猫和犬肾盂肾炎——一项范围综述。 机翻

There is a lack of consensus regarding the diagnosis and treatment of bacterial pyelonephritis (PN) in dogs and cats. The aim of this scoping review was to increase knowledge by describing published scientific literature on the subject, specifically to collate definitions and diagnostic criteria, antimicrobial regimens, reported outcomes, and to identify knowledge gaps. A structured search was performed of the bibliographic databases Biosis, CAB Abstracts, Medline, Web of Science Core Collection and Scopus. We included 176 original studies written in English or German that reported on any aspect of spontaneous PN (risk factors, signalment, history, clinical findings, diagnostic criteria, treatment, prognosis, and outcome) in cats or dogs. Three types of PN were identified in the literature: histologically diagnosed PN, clinically diagnosed PN, and pyonephrosis. Multiple information gaps regarding clinical aspects of PN in dogs and cats were detected. There were no universally accepted clinical criteria for clinical diagnosis of PN and no clinical trials evaluating antimicrobial or other treatment options. Information about short-term outcomes for dogs and cats with PN was scarce, and there were no data regarding long-term outcomes. More information regarding all aspects of clinical PN in dogs and cats is needed.

The study protocol was registered prospectively in 2023 (https://syreaf.org/protocols/).

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12917-026-06042-1.

Introduction

Bacterial pyelonephritis (PN) is usually caused by ascending bacterial infection of the kidney, though a haematogenous route is also possible. It represents a potentially life-threatening, but often treatable condition [1]. Both dogs and cats are affected, and the condition may be acute or chronic. Substantial uncertainty exists regarding both diagnosis and optimal treatment in both species. Especially chronic forms of PN represent extraordinary diagnostic challenges for clinicians because of a frequent lack of localising signs of infection. To the authors’ knowledge, no specific and universally accepted clinical definition of (diagnostic criteria for) PN in dogs and cats exists, further complicating diagnosis. For establishment of a clinical diagnosis of PN, physical examination findings and clinicopathological data are often combined with imaging of the urinary tract. Pyelonephritis is often considered a differential diagnosis in dogs and cats with renal pelvic dilation, which can also result from diuresis, ureteral obstruction, or chronic kidney disease without bacterial infection [2]. Consequently, both over- and underdiagnosis of PN are likely to be relatively common, negatively affecting patient care. Fluoroquinolones are recommended internationally for treatment of PN in both species [1]. These antibacterial substances are critically important in human medicine, and their use in animals should be restricted to mitigate the risk to the public, according to the European Medicines Agency. This represents one important reason to keep overdiagnosis of PN to a minimum. Evidence-based recommendations regarding diagnosis, choice of antibacterial substance, and duration of treatment are presently difficult to make because of a lack of high-quality evidence to answer those clinical questions.

It is important that the most relevant outcomes, those that matter to owners, patients, and clinicians, are measured in clinical studies to determine the most effective treatment and management strategies for a particular disease [3]. Different variables are often used among studies, and this limits the evidence available to help with the clinical decision-making process. Knowledge about outcomes used in previously published studies on PN in cats and dogs might be of help in future establishment of core outcome sets (agreed set of outcomes to be measured and reported as a minimum in future studies conducted relating to a particular disease). It is unclear what information is available in the literature regarding PN in dogs and cats, and because of the existence of a broad range of unanswered clinical questions regarding definition, diagnosis, treatment and monitoring, and a perceived lack of studies in the area, a scoping review was considered the most appropriate knowledge synthesis method.

Materials and methods

The aim of this scoping review was to summarise current knowledge regarding the diagnosis and treatment of PN in dogs and cats. The objectives were to describe the volume, nature, and characteristics of published scientific literature on PN in dogs and cats, and to collate case definitions and diagnostic criteria, antibacterial regimens, and reported outcomes. A secondary objective was to identify knowledge gaps and prioritise future research.

Protocol and registration

The protocol for this scoping review was drafted using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Scoping Reviews (PRISMA-ScR) reporting guidelines [4]. The final protocol was registered prospectively with SYREAF, Systematic Reviews for Animals & Food in March 2023 [5]. No substantial deviations were made from the protocol. The study was reported according to the PRISMA ScR and the PRISMA 2020 updated guideline for reporting systematic reviews [6]. The PRISMA checklist is provided as additional file 1.

