Kidney Transcatheter Arterial Embolization Reduces Kidney Cyst Infection in Patients With ADPKD

Abstract

Introduction

Kidney cyst infections are serious complications in autosomal dominant polycystic kidney disease (ADPKD). Kidney transcatheter arterial embolization (TAE) reduces kidney volume in this population. We tested the hypothesis that TAE could prevent kidney cyst infections.

Methods

This retrospective cohort study included 416 patients with ADPKD who underwent kidney TAE (primary cohort), including 70 patients with a previous kidney cyst infection. The primary outcome was the annual frequency of kidney cyst infections. Secondary outcomes included hospitalization duration. The risk factors for post-TAE infection recurrence were identified by multivariable linear regression.

Results

In the primary cohort, the frequency of cyst infections decreased from 0.291 episodes per year pre-TAE to 0.087 post-TAE (P < 0.001). In the subgroup with a previous infection, the frequency decreased from 1.73 ± 1.18 to 0.49 ± 0.96 episodes per year (P < 0.001), and hospitalization days decreased from 34.1 ± 26.1 to 9.90 ± 24.4 days per year (P < 0.001). The reduction effect was pronounced in patients with a greater total kidney volume reduction rate (TKV-RR). No infections were observed in patients who achieved a TKV-RR > 75%. Multivariable analysis revealed pre-TAE infection frequency (β = 0.23; P = 0.017), a lower TKV-RR (β = −0.01; P = 0.034), and antibiotic use at TAE (β = 0.68; P = 0.007) as independent risk factors for post-TAE infection recurrence. Even among patients who consistently used antibiotics, the infection frequency decreased post-TAE.

Conclusion

Kidney TAE is a promising technique for reducing kidney cyst infections in patients with ADPKD. TAE should be performed after full remission without antibiotic therapy, but TAE reduces cyst infection even in consistent antibiotic users. Early, extensive embolization might be particularly beneficial in patients at high-risk.

Graphical abstract

ADPKD is a prevalent inherited kidney disorder in which cyst infection is a common and serious complication.¹ Cyst infections occur in approximately 30%–50% of patients with ADPKD, accounting for 11% of all hospitalizations.^2,3 Kidney cyst infection in patients with ADPKD is considered to occur predominantly through hematogenous spread, although ascending urinary tract infection may also contribute because pyuria has been observed in a subset of cases.⁴ Even if appropriate antibiotics are introduced, cyst infections can be resistant, possibly because of the closed nature of kidney cysts, which hinders the delivery of intravascular antimicrobial agents into infected cysts.^3,4 Prolonged antibiotic therapy is an effective strategy for preventing the recurrence of kidney cyst infection, and some patients are on prophylactic antibiotics because of frequent relapses.⁵ However, the long-term administration of antibiotics is closely associated with antimicrobial resistance, and antibiotic-resistant bacteria are highly prevalent in cases of cyst infection.⁴

In 1996, we developed kidney TAE surgery for patients with ADPKD to shrink enlarged kidneys.^6,7 Kidney TAE is the transcatheter embolization of the kidney artery and its branches. It is primarily indicated for patients with ADPKD who are undergoing hemodialysis and have already stopped producing urine. Since then, we have reported the outstanding effectiveness of this procedure in this population from multiple aspects: it significantly reduces kidney size, improves nutritional status⁷ and quality of life,⁸ and increases survival rates.⁹

We have reported cases suggesting that kidney TAE may reduce the frequency of kidney cyst infections in patients with ADPKD.^10,11 In this context, we hypothesized that the TAE procedure may also reduce the frequency of kidney cyst infections by blocking hematogenous infection and reducing potential infectious foci. However, these observations are based on a limited number of cases, and it remains unclear whether they can be generalized. Furthermore, we also reported a case in which a kidney cyst infection became refractory after TAE, possibly due to impaired antibiotic delivery to the kidney.¹²

Therefore, we conducted this study to test the hypothesis that kidney TAE reduces the frequency of kidney cyst infections and identify risk factors for post-TAE infection recurrence in patients with ADPKD.

Methods

Study Design and Population

To test our hypothesis that kidney TAE reduces kidney cyst infections in patients with ADPKD, a retrospective cohort study was planned to compare the frequency of kidney cyst infections before and after kidney TAE in patients with ADPKD.

Candidates for kidney TAE included patients on maintenance dialysis who exhibited symptomatic compression from enlarged polycystic kidneys, such as dysphagia, gastroesophageal reflux, early satiety, marked changes in bowel habits, dyspnea, or orthopnea. Kidney TAE was indicated when the enlarged polycystic kidneys were considered to contribute to the compressive symptoms. Because kidney TAE typically results in anuria, the procedure was usually performed only after patients’ urine output had declined below 500 ml per day. The expected loss of residual kidney function was explained to all patients, and the decision to proceed was made only when the expected clinical benefits outweighed this disadvantage. To exclude active infection, all patients were confirmed to have no fever for at least 1 month before the procedure, and no elevation of serum C-reactive protein (CRP) or leukocyte counts from their baseline values. The procedure for kidney TAE was previously described.⁷

We included patients with ADPKD who had undergone kidney TAE at Toranomon Hospital Kajigaya (Kanagawa, Japan) between January 1, 2011, and December 31, 2022. Our institutions is a tertiary referral center that receive patients from across the country specifically for kidney TAE. The diagnosis of ADPKD was based on the criteria of Pei et al.¹³ Patients were excluded if they met at least 1 of the following criteria: (i) had not been followed for 1 year before and 1 year after kidney TAE, (ii) had undergone nephrectomy within 1 year after kidney TAE, (iii) had undergone partial (including unilateral) kidney TAE, and (iv) had previously undergone kidney TAE. Regarding criterion (iii), complete embolization was confirmed on angiography at the end of TAE procedure in all cases. Patients meeting these criteria constituted the primary cohort. Because kidney TAE as a preventive intervention is indicated primarily for patients with recurrent cyst infections, we additionally defined a subgroup of patients who had experienced at least 1 kidney cyst infection during the year preceding kidney TAE and performed a focused analysis in this subgroup.

