UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 1 By the end of this presentation I hope I was able to summarize the American Academy of Pediatrics UTI Practice Guidelines. I'll tell you a little bit about studies you've heard before, very shortly on the prevalence, diagnosis and treatment of UTI particularly because they were incorporated, adopted into the American Academy of Pediatrics UTI Guideline. We'll discuss follow-up diagnostic imaging for UTI particularly ultrasound, VCUG and DMSA. Because Steve came to Grand Rounds I'm going to talk about bladder and bowel dysfunction and the management of basically ureter reflux in light of recent publications that evaluated the role of antimicrobials. The NIDDK RIVUR Study stands for Randomized Intervention for Vesicoureteral Reflux, the NIDDK CUTIE Study that stands for Careful UTI Evaluation and the Thrasher Foundation ARKS Study or Antibiotic Resistance in Kids Study and then tell you a little bit about what we are working on in the area of urinary tract infections.

The guideline in one slide. The recommendation is to obtain both urinalysis and urine culture on catheterized human specimens or suprapubic aspiration specimens. The guideline applies only to children 1 to 24 months with a febrile urinary tract infection. If girls have two risk factors, if boys have three risk factors or boys are uncircumcised and the risk factors were the ones we described in the mid-'90s with a group at the Children's Hospital of Philadelphia, Kathy Shaw and Mark Gorelick, and are white race, age less than 12 months, temperature at least 39 degrees Centigrade, fever for at least 48 hours in girls and at least 24 hours in boys, and the absence of another source of infection. So more than two or more than three.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 2 Define UTI ss the presence of both pyuria and bacteruria, so this was new in the guideline and bacteruria was defined as 50,000 CFU/mL. Bacteruria without pyuria likely represents asymptomatic bacteruria. It recommended treating orally with 7 to 14 days of antibiotics unless the child is toxic, unable to retain oral treatment. Recommended obtaining an ultrasound after the first UTI and if abnormal hydronephrosis, scarring obstruction obtain a voiding cystourethrogram. Obtain a voiding cystourethrogram after a second urinary tract infection and obtain urine specimens with subsequent febrile illnesses.

What do providers do? Not here but in many other places children are treated for urinary tract infections without urinalysis and without urine culture. Hillary Copp did this study, published in 2013 in Pediatrics, used a very large database, claims database between 2002 and 2007, 40,000 episodes of outpatient urinary tract infection in children in about 28,000 children and what she found is that 57% of instances a culture was performed, but you know 40 some percent of time it wasn't done. Urinalysis was not done and neither urinalysis nor urine culture was done in about 20% of instances. Now if you focus in the children under age 2 years the converse of this, 68%, so 32% of children under 2 were treated without a urinalysis and without a urine culture, and that results in unnecessary referrals and unnecessary imaging for children.

Our first question was what was the prevalence of urinary tract infection in febrile infants presenting to the emergency dept. with and without an apparent source of fever and we found 5% overall prevalence of UTI higher in Caucasian girls who had a temperature greater than 39 degrees

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 3 Centigrade. We wanted to assess the validity of standardly used urinalysis methods and criteria and we used the standard UA using centrifuge, unstained urine, we compared that with an enhanced UA and showed that the enhanced UA was more sensitive, more specific and had high positive predictive value using uncentrifuged urine and gram stain smears. Then we developed that decision rule with Philadelphia looking at age, race, height of temperature, duration of fever and the absence of resource. We reported then the frequency distribution of bacterial colony counts in catheterized urine specimens. We used pyuria according to bacterial colony counts. We described the accuracy of the dip stick urinalysis and we used the pyuria and the colony counts to distinguish UTI from asymptomatic bacteruria.

So the American Academy published the Table 1 that shows all the studies put together, the sensitivities and specificities of the various components of the urinalysis, the leucocyte esterase, the nitrates, microscopy and so forth and the combination of all of them. In our hands the standard urinalysis 54% sensitivity, the bacteria was more sensitive but lacked a positive predictability so pretty much everybody on the standard analysis will have something that looks like bacteria. The enhanced urinalysis was more sensitive and had a much higher positive predictability and the leucocyte esterase was not very - was helpful but not extremely helpful, not as high as the AAP recommended, and the nitrites don't work in catheterized urine specimens because the urine needs to be in the bladder for many, many hours for the conversion from nitrates to nitrites to occur. The leucocyte esterase however when present indicates that white cells are there.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 4 On several thousand urine specimens of children presenting to the emergency dept. with fever under age 2 were reported frequency distributions of bacterial colony counts. On the horizontal axis you have colony counts, on the vertical axis you have children. The green bar indicates the presence of no growth, the gray bar indicates the presence of single pathogens and the blue, light blue bars indicate the presence of known pathogens or mixed organisms. As you can see of the 3200 specimens most children had no growth, 2900. And if you draw an imaginary line at 50,000 CFU/mL and you look below that you tend to see more non-pathogens or mixed organisms. And as you migrate over 50,000 you see mostly single pathogens.

