Acute renal intracystic hemorrhage in patients with autosomal dominant polycystic kidney disease

Renal cyst bleeding is a frequent problem in patients with autosomal dominant polycystic kidney disease (ADPKD). However, information is still limited on its frequency, causative factors, and effects on enlargement of polycystic kidneys in ADPKD. We investigated the total volume of acute renal intracystic hemorrhage and its association with total kidney volume (TKV) in a large series of patients with ADPKD on dialysis, referred for renal transcatheter arterial embolization. All patients had undergone CT scan and MRI scan before the procedure. We evaluated factors potentially associated with acute renal intracystic hemorrhage. The association between the volume of acute renal intracystic hemorrhage and the potential predisposing and associated factors was analysed by univariable and multivariable regressions. We enrolled 199 patients who underwent renal transcatheter arterial embolization from 2014 to 2018 (107 men, 92 women; mean age 59.1 ± 8.6 years). The median volume of acute renal intracystic hemorrhage was 97.3 ml (interquartile range 36.6–261.7 ml). Multivariable analysis revealed that body weight, kidney stones, systolic blood pressure, and total volume of acute renal intracystic hemorrhage were significantly associated with TKV; age, body mass index, smoking, renal cyst infection, serum alkaline phosphatase, and TKV were significantly associated with the volume of acute renal intracystic hemorrhage ; and sex, age, dialysis vintage, TKV, and total volume of acute renal intracystic hemorrhage were significantly associated with the number of microcoils required to achieve renal transcatheter arterial embolization. Total volume of acute renal intracystic hemorrhage was significantly associated with TKV (r = 0.15, p = 0.0325) and was greater in younger patients (r= − 0.32, p < 0.0001). Total volume of acute renal intracystic hemorrhage was also correlated with the number of microcoils required for renal transcatheter arterial embolization (r = 0.23, p = 0.0012). Acute renal intracystic hemorrhage is frequent among ADPKD patients on dialysis, and total volume of acute renal intracystic hemorrhage significantly associated with TKV. Total volume of acute renal intracystic hemorrhage was greater in younger patients with higher renal artery luminal size. These results suggest that renal cyst bleeding and renal artery blood flow may synergistically accelerate the enlargement of polycystic kidneys in ADPKD patients on dialysis.


Introduction
Autosomal dominant polycystic kidney disease (ADPKD) is the most frequently inherited kidney disease. ADPKD is characterized by the progressive development and growth of cysts that cause enlargement and distortion of the kidneys and impair renal function. This results in end-stage kidney disease (ESKD) in many patients. It is the fourth leading cause of ESKD in adults worldwide [1][2][3]. Generally, deterioration of renal function is proportional to the increase in kidney size, and many patients with ADPKD have renal dysfunction and massive kidneys. A recent populationbased study estimated the clinically significant incidence of ADPKD as 3.06 per 100,000 person-years [4].
The renal cysts of ADPKD show various features in imaging studies. To our knowledge, there have been few discussions on the implications of these features in ADPKD, and their associations with prognosis of ADPKD remain unknown. The occurrence of renal intracystic hemorrhage in ADPKD has long been known, although it has yet to be resolved [5]. Riyahi et al. reported that hemorrhagic renal cysts, which show high intensity in T1 weighted imaging (T1WI) using magnetic resonance, are recognized frequently, and their number correlates significantly with prognosis in patients with early ADPKD [6]. However, high intensity on T1WI is not specific to acute intracystic hemorrhage, and it is difficult to differentiate from other conditions such as infected cysts [7,8]. Additionally, the T1WI intensity of chronic intracystic hemorrhage can be similar to that of acute hemorrhage, making it difficult to differentiate the two conditions by T1WI [9]. Compared with magnetic resonance imaging (MRI), computed tomography (CT) scanning is a superior imaging modality for identifying acute hemorrhage [7], and the combination of contrast-enhanced CT and unenhanced MRI is currently considered the most reliable strategy for diagnosing acute renal intracystic hemorrhage (ARIH) in patients with ESKD. Furthermore, the frequency of ARIH in patients with ADPKD and ESKD has not been reported. Therefore, we conducted this study to investigate the association between the total volume of ARIH (identified by contrast-enhanced CT and unenhanced MRI) and total kidney volume (TKV). We evaluated potential factors that affect ARIH in patients with ADPKD and ESKD. This is the first study to clarify the volume of ARIH and its association with total kidney volume and age in patients with ADPKD and ESKD. We aimed to determine the influence of renal cyst bleeding on enlargement of polycystic kidneys and propose a new mechanism of renal cyst growth.