Patients were followed up from 1 year before the date of kidney TAE until the earliest date of all-cause death or censoring (censoring for loss to follow-up or administrative censoring occurring 1 year after the date of kidney TAE).

This study involved human participants and was approved by the internal ethics review board of Toranomon Hospital (approval number: 2565-B). The study was performed in accordance with the Declaration of Helsinki.

Diagnosis and Definition of Kidney Cyst Infection Episodes

Although definitive diagnosis of cyst infection requires cyst aspiration, this is not routinely performed in all cases. Therefore, the diagnosis of kidney cyst infection was made based on the criteria we previously reported,¹⁴ requiring at least 1 of the following abdominal magnetic resonance imaging (MRI) or computed tomography (CT) findings: gas inside the cyst, high intensity on diffusion-weighted imaging, fluid–fluid level, or cyst wall thickening, in addition to fever, abdominal pain, back pain, or tenderness localized to the cyst, serum CRP elevation, and no other source of fever detectable. Both MRI and CT studies were performed for all cases, enabling the exclusion of other potential infection sources and differentiation from intracystic hemorrhage. Acute intracystic hemorrhage was distinguished from cyst infection primarily on the basis of CT density because hemorrhagic cysts show high CT density (30–100 Hounsfield units) and often contain an irregular intracystic mass (hematoma).^14,15 In addition, sequential changes in the imaging before and after symptom onset were used to localize the infected cyst. Laboratory and clinical findings, including white blood cell count >10,000/μl, serum CRP > 15.0 mg/dl, and body temperature > 38° C, were also considered because they strongly suggested cyst infection rather than hemorrhage.¹⁴ Only cases requiring hospitalization were included in the analysis. Outpatient cases were excluded because cross-sectional imaging is often not performed, making it impossible to confirm the diagnosis of kidney cyst infection accurately or to localize the infection to the kidney. Some outpatient cases are managed both at our institutions and referring clinics, resulting in incomplete follow-up data in our institutions, and the precise onset and resolution of infection episodes are frequently unclear. A single hospitalization due to kidney cyst infection was considered 1 incidence of kidney cyst infection. Recurrence of kidney cyst infection was defined as a new hospitalization for kidney cyst infection during the observation period.

Management of Patients With Consistent Antibiotics Use

As described above, some patients have difficulty discontinuing antibiotics due to frequent cyst infection relapse; therefore, we sometimes perform kidney TAE for such patients to reduce the enlarged kidney volume. Another reason is that kidney TAE may prevent kidney cyst infection.⁵ When kidney TAE is performed under consistent antibiotic use, the procedure is undertaken only if the patient has remained afebrile for at least 1 month while receiving antibiotics. Moreover, we continued antibiotic therapy for at least 4 weeks after kidney TAE in patients with refractory cyst infections. We follow these rules because we have observed severe emphysematous kidney cyst infection just after kidney TAE when a cyst infection had not been controlled sufficiently.¹² We define “consistent antibiotics use” as continuous antibiotic therapy from the last episode of cyst infection before TAE until the time of TAE. We also examined the efficacy and safety of kidney TAE in patients who consistently used antibiotics.

Data Collection and Variables

Clinical data were retrieved from the medical records. The collected data included baseline patient demographics (age, sex, height, body weight, blood pressure, comorbidities, medications, symptoms, and hemodialysis), blood examination findings (albumin, aspartate aminotransferase, alanine aminotransferase, CRP levels, and leukocyte count), imaging findings (abdominal ultrasound scan, CT, and MRI), the number of coils used in kidney TAE, and the incidence of kidney cyst infections during the follow-up period. Sex was defined as biological sex recorded in the medical records.

Imaging Studies and Calculation of the TKV-RR

Patients who underwent kidney TAE were routinely followed at our hospital for 1 year and were instructed to visit the hospital promptly if infection was suspected. As a tertiary referral center, our institution follows patients for 1 year after kidney TAE, because kidney volume reduction plateaus within this period and careful monitoring including dry weight adjustment is required during this time. Patients are subsequently transferred back to their referring physicians; therefore the follow-up period was set to 1 year. Abdominal CT scans were routinely performed in all patients before and at 12 months after kidney TAE. CT scans were performed with a 16 multidetector row CT scanner Aquilion 16 (Toshiba, Tokyo, Japan) using a previously reported method.¹⁴

Based on the CT images, the TKV was calculated using SYNAPSE VINCENT (Fujifilm, Tokyo, Japan).

The 12-month TKV-RR was calculated as follows:

$$\text{TKV}-\text{RR}(\%)=\left(1-\frac{\mathrm{T}\text{KV}\text{at}12\text{months}\text{after}\text{kidney}\text{TAE}}{\text{TKV}\text{before}\text{kidney}\text{TAE}}\right)\times 100$$

Patients were stratified into 4 groups based on their TKV-RR: less than 25%, 25%–50%, 50%–75%, and > 75%.