Now looking at the same thing but now using white blood cells counts in the urine/cubic mm gray bars now indicate less than 10 white cells, light blue indicates at least 10 white cells per cubic mm. Most children with no growth had less than 10 white cells per cubic mm but pyuria can be a nonspecific finding too, so 2%, 3% of kids will have pyuria and they have many other diagnoses and the culture is negative. So again your imaginary line at 50,000 shows that below that you see less than 10 white cells, once you jump above 50,000 you tend to see at least 10 while tells per cubic mm. So putting the two things together Ellen and I recommended at that time 50,000 and the presence of 10 white cells per cubic mm is the way to go on a catheterized urine specimen and that was adopted into the American Academy of Pediatrics Guidelines.

We went on and did the Oral Versus IV Study. The goals of therapy are to eliminate infection, to prevent complications and reduce the likelihood of renal scarring. We compared oral 14 days of

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 5 Cefixime versus 3 days of Cefotaxime, we did DMSA scans, renal ultrasounds and VCUGs. We called them monthly, we saw them whenever they were febrile and we looked at short term and long term outcomes.

And short term outcomes and long term outcomes were similar. Urine was sterile within 24 hours in both treatment groups, no need to repeat a test of urine culture. Children defervesced within 24 hours, 24, 25 hours, there was no difference in rate of infections between the two treatment groups and no difference in renal scarring, and not even among those in the final row there that had acute pyelonephritis on DMSA at study entry. We were able to tease out any difference between oral and intravenous, about 15% of kids had renal scarring on follow-up, even among those who were the only ones at risk for renal scarring. And the extent of the renal parenchyma when they scarred was relatively small, it was 8% of the renal parenchyma.

The American Academy went on on Table 2 and showed all the antibiotics that could be used and the dosing for antibiotics for parenteral treatment. On Table 3 the antibiotics used for oral therapy for urinary tract infection. There is substantial geographic viability and susceptibility to the various antibiotics. To share with you some Pittsburgh data, Trimethoprim-sulfamethoxazole, it's about 20%. Our population if you look at the bacteria report he was a little bit biased by a lot of the transplant, a lot of the IV, a lot of the patients that are frequent flyers in the hospital so it's a little bit higher than that, but that's about what you see for pediatric outpatient urinary tract infections.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 6 First generation cephalosporins, Cefixime is an example, it's 15% resistance, Amoxicillin, Ampicillin 50% resistance, and third generation cephalosporins have been consistently 3, 4, 5% resistance over time, there has been no growth in that resistance.

The Academy went on to say that Nitrofurantoin is not appropriate for febrile infants, it doesn't achieve therapeutic concentrations in blood and parenchyma and stated that 1, 2, 3 day courses for febrile UTI had clearly inferior results and should not be used and that no data were available to use 7, 10 or 14 day courses, there have been no good studies looking at duration of therapy for children with UTI, and I'll go back to that a little bit later.

Going on to imaging, prenatal ultrasonography had reduced the prevalence of unsuspected obstructive uropathy, viability and timing and quality of prenatal ultrasound occurs. It does not rule out structural abnormality unless it's a detailed anatomic survey with measurements performed during the third trimester and interpreted by qualified individuals. It can be used during the first 2 days of therapy to identify serious complications, renal or perirenal abscesses, pyelonephrosis with obstructive uropathy, unusually severe illness or no substantial clinical improvement, but changes in size, shape and echogenicity could be attributable to edema, frequently occurring in early infection. Lack of exposure to radiation is a major advantage, recognizing its limitations.

In our hands with about 300 children we enrolled in the study 90% of children had a normal ultrasound following the urinary tract infection and all the findings shown in this slide, dilated

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 7 pelvis, pelvocaliectasis, hydronephrosis and so forth, they did not modify management in any child.

Here we see lined up all the studies that have looked at what the benefit of renal ultrasound is with regards to changing management. Some of them are retrospective, our study is second there on the list, prospective in the various settings, in emergency dept., inpatient, urology, outpatient and so forth and you can see that the proportion of children with abnormal ultrasounds is around 10 to 15%. And in most instances it doesn't modify management, the few studies that didn't determine whether it modified management or not they compared it with okay was it good at predicting who had dilated reflux, grade 3 to 5 reflux and the answer is no. 80 out of 119, 17 out of 27, so it doesn't pickup all the kids with high grade reflux.

