Effectiveness of Assessing Ureteral Patency using Pre-operative Phenazopyridine

Abstract

Objectives

To determine the efficacy of phenazopyridine when used intraoperatively to assess ureteral patency and to investigate factors that may influence its efficacy.

Methods

This is a retrospective chart review performed at the Olive View-UCLA Medical Center, a Los Angeles County teaching hospital, from January 2014 through July 2016. Patients undergoing cystoscopy at the time of gynecologic surgery were identified via department case logs. All women receiving preoperative oral phenazopyridine were included. If ureteral flow was unable to be visualized with phenazopyridine alone, the medication was deemed ineffective and sodium fluorescein was given intraoperatively. Patients were divided into a phenazopyridine effective or phenazopyridine ineffective group. Patient demographics, renal function, intraoperative fluids and urine output, estimated blood loss, timing and dose of medication administration, and complications were gathered from the chart and compared between groups using Fisher’s exact test, two sample T test, Wilcoxon test, and logistic regression for multivariable analysis. A p-value of <0.05 was determined to be significant.

Results

Preoperative phenazopyridine was effective in 190 of 207 patients (91.8%). It was ineffective in 17 patients who then required intraoperative sodium fluorescein. The group in which phenazopyridine was effective was more likely to have been given a 200mg (vs 100mg) dose (p = 0.02) and had lower intraoperative urine output (median 450mL vs 800mL, p = 0.002).

Conclusions

Preoperative oral phenazopyridine is effective in > 90% of cases to detect during gynecologic surgery. A higher phenazopyridine dose and lower intraoperative urine output were associated with increased efficacy.

Introduction

The rate of urinary tract injury at the time of gynecologic surgery in various studies ranges from 0.6% to 11% depending on the procedure.^1,2,3,4 Given this, intraoperative cystoscopy is often used to assess integrity of the bladder and ureters and has been shown to increase detection rates of lower urinary tract injuries.^1,5 Previously, visualization of ureteral jets was aided by intravenous administration of indigo carmine. However, unavailability of this medication since 2014 has necessitated exploration of alternative methods for assessing ureteral patency at the time of cystoscopy.

One medication that has emerged as potentially useful for cystoscopic assessment of ureteral flow is phenazopyridine (Pyridium – Gemini laboratories, Bridgewater, NJ), which stains the urine a dark orange. Since 2014, our institution has adopted the practice of administering preoperative oral phenazopyridine as we routinely perform intraoperative cystoscopy at the time of most major gynecologic surgeries, including hysterectomies of all routes. However, at the beginning of this study, there was little data regarding the efficacy of phenazopyridine in such clinical scenarios, particularly in the general gynecology literature. This is especially true given poorly defined pharmacokinetic properties with subsequently unpredictable excretion times, as well as unknown utility in patients with compromised renal function.^6,7

The goal of our study was to determine the efficacy of preoperative phenazopyridine in assessing ureteral patency during gynecologic surgery. Secondarily, we sought to identify factors associated with the efficacy of preoperative phenazopyridine use in this clinical setting.

Materials and Methods

After IRB approval was obtained, a retrospective review was performed on all women undergoing cystoscopy performed by surgeons in the Department of Obstetrics and Gynecology at Olive View-UCLA Medical Center, a Los Angeles County teaching hospital in northern Los Angeles, from January 1, 2014 until July 31, 2016. We chose this time period to ensure inclusion of all patients who received preoperative phenazopyridine prompted by the indigo carmine shortage in June of 2014. Consecutive patients were identified via review of departmental surgical case logs identifying all patients who underwent cystoscopy during the study period. Only patients who underwent cystoscopy with the goal to identify ureteral patency as part of a larger gynecological surgery were included in the study. Patients were excluded if they did not receive phenazopyridine or underwent cystoscopy solely for other reasons, such as evaluation of microscopic hematuria, isolated mid-urethral sling placement, or cervical cancer staging. Patient charts, operative reports, and medication administration records were reviewed. Patients were included if they underwent cystoscopy and had received a preoperative dose of phenazopyridine within twenty-four hours of the cystoscopy. Review of the medical administration records in the electronic health record was performed to confirm for each case that the patient actually took the phenazopyridine, the dose and time of drug administration. Phenazopyridine was given as a single 100 mg or 200 mg dose at provider discretion.