The MRI method for detecting kidney cyst infections was described previously.¹⁴

Outcomes

The primary outcome of this study was the annual frequency of kidney cyst infections in the year after kidney TAE compared with the year before the procedure.

The secondary outcomes included the annual hospitalization period due to kidney cyst infection and the maximum CRP level during hospitalization due to kidney cyst infection.

Statistical Analysis

First, the baseline characteristics of the primary cohort and the subgroup were quantified. Continuous variables are presented as medians and interquartile ranges or means and standard deviations, as appropriate. Categorical variables are presented as numbers and percentages.

Second, we evaluated the changes in the annual frequency of kidney cyst infections before and after kidney TAE in the primary cohort, both overall and across TKV-RR groups. In addition, the annual frequency of liver cyst infection before and after kidney TAE was also examined. Furthermore, as an exploratory analysis, the association between kidney and liver cyst infections was assessed using Fisher exact test and the chi-square test for trend.

Third, we performed a focused analysis in the subgroup of patients with a history of kidney cyst infection. In this subgroup, we evaluated the changes in TKV, the annual frequency of kidney cyst infections, the annual hospitalization period due to kidney cyst infection, and the maximum CRP value during each hospitalization. We compared these outcomes between the TKV-RR groups, and a 2-group analysis (TKV-RR < 50% vs. ≥ 50%) was also performed to assess the robustness of the findings with larger subgroups.

Finally, to identify factors associated with the recurrence of kidney cyst infection after kidney TAE in this subgroup, we performed univariable and multivariable linear regression analyses using the frequency of post-TAE kidney cyst infections as the dependent variable. Explanatory variables included patient demographics (age, male sex, and diabetes), laboratory parameters (serum albumin and CRP levels), hemodialysis duration, pre-TAE infection burden (frequency of kidney cyst infection during the year before TAE and the duration between the last episode of kidney cyst infection and the TAE procedure), and TAE procedural variables (number of coils used, TKV before and after TAE, TKV-RR, and antibiotic use at the time of TAE). For the multivariable analysis, each explanatory variable of interest was included in a separate regression model adjusted for 4 covariates: age, male sex, diabetes, and the frequency of kidney cyst infections during the year before TAE. These variables were a priori chosen based on the clinical relevance and potential confounding effects to facilitate comparisons of effect sizes.

A P-value < 0.05 was considered to indicate statistical significance.

Statistical analyses were performed using Stata/MP version 16.1 (StataCorp, College Station, TX).

Results

Study Population Selection

Figure 1 summarizes the process of study population selection. Among the 538 patients initially included, 122 met exclusion criteria for the following reasons: (i) 83 patients had insufficient follow-up period, (ii) 8 patients had undergone nephrectomy, (iii) 26 patients had partial kidney TAE, and (iv) 5 patients had a history of kidney TAE. Therefore, the primary cohort comprised 416 patients. Among them, 70 patients had experienced at least 1 kidney cyst infection during the year preceding kidney TAE and constituted the subgroup with a previous infection history. The number of patients in each TKV-RR category in the primary cohort and the subgroup was as follows: TKV-RR < 25%, 44 and 8; 25%–50%, 190 and 33; 50%–75%, 157 and 23; and > 75%, 25 and 6, respectively. Representative CT images showing pre- and post-TAE kidney appearance for each TKV-RR category are shown in Supplementary Figure S1. Patients who underwent nephrectomy after TAE had already been scheduled for nephrectomy before the procedure, and kidney TAE was performed to reduce the kidney volume and create space for the transplanted kidney. No patient died during the observation period. No patient required an increase in the dialysis frequency after TAE.

Figure 1.

Flow chart of study population selection. TAE, transcatheter arterial embolization.

Baseline Characteristics

Table 1 presents the baseline characteristics of the patients of the primary cohort. All patients were on HD via arteriovenous fistula. A greater TKV-RR was associated with younger age and a greater number of coils used for kidney TAE. Table 2 presents the baseline characteristics of the subgroup with an infection history. In this subgroup, 65 (92.9%) patients were anuric at the time of kidney cyst infection. Follow-up duration was 1 year for all patients before TAE and 364.3 ± 2.12 days after TAE. Supplementary Table S1 shows the baseline characteristics of the patients stratified by whether kidney cyst infection decreased after kidney TAE. Patients whose kidney cyst infection decreased after kidney TAE had smaller TKVs and larger TKV-RRs.

Table 1.