Reflux is unusual in normal children, it's around 1%. This study by Ransley 1.3% in 1978, in children with prenatal hydronephrosis it's around10%. It tends to be of high grade, tends to be in males and they have higher voiding pressures. VUR in children evaluated for UTI occurs in about 25 to 40%, more commonly Caucasian girls 30%, lower in African American girls, 30%-50% of reflux in children with siblings who have reflux, and 65% in affected offspring. So there is a genetic predisposition.

How do we classify reflux? Grades I, II, III, IV, V, distal ureter, proximal ureter, mild, moderate and severe dilatation of the urinary tract, and the two figures in the bottom from the American

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 8 Urologic Association Guidelines show that the lower the reflux, the lower the grade of reflux the faster it resolves, unilateral resolves faster than bilateral and the younger the children the faster it resolves than they older the children at the time of diagnosis. So lower grade, unilateral and younger is good prognosis for faster resolution of Vesicoureteral Reflux.

How did it work in our hands? If 60% of children had a normal BCG, 40% had reflux following UTI once they have reflux they tend to be of grades I, II and III, those are the bulk of the numbers and only 5 children out of the 300 had grade 4 VUR and nobody had grade 5 VUR. The figure on the right shows the proportion of renal scarring according to selected demographic and clinical characteristics. Age less than one, one to two it didn’t make any difference in the likelihood of renal scarring. Duration of fever before initiating therapy there was a trend but it was not significant, only the presence of reflux versus no reflux was significantly associated with a higher likelihood of renal scarring on follow up.

The figure on the left shows the frequency distribution of the various degrees of no reflux on grades I, II, III, IV, according to the presence of normal ultrasound or an ultrasound showing dilatation of the urinary tract. In dark blue you see the kids that had even without reflux some of them had abnormal ultrasounds, some with grade 2, and not everybody with grade 4. Only three of the five with grade 4 had an abnormal ultrasound. On the right, the same thing in a slightly different way, now we’re looking at everybody that had a normal ultrasound and everybody that had an ultrasound showing dilatation of the urinary tract according to no reflux, low grade and high grade reflux. And

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 9 overall the prevalence of reflux in the ones with normal ultrasounds an abnormal ultrasound was the same, it was identical. The ones with an abnormal ultrasound were more likely to have high grade reflux, more grade 3-5 reflux.

The DMSA scan on the left shows a normal homogenous uptake throughout the kidney with well preserved contours, the one in the middle shows decreased uptake in the upper part of the left kidney will conserved contours, and the one on the right shows a V shaped defect that has decreased uptake with loss of contours indicating renal scarring.

DMSA scan in our study showed that about 2/3 of children have findings of acute pyelonephritis when imaged shortly after the infection and if you assume that the test has about probably 90% sensitivity that makes about ¾ of children with a febrile urinary tract infection at under age 2 will have some indication of acute pyelonephritis on DMSA renal scan if we were to do it. So the consensual model is that children have urinary tract infections, some children have vesicoureteral reflux and VUR makes children more likely to have urinary tract infections and some of them end up having renal scarring that may result in endstage renal disease, hypertension and preeclampsia, bad things.

So the interventions have been to use prophylactic antibiotics to prevent recurring urinary tract infections or to correct the reflux to reduce their likelihood of urinary tract infection, and we’ll get back to this. The problem with the model is that scarring is less frequent. We observe it in children

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 10 who do not have reflux. There’s a paucity of evidence that interventions for vesicoureteral reflux are effective. And antibiotic pressure we thought to increase antimicrobial resistance. Repeated VCUGs and VUR treatment have potential harms and definite costs and parents perceive that antibiotics may not be always necessary. And the widespread use of prenatal ultrasound has resulted in early diagnosis and treatment, surgical or medial, of children with obstruction of the urinary tract.

So Nader Shaikh in 2010 published these very nice metaanalysis in pediatrics and put all the studies together that looked at renal scarring and DMSA can kind of lined them up in year of publication. As you can see this one was done in the early nineties, the likelihood of renal scarring was very high. As the years go by, this is our study that has 15% pretty much among the ones with pyelo and lower amongst everybody, and the Garin study was even lower. The same trend was noted when you lined them by year there and then the bubbles indicate the size of the various studies. So the decrease in renal scarring may be due to increased awareness and earlier diagnosis and treatment of UTI and VUR may not be necessary for scarring to occur, or may not be the most important risk factor in children with urinary tract infection.

