Abdominal Radiology (2019) 44:2835–2840 https://doi.org/10.1007/s00261-019-02017-z

KIDNEYS, URETERS, BLADDER, RETROPERITONEUM

Clinical signifcance of incidentally discovered renal cysts in pediatric patients

Ariel Botwin1 · Teerasak Phewplung1,2 · Kaiming Wu1 · Ruth Lim1 · Avram Z. Traum3 · Michael S. Gee1

Published online: 10 April 2019 © Springer Science+Business Media, LLC, part of Springer Nature 2019

Abstract Purpose To determine the clinical signifcance of incidentally discovered renal cysts in pediatric patients and identify imag- ing predictors of autosomal dominant polycystic disease (ADPKD). Methods A retrospective search of radiology reports from 2000 to 2016 was performed to identify patients < 18 years old with an imaging exam identifying at least one renal cyst and a ≥ 1-year follow-up renal imaging exam for cyst evaluation and/or diagnosis of ADPKD. Cysts with clear solid mass components were excluded. Results 84 pediatric patients with renal cysts were identifed (mean age, 9.5 years), including 76 patients with incidentally discovered cysts and 8 patients with cysts identifed from screening for ADPKD family history. Among the incidentally discovered cyst group, 7.9% were found to have ADPKD compared with 100% of patients with cysts and ADPKD family history. Maximum cyst diameter was signifcantly increased in patients with ADPKD compared to patients without ADPKD (22.0 mm vs 12.7 mm; P < 0.001, Fisher’s Exact test). Multiple cysts or bilateral cysts were imaging features associated with a signifcantly higher (P < 0.01) incidence of ADPKD, both for the entire study population and the incidentally discovered cyst group. An increase in cyst size on the follow-up study was associated with higher incidence of ADPKD (P < 0.05). No malignancies were identifed. Conclusions Incidentally discovered renal cysts in pediatric patients are associated with a small but non-zero risk of ADPKD. Among cyst characteristics, bilaterality, multiplicity, large size, and increased size on follow-up imaging were associated with a statistically signifcant elevation of ADPKD risk, and should prompt diagnostic evaluation.

Keywords Incidental · Renal cyst · Pediatric · Imaging · Autosomal dominant polycystic kidney disease · Renal malignancy · Ultrasound

Introduction

Renal cysts are among the most common incidentally dis- covered lesions on imaging in both children and adults Ariel Botwin and Teerasak Phewplung have contributed equally to [1]. While incidental renal cysts in adults are often clini- this work. cally insignifcant, their presence is more noteworthy in * Michael S. Gee the pediatric population due to potential associations with [email protected] cystic renal tumors and polycystic kidney disease (PKD). Autosomal dominant PKD (ADPKD) is the most com- 1 Department of Radiology, Massachusetts General Hospital and Harvard Medical School, 55 Fruit Street, Boston, mon inherited renal cystic disease, leading to progressive MA 02114, USA cyst development and renal enlargement associated with 2 Department of Radiology, King Chulalongkorn Memorial development of abdominal pain, hypertension, and renal Hospital, The Thai Red Cross Society, Faculty of Medicine, impairment [2]. Other less common hereditary causes of Chulalongkorn University, 1873 Rama IV Road, Pathum renal cyst formation in children include Wan, Bangkok 10330, Thailand complex, Bardet–Biedl syndrome, Beckwith–Wiedemann 3 Department of Pediatrics, Boston Children’s Hospital syndrome, and Juvenile Nephronophthisis, along with sev- and Harvard Medical School, 300 Longwood Avenue, eral others [3]. Non-hereditary causes, such as multicystic Boston, MA 02115, USA