Information sources, search and initial screening

Search strategies were drafted by the authors and two information specialists at the Swedish University of Agricultural Sciences. The bibliographic databases Biosis, CAB Abstracts, Medline (Web of Science interface), Web of Science Core Collection and Scopus were searched, as recommended [7]. There were no initial restrictions for language, year of publication, or publication type. Review texts published < 10 years of the initial search were used to screen reference lists for eligible studies. No supplementary attempts to reach grey literature were made. The initial search was performed on the 22nd of March 2023, and a second search was performed on the 5th of February 2025 to include more recent publications. Detailed information about the search process is provided in additional files 2 and 3. References in chapters relevant to PN in the most recent editions of four companion animal medicine textbooks were also reviewed [8–11].

Individual database results were imported into EndNote X9.3.3 (Clarivate) software, where duplicates were removed using the Karolinska Institute University library method [12]. Thereafter, the final reference list was imported into Rayyan, an internet-based program [13], for blinded screening. The updated screening in 2025 was conducted in Covidence [14]. Three authors (LP, KS, CE) took part in the identification of potentially eligible studies from the search based on titles and abstracts. Full texts were retrieved whenever there was uncertainty about eligibility from review of title and abstract only.

Eligibility criteria

Publications remaining after initial screening were retrieved, and full texts were reviewed independently by two reviewers (LP, KS). To be eligible for inclusion in the review, studies had to be written in English or German and report on any aspect of spontaneous PN (risk factors, signalment, history, clinical findings, diagnostic criteria, treatment, prognosis, and outcome) in cats or dogs. Studies of acute kidney injury (AKI) in dogs and cats were included if infection was the reason for AKI in any of the animals. Urinary tract infection studies were included unless clearly stated that the infections were situated in the lower urinary tract (cystitis, prostatitis, urethritis). Case reports were included. Publications were excluded if animals were experimentally infected or infected with Leptospira species. Studies in people were excluded.

Data extraction

A Microsoft Excel data extraction sheet was constructed for documentation of relevant information from included studies. Two authors independently extracted data from all references, and any inconsistencies were resolved by collective agreement. An initial pilot was conducted on 20 references, after which the form was refined. Given the heterogeneity of included sources, the pilot was expanded to 40 references to ensure that the form adequately captured variation in study design, source type, terminology, and reporting detail. Following this process, the final data charting form was applied to all included sources. The data extracted included study characteristics (title, journal, year of publication, study type) and population characteristics (species, number of animals with PN). Also, diagnostic criteria for the diagnosis of PN (if present), clinical data, treatment, and outcome data were recorded. Studies were characterised as clinical trials, observational studies, or case reports. Studies that reported clinical cases case by case were defined as case reports irrespective of the number of cases described.

Results

Included studies and study characteristics

The initial search resulted in the identification of 6443 references. After removal of duplicates, the number of unique references identified in the first search was 3994. The second search identified a further 315 unique studies, making the total number of screened studies (title and abstract) 4309. Figure 1 shows the search and study identification process in detail.

[图：Flow diagram of search and screening process]

One hundred and seventy-six studies involving approximately 1,600 cats and dogs with PN were included for data extraction [2, 15–189]. Of these animals, 817 were identified in two retrospective studies evaluating prescription data in dogs and cats diagnosed with upper or lower bacterial urinary tract infections at several veterinary practices in the USA and Canada [180, 181]. One hundred and fifty-five studies concerned dogs, 55 studies concerned cats, and 6 studies concerned both species. The studies included in this review were published between 1954 and 2025, with an increase in publication numbers in the last 10 years (Fig. 2). Most studies were conducted in the US (n = 60), followed by Germany (n = 19), India (n = 15), the United Kingdom (n = 14), Australia (n = 11), Canada (n = 10), and Korea (n = 10). The remaining 37 studies were performed in 14 other countries. The journals with the highest number of included studies were Journal of the American Veterinary Medical Association (n = 27), Journal of Veterinary Internal Medicine (n = 23), Journal of Feline Medicine and Surgery (n = 10), and Journal of Small Animal Practice (n = 9). The remaining 107 studies were published in 52 other journals. There were 95 observational studies (Tables 1 and 2) and 81 case reports (Table 3). Observational studies that used histopathological criteria for the diagnosis of PN (n = 18) and 4 studies that described pyonephrosis (accumulation of purulent material in the renal pelvis secondary to an obstruction) are reported separately in Table 2. No clinical trials were identified, but one meta-analysis investigated mortality in dogs and cats with AKI based on aetiology (listed among observational studies in Table 1). The meta-analysis suggested higher rates of overall mortality in cats and dogs with AKI due to noninfectious aetiologies relative to infectious aetiologies.