So going back to the conceptual model, some children have urinary tract infections and sometimes we’re not pretty good at identifying them quickly and treating them so there was delayed treatment, delayed diagnosis and delayed treatment. Some children are born with renal dysplasia and that may result in renal scarring and the bad things that we talked about; end stage renal disease, hypertension and preeclampsia. So the interventions would be for prompt diagnosis and treatment of urinary tract

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 11 infection and the question about whether prophylactic antibiotics prevent recurring urinary tract infection and the role of VUR remained to be determined.

Nader published these metaanalysis too in looking at renal scarring following first urinary tract infection. As you can see in our studies hidden somewhere here, as you can see the kids that didn’t have VUR versus the kids that had VUR the risk ratio with VUR was 1.45 for having acute pyelonephritis. It increases the likelihood of children having acute pyelonephritis. Similarly, no reflux and reflux increases the likelihood by risk ratio of 2.62 of children having renal scarring. And now if we look at children with low grade versus high grade reflux dilating reflux, again, risk ratio 2.1, increases the likelihood the higher the grade of reflux the higher the likelihood of renal scarring.

Now Nader again Tim Shope and others came up with this very interesting individual patient metaanalysis just published a few months ago in JAMA Pediatrics. What Nader did on these studies he defined 3 models trying to go after a little bit about what the guidelines were trying to address. The first model used kind of a clinical approach. If we use a temperature grade of 39 degrees centigrade where the child had an organism different than E-Coli and whether they had an abnormal ultrasound. And then the second model included acute phase reactants polymorphonuclear counts and CRP levels indicating inflammation. And the third model included okay we did the VCUG and now we know that the kid has grade 3 or grade 4 or grade 5 reflux. And how do these 3 models compare? Of course the third one was the best model but the first one was not far away. It was very close in the area under the ROC curve so the first one is probably the easier one to use.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 12

And then they assign points based on the odds ratio and as you can see having an abnormal ultrasound gets 2 points and having grade 4-5 reflux gets 6 points for that. And then if you look and see what happens, if you have 1 point which would be like you had temperature with a UTI greater than 39 degrees you have about baseline risk, 15% of renal scarring as indicated here in the slide. But as soon as you have an abnormal ultrasound which gives you 2 points you move onto twice the risk of renal scarring. So this indicates that you know using ultrasound in this manner might be of help.

I told you Steve came to Ground Round so I’m going to have to talk about bladder and bowel dysfunction. This used to be called voiding dysfunction. A few years later it would move onto be called dysfunction elimination syndrome and now it’s called bladder and bowel dysfunction. I think of these are Dr. Perman would always refer to unsung heroes this is like the unsung hero of urinary tract infections. If you think of something you’ve got to ask about this. This is uninhibited bladder contraction, contractions of involuntary sphincter increases the intravesical pressures with incomplete bladder emptying. Reflux then increases the likelihood of acute pyelonephritis. Acute is presented with a spin top urethra, reflux, and we can easily use the Canadian questionnaire by Farhat asking all the various questions that ask questions about constipation and about dysfunctional elimination and a score of at least 6 in girls and at least 9 in boys is positive for dysfunction of bladder and bowel dysfunction. And Nader modified these questionnaire when he did the study with Steve in the mid nineties, the Canadians I don’t understand how they understand this question, “I

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 13 miss having a bowel movement every day.” No child in the U.S. will understand that question. So it means that they are not having a bowel movement everyday.

So the study that Steve and Nader did in the mid-nineties, published in 2003 showed with the population of children we had in the UTI study, in the oral versus IV study, we surveyed them when they were school aged girls, many years later and we matched them with controls that were the next patient seen in the emergency department that did not have a UTI, had a fever but did not have a UTI. And as you can see, and they all were given the modification, the Pittsburgh modified Toronto instrument and you know it doesn’t matter how you look at it, 1 in 5 girls have bladder and bowel dysfunction whether they were in the comparison cohort, the UTI cohort, whether they had reflux or whether they didn’t have reflux, it didn’t make a difference. If you ask about this, you’re going to find it, one in every five girls with UTI. So BBD is not more common in children with prior UTIs and BBD, Bladder and Bowel Dysfunction and VUR do not segregate together.

