Abstract
A case of bilateral, upper urinary tract infection caused by haemolytic E coli in a female Birman cat is presented. Ultrasonographic examination of the kidneys documented changes in size, outline, echogenicity and architecture. Ultrasound guided fine needle aspiration of fluid from the renal pelvis was used to make the diagnosis. Fluid was submitted for culture and sensitivity and based on the results, antimicrobial therapy was initiated. The treatment was monitored over a 406-day follow-up period. Despite extensive treatment with specific antibiotics and supportive therapy, recurrence of urinary tract infection occurred.
Urinary tract infection occurs when uropathogens colonise urine or tissue of the urinary tract. However, isolation of significant numbers of bacteria from urine does not localise the site of infection to the urinary tract unless the urine sample was collected by cystocentesis. Nor does isolation of bacteria from renal pelvic urine imply that invasion of the renal parenchyma has occurred. Clinical and laboratory findings may occasionally indicate pyelonephritis, but confirmation of the diagnosis is, at best, difficult (Polzin et al 1989). Pyelonephritis may be sub classified based on duration as acute (days to weeks) or chronic (months to years). Clinical findings in patients with acute pyelonephritis may vary from mild, non-specific findings to profound clinical signs consistent with septicaemia. Signs associated with acute pyelonephritis include pyrexia, anorexia, vomiting, renal pain, lethargy, malaise and tremors. Clinical manifestations of chronic pyelonephritis are often subtle.
A satisfactory, sensitive, specific, safe, and noninvasive method for identifying renal infection in dogs and cats is not currently available. Although ultrasound guided biopsy (core biopsy as well as aspiration) is an invasive method, ultrasound has become a useful diagnostic tool for the investigation of soft tissue disorders. However, aspiration gives a lower diagnostic yield than core biopsy (Barr 1995). Continuous real-time ultrasonographic monitoring of the position of the biopsy needle minimises the risk associated with blind percutaneous sampling techniques. The procedure is considered sensitive, specific and safe in addition to being simple to perform (Barr 1995).
This case illustrates the diagnostic usefulness and safety of ultrasound guided sampling in a cat with bilateral, upper urinary tract infection, and also the difficulties encountered in treating the patient.
Case history
A 4-year-old neutered female Birman cat, weighing 3 kg, was presented to the emergency clinic at the Norwegian School of Veterinary Science. The history consisted of lethargy, anorexia and excessive vomiting of 2 days duration. The water consumption had been inadequate.
On physical examination the cat was depressed and had a rectal temperature of 40.9°C. Pulse and respiration were 106/min and 30/min, respectively. Auscultation of the heart and lungs was normal and there was no lymphadenopathy. Palpation of the kidneys was painful and provoked spontaneous vomiting. The kidneys differed in size, the left kidney was enlarged and the right was smaller than expected. Ventral dorsal and right lateral survey radiographs were taken on the day of arrival (Fig 1). The radiographs revealed an enlarged (>3.0 times the length of the second lumbar vertebral body) left kidney and a small (<2.5 times the length of the second lumbar vertebral body) right kidney (DiBartola et al 1987, Walter et al 1987).
Ventral dorsal (a) and right lateral (b) radiographs revealed an enlarged (>3 times the length of the second lumbar vertebral body) left kidney and a small (<2.5 times the length of the second lumbar vertebral body) right kidney.
Blood samples were drawn for routine haematology and standard biochemistry profile prior to administration of intravenous Ringer's solution. A urine sample was collected by cystocentesis. An aliquot of the urine was submitted for culture and sensitivity. The results of the laboratory findings at the day of arrival are presented in Tables 1–3.
The cat was hospitalised and oral treatment with 10 mg doxycyclin (Vibramycin®) twice daily was initiated. On day 3, the body temperature was below 39.0°C. Based on initial in vitro bacterial sensitivity testing, the doxycyclin treatment was changed to cephalosporin, which was administered intravenously (cephalotin, Keflin®) for 3 days and then administered orally (100 mg twice daily) for 7 weeks.
Three days after the initial examination, an ultrasound examination was performed. The left kidney was large (4 cm in diameter) and had irregular architecture of the medulla and collecting system. The findings were consistent with dystrophic mineralisation. Acoustical shadows were originating from the pelvis. The cortex was regular and even with an expected echo pattern (Fig 2). The right kidney was small (2 cm in diameter) and the only normal renal architecture was found in a small area at the caudal pole. The cranial portion had thin cortex, irregular medulla and a mixed anechoic and hypoechoic cystic change (Fig 3). Fluid was seen in the pelvis of both kidneys.
Longitudinal (a) and transverse (b) ultrasound images of the left kidney. The kidney was 4 cm in diameter with an irregular architecture of the medulla and collecting system. There was dystrophic mineralisation and acoustical shadows originating from the pelvis (b: open arrow). The cortex had normal thickness and echo pattern.
Longitudinal (a) and transverse (b) ultrasound images of the right kidney. The kidney was 2 cm in diameter. Normal kidney architecture was found in a small area at the caudal pole. The cranial portion had thin cortex (a, b: open arrowheads), and an irregular medulla with mixed anechoic and hypoechoic cystic changes (a, b: open arrows).