[图：Publication year for the 176 studies included in this review]

Diagnostic criteria and reported outcomes

Clinical diagnostic criteria for the diagnosis of PN were provided in only 19 studies. These criteria were pre-defined in the methods section in 17 of these 19 studies and presented in the results section in two studies. The specific clinical diagnostic criteria used for diagnosis of PN are provided in Table 4. The criteria were heterogeneous but typically combined elements from two or more of four domains, microbiology (most required a positive urine culture of urine obtained by cystocentesis, with a minority using pyelocentesis as a reference standard, or bacteriuria and pyuria on sediment instead of culture), clinical signs (most commonly lethargy, inappetence, vomiting, fever, and renal or abdominal pain), clinicopathology (azotaemia, leucocytosis, left shift, or increased acute phase proteins), and imaging (renal pelvic dilation, proximal ureteral dilation, hyperechoic cortices, retroperitoneal changes). In Fig. 3, a summarizing overview of diagnostic criteria used for the diagnosis of PN is presented.

[图：Overview of diagnostic criteria used for the diagnosis of PN. Unspec: unspecific signs such as lethargy, inappetence, or]

Only a few observational studies reported clinical outcomes, most commonly short-term survival. Long-term renal functional data in dogs or cats with PN are virtually absent (Table 1). Common urinary comorbidities described in dogs and cats with PN included urolithiasis (cystic, renal and ureteral), chronic kidney disease and congenital abnormalities such as ectopic ureters. A few observational studies reported descriptive proportional data regarding comorbidities within or outside the urinary tract, infecting agents, clinical signs and laboratory or ultrasound abnormalities. Data from these studies are summarised in Table 5.

Diagnostic test accuracy studies (biomarkers and imaging)

Six studies evaluated diagnostic tests for PN. Two feline studies assessed serum amyloid A (SAA) with comparator groups (Kurtz et al., Degenhardt et al.). Kurtz et al. (2024) evaluated SAA in cats with suspected/confirmed bacterial PN versus cats with lower urinary tract disease or chronic kidney disease (CKD) and concluded that an SAA > 51.3 mg/L was suggestive of PN. In this study, cats with ureteral disease, including ureteral obstruction, were excluded. Degenhardt et al. compared SAA among groups of cats with AKI of different aetiologies, azotaemic CKD, and acute-on-chronic kidney injury. Serum amyloid A could not discriminate among groups but 4/4 cats with azotaemic PN had an SAA above 100 mg/L. The remaining one cat with a diagnosis of PN in this study was not azotaemic and had an SAA within the reference interval.

Two studies explored urinary microRNA (miRNA) profiles as potential discriminatory markers for PN versus other urological diseases. Jessen et al. suggested that upregulation of miRNA16 might be PN-specific, pathogen-specific (Escherichia coli), or both. Godia et al. found 7 miRNAs that were differentially abundant between control and PN cats, and another 6 miRNAs that were relevant for discriminating PN from other urological conditions (including ureteral obstruction).

Renal imaging studies with comparator groups were limited to two [2, 147]. Renal pelvic width measured using ultrasonography in dogs and cats with normal renal function (with and without diuresis) were compared to pelvic width in animals with nonobstructive urinary tract disease (AKI, CKD, PN, ectopic ureter and urinary tract neoplasia) and partial to completely obstructive urinary tract disease [2]. A substantial overlap in pelvic width across groups was shown. In cats, Quimby et al. compared ultrasonographic features among groups of cats with stable CKD, clinically diagnosed PN, and ureteral obstruction, and similarly found considerable overlap in renal pelvic and proximal ureteral dilation among groups, indicating that these measurements alone are insufficient to distinguish PN from ureteral obstruction or CKD [147]. Taken together, these imaging studies suggest that renal pelvic or proximal ureteral width measured by ultrasound should not be used as a standalone diagnostic test for PN.

Antibacterial treatment and outcome

No prospective trials were comparing the efficacy among different antibacterial substances or treatment durations. Whitehouse et al. provided the only cohort-level antibacterial treatment data for PN, and in that study, E. coli comprised 89% of 38 isolates from 32 cats with PN [183]. In 18 cats, a potentiated amoxicillin was used, and in 12 cats, a fluoroquinolone was used. Eight of these 12 cats were given pradofloxacin and 4 were given enrofloxacin. The empirically chosen antibacterial substance was changed when culture results were available in 11 (39%) of cats. Median treatment duration was 29 days (range, 10–62). Among 12 cats with post-therapy cultures, bacteriologic cure occurred in 10 (nine of them had received a fluoroquinolone, one had received cefodoxime). Two cats remained bacteriuric (E. coli), one of these had received a fluoroquinolone, and 1 had received cefovecin. Survival to discharge was 94% (29/31), and 3-month survival among cats with follow-up was 70% (14/20). Non-survivors had higher admission serum creatinine, BUN, and phosphorus concentrations than survivors [183].