Okay, I want to take a deep breath and talk about this one. Action statement 6,7 this is the one that gets the press. No VCUG with first urinary tract infection. The rationale for this one was that infants without VUR develop acute pyelonephritis and the prophylaxis effectiveness has been challenged by some studies. If prophylaxis is not beneficial and that was an assumption in the American Academy of Pediatric Guidelines, that prophylaxis was not beneficial, VUR not required for acute pyelonephritis, what’s the rationale for doing the VCUG. So they went on and did an individual patient metaanalysis and asked for data from some authors that provided data in children 1-24

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 14 months of 6 recent randomized controlled trials. And the findings were not significant. If you look at the prophylaxis group and the no prophylaxis group, in no reflux grades I, II, III, IV the findings were not significant. They highlighted that prompt diagnosis and treatment of febrile UTI was important regardless of VUR and that the RIVUR study would provide additional evidence, they said that the VCUGs indicated if the ultrasound is abnormal and following a second urinary tract infection and that children with febrile UTI should have urine tested with subsequent febrile illnesses. There was a lot of reaction to these guidelines from the urology community, from the nephrology community and a few of us as well at that time based on the quality of the study that was used to make that determination. Now let's talk about the studies.

So we published with Ron Keren in this editorial in the New England Journal of Medicine. At the same time that the Craig study, that was the largest study to that time was being reported in the New England Journal of Medicine and what we found, and those were the studies used by the American Academy to do that individual patient metaanalysis and throw it into the blender and come up with results that then were recommended and put in a guideline. The effect of prophylaxis did make a difference in all of these studies. So it hovers around 1 the relative risk about 1, sometimes below 1, sometimes above 1 but basically showed no difference. The samples were very small in each of these studies. And this is the major limitation because the lack of effect of prophylaxis is fairly due to insufficient sample rather than lack of effectiveness of prophylaxis. The studies were not blinded, all of them except the Craig study were not blinded. And they enrolled and used bag specimens in most instances, even on uncircumcised boys. And that's a no/no even in the American Academy of

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 15 Pediatric Guidelines. So the Craig study was the only one that was decent in size to be able to look at difference, it was blinded, but it also had some limitations because 40% of the kids didn't have a VCUG done and there were a lot of other issues. But in spite of that it was the best study to date.

The Swedish group published a whole issue in the Journal of Urology, 6 or 7 papers in the Journal of Urology when their reflux study was completed a few years later. There were 23 pediatric centers, they had the goal of enrolling 330 children that were aged 1 to 2 years old, so no under 1. They had to have VUR - the Swedish don't believe in VUR grades I to II, they don't call it VUR, they only say reflux is greatly involved. They had 2 year follow-up, febrile UTI, renal scarring, resolution of reflux. They used midstream, bags, suprapubic a mix of everything, symptoms and abnormal labs. And they enrolled 203 children. They randomized them prophylaxis, they did endoscopic treatment of the second arm and surveillance which was control, there was no placebo being used, and they followed them for 2 years.

They showed increased likelihood of febrile UTI in the surveillance group in girls but not in boys. It was unrelated to the VUR severity or the DMSA at study entry. They showed increased likelihood of resistant pathogens in the prophylaxis group. They showed a shorter median time to first reinfection in the surveillance group, and interesting, this was one of the few studies that showed that renal scarring was much lower in the prophylaxis group 6% versus 18% in the surveillance group. But they had limited statistical power because they only involved 70% of the projected sample, 203 out of 330. There was no placebo, there were no children under age 1 year.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 16

So new studies that we have completed are the RIVUR study, the Randomized Intervention for VUR, 600 children age 2 to 72 months with reflux grades I through IV, had 2 year follow-up. They were randomized to receive TMP-SMZ or placebo. Their outcomes were recurrent urinary tract infection, renal scarring and the presence of bacterial resistance.

The CUTIE study, that stands for Careful UTI Evaluation, was also completed with an observation and cohort study. 200 children between children in Philadelphia and Pittsburgh, we were the top enrollers in that study, 2 to 72 months, no VUR, 2 year follow-up, identical to RIVUR on everything, definitions, except that these kids did not have VUR. So if you think about it once we publish these results this will be the natural history of reflux because we can compare kids who had UTI with no VUR, we are following the same way as the placebo patients in the RIVUR study that had also the same exact identical follow-up. So and the findings are being analyzed and reviewed.

And then we'll have the results of the ARKS study, Anabolic Resistance in Kids Study, funded by the Thrasher Foundation, to try to determine the impact on resistance and we're looking at E-coli in he stool, S. aureus in the nares and pneumonia in the nasopharynx to look at places where there is generalized increased resistance.

