Emphysematous Pyelonephritis

3471 4

http://www.indianjnephrol.org/articles/2013/23/2/images/IndianJNephrol_2013_23_2_119_109418_f6.jpgI was asked to see 74 year old man with an acute on chronic kidney injury.  He had 2 days of generalised lower abdominal pain and vomiting but no urinary symptoms or fever.  His past history included advanced CKD, benign prostatic hypertrophy and a slow growing renal cell tumour under radiological surveillance.  His vital signs were normal and he had mildly raised inflammatory markers.  I ordered a CT KUB to exclude obstruction (it was a weekend and was no ultrasound service in the hospital).  To my surprise this came back as showing emphysematous pyelitis.  Interestingly there had been a hiss of air as he was catheterised for fluid balance monitoring – a fact I had dismissed at the time!

Emphysematous UTIs are gas forming infections of the urinary tract and can manifest as cystitis (gas within the bladder wall), pyelitis (gas within the collecting system) or pyelonephritis (gas within renal parenchyma or perinephric tissues).  It is a relatively rare condition and there is a dearth of literature describing incidence.  Diabetes and urinary tract obstruction are major risk factors, present in around 80% and 20% of patients respectively. Causative organisms are most commonly E. Coli and Klebsiella pneumoniae, with Candida being involved less frequently. Presentation is usually similar to acute severe pyelonephritis with fever, flank pain and vomiting. 50% of patients have an associated bacteraemia. Diagnosis is usually made by CT which shows the extent of gas within the urinary tract and any obstruction.
Treatment depends on the extent of infection.  It ranges from parenteral antibiotics alone for patients where gas is limited to the collecting system with no obstruction, to percutaneous drainage of purulent material and antibiotics if there is abscess formation or extension of gas into the perinephric space, to nephrectomy in patients with diffuse gas and extensive renal destruction.

In the above case, the urine sample was initially reported as ‘no significant bacteriuria’ but subsequently grew a resistant E. Coli >1,000 – <10,000 cfu/ml.  Urology felt that surgical intervention was not required as the renal parenchyma was not involved and he had no abscess formation.  The patient completed 2 weeks of ertapenem and his renal function returned to baseline.
I took several learning points away from this case:
  • As someone who spends a lot of time signing off patients’ results, I realise that ‘no significant bacteriuria’ is not the same as ‘no growth’, and in this case the difference was substantial.  The wording of how we report things and how we interpret that is crucial.
  • A high index of suspicion is required to diagnose emphysematous UTIs and the most appropriate imaging modality should be considered.  Ultrasound is generally the first line investigation for urinary obstruction in patients with acute kidney injury or febrile urinary tract infection due to high sensitivity for hydronephrosis, lack of ionising radiation and lower cost than CT. Ultrasound appearances in emphysematous UTIs can be difficult to interpret however: gas, calculi and calcifications are hard to distinguish and there is often variability in how they are reported. CT is able to precisely localise the presence of gas within the urinary tract and determine whether there is involvement of the renal parenchyma and perinephric tissues.  It can also identify any concomitant pathology or alternate diagnosis e.g. renal calculi.  CT is therefore preferable for diagnosis and subsequent severity staging.
  • Pneumaturia has been described as a presenting feature of emphysematous UTI.  Other causes include vesicovaginal or vesicoenteric fistulae, renal tumour infarction and recent instrumentation.  The unexpected air hiss when catheterising this gentleman was a warning of a more serious pathology and should prompt further investigation.

While the outcome of emphysematous pyelitis is better than that of pyelonephritis (which has a mortality of 18-70% depending on extent of involvement), it is still not a condition to be taken lightly. 
Post by Ailish Nimmo, Royal Infirmary of Edinburgh


  1. @Mohammad A Kaballo,
    NO not yet. One case was indicated but the family members of that patient didnot give consent.Finally that patient expired.

  2. @BIRDEM Nephrology BD
    EPN survival rates in your centre appear to be better than those previously described in the literature. Just curious to know if any of them required a nephrectomy?

  3. Very interesting case. I saw few cases during my training and their severity ranged from mild to fulmminant. One of the diagnostic dilemmas frequently encountered in mild cases, i.e. Class-1 (as per Huang et al), is whether the gas in the collecting system is due to a genuine emphysematous infection or is it just a result of bladder catheterization. As you have already mentioned, pneumaturia suggests the former rather than the latter. Personally, I have a very low threshold to start antibiotic therapy should imaging shows air in the collecting system. By the way, is the figure attached to this post belongs to your patient? I presume not!! The emphysema involving the right kidney is quite severe! Thanks for the post.

  4. I am working in nephrology department of a diabetic hospital. we have encountered around 35 cases of emphysematous pyelonephritis in last 6-7 years. Maximum patients were female and all cases had uncontrolled blood glucose level.No obstructed uropathy was found. E.coli was the commonest organism causing EPN.Ct scan was the preffered method of diagnosing EPN. 90% of cases survived with medical management with or without open drainage.

Leave a Reply