Pyelonephritis & Renal Infection
- Pyelonephritis is a kidney infection — usually a urine infection that climbed upstream from the bladder. It's a clinical diagnosis; imaging is for the sick, the not-improving, and the "is there a complication?" crowd.
- On contrast CT the classic look is a wedge of kidney that lights up less and later — striated stripes pointing toward the renal sinus. There is no widely accepted numeric threshold for calling it; the pattern is the diagnosis.
- The findings you must not miss are the complications: an obstructing stone (unenhanced CT finds ureteral stones with sensitivity 97% and specificity 96% in acute flank pain), a walled-off abscess, and gas in the kidney.
- Gas gets graded: the Huang–Tseng classes run from class 1 (gas in the collecting system only, best prognosis) to class 4 (bilateral, or a solitary kidney).
- A normal CT does not rule out pyelonephritis. The scan exists to find the "and," not to confirm the diagnosis you already made at the bedside.
Most kidney infections never need a scanner. Someone shows up with a fever, a sore flank, and urine that could be classified as a biohazard, and the diagnosis is made before radiology even wakes up. So why do we image at all? Because every so often the kidney isn't just infected — it's infected and something else is wrong. Our job is to find the "and."
What's actually happening
Think of the urinary tract as a one-way river that's supposed to flow downhill: kidney to ureter to bladder to exit. The trouble starts when bacteria decide to be salmon and swim upstream from the bladder back into the kidney. That's the usual route, called ascending infection, and it's why a simple bladder infection that gets ignored can graduate into a full kidney infection. The rarer route is hematogenous seeding — bacteria arriving by the bloodstream, which tends to plant small round abscesses in the cortex rather than the wedge-shaped, tubule-following pattern of ascending disease.
The medical name, pyelonephritis, just bolts together "pyelo" (the renal pelvis, the funnel where urine collects) and "nephritis" (inflamed kidney). So: inflamed kidney plus inflamed funnel. The kidney swells, gets angry, and the inflamed tubules stop moving filtrate — and contrast — as smoothly as they should. That last part is the entire basis of every CT sign on this page.
When do we even image?
Here's the honest version: an otherwise healthy adult with classic symptoms gets antibiotics, not a CT. Imaging earns its keep when the picture is murky or worrying — and the people we worry about are fairly predictable.
| Image when... | Because we're hunting for... |
|---|---|
| Not improving on appropriate antibiotics | A walled-off pocket of pus (abscess) |
| Known or suspected kidney stones | An obstruction trapping infected urine |
| Diabetes, immunosuppression, or looking septic | Severe, complicated, or gas-forming infection |
| Diagnosis genuinely unclear | Something else entirely (appendicitis, etc.) |
An infected kidney behind an obstruction — say, a stone wedged in the ureter with pus building up above it — is an emergency. That's a closed, pressurized space full of bacteria (pyonephrosis), and it can tip someone into sepsis fast. It usually needs decompression, not just antibiotics. If you see a stone plus a swollen, dilated collecting system plus a fever, say so loudly.
Findings by modality
Ultrasound
Ultrasound is often the first test in children and pregnancy because it skips the radiation, but it is the wrong tool for seeing uncomplicated pyelonephritis: most early infections look completely normal. What it does well is the complication hunt. Look for a dilated collecting system (obstruction), echogenic debris or layering material inside that dilated system (pyonephrosis rather than plain hydronephrosis), a rounded hypoechoic or complex collection with no internal flow on color Doppler (abscess), and — rarely, but unmistakably — bright echogenic foci with dirty shadowing or ring-down artifact from gas. A swollen, globally enlarged kidney with a smudged corticomedullary junction is a nonspecific "this kidney is unhappy" sign; a focal hypoechoic or hyperechoic wedge with reduced color flow is the closest ultrasound gets to the CT picture.
CT
The workhorse is contrast-enhanced CT, read in the nephrographic phase, when normal parenchyma should be uniformly bright. Infected, inflamed tubules don't take up contrast as eagerly or as evenly as healthy ones. Imagine pouring dye into a sponge that's partly waterlogged — the dye seeps in slowly and unevenly. On CT that shows up as the striated nephrogram: alternating bright and dull bands fanning from the renal sinus to the cortex, like someone dragged a comb through the kidney. The affected zones are typically wedge-shaped, apex at the sinus and base at the capsule, because that's the shape of the tubule-and-vessel territory the bacteria climbed.