The RIVUR study was sponsored by NIDDK, it was a multicenter randomized placebo controlled study, it was conducted at 5 clinical treatment centers or core sites, and those were Pittsburgh,

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 17 Philadelphia, Detroit, Buffalo and Baltimore. Everybody except us, and that scared us very much at the beginning, was actually a conglomerate of 5 or 6 major children's hospitals, whether it was nephrology or urology or inpatient services we were the only ones that went alone into this study, and we ended up enrolling the most patients in the study thanks to all of you. There were 19 clinical sites, they included primary care, urology, nephrology, emergency dept. and inpatient services, and the data coordinating center was the University of North Carolina in Chapel Hill.

Inclusional criteria were children 2 months to 5 years at the time of randomization with the first or second febrile or symptomatic urinary tract infection, and it was defined as the presence of pyuria, a positive urine culture by cath, suprapubic or clean void depending on the age, bag specimens were not permitted. Who had grade I to IV VUR. We excluded them if the UTI was diagnosed more than 112 days prior to randomization, if they had any other comorbid urologic anomalies, any contraindications to use of TMP-SMZ, or any other selected medical conditions. Follow-up included phone calls every 2 months, visit every 6 months and scheduled visits to rule out urinary tract infections, DMSA renal scans were done at entry, at 12 months, 4 months after treatment failure and at 2 years at 24 months.

Pyuria was defined in any way pyuria can be defined, 10 white cells per cubic mm, 5 white cells per how power field or positive leucocyte esterase, a culture proven infection was required greater than 50,000 for cath or suprapubic and greater than 100,000 for clean voided specimens. Children could have a fever, more than 38 degrees, or they could have symptoms related to the urinary tract,

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 18 suprapubic, abdominal, flank pain or tenderness, urgency, frequency, hesitancy, dysuria, foul smelling urine and young infants failure to thrive, dehydration or hypothermia.

Primary outcome was the presence of recurrence of febrile or symptomatic urinary tract infection, secondary outcomes included the prevalence, the extent and new renal scars on the outcome DMSA scan. Treatment failure was the defined as the recurrence of two febrile urinary tract infections, one febrile and three symptomatic urinary tract infections, it was pretty stringent, or four symptomatic urinary tract infections or new or worse scar at the 12 month follow-up. Bacterial resistance, we look at stool colonization with E-coli, resistant to TMP-SMZ and recurrent febrile symptomatic UTI caused by a resistant pathogen. And the covariates were a VUR grade, bladder and bowel dysfunction, constipation and adherence to study medication.

We obsessively evaluated the DMSA renal scans. DMSA scan were done as I said less than 112 days from the index UTI, they were repeated at 12 months, 4 months after treatment failure, 24 months. We looked at definition of renal scarring after increased uptake associated with loss of contours or cortical thinning. Each kidney was divided into 12 segments as you see on the picture there. There was a 5 level grading system used for qualifying the grade of scarring. There were two nuclear medicine physicians, one of them Massoud Majd who is the world expert on DMSA who reviewed and adjudicated all the DMSA renal scans and CVR scarring was defined as grade 3 or 4 in at least one kidney which is basically a lot of kidney areas involved or global atrophy characterized by diffuse scarred and shrunken kidney.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 19

11,000 children, close to 11,000 children were assessed for eligibility. Most were not eligible because of not meeting UTI criteria, because of not having had a BCUG or not having reflux or having 61 with high grade reflux. About 13%, 1426 were eligible to participate. And of those 43%, 607, were randomized over a 6 year period. And about half, 302 and 305, were included in the analysis in each of the two treatment groups and they were assessed at the one year and the two year visits. There were no differences between the treatment groups with regards to selected baseline demographic characteristics, median age was 12 months, most were girls, 92% were girls in each of the two treatment groups. The boys, the few boys that we had mostly were uncircumcised, most were Caucasian, about 80%, and about 10, 15% were of Hispanic ethnicity.

There were no differences between the treatment groups with regards to selected clinical characteristics. For most children over 90% it was the first UTI, for most of them, 85% or 88%, it was afebrile urinary tract infection. When they were toilet trained at entry more than half had bladder and bowel dysfunction, that's very high. A very small proportion had hydronephrosis on the renal ultrasound, about 5%. And the grades of reflux that we found was mostly grades II and III, about 80%. A little over 10% had grade I and a little under 10% had grade IV VUR. And of course nobody had grade V because there was an exclusion criteria, and about half the kids had bilateral reflux and renal scarring was very infrequent at study entry, 3 to 4%.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 20 Now to the main results. Antimicrobial prophylaxis reduced in half the likelihood of febrile or symptomatic urinary tract infection recurrences, 25% to 12%, with very tight confidence intervals. Twice as many children in the placebo group had treatment failure, 9.6% versus 5%. We found no differences in renal scarring at 2 years. No difference in severe renal scars and no differences in new renal scars between the two treatment groups. Once we looked at E-coli in the stool of children there was a trend towards high likelihood of resistant E-coli among those who were receiving prophylaxis but the trend was not significant. Once we looked at reinfections the reinfections among children who were receiving prophylaxis were most likely and significantly more likely to be caused by a pathogen that was resistant to the antibiotic that we were using for prophylaxis.