Methods
This was a retrospective, single-center study designed to investigate the association between ARIH and total kidney volume and the age of patients with ADPKD on dialysis. This retrospective study was reviewed and approved by the ethics committee of Toranomon Hospital in June 2020. We studied the patients who underwent renal transcatheter arterial embolization (TAE) at our hospital, because the number of patients was large, and all of them had symptomatic enlarged polycystic kidneys. Additionally, before renal TAE, most of them underwent contrast-enhanced CT and unenhanced MRI, which confirmed the diagnosis of acute renal intracystic hemorrhage.

Patients
All patients with ADPKD who underwent renal TAE at Toranomon Hospital from January 2014 through December 2018 were selected in this study. These patients were identified from the database of the Toranomon Hospital Department of Nephrology (Kawasaki, Japan), which was updated by a research assistant each time renal TAE was performed. All patients who underwent renal TAE were confirmed from the medical records of Toranomon Hospital (Tokyo and Kawasaki, Japan). All patients were adults (≥ 20 years) who met the criteria for diagnosis of ADPKD described by Pei et al. [10] and by the Progressive Renal Disease Research report from the Ministry of Health, Labour and Welfare of Japan (Supplementary Table S1). Candidates for renal TAE were patients who were on maintenance dialysis and had compression symptoms related to enlarged polycystic kidneys, including dysphagia, gastroesophageal reflux, early satiety, marked changes in bowel habits, dyspnea, and orthopnea. All patients accepted to undergo renal TAE and provided written consent after being fully informed about the procedure, including potential complications such as fever, pain, and anuria. The procedures were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration. We excluded patients who had previously undergone nephrectomy, renal TAE, or renal cyst drainage. We also excluded patients who did not undergo contrast-enhanced CT and unenhanced MRI.

Clinical and laboratory assessments
Clinical features, including height and weight, medical history, and smoking history, were recorded before performing renal TAE. Dry weight was used as the body weight of dialysis patients. Systolic blood pressure (BP) was measured with an automatic device in the sitting position, and the average of three systolic BP readings measured in the morning on a nondialysis day was used for analysis. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Laboratory tests were performed before renal TAE. Blood tests were performed at the start of dialysis in patients on hemodialysis. All laboratory tests were performed by automated standardized methods at our hospital within 24 h of collecting blood samples.

