AbstractObjective

To determine the impact of antibiotic therapy (ATBT) on outcomes of renal cyst infection (CyI) in patients with polycystic kidney disease.

Patients and Methods

We undertook a single-center retrospective study of CyI in autosomal dominant polycystic kidney disease (January 1, 2000, through December 31, 2018). Cyst infections were classified as definite (microbiologically proven), probable (radiologic signs), or possible (clinical or biologic signs only). We studied the determinants of ATBT failure (persistence of infection beyond 72 hours of microbiologically adequate initial ATBT, with requirement for ATBT change, cyst drainage, or nephrectomy) and recurrences (>14 days after the end of ATBT).

Results

Among 90 patients, 139 CyIs (11 definite, 74 probable, 54 possible) were compiled. Cultures were positive in 106 of 139 (76%) episodes, with Escherichia coli found in 89 of 106 (84%). Treatment failures and recurrences within 1 year of follow-up were more frequent in definite/probable CyI (20/85 [34%] and 16/85 [19%]) than in possible CyI (2/54 [4%] and 4/54 [7%]; P<.01 and P=.08, respectively). Male sex (odds ratio [OR], 7.79; 95% CI, 1.72 to 46.68; P<.01), peak C-reactive protein level above 250 mg/L (OR, 7.29; 95% CI, 1.78 to 35.74; P<.01; to convert C-reactive protein values to nmol/L, multiply by 9.524), and cyst wall thickening (OR, 7.70; 95% CI, 1.77 to 43.47; P=.01) but not the modalities of initial ATBT were independently associated with higher risk of failure. In a Cox proportional hazards model, kidney transplant recipients exhibited higher risk of recurrence (hazard ratio, 3.76; 95% CI, 1.06 to 13.37; P=.04), whereas a total duration of ATBT of 28 days or longer was protective (hazard ratio, 0.02; 95% CI, 0.00 to 0.16; P<.001), with an inverse correlation between duration and recurrence (81% for treatment <21 days, 47% for 21 to 27 days, 2% for ≥28 days; P<.0001).

Conclusion

Initial first-line ATBT had no significant effect on renal CyI treatment failure. Treatment duration of 28 days and longer reduced recurrences.

Autosomal dominant polycystic kidney disease (ADPKD), the most common inherited kidney disorder, affects 1 in 1000 to 1 in 2500 live births.1Lanktree M.B. Haghighi A. Guiard E. et al.Prevalence estimates of polycystic kidney and liver disease by population sequencing. It is responsible for the development of numerous cysts arising from various renal tubule segments and leading to the destruction of the renal parenchyma, with end-stage kidney disease typically reported in the sixth decade. Other complications include intracystic bleeding and cyst infection (CyI).2Autosomal dominant polycystic kidney disease. Renal CyI has long been identified as a major complication of ADPKD. It is associated with worsening of kidney function3Cornec–Le Gall E. Audrézet M.P. Rousseau A. et al.The PROPKD score: a new algorithm to predict renal survival in autosomal dominant polycystic kidney disease. and significant mortality, reported in up to 7% per episode in previous studies.4Lantinga M.A. Casteleijn N.F. Geudens A. et al.Management of renal cyst infection in patients with autosomal dominant polycystic kidney disease: a systematic review.As cysts are frequently disconnected from the urinary tract, urine culture is often negative and bacterial documentation is challenging. Cyst puncture is considered the “gold standard” for diagnosis but is rarely performed. In this context, most studies published on CyI have focused on alternative diagnosis methods based on algorithms that include a combination of evocative clinical, biologic, and radiologic signs.5Sallée M. Rafat C. Zahar J.R. et al.Cyst infections in patients with autosomal dominant polycystic kidney disease., 6Lantinga M.A. Drenth J.P.H. Gevers T.J.G. Diagnostic criteria in renal and hepatic cyst infection., 7Suwabe T. Ubara Y. Ueno T. et al.Intracystic magnetic resonance imaging in patients with autosomal dominant polycystic kidney disease: features of severe cyst infection in a case-control study., 8Jouret F. Lhommel R. Devuyst O. et al.Diagnosis of cyst infection in patients with autosomal dominant polycystic kidney disease: attributes and limitations of the current modalities., 9Infected cyst in patients with autosomal dominant polycystic kidney disease: analysis of computed tomographic and ultrasonographic imaging features. Medical imaging procedures can also notably help the clinician distinguish renal CyI, interstitial pyelonephritis, and hepatic CyI, given the high prevalence of concomitant hepatic cysts reported in up to 80% of patients with ADPKD.It has long been suggested that diffusion of antibiotics (ATBs) into the renal cyst would depend more on transepithelial transport than on glomerular filtration.10Polycystic kidney disease: a predominance of giant nephrons.,11Jacobsson L. Lindqvist B. Michaelson G. Bjerle P. Fluid turnover in renal cysts. However, cyst wall properties limit or delay diffusion of most antibacterial drugs, like β-lactams and aminoglycosides.12Bennett W.M. Elzinga L. Pulliam J.P. Rashad A.L. Barry J.M. Cyst fluid antibiotic concentrations in autosomal-dominant polycystic kidney disease., 13Cyst fluid antibiotic concentrations in polycystic kidney disease: differences between proximal and distal cysts., 14Bonadio M. Marino O. Catania B. Giannotti P. Penetration of piperacillin into renal cysts. This is likely to be responsible for a high level of treatment failures and recurrences.4Lantinga M.A. Casteleijn N.F. Geudens A. et al.Management of renal cyst infection in patients with autosomal dominant polycystic kidney disease: a systematic review. One would therefore expect better efficiency of lipid-soluble drugs, such as fluoroquinolone and cotrimoxazole.12Bennett W.M. Elzinga L. Pulliam J.P. Rashad A.L. Barry J.M. Cyst fluid antibiotic concentrations in autosomal-dominant polycystic kidney disease.,13Cyst fluid antibiotic concentrations in polycystic kidney disease: differences between proximal and distal cysts. This hypothesis has never been investigated, however, and data regarding the optimal ATB therapy (ATBT) of these infections are still lacking. We therefore undertook a retrospective study of CyI in patients to study the determinants of treatment failures and recurrences in CyI.Patients and MethodsStudy Population and Data Collection