Baseline characteristics of primary cohort

Characteristic	Total (n = 416)	< 25% (n = 44)	25%–50% (n = 190)	50%–75% (n = 157)	> 75% (n = 25)
Age, yr	58.8 ± 8.9	62.3 ± 7.9	61.0 ± 8.7	56.0 ± 8.5	52.9 ± 8.0
Male, n (%)	204 (49.0)	20 (45.5)	109 (57.4)	64 (40.8)	11 (44.0)
BMI, kg/m2	22.5 ± 3.2	22.6 ± 3.2	22.8 ± 3.0	22.4 ± 3.5	21.2 ± 2.6
Corrected BMI, kg/m2	20.8 ± 3.1	21.1 ± 3.0	20.9 ± 2.9	20.7 ± 3.4	19.4 ± 2.5
Ever smoker, n (%)	157 (37.7)	16 (36.4)	69 (36.3)	62 (39.5)	10 (40.0)
Hypertension, n (%)	383 (92.1)	40 (90.9)	177 (93.2)	142 (90.4)	24 (96.0)
Diabetes, n (%)	18 (4.3)	3 (6.8)	9 (4.7)	6 (3.8)	0 (0.0)
Hemodialysis vintage, yr	5 (2–10)	9 (5–14)	6 (3–11)	4 (2–7)	2 (1–5)
Anuria, n (%)	310 (74.5)	39 (88.6)	156 (82.1)	104 (66.2)	11 (44.0)
Alb, g/dl	3.30 (3.10–3.60)	3.40 (3.10–3.60)	3.30 (3.10–3.60)	3.30 (3.10–3.50)	3.40 (3.10–3.50)
AST, U/l	11 (9–15)	12 (9–15)	12 (9–15)	11 (9–14)	11 (9–14)
ALT, U/l	8 (6–11)	8 (6–12)	8 (6–11)	8 (6–11)	7 (5–11)
Alkaline phosphatase, U/l	186 (126–273)	177 (106–270)	190 (126–275)	186 (129–261)	179 (143–281)
γ-glutamyl transpeptidase, U/l	30 (18–58)	24 (19–57)	34 (19–59)	30 (19–57)	21 (16–53)
Cholinesterase, U/l	204 (174–238)	196 (181–230)	206 (178–240)	203 (171–233)	210 (180–245)
CRP, mg/dl	0.2 (0.0–0.6)	0.2 (0.1–0.9)	0.2 (0.1–0.5)	0.1 (0.0–0.6)	0.3 (0.0–1.0)
WBC, ×103/μl	4.70 (3.90–5.62)	4.55 (3.80–5.20)	4.70 (3.90–5.70)	4.80 (3.90–5.60)	5.50 (4.60–5.90)
Immunosuppressive drugs use, n (%)	9 (2.2)	0 (0.0)	6 (3.2)	2 (1.3)	1 (4.0)
Number of coils	37 (30–46)	32 (24–44)	37 (29–44)	37 (30–48)	44 (32–56)
TKV pre-TAE, ml	4114 (2860–6260)	4308 (2268–5339)	4410 (2942–6703)	3794 (2695–5783)	4575 (3448–5896)
TKV post-TAE, 12 mo, ml	2070 (1251–3322)	3594 (1899–4858)	2628 (1830–3946)	1493 (1014–2338)	928 (687–1249)
TKV-RR (%)	47.2 (34.5–59.5)	18.2 (13.0–21.6)	39.2 (33.1–44.2)	59.6 (54.8–66.0)	79.1 (76.7–81.5)
TLV, ml	2505 (1493–4600)	2055 (1639–2973)	2436 (1458–4672)	2802 (1574–4973)	2114 (1655–3680)

Alb, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CRP, C-reactive protein; TKV, total kidney volume; TKV-RR, total kidney volume reduction rate; TLV, total liver volume; WBC, white blood cell count; yr, year(s).

All patients were on hemodialysis via arteriovenous fistula. Corrected BMI was calculated by subtracting the estimated cyst weight (TKV in ml, assuming a cyst fluid density of 1 g/ml) from body weight before BMI calculation.

Table 2.

Baseline characteristics of subgroup with previous kidney cyst infections

Characteristic	Total (n = 70)	TKV-RR < 25% (n = 8)	TKV-RR 25–50% (n = 33)	TKV-RR 50–75% (n = 23)	TKV-RR > 75% (n = 6)
Age, yr	58.9 ± 10.1	60.9 ± 10.2	62.8 ± 7.25	55.3 ± 11.0	48.3 ± 9.03
Male, n (%)	46 (65.7)	8 (100.0)	22 (66.7)	12 (52.2)	4 (66.7)
BMI, kg/m2	22.5 ± 3.31	23.5 ± 3.03	22.4 ± 2.76	22.2 ± 4.23	23.0 ± 2.87
Corrected BMI, kg/m2	20.7 ± 3.0	21.7 ± 2.3	20.5 ± 2.4	20.7 ± 3.9	21.0 ± 3.1
Ever smoker, n (%)	28 (40.0)	5 (62.5)	14 (42.4)	7 (30.4)	2 (33.3)
Hypertension, n (%)	64 (91.4)	8 (100)	31 (93.9)	20 (87.0)	5 (83.3)
Diabetes, n (%)	2 (2.86)	0 (0)	0 (0)	1 (4.3)	1 (16.7)
Hemodialysis vintage, yr	8 (3–13)	9 (5–12)	9 (4–15)	6 (2–12)	2 (1–6)
Anuria, n (%)	65 (92.9)	8 (100)	32 (97.0)	21 (91.3)	4 (66.7)
Alb, g/dl	3.05 (2.80–3.40)	3.00 (2.95–3.17)	3.10 (2.90–3.40)	3.00 (2.70–3.40)	2.90 (2.67–3.20)
AST, U/l	13 (10–16)	14 (10–16)	13 (10–17)	11 (10–14)	14 (12–20)
ALT, U/l	7 (5–11)	8 (6–9)	7 (5–11)	7 (6–10)	10 (6–14)
Alkaline phosphatase, U/l	174 (116–268)	294 (110–406)	174 (113–245)	164 (119–262)	188 (155–367)
γ-glutamyl transpeptidase, U/l	32 (19–62)	44 (22–83)	24 (18–61)	33 (16–54)	57 (27–90)
Cholinesterase, U/l	198 (163–230)	204 (163–231)	201 (193–236)	185 (158–222)	154 (135–173)
CRP, mg/dl	0.34 (0.10–1.25)	0.52 (0.20–3.10)	0.29 (0.10–0.80)	0.30 (0.07–1.60)	2.35 (1.24–4.08)
WBC, ×103/μl	4.75 (3.90–5.67)	4.55 (3.72–5.62)	4.70 (3.90–5.70)	4.70 (3.90–5.60)	5.55 (4.82–6.65)
Duration from the discharge date of the last kidney cyst infection before TAE to the date of the kidney TAE, days	68 (43–116)	101 (56–132)	66 (44–96)	91 (42–120)	45 (23–60)
Kidney cyst drainage for the last kidney cyst infection before TAE, n (%)	10 (14.3)	1 (12.5)	3 (9.09)	4 (17.4)	2 (33.3)
Immunosuppressive drugs use, n (%)	2 (2.86)	0 (0)	0 (0)	1 (4.3)	1 (16.7)
Antibiotics use, n (%)	23 (32.9)	3 (37.5)	13 (39.4)	5 (21.7)	2 (33.3)
Number of coils	38 (30–50)	30 (27–33)	36 (28–43)	41 (34–54)	58 (40–65)
TKV pre-TAE, ml	4482 (2965–6348)	4410 (2997–6358)	4664 (3455–6372)	3806 (2828–5521)	5612 (3192–7741)
TKV post-TAE, 12 mo, ml	2026 (1396–3507)	3875 (2639–5138)	3047 (1951–4156)	1493 (1174–2012)	900 (690–1379)
TKV-RR (%)	46.4	15.5	37.9	60.4	80.7
TLV, ml	2244 (1687–3489)	2102 (1728–2732)	2235 (1662–3405)	2342 (1597–3531)	2700 (1841–4167)