Vol.:(0123456789)1 3 2836 Abdominal Radiology (2019) 44:2835–2840 dysplastic kidney, simple cyst, and cysts associated with Data collection long-term dialysis, may be observed as well [3, 4]. Ultra- sound (US) plays an important role in screening children Demographic information including patient age, gender, and with a family history of ADPKD, with cysts on imaging whether patients had a family history of ADPKD was col- considered part of disease diagnosis [2, 5]. However, it lected. The imaging modality in which renal cysts were frst is unclear what the likelihood of ADPKD is for inciden- detected and the indication for these studies were also col- tally discovered cysts in the general pediatric population. lected. Several renal cyst features were recorded on initial Although ADPKD is typically inherited from a parent with imaging. These included whether cyst(s) were unilateral or known family history, there is highly variable disease pen- bilateral, single or multiple, and simple or complex. Simple etrance and family members might not be aware of the cysts were defned as homogeneously anechoic lesions with presence of the disease [6]. In addition, sporadic ADPKD posterior acoustic enhancement on US and fuid-attenuation can arise due to the presence of disease-causing mutations or fuid-signal intensity lesions with a sharp interface with the in patients with no family history [7]. adjacent renal parenchyma on computed tomography (CT) or Only a few studies in the literature focus on the clini- magnetic resonance imaging (MRI), respectively. Cysts with cal prognosis of renal cysts in the pediatric population. A thin internal septations or complex fuid (echogenicity on US, 1991 study demonstrated no clinical signifcance of simple attenuation of greater than 20 Hounsfeld units on CT without renal cysts in children discovered by US, although patients enhancement, or high T1-weighted signal intensity without with a family history of ADPKD were specifcally excluded enhancement on MRI) were recorded as complex cysts. Any from this analysis [8]. Another study from 2008 focusing cysts with thick or enhancing septa, nodular/solid components, on the risk of malignancy associated with complex renal or increased Doppler vascularity were excluded from the study. cysts in pediatric patients found a high malignancy risk only The maximum cyst diameter based on caliper measurement in complex cysts with thick or irregular septa, a thickened in centimeters (cm) was recorded for both initial and follow-up outer wall, or nodular or solid internal components [9]. To imaging. Renal cyst size on follow-up imaging was classifed our knowledge, there are no published data that assess the as either decreased, stable, or increased. All images in this incidence of ADPKD in pediatric patients with incidentally study were reviewed separately, at which time renal cysts were discovered renal cysts. measured and other renal cyst features were recorded. ADPKD The purpose of this study was to (1) determine the inci- genetic testing and status was also retrieved from the electronic dence of both ADPKD and cystic renal tumors arising from medical records along with renal biopsy/histology results. The incidentally discovered renal cysts in pediatric patients with- reference standard for the diagnosis of ADPKD was based out suspicious features and (2) identify renal cyst features on genetic testing. Patients with a positive family history as associated with ADPKD. well as the development of multiple renal cysts on subsequent follow-up imaging were considered positive for ADPKD in the absence of genetic testing. Patients were also considered negative for ADPKD in the absence of genetic testing if they Methods had a single cyst that either was stable or decreased in size on follow-up imaging, and had no family history of ADPKD as Patients documented in the medical record. All patients had pediatric nephrology consultation. This study was approved by the institutional review board (IRB) and a waiver of informed consent was obtained. Statistical analysis A Health Insurance Portability and Accountability Act (HIPAA) compliant retrospective search of radiology reports Fisher’s exact test was used to assess for statistical signif- from 2000 to 2016 was performed. The keywords used for cance of imaging feature proportions (Stata version 14.2, Col- the query were renal cyst, renal cystic tumor, and renal mass. lege Station, TX). A P value of less than 0.05 was considered Patients under 18 years of age with at least one renal cyst signifcant. identifed on imaging were identifed. The electronic medical record was then queried to identify those patients who had a follow-up imaging study at least one year afterward to assess for cyst stability and change in size. Of note, patients who had a diagnosis of ADPKD made after an imaging study demonstrating a renal cyst were included in the study, even if a follow-up imaging study depicting the renal cyst was not available.