Imaging studies
Abdominal contrast-enhanced CT and unenhanced MRI were routinely performed in all patients just prior to renal TAE. CT was performed with a 16-MDCT scanner and MRI was performed using a 1.5-T apparatus using a previously reported method [7]. All patients were on dialysis. Therefore, gadolinium contrast agent for MRI was contraindicated and unenhanced MRI was performed. Kidney volumes were determined using Vincente software, version 4 (Fujifilm Co., Tokyo, Japan) by a single group of medical staff. ARIH was defined as a cyst containing a poorly demarcated mass with high CT density (30-100 Hounsfield units [HU]) without enhancement by contrast media at CT imaging, as well as with high intensity in T1WI and diffusion weighted magnetic resonance image (DWI) studies. The volume of ARIH (30-100 HU) was also determined using Vincente software, version 4 by a single group of medical staff as shown in Fig. 1. MR images of the same slice were referenced, and Image of polycysƟc kidney on CT Renal intracysƟc area with CT density (30-100 HU) in the right kidney Fig. 1 An example of the measurement of the volume of acute intracystic hemorrhage using Vincente software, version 4. The renal intracystic area with CT density (30-100 HU) (shown in green) was determined for each slice. The total volume of acute intracystic hemorrhage in each kidney was calculated by summing the green areas with CT density (30-100 HU) in all slices of each kidney   regions without high intensity in T1WI and DWI examinations were manually removed. We had previously compared and reported the CT density of bleeding cysts, normal cysts, and infected cysts in patients with ADPKD. The most useful cutoff value to distinguish them was 25 HU [7]. However, to increase the specificity of diagnosis in this study, we defined the CT density of acute intracystic hemorrhage as 30-100 HU. Generally, the CT densities of hyperacute, acute, and subacute hematomas are reported to be approximately 30-100 HU [11]. It has been reported that CT is the optimal method for its evaluation, although MRI may supplement CT findings because it detects more hemorrhagic cysts than SD standard deviation, IQR interquartile range (25-75%), LMW heparin low molecular weight heparin 11 (9,14) 10 (7, 13) 12 (9, 15) 11 (9, 13) 11 (8,14) 0.0719 CT and helps distinguish them from carcinomas [12]. We consider imaging studies combined with contrast-enhanced CT and unenhanced MRI to be the most reliable imaging tools for the definitive diagnosis of ARIH of the kidneys in patients with ESKD. Their combination can distinguish other diseases such as renal carcinoma and hemangioma.

Renal TAE procedure
The procedure for renal TAE has changed since 1996 [13], although it has not been altered since January 2006 (including the microcoils used) [14]. After the femoral artery was cannulated, aortography was performed using a pigtail catheter, and selective renal artery angiography was performed using a shepherd's hook catheter. A microcatheter was inserted into the peripheral branches of the renal artery by employing a guide wire. Platinum microcoils were then advanced with a pusher. We used one type of platinum microcoil during the study period. Microcoils are covered by health insurance for arterial embolization in Japan, and we are familiar with performing renal TAE using microcoils as the embolic material. To avoid recanalization, coils were inserted as peripherally as possible into small renal artery branches, and both peripheral and proximal renal artery occlusion were performed as completely as possible on both sides simultaneously (Supplementary Figure S1). It is desirable to embolize as many branches of the vessels supplying both kidneys as is feasible, including additional renal arteries and capsular arteries. This is because high blood flow in any remaining arteries increases the risk of hemorrhage secondary to volume overload.

Statistical analysis
Clinical features and laboratory data of patients at renal TAE were used for analysis. Normally distributed baseline variables are presented as mean ± standard deviation, and non-normally distributed numeric baseline variables are presented as median and interquartile range. Univariate and multivariate regression models were used to analyze factors with an influence on TKV, ARIH volume, and the number of microcoils used for renal TAE. Predictive variables for the multivariable analysis were chosen using the stepwise selection method. Because of skewed variance, logarithmic transformation was performed for data on TKV, ARIH volume, prothrombin time-international normalized ratio, serum aspartate aminotransferase, serum alanine aminotransferase, serum alkaline phosphatase (ALP), serum gamma-glutamyl transferase, and serum C-reactive protein (CRP). A probability (p) value of less than 0.05 was considered to indicate significance. To graphically evaluate the association between TKV and ARIH volume, ARIH volume and age, the number of microcoils and age, and the number of microcoils and ARIH volume, a scatter plot and regression line were created. Logarithmic transformation was performed for data on ARIH volume and TKV on the scatter plot. All analyses were performed using SAS software, version 9.3 (SAS Institute Inc., Cary, NC, USA), and p < 0.05 was considered to indicate significance.