The supporting cast: focal or global renal enlargement, perinephric fat stranding (inflammation smudging up the normally crisp fat), thickening of Gerota fascia and of the renal pelvic wall, and on delayed images a persistent or increasing nephrogram in the affected zones as the sluggish tubules finally fill. The unenhanced series matters too — it's where the obstructing stone hides once everything else turns white.
Two complications have their own CT signatures. A renal abscess is a rounded, fluid-attenuation collection with a thick enhancing rim and a center that never enhances, because liquefied pus has no blood supply to carry contrast. Emphysematous pyelonephritis is gas within the renal parenchyma — near-black foci or streaks where solid kidney should be — and its extent is what the Huang–Tseng classes grade (see below). Both get the full treatment on the renal infection spectrum page.
Plenty of pyelonephritis looks completely normal on imaging — especially uncomplicated cases scanned early. A clean CT does not rule out a kidney infection. The scan is there to find complications, not to confirm the diagnosis you already made at the bedside.
MRI
MRI is the radiation-free option when the question is complication rather than diagnosis (pregnancy, children, repeated imaging). The infected wedges are T2-hyperintense and swollen, show restricted diffusion on DWI, and enhance less than their neighbors after gadolinium in exactly the same striated, wedge-shaped distribution CT shows. An abscess restricts diffusion centrally — bright on high b-value DWI, dark on the ADC map — which is a useful way to separate pus from a simple cyst or a bland fluid collection. Gas is a signal void on every sequence and is much easier to appreciate on CT.
Nuclear medicine
A technetium DMSA cortical scan shows acute pyelonephritis as focal or diffuse areas of reduced cortical tracer uptake; its main clinical home is pediatric imaging, where the question is whether an infection has left permanent scars (a defect that persists on a follow-up study). The DMSA cortical scan page covers it properly.
The numbers
| What | Threshold / value | Why it matters |
|---|---|---|
| Adult kidney length at sonography (median) | 11.2 cm left, 10.9 cm right | The baseline against which "swollen" and "small and scarred" are judged; varies with age, sex, and habitus |
| Adult kidney volume at sonography (median) | 146 cm³ left, 134 cm³ right | Same baseline, for labs that report volume |
| Unenhanced CT for ureteral stone in acute flank pain | Sensitivity 97%, specificity 96%, accuracy 97% | The obstructing stone is the complication that turns pyelonephritis into an emergency; unenhanced CT is the test that finds it |
| Emphysematous pyelonephritis, Huang–Tseng class 1 | Gas confined to the collecting system | The best-prognosis class; class rises with extension into parenchyma and beyond |
| A numeric CT or ultrasound threshold for "this is pyelonephritis" | None widely accepted | The diagnosis is the striated, wedge-shaped pattern plus the clinical picture, not a measurement |
How good is the test
There is no robust pooled sensitivity or specificity figure for CT, ultrasound, or MRI in diagnosing acute pyelonephritis, and that is the honest state of play: imaging is supportive, not diagnostic, and a normal study never excludes infection. The number that is robust is the one that matters for the emergency: unenhanced CT detects the ureteral stone in acute flank pain with sensitivity 97% and specificity 96%, so if the question is "is this kidney infected and obstructed," CT answers it.
Grading the gas
When gas is present, the Huang–Tseng CT classification stages emphysematous pyelonephritis by how far it has spread. Here is the abbreviated version; the full table with management context lives on the renal infection spectrum page.
| Huang–Tseng class (CT, 2000) | Where the gas is |
|---|---|
| Class 1 | Collecting system only (best prognosis) |
| Class 2 | Renal parenchyma, without extension to the extrarenal space |
| Class 3A / 3B | Gas or abscess extending to the perinephric (3A) or pararenal (3B) space |
| Class 4 | Bilateral EPN, or EPN in a solitary kidney |
Higher class correlates with worse outcome, which is why the report should say where the gas is, not just that it exists.