1 3 Abdominal Radiology (2019) 44:2835–2840 2837

Table 1 Patient demographics, imaging characteristics, and renal cyst Results size data Total number of patients 84 Patient demographics and imaging data

Males:Females 38:46 A total of 84 pediatric patients (38 males, 46 females) with Mean age in years (range) 9.9 (2 days to 18 years) renal cysts were included in this study. The mean age of the Indication for study study sample was 9.9 years (range 2 days to 17.9 years). Of Abdominal pain 24 (28.6%) these patients, 76 had incidentally discovered renal cysts Source of infection 17 (20.2%) and 8 had cysts identifed during ADPKD screening because Prenatal abnormality 12 (14.3%) of a family history of the disease. A total of 42 patients Evaluation for congenital abnormality 9 (10.7%) were excluded from the initial query due to a lack of follow- Screening for FH of PKD 8 (9.5%) up imaging. The most common imaging study indication Malignancy workup 5 (6.0%) was abdominal pain (28.6%), followed by infection (20.2%) HTN 3 (3.6%) and prenatal abnormality (14.3%). US (79.8%) was the Trauma 2 (2.4%) most common imaging modality for initial cyst identifca- Initial imaging modality tion, with CT (17.9%) and MRI (2.4%) much less common US 67 (79.8%) (Table 1). CT 15 (17.9%) MRI 2 (2.4%) Renal cyst clinical outcomes Mean time between studies (months) 36.8 + 29.1 Change in cyst size on second imaging study One cystic tumor was identifed and found to be multilocular Decrease 28 (37.8%) cystic nephroma by histology (Fig. 1). No malignant tumors Stable 40 (54.1%) were identifed in this study. A total of 14 patients (16.7%) Increase 6 (8.1%) were diagnosed with ADPKD, including 10 patients who FH family history, PKD polycystic kidney disease, HTN hyperten- were diagnosed following the frst imaging study without sion, US ultrasound, CT computed tomography, MRI magnetic reso- follow-up imaging. Of these, 8/14 cases were confrmed by nance imaging genetic testing, while 6/14 cases were diagnosed based on positive family history as well as the development of multi- ple renal cysts on subsequent follow-up imaging (Table 2).

Fig. 1 Images of an incidentally discovered right complex renal the abdomen two years later, an axial T1-weighted image after gado- cyst in a 17-year-old girl who presented with severe dysmenorrhea. linium contrast administration demonstrated increased cyst size and a Initial axial post-intravenous-contrast CT of the abdomen showed complexity with enhancing internal septa. This was found to be a a 3.5-cm septated cyst in the right kidney. b On follow-up MRI of multilocular cystic nephroma by histology

1 3 2838 Abdominal Radiology (2019) 44:2835–2840

Table 2 Genetic testing status for positive and negative cases of auto- somal dominant polycystic kidney disease Genetic testing status ADPKD-positive ADPKD- cases negative cases

Genetic testing 8 5 No genetic testing 6a 65b a Patients with a positive family history as well as the development of multiple renal cysts on subsequent follow-up imaging were consid- ered positive for ADPKD in the absence of genetic testing b Patients were considered negative for ADPKD in the absence of genetic testing if they had a single cyst that either was stable or decreased in size on follow-up imaging, and had no family history of ADPKD

Furthermore, 8/14 (57.1%) were diagnosed after undergoing a screening US because of a family history of the disease and 6/14 (42.9%) were diagnosed incidentally after undergoing imaging performed for other reasons (Figs. 2, 3). Of the 74 patients with follow-up imaging, the renal cyst was stable in size in 40 patients (54%), decreased in size in 28 patients (38%), and increased in size in 6 patients (8%).

Renal cyst imaging features associated with ADPKD Fig. 3 US images of the right kidney performed for evaluation of abdominal pain in a 4-year-old boy with no known family history of We next evaluated renal cyst features associated with ADPKD. a US with the patient in the supine position showed a 0.9- ADPKD (Table 3). Mean maximum cyst diameter in cm simple cyst in the right kidney. b Follow-up US 12 months later patients with ADPKD was 22.0 ± 11.2 mm, compared with with the patient in the prone position revealed an increase in cyst size 12.7 + 8.2 mm for patients without ADPKD, a statisti- and the development of a new simple cyst in the right kidney. The patient was later diagnosed with ADPKD based on genetic testing cally signifcant diference (P < 0.001). In patients with a