I want to pause here because these are frequent questions I get. Okay, so we are increasing resistance and in some ways the point about this resistance is that we expect to find a resistant pathogen in a child who is on prophylaxis and has a breakthrough infection. That's exactly what we expect. And I worry about the ones that are not having a resistant pathogen because it tells me that they were not adherent with study medication, So we never use, or we rarely use TMP-SMZ for treatment of urinary tract infection. If they are on prophylaxis and they have a breakthrough we move on to a third generation cephalosporin where resistance has not consistently changed over the years.

So the Kaplan-Meier curves. On the horizontal axis you have time, on the vertical axis you have the proportion of children with urinary tract infection, the dotted line represents placebo, the sold line

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 21 represents antimicrobial prophylaxis. You have confidence intervals every 6 months, a bar is every 6 months, and at the bottom you have the number of children at risk at any given time point. Antimicrobial prophylaxis reduced in half the proportion of children with febrile or symptomatic urinary tract infection. This is different from the Craig study, the Craig study showed difference early on and then both difference kind of vanished after 6 months. In our study those differences appeared early on and actually widened progressively over time as you see that the curve continued to diverge.

Now going onto the subgroup analysis, again overall we have - on the ratios and confidence intervals we have a 50% reduction overall and there were a total of 111 children who had 171 recurrences and of those 171 recurrences 72% were febrile recurrences. And you can see in this slide that the path of ratios here on the middle column consistently favor antimicrobial prophylaxis irrespective of sex, age, degree of VUR and whether the infection, the index infection was caused by a sensitive or a resistant bacteria.

Interestingly the children who had grade II to IV VUR tend to have more reinfections and that happened in our study as well. They had more reinfections but children who had grade I and II VUR, which in the Swedish mind is not even VUR, and they did not even enroll them in the study, those children derived the most benefit from antimicrobial prophylaxis by having 70% protection, that has a ratio of .3 that had 70% reduction in the likelihood of recurrent urinary tract infections. And the two groups that derived the most benefit from antimicrobial prophylaxis were the ones that

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 22 had a fever at the time of the UTI, and the ones that had bladder and bowel dysfunction at study entry. So those there was a 60% reduction and an 80% reduction of the likelihood of UTI whenever there were on antimicrobial prophylaxis.

Now we showed the same Kaplan-Meier, but now we stick with the children 1 to 24 months, this is on the supplemental material in the New England Journal of Medicine paper. We find that the numbers get smaller, but the differences continue to be significant between these are the kids that are less than 2 years who had a first UTI and the UTI was febrile. The confidence interval bars start to overlap, the findings are still significant between a prophylaxis and placebo. And on the right we see the dotted red and solid red curves, those are the children who have high grade reflux and these are the ones on black that have the lower grade reflux grades I to II, so this is dilated reflux. You can see that there is less differences among those, they had more reinfections but there was less benefit of prophylaxis and the ones that had the lower grade reflux there is more distance between the two curves and the findings become significant. And those are the children at risk of each of the time intervals.

So how does that modify the AAP imaging recommendations? The variables that we need to take into account are the variables that influence the decision for individual children, and they include family history, the age of the child, the gender, the race that determine the probability of having this ureteral reflux. How was the UTI clinical course, how sick was the child whenever they had the UTI? Do they have bladder and bowel dysfunction? What is the discomfort and cost of a VCUG?

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 23 What's the importance for parents, parents place on preventing recurrences, how - it goes back to how bad was the infection whenever the had the initial infection? What is the perceived potential adherence to a prophylaxis longer term? And what are the potential effects of prophylaxis resistance and effects on the microbiome? And most importantly, what is the actual effectiveness of prophylaxis? Once we didn't know the effectiveness of prophylaxis and the AAP had all those studies that I showed you before in that editorial in the New England Journal of Medicine that showed no difference, holding the horses was okay. In some way now that we know that the prophylaxis reduces in half the likelihood of reinfections we may need to reconsider some of these things. And we are conducting decision analysis and cost effective analysis to try to enable more informed decisions on these areas as well.