Alb, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CRP, C-reactive protein; mo, month; TAE, transcatheter arterial embolization; TKV, total kidney volume; TKV-RR, total kidney volume reduction rate; TLV, total liver volume; WBC, white blood cell count; yr, year(s).

All patients were on hemodialysis via arteriovenous fistula. Corrected BMI was calculated by subtracting the estimated cyst weight (TKV in ml, assuming a cyst fluid density of 1 g/ml) from body weight before BMI calculation.

Supplementary Table S2 shows the baseline characteristics of the patients stratified by whether they experienced at least 1 kidney cyst infection within the year after kidney TAE. Patients who experienced kidney cyst infection within the year after kidney TAE were older and had a higher rate of antibiotic use, a larger TKV at 12 months post-TAE, and a lower TKV-RR. All antibiotic agents administered at the time of kidney TAE were initiated during the last episode of kidney cyst infection before kidney TAE.

Overall Analysis

In the primary cohort, the annual frequency of kidney cyst infection decreased from 0.291 to 0.087 episodes per year after kidney TAE (P < 0.001). The frequency of liver cyst infection also decreased from 0.286 to 0.129 episodes per year (P = 0.020) (Supplementary Table S3). In an exploratory analysis, patients with kidney cyst infections were significantly more likely to also have liver cyst infections (21.4% vs. 2.6% [odds ratio = 10.21]) (P < 0.001), with a dose-dependent relationship (kidney infection 0 episodes: 2.6%; 1 episode: 16.7%; and 2 episodes: 29.4% [P-value for trend < 0.001]).

Subgroup With a Prior Infection History

Figure 2 summarizes the results. The 12-month TKV-RR (Figure 2a) was 46.4% (pre-TAE: 4931 ± 2520 ml and post-TAE: 2645 ± 1778 ml) (P < 0.001). The annual frequency of kidney cyst infection during the 1-year period pre- and post-TAE (Figure 2b) was 1.73 ± 1.18 per year and 0.49 ± 0.96 per year, respectively, showing a significant decrease after kidney TAE (P < 0.001). Infection frequency decreased significantly as early as the first 6 months after TAE compared with the year before TAE (annualized rate: 1.73 vs. 0.60 episodes per year) (P < 0.001). The annual hospitalization periods due to kidney cyst infection during the 1-year period before and after kidney TAE (Figure 2c) were 34.1 ± 26.1 days per year and 9.90 ± 24.4 days per year, respectively, showing a significant decrease after kidney TAE (P < 0.001). The maximum CRP value during each hospitalization due to kidney cyst infection during the year pre- and post-TAE (Figure 2d) was 16.8 ± 9.61 mg/dl and 13.3 ± 7.73 mg/dl, respectively (P = 0.34). Among the subgroup, 65 out of 70 patients were anuric at the time of the first episode of kidney cyst infection during the study period. Of the 5 patients with residual urine output, urinalysis was performed for all 7 episodes of kidney cyst infection before TAE, of which 1 was positive for pyuria.

Figure 2.

Results of the subgroup analysis. The TKV, frequency of kidney cyst infections per year, annual hospitalization period due to kidney cyst infection, and maximum CRP values during hospitalization were compared between the 1-year periods pre- and post- TAE. The TKV significantly decreased (P < 0.001). The frequency of cyst infections and hospitalization days per year also significantly decreased after kidney TAE (both P < 0.001). No significant difference was observed in the maximum CRP values. CRP, C-reactive protein; TAE, transcatheter arterial embolization; TKV, total kidney volume; TKV-RR, TKV reduction rate.