Demographic data including age, race, body mass index (BMI), parity, American Society of Anesthesia (ASA) classification score, preoperative diagnosis, surgical procedure, renal function via serum creatinine and glomerular filtration rate were recorded. Nursing notes, operative reports, discharge summaries, and clinic postoperative visit notes were reviewed to determine the time and dose given of the preoperative phenazopyridine, the intraoperative intravenous fluids given, the estimated blood loss, the intraoperative urine output for the entire procedure (as we did not have the exact timing of the cystoscopy), administration and timing of additional agents for evaluation of ureteral flow (i.e. sodium fluorescein, methylene blue), and intraoperative or postoperative complications that occurred up to six weeks postoperatively. There was not official protocol for what to do if phenazopyridine was ineffective, and alternative medications or procedures were selected per provider preference. To verify that sodium fluorescein had not been given, we cross checked our data with a pharmacy audit. Because of the retrospective nature of this study, we were unable to determine the exact time of intraoperative cystoscopy. Instead we used the surgery end time as a surrogate marker as it is our typical practice to perform our cystoscopies at the end of the case. During uterosacral ligament vaginal vault suspensions, we perform cystoscopy after placing the suspension sutures and again after suspending the vaginal vault, the latter of which occurs near the end of the surgery. If a patient underwent more than one cystoscopy during the surgery, the last cystoscopy was the only one included in this study.

Descriptive analysis was used to characterize the patient population. Phenazopyridine effectiveness was derived by comparing two cohorts of patients: a phenazopyridine effective and a phenazopyridine ineffective group. Differences between the groups were tested with Fisher’s exact tests for categorical values and Wilcoxon tests for continuous values. Factors studied in univariate analysis for association with phenazopyridine effectiveness included age, race, gravity, parity, American Society of Anesthesia (ASA) classification score, height, weight, body mass index, serum creatinine, glomerular filtration rate, phenazopyridine dose, preoperative diagnosis, primary surgical procedure, intraoperative intravenous fluids, estimated blood loss, intraoperative urine output, net intraoperative fluid status, time from phenazopyridine dose to end of surgery and complications. Factors found to be significant by univariate analysis at the P ≤ 0.25 significance level were tested using multivariate regression models.

Results

We identified 482 patients who underwent cystoscopy during gynecologic surgery at our institution between January 2014 and July 2016. Of those, 207 patients met inclusion criteria; 275 patients were excluded because they either did not receive phenazopyridine (N= 123) or underwent cystoscopy solely for reasons other than evaluation of ureteral patency following gynecological surgery (N= 152). During the early study period leading up to the national shortage of indigocarmine, there were 42 patients who received indigocarmine, 42 patients who received methylene blue and 13 patients who received no agent to aid detection of detect ureteral patency. Following December of 2014, none of our patient received indigocarmine and phenazopyridine became the predominant agent given for cystoscopy to assess ureteral patency. There were 25 patients who did not receive phenazopyridine after December 2014 (4 methylene blue, 7 sodium fluorescein, 14 no agent). Of the two hundred and six patients received pre-operative oral phenazopyridine (either 100mg or 200mg) on the morning of scheduled surgery. One patient received 200mg of phenazopyridine the night before surgery. Demographic data, indication for surgery and type of surgery is found in Table 1. Surgical indications included pelvic floor disorders, gynecologic malignancy or premalignancy, symptomatic fibroid uterus or abnormal uterine bleeding, and chronic pelvic pain. The surgeries were performed by Urogynecology in 72 cases (35%) and Gynecology or Gynecologic Oncology in 135 cases (65%).

Table 1.

