to void [16], at a bladder volume of <350 mL, with no DO elicited during the urodynamic study. The criteria of IBS was based on the findings of normal subjects who had a bladder volume at ‘strong desire to void’ of ≈500 mL and ‘first desire to void’ at ≈70% of the maximum bladder capacity [17]. If patients had a ‘strong desire to void’ at a volume less than the bladder volume of ‘first desire to void’ in normal subjects, they were classified as having IBS.

The urine samples were collected when the bladder was ‘extremely full’ and participants had a strong desire to void. Voided urine was placed on ice immediately and transferred to the laboratory for preparation for NGF measurement. The urine samples were centrifuged at 3000g for 10 min at 4°C. The supernatant was separated into aliquots in 1.5 mL tubes and preserved in a freezer at −80°C. At the same time, 3 mL of urine was taken to measure the urinary creatinine (Cr) level.

Urinary NGF concentration was determined using an immunoassay system (Emax®, Promega, Madison, WI, USA) with a specific and highly sensitive ELISA kit, which had a minimum sensitivity of 7.8 pg/mL. Assays were conducted according to the manufacturer’s instructions. The detailed procedure was described previously [8]. Generally, urine samples were not diluted in the ELISA assay. When the urinary NGF concentration was higher than the upper detection limit (250 pg/mL) the urine samples were diluted to fit the detection limit. For urine samples with NGF concentrations lower than the detectable limit but above zero, they were concentrated using a column-protein concentrate kit (Amicon Ultra-15, Millipore, USA) before measuring the NGF value. All samples were run in triplicate, and urinary NGF levels without a consistent value in three measures were repeated and the values were averaged. The criterion for defining consistent values was that the coefficient of variation (sd/mean) of the three absorbance values was <0.10. If the coefficient of variation of the samples was >0.10, data of the first run were discarded and the samples were re-run in a triple repeat. The total urinary NGF levels (pg/mL) were further normalized by the concentration of urinary creatinine (mg/dL), and the ratio of NGF/Cr was used as a normalized urinary NGF level.

Urinary PGE2 was measured using a high-sensitivity ELISA kit (R and D Systems, Minneapolis, MN, USA), according to manufacturer’s instructions. Briefly, calibrator diluent buffer was added to antibody pre-coated plates at the zero standard (B0, 100 µL) and non-specific binding (150 µL) wells. Then 100 µL of high-sensitivity PGE2 standards (19.6–1250 pg/mL) and three-times diluted urine samples were pipetted into the remaining wells. Standards and all samples were assayed as a triple repeat. The concentration of PGE2 was determined by measuring the absorbance with a microplate reader (SynergyTM, HT Bio-Tek Instruments, Vermont, USA) at 450 nm and 540 nm. The total urinary PGE2 level (pg/mL) was normalized by the urinary Cr and the ratio of PGE2/Cr used as a normalized urinary PGE2 level.

Urinary NGF/Cr and PGE2/Cr levels were compared among control, patients with IC/BPS, IBS and DO subgroups using one-way anova test. The Mann–Whitney U-test was used for statistical analysis between two subgroups, without assuming normality and univariate relationship. The correlation between biomarkers was calculated using Pearson’s correlation coefficient; in all tests P < 0.05 was considered to indicate statistical significance.


The participants comprised 40 women with IC/BPS, 29 with OAB-dry, 25 with OAB-wet and 27 controls. After the video-urodynamic study, IBS was diagnosed in 31 women and DO in 23. The mean (sd, range) age of the women was 49.5 (14.2, 27–79), 57.3 (13.6, 32–87), 64.1 (11.1, 42–81) and 57.6 (13.3m 18–79) years, respectively (P > 0.05).

The urinary NGF and PGE2 levels in all subgroups are shown in Table 1. Among IC/BPS and the urodynamic subgroups, the urinary NGF/Cr level of patients with IC/BPS was not significantly different from that of DO (P = 0.44), but was significantly higher than IBS (P = 0.01) or the control group (P = 0.01). The mean urinary NGF/Cr levels were greater in patients with IBS than in controls, but this difference was not significant. Urinary PGE2/Cr levels were not significantly different among any of the subgroups.