This retrospective observational study was conducted in the university hospital Necker-Enfants Malades, a reference center for ADPKD in Paris, France. We used a computerized database to screen all cases of renal CyI in patients hospitalized between January 2000 and December 2018, with at least 1 imaging procedure performed at the time of infection. Medical records were reviewed, and relevant data were collected.

DefinitionsAutosomal dominant polycystic kidney disease was diagnosed according to criteria previously described by Cornec–Le Gall et al.15Cornec–Le Gall E. Alam A. Perrone R.D. Autosomal dominant polycystic kidney disease. Diagnostic criteria for renal CyI and CyI classification were adapted from 1 of the more frequently used clinical studies by Sallée et al.5Sallée M. Rafat C. Zahar J.R. et al.Cyst infections in patients with autosomal dominant polycystic kidney disease. Indeed, CyI was suspected in the presence of acute fever (temperature ≥38 °C) for 48 hours or longer and increased C-reactive protein concentration of 50 mg/L (to convert C-reactive protein values to nmol/L, multiply by 9.524) or more (Figure 1). A lower fever cutoff and duration were retained to match the World Health Organization definition of fever and also to address the possibility of milder clinical symptoms reported in the context of maintenance dialysis or kidney transplant.

Figure 1Flowchart. Among 377 episodes of suspected renal cyst infection (CyI) screened in a computerized database between January 2000 and December 2018, 139 cases met the definition of definite CyI (n=11), probable CyI (n=74), and possible CyI (n=54). ADPKD, autosomal dominant polycystic kidney disease.

Cyst infections were classified as definite when a renal cyst aspiration could evidence neutrophil debris (pus) or microorganisms at direct examination or bacterial culture. They were otherwise classified as probable or possible.

Cyst infections were classified as probable when no cyst aspiration could be performed but at least 1 imaging procedure was considered suggestive of CyI.5Sallée M. Rafat C. Zahar J.R. et al.Cyst infections in patients with autosomal dominant polycystic kidney disease.,7Suwabe T. Ubara Y. Ueno T. et al.Intracystic magnetic resonance imaging in patients with autosomal dominant polycystic kidney disease: features of severe cyst infection in a case-control study.,8Jouret F. Lhommel R. Devuyst O. et al.Diagnosis of cyst infection in patients with autosomal dominant polycystic kidney disease: attributes and limitations of the current modalities.,16Jouret F. Lhommel R. Beguin C. et al.Positron-emission computed tomography in cyst infection diagnosis in patients with autosomal dominant polycystic kidney disease.•

Computed tomography (CT) scan or magnetic resonance imaging with thickening of cyst wall, intracystic fluid/fluid level or gas, or pericystic fat infiltration in 1 or more renal cysts

•

18Fluorodeoxyglucose positron emission tomography/CT with focally increased uptake around 1 or more renal cysts (distinct from both physiologic accumulation in parenchyma and pelvicalyceal physiologic excretion)