Demographic Information

	Phenazopyridine Effective	Phenazopyridine Ineffective	Total	P valuea
Number	190	17	207
Mean age (range)	51.2 (33 – 77)	54.5 (39 – 66)	51.4 (33 – 77)	0.1388
Race/Ethnicity (n, %)				1.0000
Asian	10 (5)	0 (0)	10 (5)
Black	5 (3)	0 (0)	5 (2)
Caucasian	32 (17)	3 (18)	35 (17)
Hispanic	143 (75)	14 (82)	157 (76)
Parity, median (range)	3 (0 – 12)	3 (1–7)	3 (0 – 12)	0.2019
BMI mean (range)	30.2 (17.2 – 52.9)	30.7 (21.5 – 43.0)	30.1 (17.2 – 52.9)	0.6360
ASA Classification, median (range)	2 (1–4)	2 (2–3)	2 (1–4)	0.2852
Preoperative Dx (%)b				0.2348
AUB	63 (94)	4 (6)	67 (100)
Malignancy	58 (95)	3 (5)	61 (100)
POP	62 (86)	10 (14)	72 (100)
Other	7 (100)	0	7 (100)
Primary Procedure (%)c				0.0682
AH	59 (97)	2 (3)	61 (100)
LH	60 (95)	3 (5)	63 (100)
VH	57 (84)	11 (16)	68 (100)
Pro	13 (93)	1 (7)	14 (100)
Other	1 (100)	0	1 (100)

^aFisher’s exact tests were used for categorical variables. Wilcoxon tests were used for numerical variables.

^bPrimary preoperative diagnosis. AUB = Abnormal uterine bleeding, including symptomatic fibroids, regardless of predominantly bleeding or bulk symptoms. Malignancy includes confirmed gynecologic malignancies as well as premalignant conditions, including complex atypical hyperplasia, risk reducing surgery for BRCA or Lynch Syndrome patients, high grade cervical dysplasia, and adnexal masses (regardless of final pathologic diagnosis). POP = pelvic organ prolapse. Other includes pain, endometriosis, urinary incontinence without prolapse, or Gartner’s duct cyst.

^cAH = Abdominal hysterectomy, including total abdominal hysterectomy, supracervical hysterectomy, modified radical hysterectomy. LH = Laparoscopic hysterectomy, including total laparoscopic hysterectomy and laparoscopic assisted vaginal hysterectomy. VH = Vaginal hysterectomy, may include prolapse or incontinence procedures. Pro = Prolapse repair without hysterectomy, includes uterosacral ligament suspension, anterior repair, posterior repair, sacrocolpopexy, colpocleisis. Other = left salpingo-oophorectomy.

Of 207 patients, 190 needed only preoperative phenazopyridine to confirm ureteral flow, corresponding to 91.8% efficacy [Table 2]. Two urinary tract injuries – one cystotomy and one ureteral ligation during a vaginal hysterectomy – were identified with use of preoperative phenazopyridine alone. The ureteral ligation was confirmed intraoperatively with a retrograde pyelogram.

Table 2.

Efficacy of preoperative phenazopyridine in visualizing ureteral flow

	Number of Patients (%)	Urinary tract injuries identified (%)
Phenazopyridine effective	190 (91.8)	2 (1.1)
Phenazopyridine ineffective	17 (8.2)	0 (0)
Total patients	207 (100)	2 (0.9)

The other 17 patients (8.2%) required intraoperative sodium fluorescein because phenazopyridine was ineffective in identifying flow in either of the ureters [Table 2]. Sodium fluorescein given intraoperatively at the time of cystoscopy was effective in 15 (88%) of these patients. The remaining two patients had other abnormalities (one had “sluggish” ureteral flow, the other had a history of a reimplanted ureter) that required additional interventions (intravenous pyelogram and intraoperative urology consult, respectively) to visualize the ureteral flow. Neither had intraoperative urinary tract injuries.

Univariate analysis of potential factors that might be associated with reduced effectiveness of phenazopyridine were analyzed by comparing the phenazopyridine effective and ineffective groups. A 200mg versus 100mg phenazopyridine dose (p = 0.0224) and lower intraoperative urine output (median 450mL vs 800mL, p = 0.0023) were statistically significant factors associated with phenazopyridine success. Longer time between phenazopyridine administration and surgical end time trended towards an association with phenazopyridine efficacy, but was not statistically significant (0.1419). The other factors were not significant, including phenazopyridine to surgery start time.