Table 1.  Urinary NGF and PGE2 levels in the control and patients of urodynamic subgroups and IC/BPS
No. of women27312340
Cr, mg/dL19.02 (2.10)19.14 (2.24)28.97 (3.83)57.78 (6.88)
P   <0.001 (vs all)
NGF/Cr0.09 (0.04)0.25 (0.10)1.93 (0.77)1.35 (0.36)
P 0.14 (vs control)0.01 (vs control)0.01 (vs control)
   0.02 (vs IBS)0.01 (vs IBS)
    0.44 (vs DO)
PGE2/Cr9.95 (1.03) 11.55 (1.51) 11.51 (1.60)10.27 (2.78)
P 0.56 (vs control)0.59 (vs control)0.93 (vs control)
   0.98 (vs IBS)0.72 (vs IBS)
    0.76 (vs DO)

Figure 1 shows the distribution of urinary NGF/Cr and PGE2/Cr levels in controls and patients with IBS, DO and IC/BPS. There was no significant correlation between urinary NGF/Cr and PGE2/Cr levels in any of the subgroups. The urinary PGE2/Cr levels were also not significantly different between any of the subgroups. Using a receiver operating characteristic curve for assessing urinary NGF/Cr levels in patients with IC/BPS and IBS, IC/BPS could be diagnosed with a sensitivity and specificity of 75% and 65.5%, respectively, based on a urinary NGF/Cr threshold of 0.015.

Figure 1.

Scatter plots of (A) urinary NGF/Cr and (B) PGE2/Cr levels in controls, and patients with DO, IBS and IC/BPS.


The results of this study showed that women with either IC/BPS or DO had an higher urinary NGF/Cr level, significantly greater than that of women with IBS and the controls. Although PGE2 has been suggested to act as a biomarker for the diagnosis of OAB, we did not find a significant difference among IC/BPS, DO, IBS and the controls.

The clinical symptoms of IC/BPS and OAB are similar except that bladder pain typically presents in IC/BPS, and urgency and urgency incontinence present in OAB-dry and OAB-wet, respectively. However, patients with IC/BPS usually cannot tolerate a full bladder before the maximum bladder capacity is reached, and therefore patients with IC/BPS in an early phase might not characteristically present with bladder pain but merely with frequency and urgency, which are similar to the symptoms of OAB-dry.

IC/BPS is a difficult disease to diagnose and treat. Objective information about IC/BPS is only available in patients with end-organ disease through invasive procedures, such as cystoscopic hydrodistension and tissue biopsy [3]. The development of biomarkers for the diagnosis and prediction of IC/BPS progression is therefore a high priority. Previous histological studies of tissue biopsy from IC/BPS patients documented infiltration of mast cells, macrophages and eosinophils as consistent findings, and the significant associations for T and B cell staining were similar to those for overall inflammation [18]. Therefore, an increased NGF level either in the bladder or urine indicates the presence of chronic inflammation in patients with IC/BPS.

Recent investigations have suggested that OAB might be a condition reflecting bladder inflammation. Histological investigation of bladder urothelium and suburothelium found that signs of chronic inflammation were present in 60% of baseline biopsies of patients with OAB [19]. OAB could be a subtype of neurogenic inflammation characterized by a series of vascular and nonvascular inflammatory responses, triggered by the activation of primary sensory neurones and the subsequent release of inflammatory neuropeptides, including substance P and calcitonin gene-related peptide [20]. The inflammatory response induced overexpression of TRPV1 receptors in the suburothelium as well as c-fos protein in the dorsal root ganglia in rat models of OAB and in human bladder biopsies [21]. Urinary inflammatory biomarkers might also be elevated in DO as well as in patients with IC/BPS.