The cat was mildly sedated with xylazin (Narcoxyl® vet.) 2 mg/kg bodyweight and a biopsy site was clipped and prepared aseptically. The transducer was covered with a sterile sleeve, and sterile acoustic gel was used to maintain contact between the transducer and the skin. Ultrasound guided percutaneous fine needle aspirates were obtained from the renal pelvis of both kidneys. The sampling procedure was performed freehand, using a 21-gauge needle and a syringe. Microscopic evaluation of the aspirates from the pelvis of both kidneys revealed numerous polymorph neutrophils and Gram-negative rods. After culturing a sample from the aspirate, the bacteria were subsequently classified as haemolytic E coli. A diagnosis of bilateral, upper urinary tract infection (E coli) with probable secondary uraemia was made.
Haemogram on the day of arrival
Central Laboratory, The Norwegian School of Veterinary Science
Serum biochemistry results on the day of arrival
Central Laboratory, The Norwegian School of Veterinary Science
The cat was hospitalised and treated intensively for 3 weeks and its general condition improved. Six weeks after arrival, the cat was released from the clinic but treatment with cephalotin was continued at home for another 2 weeks. During the 406 days follow-up period, urine and blood samples were collected regularly to monitor the condition (Fig 4 and Table 3). A recurrence with a moderate deterioration in the clinical condition occurred 371 days after the initial diagnosis and E coli was cultured from a cystocentesis sample (Table 3). Based on culture and sensitivity testing, cephalosporin was again the drug of choice and prescribed for 3 weeks (100 mg twice daily). The general condition of the cat normalised within a few days and remained unremarkable for the remainder of the observation period. At the 406-days follow-up ultrasound examination, the left kidney was reduced in size to approximately 2 cm in diameter. The right kidney remained unchanged. No bacteria were found on cytology and culture of the urine, however, serum concentrations of urea and creatinine remained 2.0 and 1.5 times the upper reference limit, respectively (Fig 4).
Monitoring of serum urea and creatinine during the 406 days follow-up period.
Urinalysis (samples collected by cystocentesis) on the day of arrival and during the follow-up period
High power field.
Urine specific gravity was measured using a refractometer. Measurements of pH, bilirubin, protein, glucose, ketones and occult blood were by a dipstick method (Multistix®, Bayer Corporation, Elkhart, IN, USA). The results for glucose and ketones were normal in all samples and therefore omitted from the table.
Colour and turbidity were also unremarkable in all samples and no casts or crystals were revealed.
White blood cells (WBC), red blood cells (RBC) and bacteria were quantified by microscopic examination of urine sediment.
Discussion
The ultrasonographic changes that may be seen associated with lesions of the kidneys are often non-specific. Cytological or histological examination is therefore frequently indicated if a definitive diagnosis is required. Samples of fluid or tissues from the kidneys may be collected during explorative surgery of the abdominal cavity. However, percutaneous sampling techniques may be preferred due to less trauma and because general anaesthesia is usually not necessary. Ultrasound can be used to guide the needle as it is introduced into the kidney. The major contraindications for percutaneous renal biopsy are blood-clotting disorders.
The importance of bacterial infection of the kidney as a cause of renal disease in the cat is unclear and the pathogenesis of pyelonephritis is not known (Gaskell 1988).
The prevalence of bacterial pyelonephritis in the cat is unknown since only isolated surveys on small populations have been made (Finco & Barsanti 1979). In a study of 74 cats with chronic renal disease 9.5% (seven cats) had chronic pyelonephritis (DiBartola et al 1987). Two cases of distal renal tubular acidosis associated with chronic pyelonephritis in a cat were reported (Drazner 1980, Watson et al 1986).
Studies in dogs indicate that vesicoureteral reflux is a factor in the development of bacterial pyelonephritis (Tsuchida et al 1973). However, the question arises concerning whether the vesicoureteralreflux, the infection, or other factors are responsible for the decreased size and function of the kidneys. Experiments with intravenous injections of bacteria into rodents indicate that naturally occurring infection by this route is possible (Fry et al 1975). Patchy pyelonephritis resembling the naturally occurring lesion has been reproduced experimentally in cats by intravenous injection of a feline strain of E coli and temporary obstructions of the ureters. The lesions were less extensive and, over many months did not progress to chronic end-stage kidney disease (Kelly et al 1979). Another attempt to induce pyelonephritis failed in four cats that had been inoculated intravenously with a strain of E coli isolated from kittens with pyelonephritis (Hara et al 1976). Therefore, it remains unclear if infection in natural occurring disease in cats is by haematogenous or ascending routes. The reason for the apparent progression of pyelonephritis in naturally occurring cases is obscure, although repeated infection of a previously damaged kidney, made susceptible as a result of scar tissue and intrarenal obstruction, may be responsible (Gaskell 1988). A variety of organisms have been isolated from cases of feline pyelonephritis, including E coli, Proteus spp., and Staphylococcus spp (Forrester 1997).
This report gives an example of the usefulness of real time ultrasound guided fine needle aspirates in a case of bilateral, upper urinary tract infection in a cat. Initially the cat was severely depressed mainly due to kidney dysfunction and intoxication. The biochemistry profile showed results consistent with reduced renal function (elevated serum concentrations of urea, creatinine, amylase) and inflammation (hyperglobulinaemia and increased number of neutrophils). Despite intensive, long-lasting treatment a recurrent bacteriuria was experienced 10 months after the first episode. No predisposing causes for either the initial or recurrent infection were identified. The genetic homogeneity of the bacterial strains of E. coli isolated initially and after the recurrent infection was not evaluated. Therefore, it is not possible to decide if the second episode of bacteriuria was a relapse or a recurrent infection.