Cyst infections were classified as possible when those criteria were not retained but all the following clinical criteria could be evidenced:•

Abdominal or lumbar pain

•

Exclusion of other extrarenal cause of febrile abdominal pain

•

Exclusion of renal cyst hemorrhage, except for cases with documented microorganism in blood or urine culture specimens suggestive of secondary infection of cyst hemorrhage

Unlike Sallée et al, we choose to distinguish probable and possible cases of CyI using imaging procedures to better delineate renal CyI and unrelated pyelonephritis8Jouret F. Lhommel R. Devuyst O. et al.Diagnosis of cyst infection in patients with autosomal dominant polycystic kidney disease: attributes and limitations of the current modalities. and to exclude concomitant hepatic CyI, which are frequently reported.

Antibiotic therapy was defined as initial therapy for the empirical therapy introduced at the first day of medical care and follow-up therapy thereafter.

Initial ATBT failure was defined by the persistence of clinical or biochemical signs of infection (fever, sepsis, persisting high C-reactive protein level, persisting positive microbiologic cultures) beyond 72 hours of microbiologically adequate initial ATBT that would require ATBT change, cyst drainage, or partial or total nephrectomy.

Recurrent infection was defined by the recurrence of clinical signs of CyI more than 14 days after the end of the ATBT. All cases with subsequent documentation of a new microorganism were classified as a new episode.

Statistical Analyses

Frequency differences for qualitative variables were compared in χ2 tests with Pearson correction. Mann-Whitney tests were used to compare means between pairs of groups for quantitative variables. All tests were 2 tailed, and a P value of less than .05 was considered statistically significant. Follow-up data were censored at the date of death, loss to follow-up, cyst drainage, or nephrectomy. Multivariate Cox models were used to assess risk factors of recurrent infection at day 365. Logistic regression was used to assess risk factors of initial treatment failure. Models were built using a conditional backward stepwise variable selection process based on variable influence in univariate analysis. Critical entry and exit P values were .20 and .10, respectively. No variable was forced into the final model if it was not previously selected. Correlation and interaction were checked within final models, as were checked assumption for log-linearity of continuous variables and proportional hazard assumptions for survival models. Data are given as odds ratio (OR [95% CI]) or hazard ratio (HR [95% CI]) according to the model. Statistical analyses were performed using R version 3.6.2 (R Foundation for Statistical Computing), survival, cmprisk, survey, ggplot2, and finalfit packages.

Ethical Issues

This study was performed in accordance with the Declaration of Helsinki. According to French law, the study was declared to the French Comité National de l’Informatique et des Libertés (2142688), and given its retrospective nature, it did not require Institutional Review Board approval. As requested in retrospective studies, tacit agreement was collected after written information for all cases.