Antibacterial treatment and outcomes were reported in 93 animals across 81 case reports (Table 3). Amoxicillin/clavulanic acid and fluoroquinolones were the most frequently used antibacterials. Enrofloxacin was the most prescribed fluoroquinolone substance, used in both dogs and cats (Table 3). Reported treatment durations typically ranged from 2 to 6 weeks. No reliable comparisons could be made about treatment efficacy among substances because cases varied widely in severity, comorbidities, and reporting detail.

Histopathology studies and pyonephrosis

Studies describing PN diagnosed by histopathology and pyonephrosis were two distinct themes that stood out among included studies (Table 2). In histopathology studies, necropsy or biopsy series of cases in which PN was diagnosed by tissue diagnosis were described. These reports included both acute suppurative and chronic renal interstitial patterns, frequently together with other urinary tract pathology, such as ectopic ureters. In some of these studies, culture or bacteriologic data implicated Enterobacterales (E. coli, Proteus spp.) as infecting agents. Because case identification was often post-mortem and clinical information and antibacterial data were limited, some cases are likely to have been subclinical. Nevertheless, these studies present plausible pathophysiologic pathways (ascending infection, reflux, anatomic abnormalities) and suggest that PN can be present without clear ante-mortem localizing clinical signs.

The pyonephrosis studies focused on obstructive, pus-filled renal pelves and emphasised imaging characteristics and interventions. Procedural series described renal-sparing approaches such as retrograde ureteropelvic lavage with ureteral stenting in dogs and subcutaneous ureteral bypass devices in cats. Only short-term clinical resolutions were documented.

Discussion

In this study, a systematic search of databases was performed to identify published scientific studies of all aspects of PN in dogs and cats. In the 176 studies that were finally included, three distinct entities were identified: histologically diagnosed PN, clinically diagnosed PN and pyonephrosis. Histologically diagnosed and clinically diagnosed PN represent different diagnostic approaches and may therefore identify somewhat different patient populations. Pyonephrosis, accumulation of purulent material in the renal pelvis secondary to an obstruction, could be described as a separate form of PN. Interestingly, pyonephrosis may also develop when accumulated pus secondarily obstructs the ureter in cases of PN. Pyonephrosis is probably usually less difficult to diagnose clinically than unobstructed PN, at least with access to relevant imaging techniques, because of the accumulation of pus in the renal pelvis. Therefore, there should probably be different clinical criteria developed for diagnosis of pyonephrosis compared to nonobstructive PN.

Overall, few studies described clinical data for dogs and cats diagnosed with PN, and when present, clinical diagnostic criteria for this diagnosis varied and were often extensive. This variation leads to different clinical populations in different studies, making evidence synthesis and generalisation of results difficult. In some studies, PN was further characterised as acute or chronic, but often no distinction between acute or chronic was made. The long-term impact of PN on renal health is unclear. While chronic or subclinical PN could lead to tubulointerstitial damage and contribute to CKD development or progression, prospective data linking clinically diagnosed PN to subsequent decline in kidney function are lacking. Conversely, pre-existing CKD may predispose to PN, making causal direction difficult to determine in retrospective cohorts.

In humans, the situation regarding definition of, and diagnostic criteria for, PN seems similarly unclear [190, 191]. The definitions of human UTIs have been proposed to change from uncomplicated or complicated UTI, respectively, to local (cystitis) or systemic (PN or prostatitis with or without sepsis), the latter requiring more intensive diagnostic evaluation [192]. In recent human guidelines, however, the terms uncomplicated UTI and complicated UTI are still used [193, 194].

The feline studies evaluating the diagnostic value of serum SAA and urinary mRNAs for the diagnosis of PN [22, 85] indicate that there is a perceived need for PN-specific biomarkers for diagnosis. In humans, markers of systemic inflammation such as C-reactive protein and procalcitonin can be used to evaluate involvement of the renal parenchyma. In a Cochrane review from 2020, none of these tests were considered accurate enough to allow clinicians to confidently differentiate upper from lower urinary tract disease in children, mainly because of a low specificity [195]. More specific urinary markers of renal involvement (for example, urinary interleukins) have been evaluated in people but are not yet used in clinical practice [196, 197]. Further studies of sensitive and specific biomarkers for the diagnosis of PN seem to be needed in both human and companion animal medicine.