Results
A total of 226 dialysis patients underwent renal TAE for symptomatic renomegaly from January 2014 through December 2018. Among them, 27 patients were excluded vor various reasons reported, in Fig. 2. The remaining 199 patients were analyzed: 107 men and 92 women with mean age of 59.1 ± 8.6 years ( Table 1). The median volume of ARIH was 97.3 ml and the interquartile range (25-75%) was 36.6-261.7 ml.
Univariate analysis revealed that sex, height, body weight, BMI, systolic BP, antiplatelet medicine, ARIH volume, and number of microcoils required to achieve renal TAE were significantly associated with TKV (Table 2). ARIH volume differed significantly among the four groups stratified according to ARIH quartile volume (Table 1). Logarithmic ARIH volume was weakly correlated with logarithmic TKV (r = 0.15, p = 0.0325) (Fig. 3). Multivariate analysis with stepwise elimination confirmed that body weight, systolic BP, kidney stones, and ARIH volume were significantly associated with TKV (Table 2).
Univariate analysis revealed that age, height, BMI, smoking, diabetes mellitus, renal cyst infection, number of microcoils required to achieve renal TAE, and TKV were significant factors for ARIH volume (Table 3). Age was inversely correlated with logarithmic ARIH volume (r = −0.32, p < 0.0001) (Fig. 4). Multivariate analysis with stepwise elimination confirmed that age, BMI, smoking, renal cyst infection, serum ALP, and TKV were significant factors for ARIH volume (  Fig. 3 The correlation between log (acute intracystic hemorrhage volume) and log (total kidney volume). Log (total kidney volume) was significantly greater in patients with greater log (acute intracystic hemorrhage volume) (r = 0.15, p = 0.0325) partial thromboplastin time, serum CRP, TKV, and ARIH volume were significant factors for the number of microcoils. Age was inversely correlated with number of microcoils (r = − 0.25, p = 0.0004) (Fig. 5). Multivariate analysis with stepwise elimination confirmed that sex, age, dialysis vintage, TKV, and ARIH volume were significant factors for the number of microcoils ( Table 4). The number of  Figure S2).