DiscussionWe here present the results of a large study focusing on renal CyI in ADPKD. The main challenge of CyI studies lies in the fact that definite diagnosis through an invasive procedure is rarely performed. Using definitions (definite, probable, and possible) based on a consensual set of clinical, biochemical, and radiologic variables,5Sallée M. Rafat C. Zahar J.R. et al.Cyst infections in patients with autosomal dominant polycystic kidney disease.,8Jouret F. Lhommel R. Devuyst O. et al.Diagnosis of cyst infection in patients with autosomal dominant polycystic kidney disease: attributes and limitations of the current modalities. we collected 139 cases during an 18-year study period. Important conclusions could be drawn.First, initial treatment failure appeared frequent in our study (16%) although much less than previously reported in the meta-analysis by Lantinga et al4Lantinga M.A. Casteleijn N.F. Geudens A. et al.Management of renal cyst infection in patients with autosomal dominant polycystic kidney disease: a systematic review. (61%). However, this meta-analysis mainly included case reports, with a bias toward higher rate of atypical pathogens and outcome, as reflected by the high reported mortality (7%). In addition, the definition retained in this study encompassed any treatment adaptation regardless of the clinical course (including changes of “treatment modalities”). In the setting of a retrospective review of individual files, we defined more stringent criteria that helped refine the rate of treatment failure. By multivariable logistic regression analysis, we found that male sex, higher peak C-reactive protein levels, and presence of a thickened cyst wall were independently associated with higher risk of treatment failure. Of these parameters, only the biologic inflammatory syndrome had previously been identified by Lantinga et al.17Lantinga M.A. de Sévaux R.G.L. Gevers T.J.G. et al.Clinical predictors of escalating care in hepatic and renal cyst infection in autosomal dominant polycystic kidney and liver disease. Indeed, in a retrospective cohort that mixed 28 hepatic CyIs and 49 renal CyIs, these authors reported that increasing white blood cell count was associated with escalating care, defined by the need of breaching the infected cyst wall (percutaneous access or surgery). The inflammatory syndrome might correlate with the bacterial inoculum and was also identified as an independent risk factor of early treatment failure in women with community-onset acute pyelonephritis.18Wie S.H. Ki M. Kim J. et al.Clinical characteristics predicting early clinical failure after 72 h of antibiotic treatment in women with community-onset acute pyelonephritis: a prospective multicentre study. Lantinga et al also found that atypical pathogen isolation (ie, other than E. coli) was predictive of escalating care. In our study, we also found constant initial treatment failure in the small subset of patients with CyI involving S. aureus or Staphylococcus lugdunensis. This might reflect the propensity for these pathogens to form abscesses (75% in this series). Nevertheless, this association did not reach significance after adjustment for multiple variables. Cyst wall thickening was the only morphologic feature found to be independently associated with treatment failure. Indeed, inflammatory signals might trigger pericystic fibrosis, with extracellular matrix deposition and thickening of cyst basal membrane that could impair ATB diffusion. Interestingly, larger cyst diameter, as suggested by others,5Sallée M. Rafat C. Zahar J.R. et al.Cyst infections in patients with autosomal dominant polycystic kidney disease. was not associated with treatment outcomes. Also, despite theoretical concerns regarding intracyst diffusion, initial ATBT modalities (ie, lipid solubility, bitherapy, intravenous delivery) did not appear to be associated with early CyI outcomes.Second, we also evidenced a 14% rate of recurrent CyI within 1 year of follow-up. This was higher than previously reported in the meta-analysis of Lantinga et al4Lantinga M.A. Casteleijn N.F. Geudens A. et al.Management of renal cyst infection in patients with autosomal dominant polycystic kidney disease: a systematic review. (7%) and suggested that prolonged follow-up might be relevant for patients with CyI.For the first time, we could evidence the benefits of long ATBT duration of 28 days or more, which appeared to be independently associated with a major decrease of subsequent recurrence. Indeed, patients treated for at least 28 days experienced almost no recurrent CyI within 1 year of follow-up. This may reflect the peculiar characteristics of CyI, with a large inoculum and structural properties that hamper diffusion of ATBs. Renal cysts consist of a single layer of flat epithelial cells, with short (100 μm), or high cuboidal epithelial cells, with longer (>50 μm) and tight intercellular junctions (19Composition of fluid in twelve cysts of a polycystic kidney. 1969., 20Huseman R. Grady A. Welling D. Grantham J. Macropuncture study of polycystic disease in adult human kidneys., 21Cuppage F.E. Huseman R.A. Chapman A. Grantham J.J. Ultrastructure and function of cysts from human adult polycystic kidneys. As a result, when Bennett et al12Bennett W.M. Elzinga L. Pulliam J.P. Rashad A.L. Barry J.M. Cyst fluid antibiotic concentrations in autosomal-dominant polycystic kidney disease. studied the concentration of ampicillin inside cysts after daily injections, they evidenced a progressive increase of ampicillin level, which was undetectable after 24 hours of treatment and reached a cyst/blood concentration ratio of 4.6 after 8 days of treatment. Longer treatment would possibly overcome such delayed diffusion of ATBs and hence better suit CyI. A prolonged ATBT is in accordance with management guidelines based on recommendations of CyI experts.22

Jouret F, Hogan MC, Chebib FT. A practical guide for the management of acute abdominal pain with fever in patients with autosomal dominant polycystic kidney disease. Nephrol Dial Transplant. Published online February 11, 2021. https://doi.org/10.1093/ndt/gfab040.