We always try to learn new things, and in the context of the RIVUR study we did a mini study, a study within a study, where we asked 120 parents in Pittsburgh, half of whom agreed to participate and half of whom declined participation in the RIVUR study to complete an anonymous survey trying to evaluate the reasons to consent and not to consent for participation in a clinical trial. The conceptual model we used incorporated child characteristics, parental characteristics, parental understanding of randomization, blinding and so forth, parental perception about the study, risks, benefits, benefit to others and so forth, how the decision making occurred and that resulted in the decision to participate or not to participate. We couldn’t evaluate the study characteristics because we did just one study, this is to compare different studies.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 24 We concluded and they are showed in bold here, bold typeface, hard to read but they are bolded the ones that were significantly different between the ones who consented and declined participation in the clinical trial.

We concluded that parents who declined consent had a relatively higher

socioeconomic status, had more anxiety about their decision and found it harder to make their decision compare with consenting parents, who had higher levels of trust and altruism, perceived the potential for enhanced care, reflected better understanding of randomization and exhibited low decision and uncertainty. So consideration of these factors in the model should enhance the quality of the informed consent process and improve participation in pediatric clinical trials.

So the next steps we are currently conducting the SCOUT study, Short Course Therapy for UTI in Children sponsored by NIAID involving 746 children in Pittsburgh and Children's Hospital in Philadelphia, 2 months to 10 years. It's a 4 1/2 year study. We are comparing 5 days versus 10 days of antibiotics and the antibiotics are listed in the slide, and the outcomes, treatment failure at day 14, the presence of recurrences and the occurrence of resistance.

Nadar Shake is conducting a study looking at biomarkers comparing acute pyelonephritis versus cystitis, also sponsored by NIAID and enrolling 160 children. He only needs I think 20 more he told me yesterday, 2 months to 10 years and looking at urine interleukins, procalcitonin and correlation with DMSA scan.

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 25 He is also conducting the STAARS study, Steroids to Actively Reduce Renal Scarring funded by NIDDK looking at 390 febrile children, 2 months to 6 years, it's a 4 year study comparing Dexamethasone or placebo for 3 days plus antibiotics and looking at outcomes, DMSA scans for evaluation of renal scarring.

And we are in the process many of us in developing a proposal to study antenatal hydronephrosis and it's a collaboration of Journal of Academic Pediatrics, nephrology, urology, neonatology, imaging at Magee Women's Hospital and being developed and it will be submitted February to NIDDK.

So what I think I told you is that 60% of children have a febrile urinary tract infection, when they have a febrile UTI have acute pyelo. And 15% of them will scar. Urine culture is sterile within 24 hours so no need for test of cure. There is high levels of resistance to frequently used antibiotics. Outpatient management is safe and effective, short courses, 2 to 4 days might be okay for presumed cystitis, longer courses 10 days for presumed acute pyelonephritis. We have to maintain a heightened vigilance for UTI and always consider bladder and bowel dysfunction. It's likely that early diagnosis and treatment prevents most UTI related renal scarring and a single approach may not be appropriate to evaluate our children. We continue to do ultrasound with the first febrile UTI, if VCUG or ultrasound is abnormal or if they have greater than 39 degrees or an unusual organism in non-E-coli pathogen. The AAP watchful waiting approach may be valid regarding imaging but it's definitely not valid regarding the benefits of antimicrobial prophylaxis and again we always need to

UTI AND VUR IN CHILDREN: WHAT WE LEARNED FROM THE RIVUR STUDY, ALEJANDRO HOBERMAN, MD 26 consider the hazards of basing any kind of practice changes based on methodologically flawed studies. In children with VUR following UTI antimicrobial prophylaxis substantially reduce the risk of recurrence but not of renal scarring.

So as you will see, it took 8 years in a village to conduct this trial, and I'm indebted to all our research participants and their families, to my mentors in research, Jack Paradise, Ken Rogers and Ellen Ward, to David Promater for all his support, to all faculty members of the Division of General Academic Pediatrics, to our community practitioner partners, pediatric fitness and the Clinical Presentation of Science Institute and NIDDK, NAID, Thrasher Foundation and to the entire research team. And I want to put names to the faces. Everybody knows Nadar Shake, he's always in the emergency dept., Susan Cappelli, Mary Ann Harlan, Michelle Todd, Amy Messick hiding over there, Jennifer, she was trying not to get in the picture, and Marsha Post, Madma Kubbich, Judy Martin, Diana Kerney, Tim Shultz, Amada Zemick and Steven Saylor, all of whom part of this research team that allow us to conduct these and many other studies on a daily basis that I'm mostly proud of. And finally to all of you for your attention and all the folks that participated in the RIVUR study are listed in this slide.