To control for potential confounding factors, multivariable analysis was performed including factors that trended towards significance with a p value < 0.25 from univariate analysis [Table 4]. Factors included in the model were patient age, phenazopyridine dose, intraoperative urine output, phenazopyridine to surgery end time, intraoperative urine output, operating room net fluid, phenazopyridine to surgery time. Phenazopyridine doses were considered to select the best logistic regression model for phenazopyridine failure using stepwise selection. Multivariate analysis revealed that a lower dose of phenazopyridine was associated with increased phenazopyridine failure. The odds of phenazopyridine failure decrease by 70% when using a phenazopyridine dose of 200mg versus 100mg. Similarly, increased urine output was statistically significant increased risk of phenazopyridine failure. For every 100 mL increase in intraoperative urine output, the odds of phenazopyridine failure increased by 10%. Finally, there was a trend towards decreased phenazopyridine failure with increased time between the phenazopyridine administration and the end of surgery. For every 10 minutes increase in time from phenazopyridine administration to the surgery end time, the odds of phenazopyridine failure decreased by 4%. In summary, phenazopyridine is highly effective in detecting ureteral flow, but becomes less effective when a lower dose is given or when the urine is more dilute. Phenazopyridine appears to be less effective when given close to the surgery start time. We did not have any patients with phenazopyridine failure when phenazopyridine was given at least 4.5 hour prior to surgery start time.

Table 4.

Multiple logistic regression of potential factors associated with phenazopyridine failure in univariate analysis

Risk Factor	Odds Ratio Estimates	Wald 95% Confidence Limits	P Value
Phenazopyridine dose: 200mg vs 100mga	0.291	0.087 – 0.978	0.0459
Intraoperative urine output increase by 100 mLb	1.095	1.012 – 1.185	0.0235
Phenazopyridine to surgery end time increase by10 minc	0.956	0.900 – 1.016	0.1504

^aThe odds of phenazopyridine failure decrease by 70% when using a phenazopyridine dose of 200mg vs 100mg.

^bOdds ratio of phenazopyridine failure for every 100 mL increase in intraoperative urine output is 1.095 (95% CI 1.012 – 1.185). For 100 mL increase in intraoperative urine output, the odds of phenazopyridine failure increase 10%.

^cOdds ratio for 10 minutes increase in time from when phenazopyridine was given to the surgery end time is 0.956 (95% CI 0.900 – 1.016). For every 10 minutes increase in time from when phenazopyridine was given to the surgery end time, the odds of phenazopyridine failure decreases 4%.

We encountered 2 (0.97%) intraoperative injuries and that both of these injuries were detected with phenazopyridine alone. There were no delayed urinary tract injuries in the study population after a mean follow-up of 125 days. Additionally, there were no differences in adverse events between the two groups.

Comment

The findings of our study support the observation that preoperative oral administration of phenazopyridine is an effective and clinically useful first line strategy to facilitate evaluation of ureteral patency during gynecologic surgery, especially if given as a 200 mg dose at least 4.5 hours prior to surgery.

We identified two factors that increase the efficacy of phenazopyridine. First, using a dose of 200mg rather than 100mg preoperatively significantly improves success of phenazopyridine. Prior studies have used either 100 or 200mg without significant evidence to support either doses.^7,8,9,10 Our results support the routine use of 200mg to improve outcomes. We also found that lower urine output intraoperatively is associated with increased efficacy of phenazopyridine. We presume this is due to the production of more concentrated urine that allows for easier identification of orange colored jets. As urine output is likely the most reliable reflection of the patient’s intravascular volume, which is dynamic strongly influenced by preoperative and intraoperative hydration, ensuring that patients are not overhydrated preoperatively or intraoperatively could improve success when using phenazopyridine. In light of new data supporting goal-directed intraoperative fluid resuscitation in the context of enhanced recovery protocols in gynecologic surgery, the recommended decrease in intraoperative IV fluid administration should also optimize phenazopyridine efficacy.¹¹