renal cyst greater than 25 mm on the initial imaging study, 75% (6/8) had ADPKD. All 12 patients with multiple cysts (100%) were found to have ADPKD, compared with only 2/72 (2.8%) patients with a single cyst (P < 0.001). Seven of the patients with multiple cysts were being screened for ADPKD based on a known family history, while 5 patients had multiple cysts discovered incidentally with no known family history. The presence of bilateral renal cysts was also signifcantly associated with ADPKD, both in the entire study population (11/11 patients with bilateral cysts had ADPKD as opposed to 3/73 patients with unilateral cysts; P < 0.0001) as well as those without a known family his- tory (5/5 and 1/71, respectively; P < 0.0001). Among the 74 patients who had follow-up renal imaging available, 4/6 (66.7%) patients who had their renal cyst increase in size were found to have ADPKD, compared with 0/68 (0.0%) Fig. 2 Axial post-intravenous-contrast CT of the abdomen in a patients who demonstrated either stable or decreased cyst 15-year-old boy who presented with blunt abdominal trauma follow- size on follow-up (P < 0.0001). Of the 4 patients with ing a fall from a height demonstrates multiple simple cysts of various ADPKD who had enlarging cysts, 2/4 (50%) had multiple sizes scattered throughout both kidneys. He was subsequently diag- cysts on the baseline ultrasound. Complex cyst features nosed with ADPKD based on these imaging fndings and detailed family history. Follow-up US showed increase in size of multiple (thin septations or complex fuid) were present in 6/84 cases cysts (7.1%) overall, were present in equal proportions in ADPKD

1 3 Abdominal Radiology (2019) 44:2835–2840 2839

Table 3 Renal cyst imaging features associated with autosomal domi- family history. More importantly, patients with multiple nant polycystic kidney disease cysts, either in one kidney or in both kidneys, all were found Pertinent fndings No ADPKD ADPKD P value to have ADPKD in our study population. This applies both to patients with no known family history of ADPKD who Mean cyst size (mm) 12.7 ± 8.2 22.0 ± 11.2 < 0.001 have renal cysts discovered incidentally, as well as those FH of PKD 0 8 0.0003* with a known family history of ADPKD. In addition, renal No FH of PKD 70 6 cysts in patients found to have ADPKD (mean 2.2 cm) were Entire study population < 0.0001* signifcantly larger than in those without (mean 1.3 cm), Multiple renal cysts 0 12 with cyst size greater than 2.5 cm found to have a 75% inci- Single renal cyst 70 2 dence of ADPKD. No FH < 0.0001* In the 74 patients without a known history of ADPKD Multiple renal cysts 0 5 who had follow-up renal imaging at least one year after ini- Single renal cyst 70 1 tial cyst diagnosis, our results demonstrate a signifcantly Entire study population < 0.0001* higher risk of ADPKD in patients whose cyst size increases Bilateral renal cyst(s) 0 11 on follow-up. Our results suggest a role for one-year follow- Unilateral renal cyst(s) 70 3 up renal US in children with incidentally discovered renal No FH < 0.0001* cysts, with increased renal cyst size carrying an increased Bilateral renal cyst(s) 0 5 risk of ADPKD. Previous studies have suggested no follow- Unilateral renal cyst(s) 70 1 up is required for benign appearing cysts because of the Entire study population < 0.0001* low risk of malignancy, but our study establishes a role for Increase in renal cyst size 2 4 surveillance in terms of ADPKD risk assessment [13]. Our on FU study did not reveal any renal malignancies, consistent with Stable or decrease size on 68 0 FU the benign imaging features of cysts included in our study population [14]. Our study did include one case of benign Asterisks indicate statistical signifcance; Fisher’s Exact Test multilocular cystic nephroma that presented initially as a ADPKD autosomal dominant polycystic kidney disease, FH family complex cyst with thin septations and subsequently dem- history, FU follow-up onstrated an increase in renal size and complexity on fol- low-up imaging. While several patients in our study did not (1/14) and non-ADPKD (5/70) cases, and were not associ- undergo genetic testing for ADPKD, it is important to note ated with ADPKD. that many insurance companies do not reimburse the cost of genetic testing. This explains why many families in our cohort declined testing if there were clear family history and Discussion imaging data to suggest presence or absence of the disease. We believe this study adds to our knowledge of the Although they occur uncommonly, cysts are the most fre- natural history of renal cysts in pediatric patients. A 2015 quent incidentally discovered renal lesions in children [1, study of 36 pediatric patients with simple cysts seen on US 10–12]. Our study focuses on the outcomes of 84 cases of focused on symptomatic cysts and requirement for drainage, renal cysts in pediatric patients identifed in our institution. rather than the risk of ADPKD or malignancy [15]. A sec- The vast majority of these cases (76/84) were detected inci- ond study from 2015 of 212 children with solitary renal cysts dentally, while the remaining cases (8/84) were detected dur- discovered on US included cysts with suspicious features ing screening US examinations for patients with a family and focused on risk of renal malignancy [13]. Our study history of ADPKD. We believe that this mix of incidental addresses the risk of ADPKD associated with renal cysts in and screening patients is a fair representation of the renal children, including establishing a high probability associated cysts observed in a large pediatric patient population. This with multiple renal cysts that previously was restricted to purpose of this study was to assess the clinical signifcance patients with a known ADPKD family history [2], as well of renal cysts in the pediatric patient population. We specif- as an association between large cyst size and ADPKD. In cally excluded renal cysts with imaging features suspicious addition, our results establish a role for follow-up imaging for malignancy (thick septations, nodularity, or increased in ADPKD risk assessment for incidentally discovered renal Doppler vascularity) because these features are known to cysts. trigger more extensive imaging and clinical workup [13]. Our study has limitations, including its retrospective Our results confrm that patients with renal cysts and a nature that may introduce selection bias by excluding family history of ADPKD had a signifcantly higher inci- patients with renal cysts who did not have follow-up imag- dence of ADPKD compared with patients without a known ing performed. This stringent inclusion requirement also