The mimics
| Mimic | Looks similar because | Tell them apart by |
|---|---|---|
| Renal infarct | Also a wedge of non-enhancing kidney pointing at the hilum | Infarct margins are sharp and geographic, the wedge is uniformly dark rather than striated, a cortical rim sign may be present, and the patient is not febrile with pyuria; think embolic source, not urine culture |
| Hypovascular renal cell carcinoma or lymphoma | A focal, mass-like area of reduced enhancement | Masses are rounded, bulge or distort the renal contour, and persist on follow-up; infection follows a tubular wedge, respects the contour, and resolves after treatment — re-image before calling a mass |
| Xanthogranulomatous pyelonephritis | Enlarged, poorly functioning infected kidney | A central obstructing stone, a non-functioning kidney, and low-attenuation "bear paw" collections replacing the parenchyma; it is chronic, not acute |
| Renal vein thrombosis | Swollen kidney with delayed, patchy enhancement and stranding — and it can produce a striated nephrogram of its own | Thrombus in the renal vein, a globally swollen kidney with a delayed and persistent nephrogram, and no pyuria; the stripes don't separate them, the vein does — look at it before you blame bacteria |
| Post-procedural gas in the collecting system | Black foci in the kidney | Gas confined to the pelvis and calyces after a catheter, stent, or instrumentation in a well patient is plumbing, not emphysematous pyelonephritis; gas in the parenchyma is the alarm |
Gas inside the kidney parenchyma is the one that should make your stomach drop. Emphysematous pyelonephritis classically strikes people with diabetes, and higher Huang–Tseng class means worse prognosis. Air where air has no business being — that's the alarm bell.
What's left behind
Most infections clear and the kidney recovers. But repeated or severe infection can leave permanent dents in the kidney's surface — cortical scars overlying blunted calyces — relevant later when a scarred kidney gets confused for something else, or when you're sorting out a renal mass. A focal, mass-like bacterial infection (sometimes called focal bacterial nephritis or lobar nephronia) can even mimic a tumor; the history of fever and infection, and resolution on a follow-up scan, is what keeps you honest.
Reporting
The referrer is silently asking a handful of yes/no questions. Answer them in order.
- Is the kidney obstructed? Hydronephrosis, and its cause and level (a stone, with its size and location). Infected plus obstructed is the emergency line of the report.
- Is there a drainable collection? Location, size in three planes, whether it is intrarenal or perinephric, and whether the wall enhances — that is what the interventional team needs to decide on percutaneous drainage.
- Is there gas, and where exactly? Collecting system, parenchyma, perinephric or pararenal space, one kidney or both — in other words, the Huang–Tseng class.
- How much kidney is involved? Focal versus multifocal versus diffuse, unilateral versus bilateral.
- Anything that argues against infection? Sharp geographic margins (infarct), a contour-deforming mass, or venous thrombus.
- The disclaimer that protects the patient: a normal study does not exclude pyelonephritis.
The one-sentence version: pyelonephritis is a clinical diagnosis, and the scanner's real job is to catch the obstruction, the abscess, or the gas before they catch the patient.
References
- Huang JJ, Tseng CC. Emphysematous pyelonephritis: clinicoradiological classification, management, prognosis, and pathogenesis. Arch Intern Med 2000;160(6):797–805. Used for: the Huang–Tseng class 1 to class 4 definitions and the statement that higher class correlates with worse outcome, in "Grading the gas," "The numbers," the Key Points, and "Reporting."
- Smith RC, Verga M, McCarthy S, Rosenfield AT. Diagnosis of acute flank pain: value of unenhanced helical CT. AJR Am J Roentgenol 1996;166(1):97–101. Used for: the sensitivity 97%, specificity 96%, and accuracy 97% of unenhanced helical CT for ureteral stones in acute flank pain, in the Key Points, "The numbers," and "How good is the test."
- Emamian SA, Nielsen MB, Pedersen JF, Ytte L. Kidney dimensions at sonography: correlation with age, sex, and habitus in 665 adult volunteers. AJR Am J Roentgenol 1993;160(1):83–86. Used for: the median adult kidney lengths (11.2 cm left, 10.9 cm right) and volumes (146 cm³ left, 134 cm³ right) in "The numbers" and the Measurement chip.