Ultrasound evaluation of the urinary tract in people with PN is performed to rule out urinary tract abnormalities such as voiding dysfunction, reflux, and renal abscesses or masses [193, 197]. In dogs and cats, ultrasound is also used to rule in PN. The specificity of this diagnostic test is not known but suspected to be low. Two studies compared renal pelvis width among cats with different renal conditions and the overlap was large, supporting further evaluations of the specificity of ultrasound for the diagnosis of PN in dogs and cats.

The choice of antibacterial substance varied across studies included in this review, and this is a controversial area in veterinary medicine. Certain antibacterial classes (for example, fluoroquinolones) are considered more appropriate for the treatment of PN in dogs and cats than, for example, beta-lactams [1]. Interestingly, a recent human systematic review and meta-analysis showed that clinical cure of PN was comparable for different classes of antibacterials [198]. The authors concluded that local resistance patterns, patient factors and toxicity profiles should guide antibiotic selection for PN, rather than presumed class superiority. Ultimately, blood concentrations of the chosen substance need to be high enough to achieve adequate levels in the renal parenchyma to eradicate bacteria.

Across the veterinary literature mapped in this scoping review, reported treatment durations for PN in dogs and cats were commonly 2–6 weeks in observational cohorts and case reports, with no prospective trials comparing different durations. Previous veterinary guidelines recommended treatment durations of 4–6 weeks [199], whereas the current recommendation of antibacterial therapy for PN is 10–14 days [1]. However, prescribing data indicate that shorter courses were already used in some settings before the 2019 ISCAID update. In a North American multicentre study of cats with urinary tract infections from 2016–2018 [180], the median duration of treatment for PN was 14 days with no identifiable association between duration and clinic type or country (USA vs. Canada). In Sweden, this recommendation has been adopted in national guidance since 2022, and a multicentre retrospective cohort from Sweden found that most cats and dogs with PN during the years 2022–2024 were treated with antibacterials for ≤ 14 days, indicating guideline compliance (manuscript in preparation). Antibacterial treatment durations for PN have been investigated in several trials in human medicine. Multiple systematic reviews in adults with acute uncomplicated PN indicate that 5–7 days of treatment are as effective as longer courses [200–202]. Guideline recommendations are regimen-specific, with fluoroquinolones typically given for 5–7 days, whereas trimethoprim-sulfamethoxazole or beta-lactams are often prescribed for 10–14 days [194, 203]. In infants and children, 10–14 days of treatment remains common [197, 204]. Taken together, current veterinary recommendation of 10–14 days aligns with paediatric practice and with adult non-fluoroquinolone regimens but exceeds the shortest effective adult fluoroquinolone courses. Randomised studies comparing shorter versus longer durations (for example, 7 vs 14 days) of treatment for PN, stratified by drug class and complicating factors and using standardised outcome measures and follow-up protocols, are a clear research priority in veterinary medicine. Little is also known about optimal monitoring of individuals treated for PN, and there is no consensus regarding this in veterinary medicine. In children who have recovered clinically from an episode of acute PN, no follow-up urine culture is currently recommended because of lack of evidence to support its utility in clinical practice [197].

One objective of this scoping review was to evaluate the feasibility of a subsequent systematic review and, where appropriate, meta-analysis. Mapping identified a few domains with reasonably comparable data to consider quantitative evidence synthesis. Specifically, a prevalence meta-analysis appears feasible for short-term mortality outcomes in PN cohorts and for the proportion of AKI cases attributable to PN within defined AKI populations. In addition, a few diagnostic studies reported extractable data for potential pooling of clinical, biomarker, and imaging indicators.

The study has some limitations. Because of the language restriction, information from studies written in languages other than English or German could not be included in the review. Also, no attempt to reach grey literature was made, and therefore useful information may have been missed.

Conclusions

Multiple information gaps regarding the clinical aspects of PN in dogs and cats were detected. There are no universally accepted clinical criteria for the diagnosis and no clinical trials evaluating antibacterial (or other) treatment options. Information about short-term outcomes for dogs and cats with PN is scarce, and long-term outcomes do not seem to be investigated at all. Consequently, more research is needed, preferably starting with the development of clinical criteria for diagnosis and core outcomes. Thereafter, controlled clinical trials evaluating different antibacterial treatment options and treatment durations should be performed to improve care of cats and dogs with PN in the future.

Supplementary Information

[内容请查看原文]

[内容请查看原文]

[内容请查看原文]