There could be an overlap among DO, IBS and IC/BPS. Patients with OAB-wet might not have urodynamic DO and patients with IBS might be in the early phase of IC/BPS. The OAB symptoms might change from dry to wet, or from wet to dry with time, suggesting that OAB is a dynamic condition [22]. This fluctuating condition might be considered as a chronic inflammation of the bladder. As the urinary NGF level was significantly elevated in patients with DO but not in IBS, there could be different inflammatory degrees present in these two OAB conditions. It is not known whether patients with IBS can progress to DO, or progress from IBS to IC/BPS with time; a longitudinal study over time could answer this question.

Although urinary NGF levels were not significantly different between patients with IC/BPS and DO in this study, patients with IC/BPS had a significantly greater urinary NGF level than patients with IBS. The symptoms of IC/BPS and DO are distinct; patients with IC/BPS have bladder pain and those with DO usually have urgency incontinence. Thus, there is no need for a biomarker to make the differential diagnosis between IC/BPS and DO. By contrast, the differential diagnosis between IC/BPS and IBS based on symptoms is usually difficult. Comparison of the levels of urinary NGF in patients with IC/BPS and IBS in this study suggests its potential for use as a biomarker for noninvasive differential diagnosis. The diagnosis of IC/BPS by cystoscopic hydrodistension under general anaesthesia could thus be avoided in most patients with urgency/frequency syndrome.

Prostanoids are synthesized by cyclooxygenase in the bladder. The synthesis of prostanoids is initiated by various physiological stimuli and injuries of the urothelium, nerve stimulation, and by ATP and mediators of inflammation. Intravesical instillation with PGE2 facilitated micturition and increased basal intravesical pressure by releasing tachykinins, and contributed to urge and bladder hyperactivity of lower urinary tract [23]. In the present study, there was no significant difference in urinary PGE2/Cr levels among patients with IC/BPS, OAB and controls, suggesting that PGE2 production in the bladder wall might not be the main result of DO or IC/BPS. Instead, prostaglandin receptor EP3 expression or susceptibility to PGE2 might be more important than PGE2 production in response to bladder inflammation, as in IC/BPS or DO. Therefore, measuring urinary PGE2 levels provides no further advantage in the differential diagnosis between IBS and DO, or between IC/BPS and frequency/urgency syndrome.

The present urine samples were collected when participants had a strong desire to void. Our previous investigation showed that the urinary NGF level increases physiologically in response to bladder stretching [24]. However, because most patients with IC/BPS are used to drinking little water in their daily life, their urinary Cr concentrations were significantly higher than those of controls and patients with IBS or DO, as were their urinary NGF and PGE2 levels. Therefore, it is important to correct NGF and PGE2 levels by the constant Cr concentration in urine to prevent analytical bias due to this factor in patients with IC/BPS.

Although urinary NGF/Cr levels in IC/BPS and DO are similar, the underlying inflammatory protein expression responsible for bladder pain in IC/BPS and urgency incontinence in DO need further clarification. Analysis of multiple urinary proteins is a convenient approach to monitor the activation of inflammatory cells in bladder tissue [25]. Although it might be ideal to discover a single protein specific for IC/BPS compared to OAB, at the same time it is unrealistic to expect the complex pathology of IC/BPS to be defined by the urinary level of one protein. It is expected that an interplay of various factors from a cluster of inflammation-associated proteins might provide a unique signature for IC/BPS. If we can find a cluster of urinary proteins which is greater in IC/BPS than DO, a detailed diagnosis by proteomic study using a urine sample might be feasible.

In conclusion, urinary NGF/Cr levels are elevated in patients with IC/BPS or DO, but not in those with IBS. The results of urinary NGF measurement in this study suggest its potential value as a biomarker in the differential diagnosis of IC/BPS and IBS. By contrast, urinary PGE2/Cr levels showed no significant difference among patients with DO, IBS, IC/BPS and controls, and might not be a useful biomarker in the diagnosis of bladder dysfunction.


None declared.