Blood cultures were obtained in all patients at the time of infection; 13 of 70 patients (18.6%) had positive blood cultures in the last pre-TAE episode, most commonly Escherichia coli (n = 5) and Klebsiella pneumoniae (n = 5). Urine cultures were routinely performed in patients with residual urine output, and all were negative.

A total of 47 patients underwent kidney TAE while maintaining remission from kidney cyst infection without antibiotic use. Among them, 6 patients experienced 9 episodes of kidney cyst infection within the year after TAE. All 9 infections occurred in different cysts from the last pre-TAE infection based on MRI findings. In contrast, 23 patients required consistent antibiotic use at the time of TAE. Among them, 9 patients developed 16 episodes of kidney cyst infection within the year after TAE. Of these, 5 infections occurred in the same cyst as the last pre-TAE infection on MRI, suggesting the possibility of recurrent infection, whereas the remaining 11 infections occurred in different cysts from the last pre-TAE infection.

As a supplemental analysis, we also performed a 2-group analysis (TKV-RR < 50% vs. ≥ 50%) because the TKV-RR < 25% and > 75% groups contained relatively few patients. Both groups showed a significant decrease in the frequency of kidney cyst infection and hospitalization period post-TAE. Furthermore, the post-TAE infection frequency and hospitalization period were significantly lower in the TKV-RR ≥ 50% group than in the TKV-RR < 50% group (Supplementary Table S4).

Relationship Between TKV-RR and Kidney Cyst Infection in the Subgroup

Hereafter, patients were stratified into 4 groups based on their 12-month TKV-RR: < 25% (n = 8), 25%–50% (n = 33), 50%–75% (n = 23), and > 75% (n = 6). Figure 3 summarizes the results. The TKV-RRs of each group were 15.5% (pre-TAE: 4611 ± 2370 ml and post-TAE: 3916 ± 2030 ml), 37.9% (pre-TAE: 5240 ± 2650 ml and post-TAE: 3268 ± 1803 ml), 60.4% (pre-TAE 4361 ± 2267 ml and post-TAE 1714 ± 1035 ml), and 80.7% (pre-TAE 5840 ± 2983 ml and post-TAE 1091 ± 537 ml), respectively. The annual frequency of kidney cyst infection pre- and post-TAE in each group was 1.13 ± 0.35 and 0.50 ± 0.53 (P = 0.049), 1.88 ± 1.29 and 0.79 ± 1.19 (P < 0.001), 1.78 ± 1.28 and 0.17 ± 0.65 (P < 0.001), and 1.50 ± 0.55 and 0 (P < 0.001), respectively. The annual hospitalization periods due to kidney cyst infection before and after kidney TAE in each group were 20.1 ± 13.7 days per year and 8.00 ± 18.6 days per year (P = 0.0085), 35.0 ± 31.3 days per year and 17.5 ± 32.2 days per year (P = 0.016), 35.6 ± 21.8 days per year and 2.17 ± 8.35 days per year (P < 0.001), and 42.0 ± 18.8 days per year and 0 days per year (P < 0.001), respectively.

Figure 3.

Relationship between TKV-RR and kidney cyst infection. The effects of kidney TAE on TKV, kidney cyst infection frequency, and hospitalization duration were compared between the 1-year period pre- and post-TAE in 4 groups based on their 12-month TKV-RR: <25%, 25%–50%, 50%–75%, and >75%. Significant reductions in kidney cyst infection frequency and hospitalization periods were observed in all groups, with more pronounced effects in patients who achieved greater TKV-RR. Notably, no kidney cyst infections occurred post-TAE in the group with a TKV-RR >75%. TAE, transcatheter arterial embolization; TKV, total kidney volume; TKV-RR, TKV reduction rate.

Regardless of the TKV-RR, both the frequency of kidney cyst infections and the number of hospitalization days due to kidney cyst infections decreased after kidney TAE. However, the effect was more pronounced in patients with a higher TKV-RR, and notably, no kidney cyst infections occurred after TAE in the group with a TKV-RR > 75%.

Factors Associated With Post-TAE Recurrence of Kidney Cyst Infection

Table 3 summarizes the results of the regression analysis. In the univariable analysis, age (regression coefficient [β] = 0.02 [95% confidence interval [CI], 0.001–0.05]) (P = 0.043), pre-TAE frequency of cyst infection (β = 0.27 [95% CI, 0.09–0.46]) (P = 0.0044), TKV at 12 months after kidney TAE (β = 0.0001 [95% CI, 6.15 e⁻⁷ to −0.0003]) (P = 0.049), TKV-RR (β = −0.01 [95% CI, −0.03 to −0.003]) (P = 0.017), and antibiotic use at the time of TAE (β = 0.83 [95% CI, 0.38–1.28]) (P < 0.001) were significantly associated with post-TAE recurrence of kidney cyst infection.

Table 3.