Lipid-soluble ATBs in follow-up therapy, mainly fluoroquinolones and cotrimoxazole, were also associated with a decrease of recurrence in univariate analysis, but this protective effect was no longer significant after multivariable adjustment. A few authors have recommended lipid-soluble ATBs in renal CyI because of their higher ability to cross cell membranes.12Bennett W.M. Elzinga L. Pulliam J.P. Rashad A.L. Barry J.M. Cyst fluid antibiotic concentrations in autosomal-dominant polycystic kidney disease.,13Cyst fluid antibiotic concentrations in polycystic kidney disease: differences between proximal and distal cysts. Indeed, previous case series performed in noninfected patients reported higher renal cyst/blood concentration ratio for lipid-soluble drugs than for non–lipid-soluble ones: 3.9 for trimethoprim (n=8); 2.5 for ciprofloxacin (n=7); 0.96 for levofloxacin (n=1) vs 0.12 for ticarcillin (n=1); 0.04 to 0.13 for meropenem (n=2); 0.26 for gentamicin (n=3); and undetectable cyst level for tobramycin (n=4), amikacin (n=2), and cefotaxime (n=3).12Bennett W.M. Elzinga L. Pulliam J.P. Rashad A.L. Barry J.M. Cyst fluid antibiotic concentrations in autosomal-dominant polycystic kidney disease., 13Cyst fluid antibiotic concentrations in polycystic kidney disease: differences between proximal and distal cysts., 14Bonadio M. Marino O. Catania B. Giannotti P. Penetration of piperacillin into renal cysts.,23Elzinga L.W. Golper T.A. Rashad A.L. Carr M.E. Bennett W.M. Trimethoprim-sulfamethoxazole in cyst fluid from autosomal dominant polycystic kidneys., 24Elzinga L.W. Golper T.A. Rashad A.L. Carr M.E. Bennett W.M. Ciprofloxacin activity in cyst fluid from polycystic kidneys., 25Ohkawa M. Motoi I. Hirano S. Okasho A. Hisazumi H. Biochemical and pharmacodynamic studies of simple renal cyst fluids in relation to infection., 26Hamanoue S. Suwabe T. Ubara Y. et al.Cyst infection in autosomal dominant polycystic kidney disease: penetration of meropenem into infected cysts., 27Hiyama L. Tang A. Miller L.G. Levofloxacin penetration into a renal cyst in a patient with autosomal dominant polycystic kidney disease. This should, however, be pondered with the increased reported diffusion of non–lipid-soluble drugs in infected renal cysts, possibly as a consequence of enhanced pericystic capillary wall permeability. For instance, Ohkawa et al25Ohkawa M. Motoi I. Hirano S. Okasho A. Hisazumi H. Biochemical and pharmacodynamic studies of simple renal cyst fluids in relation to infection. reported that amikacin concentration in cyst fluid after a single 200-mg intramuscular injection was below detection limit in 53 noninfected cysts but reached half the serum values in 2 patients with infected cysts. The potential adverse effects of these drugs in patients with ADPKD, often with chronic kidney failure, and their impact on the gut microbiota and emergence of antimicrobial resistance as observed in our cohort should also be kept in mind.28Wise B.L. Peloquin C. Choi H. Lane N.E. Zhang Y. Impact of age, sex, obesity, and steroid use on quinolone-associated tendon disorders., 29Lee C.C. Lee M.T. Chen Y.S. et al.Risk of aortic dissection and aortic aneurysm in patients taking oral fluoroquinolone., 30Declining susceptibilities of gram-negative bacteria to the fluoroquinolones: effects on pharmacokinetics, pharmacodynamics, and clinical outcomes., 31Suwabe T. Araoka H. Ubara Y. et al.Cyst infection in autosomal dominant polycystic kidney disease: causative microorganisms and susceptibility to lipid-soluble antibiotics. Altogether, this series does not favor the use of lipid-soluble over non–lipid-soluble drugs in follow-up therapy, provided treatment duration is adequate. Prospective studies are now needed to confirm these results and to draw definite conclusions about lipid-soluble ATBs in the prevention of recurrent CyI.Importantly, the risk of treatment failure and recurrent infection was much higher in the patients with definite and probable CyI (ie, proven by cyst puncture or suspected with criteria that included evocative images) than in those with possible CyI (ie, suspected only on the basis of clinical and biochemical signs). This less stringent definition might have also pooled interstitial pyelonephritis, which follows other pathophysiologic patterns and requires shorter treatment duration.32Caron F. Galperine T. Flateau C. et al.Practice guidelines for the management of adult community-acquired urinary tract infections. In accordance with this hypothesis, patients with possible CyI more often had urinary tract infection symptoms, leukocyturia, and positive urinary culture and also less important inflammatory syndrome. The homogeneity of patients with definite and probable infections also provides further evidence of the usefulness of medical imaging in the diagnosis, prognosis, and management of CyI.8Jouret F. Lhommel R. Devuyst O. et al.Diagnosis of cyst infection in patients with autosomal dominant polycystic kidney disease: attributes and limitations of the current modalities.,33Neuville M.F. Lovinfosse P. Jadoul A. et al.The use of a visual 4-point scoring scale improves the yield of 18F-FDG PET-CT imaging in the diagnosis of renal and hepatic cyst infection in patients with autosomal dominant polycystic kidney disease.