Discussion
ARIH was common among patients with ADPKD on dialysis, and ARIH was independently associated with TKV. Furthermore, ARIH volume was greater in younger patients with ADPKD on dialysis. These results may suggest that renal intracystic bleeding accelerates polycystic kidney enlargement. Moreover, ADPKD progression was faster in patients who have experienced greater renal cyst bleeding. This observation is consistent with the studies by Riyahi et al., which showed that renal cysts with high intensity on T1WI were correlated with faster progression of ADPKD [6]. Approximately 42% to over 50% of patients with ADPKD experience at least one episode of gross hematuria. Gross hematuria may be an initial diagnostic clue in up to 20% of patients with ADPKD [15]. Microhematuria is often observed in patients with ADPKD. We frequently see patients who have elevated serum CRP, even though they do not have any infection. These observations suggest that renal cyst bleeding is more frequent in ADPKD than previously expected, and cyst bleeding may play an important role in progression of ADPKD. In fact, in this study, only four patients (2.0%) presenting with a suggestive clinical picture had no acute intracystic hemorrhage in either kidney. However, all of the study patients were on dialysis, and anticoagulant therapy in hemodialysis could increase the rate of cystic bleeding. It is therefore necessary to investigate the frequency of ARIH in non-dialysis patients. Renal cysts in ADPKD reportedly have various appearances on MRI, and this is presumably because of frequent cyst bleeding. The phase of bleeding differs considerably among renal cysts, which may cause various appearances on MRI. It may be important to prevent renal cyst bleeding to slow the progression of ADPKD. Renal intracystic bleeding increases the tension inside the cysts, which leads to increased cyst size and can cause an increase in overall kidney volume. However, the cause of cyst bleeding remains unknown. It is also unknown whether renal cyst bleeding results from environmental factors, genetic factors, or both. It is necessary to clarify the causative factors and mechanism of renal cyst bleeding and its role in the progression of ADPKD. Noninherited factors such as hypertension and a urologic event before 35 years of age were reported to be significant factors in influencing prognosis of ADPKD, in addition to genetic factors, which may also influence cyst bleeding [16]. Hypertension is generally a risk factor for bleeding such as intracerebral hemorrhage [17][18][19]. Likewise, hypertension may also increase intracystic hemorrhage in ADPKD, and hypertension and intracystic hemorrhage may synergistically influence progression of ADPKD. In fact, rigorous BP control was reported to be effective in slowing the progression of ADPKD [20]. However, contrary to our expectation, BP and antihypertensive medicine were not significantly associated with ARIH volume in this study. Additionally, interestingly, average systolic BP and the rate of patients on antihypertensive medicine were not different among four age groups stratified by age quartiles (Supplementary Table S2).
These results suggest that, rather than BP, inherited factors may influence the ease of bleeding, such as the size of the renal artery. The mechanism underlying renal cyst bleeding has not been proposed or discussed, and further research is needed on this issue. Other environmental factors may also influence cyst bleeding in ADPKD. Rigorous exercise, such as martial arts, can cause renal intracystic bleeding and should be avoided. Therefore, guidelines on exercise for patients with ADPKD should be developed. Antiplatelet or anticoagulant medications may be risk factors for cyst bleeding, although they were not statistically significant factors for acute intracystic hemorrhage in this study. Future studies are needed to determine the influence of these medications in patients with ADPKD. The number of microcoils used for renal TAE was independently associated with patient age. More microcoils were used in younger patients. Additionally, the number of microcoils for renal TAE was correlated with the volume of ARIH. We considered that the number of microcoils may reflect renal artery luminal volume. Therefore, these results suggest that progression of ADPKD is faster in patients who have greater renal artery luminal volume, and renal cyst bleeding is more likely to occur in such patients. Additionally, dialysis vintage was shorter in younger age groups than in older age groups (Supplementary Table S2), implying that the luminal volume was larger in younger patients with a shorter dialysis duration, and renal artery narrowing occurs with aging. This result is consistent with that of our previous study and with a report that renal blood flow decreases with age in patients with ADPKD [14,21]. Renal artery luminal volume was also correlated with the volume of ARIH. Therefore, renal artery size and renal cyst bleeding may be associated and synergistically affect ADPKD progression. We previously reported that renal artery size is substantially increased in patients with ADPKD, and an abundant renal blood supply may be important for the growth of polycystic kidneys [13,14]. Clarifying the mechanism associated with cyst bleeding and renal artery size is a subject for future investigation.
This study had limitations. First, this is a single-center retrospective study. Second, all enrolled patients were of Japanese ethnicity, which may reduce the generalizability of the findings. Third, this study analyzed the association between the number of microcoils required to achieve renal TAE and patient age. We used the same type of microcoils for renal TAE in all patients, although the number of microcoils may not correctly represent the renal artery luminal volume. Furthermore, this was a study in patients on dialysis. Anticoagulant therapy in hemodialysis patients could be a potential source of bias and could have influenced the rate of cystic bleeding. Therefore, additional studies in nondialysis patients are required to verify the association between renal cyst bleeding and prognosis of ADPKD.
This study had some strengths. ARIH was diagnosed using both contrast-enhanced CT and unenhanced MRI, which when combined are more accurate for diagnosing acute hemorrhage than enhanced CT alone. We measured the volume of intracystic hemorrhage, which may be more accurate for reflecting the amount of bleeding, compared with the number of cysts with high intensity in T1WI magnetic resonance studies as described previously [6]. The number of enrolled patients was large enough to allow detailed statistical analysis. Furthermore, ARIH volume and TKV were measured by a single researcher using specific, validated software, which minimized measurement error.
In conclusion, this study revealed that ARIH volume was significantly associated with TKV in patients with ADPKD on dialysis, and ARIH volume was significantly greater in younger patients on with shorter dialysis vintage. The number of microcoils for renal TAE was also significantly greater in younger patients than in older patients. These results suggest that renal cyst bleeding and renal artery size synergistically accelerate the enlargement of polycystic kidneys in patients with ADPKD. Further research is needed on this new proposed mechanism of accelerated renal cyst growth through cyst bleeding.