Factors associated with the recurrence of cyst infection after kidney TAE in the subgroup with a history of infection

Variable	Univariable			Multivariable
	β	95% CI	P-value	β	95% CI	P-value
Age, yr	0.02	0.001–0.05	0.043a	0.02	−0.002 to 0.04	0.079
Male (%)	0.23	−0.25–0.71	0.34	0.43	−0.04 to 0.90	0.08
Diabetes (%)	1.04	−0.32 to 2.40	0.13	1.21	−0.09 to 2.51	0.07
Duration of hemodialysis, yr	0.005	−0.03 to 0.04	0.78	−0.02	−0.06 to 0.02	0.36
Alb, g/dl	0.08	−0.43 to 0.59	0.76	0.01	−0.46 to 0.49	0.96
CRP, mg/dl	−0.008	−0.10 to 0.08	0.86	0.06	−0.04 to 0.15	0.23
Frequency of kidney cyst infection during the yr periods before kidney TAE, /yr	0.27	0.09–0.46	0.0044a	0.23	0.04 to 0.41	0.017a
Duration from the discharge date of the last kidney cyst infection before TAE to the date of the kidney TAE, d	−0.001	−0.004 to 0.002	0.57	0.001	−0.002 to 0.004	0.51
Number of coils	−0.005	−0.02 to 0.008	0.46	0.003	−0.01 to 0.02	0.77
TKV pre-TAE, ml	0.00004	−0.00005 to 0.0001	0.40	0.00006	−0.00003 to 0.0002	0.20
TKV post-TAE, 12 mo, ml	0.0001	6.51e-07 –0.0003	0.049a	0.0001	4.89e-06 –0.0003	0.042a
TKV-RR (%)	−0.01	−0.03 to −0.003	0.017a	−0.01	−0.03 to −0.001	0.034a
Antibiotics use (%)	0.83	0.38–1.28	< 0.001a	0.68	0.19–1.17	0.007a

Alb, albumin; CRP, C-reactive protein; d, day; mo, month; TAE, transcatheter arterial embolization; TKV, total kidney volume; TKV-RR, total kidney volume reduction rate.

^aP < 0.05 was considered statistically significant.

In the multivariable model adjusted for age, male sex, diabetes, and pre-TAE cyst infection frequency, pre-TAE cyst infection frequency (β = 0.23 [95% CI, 0.04–0.41]) (P = 0.017), TKV-RR (β = −0.01 [95% CI, −0.03 to −0.001]) (P = 0.034), and antibiotic use at the time of TAE (β = 0.68 [95% CI, 0.19–1.17]) (P = 0.007) were significantly associated with post-TAE recurrence of cyst infection.

Efficacy and Safety of Kidney TAE in Consistent Antibiotics Users

Since consistent antibiotic use was identified as a risk factor for post-TAE recurrence of kidney cyst infection, we conducted an evaluation of the efficacy and safety of kidney TAE specifically in consistent antibiotics users.

The TKV-RR of the group (Supplementary Figure S2A) was 56.6% (pre-TAE: 4905 ± 2577 ml and post-TAE: 2777 ± 1727 ml). The annual frequency of kidney cyst infection pre- and post-TAE (Supplementary Figure S2B) was 2.48 ± 1.50 and 1.04 ± 1.30 (P = 0.0004). The annual hospitalization periods due to kidney cyst infection pre- and post-TAE (Supplementary Figure S2C) were 49.1 ± 29.6 days per year and 18.7 ± 27.7 days per year (P = 0.0001). The maximum CRP value during each hospitalization due to kidney cyst infection during the year before and after kidney TAE (Supplementary Figure S2D) was 17.4 ± 10.9 mg/dl and 12.9 ± 8.71 mg/dl, respectively (P = 0.33).

Among consistent antibiotics users (as well as among all patients included in this study), no cases of severe cyst infections such as emphysematous kidney cyst infection were observed after kidney TAE.

Discussion

This study investigated the association between kidney TAE and kidney cyst infection in patients with ADPKD, using the largest cohort of studies on this topic. Our findings can be summarized as follows:

First, both the frequency of kidney cyst infection and the hospitalization duration due to kidney cyst infection decreased after TAE. In the primary cohort, the incidence of kidney cyst infection significantly decreased after kidney TAE, indicating that the benefit was not confined to patients with an infection history and that reducing kidney blood supply through TAE does not exacerbate the risk of infection. The infection frequency in the subgroup with an infection history exceeds the reported lifetime prevalence of 30%–50% in the general ADPKD population,^2,3 reflecting the severely enlarged kidneys of patients indicated for TAE.

Because kidney cyst infections are considered to occur predominantly through hematogenous bacterial spread, blood flow obstruction by TAE likely contributed to this reduction. This is supported by the significant decrease in infection frequency as early as the first 6 months after TAE, before nutritional improvement is expected to occur.⁷ Nutritional improvement after kidney volume reduction may also have contributed synergistically in the longer term.

In an exploratory analysis that should be interpreted cautiously, liver cyst infection frequency also decreased after kidney TAE. Patients with kidney cyst infections were significantly more likely to experience liver cyst infections, with a dose-dependent relationship, suggesting a biological link plausibly mediated by bacteremia, consistent with the shared bacteriological profiles previously reported.⁴ However, this analysis was not prespecified, the number of liver cyst infection episodes was small, and confounding factors such as changes in antibiotic therapies or nutritional improvement cannot be excluded.

Second, the infection reduction was pronounced in patients with larger TKV-RRs. Notably, no infections occurred post-TAE in the TKV-RR > 75% group. Consistent with TKV-RR being an independent predictor of recurrence, patients with post-TAE infections had significantly larger 12-month TKVs than those who remained infection-free. These findings support that both perfusion reduction and volume shrinkage prevent infection, aligning with previous hepatic cyst research.¹⁶ Therefore, maximizing shrinkage through sufficient embolization is critical. The variability in TKV-RR despite complete embolization is influenced by patient-related factors, including cyst wall thickening from previous hemorrhage or infection, arteriosclerosis, and interstitial fibrosis associated with advanced age and longer dialysis duration.¹⁷ Considering that younger patients with shorter dialysis durations achieve greater shrinkage, early intervention may particularly benefit high-risk individuals with frequent previous infections.