Our data also suggest that increased time between phenazopyridine administration and cystoscopy could improve success in identifying ureteral jets. Though statistically insignificant, results trend toward this finding. We did not have any phenazopyridine failures when the phenazopyridine was given at least 4.5 hours prior to surgery start time. A recent prospective randomized controlled trial evaluating the efficacy of different agents to aid in visualization of ureteral jets at time of cystoscopy found phenazopyridine allowed for less visibility and less surgeon satisfaction than other methods. In this study, phenazopyridine was administered “at least 30 minutes” before surgery.¹⁰ Our data suggests that the short interval between medication administration and surgery could account for decreased visibility and satisfaction with phenazopyridine in this study. This is further supported by a recent study that found the median excretion time of phenazopyridine was 70 minutes with a range of 59 to 127 minutes.⁷

A major strength of our study is its large size and generalizability with this being the largest cohort that we are aware of in which efficacy of phenazopyridine for evaluating ureteral patency has been evaluated in gynecologic surgery.^8,9,10. Our study included patients undergoing a variety of gynecologic surgeries, including procedures for benign gynecology, oncologic procedures, and pelvic reconstruction. Furthermore, cystoscopy was done by a wide variety of providers, most commonly by trainees under the guidance of general gynecologists as resident physician participate in all surgical cases at our institution.

The major weakness of this study is its retrospective nature, inconsistent documentation, inability to determine exact time of cystoscopy, and lack of protocol for phenazopyridine failures, which may have led to an overestimation of phenazopyridine ineffectiveness.

Conclusions

This trial shows that phenazopyridine is effective for intraoperative evaluation of ureteral patency and identification of urinary tract injury. Its efficacy can be improved by administration of 200mg rather than 100mg of phenazopyridine preoperatively at least 4.5 hours prior to surgery. Successful visualization of ureteral flow is correlated with lower intraoperative UOP, likely secondary to less dilute urine. Phenazopyridine use does not impede use of additional interventions (sodium fluorescein or intravenous pyelogram) when needed for further ureteric assessment and can be used effectively by all obstetrician-gynecologists, regardless of level of training.

Table 3.

Univariate analysis of potential factors associated with reduced phenazopyridine efficacy (N=207).

Risk Factor		Phenazopyridine Effective	Phenazopyridine Ineffective	All Subjects	p Value
Phenazopyridine Dose (mg) (N; %)a	100 mg	17 (9)	5 (29)	22 (11)	0.0224
	200 mg	173 (91)	12 (71)	185 (89)
Creatinine (mg/dL)b		0.63 (0.31 – 1.15)	0.64 (0.46 – 1.05)	0.63 (0.31 – 1.15)	0.8997
GFR (mL/min)b		101 (51 – 215)	99 (56 – 143)	101 (51 – 215)	0.8459
EBL (mL)b		200 (10 – 1900)	200 (25 – 800)	200 (10 – 1900)	0.6682
Intraoperative IVF (mL)b		2700 (150 – 8547)	2800 (2000 – 5000)	2700 (150 – 8547)	0.6217
Intraoperative UOP (mL)b		450 (20 – 5000)	800 (200 – 1500)	450 (20 – 5000)	0.0023
OR Net Fluid (mL)b,c		1952.5 (−1248 – 6524)	1952 (903 – 3202)	1952 (−1248 – 6524)	0.4737
Surgical Time (min)b		207.5 (62 – 511)	205 (112 – 280)	207 (62 – 511)	0.8874
Phenazopyridine administration to surgery start time (min)b		129.5 (4 – 1158)	118 (47 – 263)	127 (4 – 1158)	0.1762
Phenazopyridine administration to surgery end time (min)b		350.5 (160 – 1349)	322 (213 – 500.6)	347 (160 – 1349)	0.1419

^aFisher’s exact test

^bWilcoxon tests.

GFR = glomerular filtration rate, EBL = estimated blood loss, IVF = intravenous fluids, UOP = urine output, OR = operation room, mg = milligram, dL = deciliter, mL = milliliter, min = minutes

^cOR Net Fluid = Intraoperative IVF - (EBL + UOP)

Data points were missing as follows: Creatinine/GFR 2 patients.

Single sentence summary.

Preoperative oral administration of phenazopyridine is an effective and clinically useful first line strategy to facilitate evaluation of ureteral patency during gynecologic surgery, especially if given as a 200 mg dose at least 4.5 hours prior to surgery.