Third, patients who developed kidney cyst infections after TAE had a higher rate of consistent antibiotic requirements at the time of TAE. Antibiotic use at TAE was independently associated with increased post-TAE recurrence of cyst infection. When TAE is performed while antibiotics are still required for infection suppression, embolization may impair antibiotic delivery to the infected cyst, leading to reactivation in the same focus.

Although these results suggest that any infection should ideally be completely resolved and the patient should be antibiotic-free before undergoing TAE, our clinical data also show the practical benefit of TAE even in antibiotic-dependent cases. In clinical practice, some patients cannot avoid taking antibiotics because cyst infections recur immediately upon discontinuation. Previous case reports have described emphysematous kidney cyst infections after kidney TAE when the procedure was performed before the infection was sufficiently controlled.¹² Such severe cyst infection can occur just after kidney TAE because the kidney blood flow decreases after TAE and antibiotics are unlikely to reach the infected cysts through the blood flow. This is the most crucial factor that we consider when we perform kidney TAE in patients with refractory kidney cyst infection. However, no such severe progression was observed in our cohort. This suggests that the requirement of our protocol for consistent antibiotic users to be afebrile for at least 1 month before performing TAE ensures procedural safety. Furthermore, although consistent antibiotic use was identified as a risk factor for recurrence compared with nonuse, we revealed that these high-risk patients still experienced a significant reduction in cyst infection. Although achieving an antibiotic-free state is ideal, kidney TAE provides substantial reduction in cyst infection even for consistent antibiotic users.

As an alternative to kidney TAE, bilateral nephrectomy is a potential intervention for refractory cyst infections. However, the mortality rate after bilateral open nephrectomy has been reported to be high (3% operative mortality rate and 7% 1-year death rate).^18,19 Moreover, bilateral nephrectomy carries a significant risk of intractable postoperative hypotension due to the sudden and complete loss of the kidney-derived renin-angiotensin system.²⁰ In contrast, kidney TAE is a less invasive bilateral intervention that enables comprehensive infection prevention across both kidneys. Notably, none of the patients enrolled in this study died during the observational period, even though patients with ADPKD requiring consistent antibiotics often have poor performance status and limited surgical candidacy. This approach is more physiologically agreeable than nephrectomy. Its gradual TKV and blood pressure reduction avoids the intractable hypotension triggered by sudden depletion of the renin-angiotensin system.

A key strength of this study is the inclusion of one of the largest cohorts of patients with ADPKD who underwent kidney TAE. All patients were followed for 1 year after the procedure using a standardized protocol, ensuring consistent data collection and minimizing information bias. Kidney cyst infections were diagnosed based on uniform criteria, and the infected cyst was identified by MRI, enabling accurate and reproducible diagnosis of kidney cyst infections. These factors contributed to the reliable evaluation of the frequency of cyst infections. In addition, precise and serial assessment of the TKV using uniform imaging methodology allowed for robust quantification of the TKV-RR. Taken together, the sample size, standardized diagnostic and follow-up framework, and high-quality imaging data provide methodological rigor and enhance the generalizability of our findings.

This study has several limitations. First, owing to its retrospective and observational design without a non-TAE control group, causal inference is limited, although the within-subject comparison partially mitigates between-patient confounding. Second, only hospitalized cases were included, which may select for more severe infection episodes. Third, although we applied standardized diagnostic criteria with mandatory imaging findings, including CT density-based differentiation from intracystic hemorrhage, definitive diagnosis of cyst infection by cyst aspiration was not performed. Fourth, pathogen-specific analysis was challenging because of frequent negative cultures. Fifth, the 1-year follow-up may not capture longer-term outcomes, including whether the infection-preventive effect of TAE is sustained beyond this period. Sixth, hematogenous and ascending infections could not be fully distinguished. Although pyuria was rare, most of the patients were already anuric, precluding urinalysis to exclude ascending infection. Nevertheless, the substantial reduction in kidney cyst infection frequency after TAE suggests that hematogenous infections predominated in this cohort.

In conclusion, kidney TAE was associated with a significant reduction in kidney cyst infections, particularly in patients who achieved substantial TKV reduction. Kidney TAE is also one of the treatment options for patients with refractory kidney cyst infection in terms of reducing kidney volume and reducing recurrence. Kidney TAE may be a valuable intervention for preventing kidney cyst infections in this patient population.

Disclosures

This study was supported by a grant-in-aid for Progressive Renal Disease Research from the Ministry of Health, Labour and Welfare of Japan and the Okinaka Memorial Institute for Medical Research, Tokyo, Japan. The authors declare no competing interests.

Supplementary Material

Supplementary File (PDF)

Figure S1. Representative CT images showing pre- and post-TAE kidney appearance for each TKV-RR category. Figure S2. Efficacy and safety of kidney TAE in consistent antibiotics users. Table S1. Baseline characteristics of patients stratified by whether kidney cyst infection decreased after kidney TAE. Table S2. Baseline characteristics of the patients stratified by whether they experienced at least one kidney cyst infection within the first year after kidney TAE. Table S3. Annual frequency of liver cyst infection before and after kidney TAE. Table S4. Two-group analysis of kidney cyst infection by TKV-RR (< 50% vs. ≥ 50%).