1 3 2840 Abdominal Radiology (2019) 44:2835–2840 accounts for the relatively low number of patients included. England) 369 (9569):1287-1301. doi:https​://doi.org/10.1016/ Also, our population included both patients with inciden- s0140​-6736(07)60601​-1 3. Dell KM (2011) The spectrum of polycystic kidney disease in tally discovered cysts and those who were being screened for children. Advances in chronic kidney disease 18 (5):339-347. ADPKD based on family history. We believe it is important doi:10.1053/j.ackd.2011.05.001 to include both populations because both are part of a gen- 4. Avni FE, Hall M (2010) Renal cystic diseases in children: new eral pediatric population that develops renal cysts. The fact concepts. Pediatric radiology 40 (6):939-946. doi:https​://doi. org/10.1007/s0024​7-010-1599-5 that we identifed renal cyst features signifcantly associated 5. Ravine D, Gibson RN, Walker RG, Shefeld LJ, Kincaid-Smith P, with ADPKD that apply to both populations is important and Danks DM (1994) Evaluation of ultrasonographic diagnostic cri- validates the clinical importance of these features. Although teria for autosomal dominant polycystic kidney disease 1. Lancet our study did not address features beyond renal cyst size (London, England) 343 (8901):824-827 6. Rossetti S, Harris PC (2007) Genotype-phenotype correlations in and number associated with ADPKD, we believe that future autosomal dominant and autosomal recessive polycystic kidney studies investigating other features associated with ADPKD, disease. Journal of the American Society of Nephrology : JASN such as overall renal size and renal cyst location, would be 18 (5):1374-1380. doi:https​://doi.org/10.1681/asn.20070​10125​ benefcial. Finally, we acknowledge that many pediatric 7. Fujimaru T, Mori T, Sekine A, Mandai S, Chiga M, Kikuchi H, Ando F, Mori Y, Nomura N, Iimori S, Naito S, Okado T, Rai T, patients with ADPKD will have a known family history of Hoshino J, Ubara Y, Uchida S, Sohara E (2018) Kidney enlarge- disease and undergo routine US screening. However, the ment and multiple liver cyst formation implicate mutations in incomplete penetrance of the disease and the existence of PKD1/2 in adult sporadic polycystic kidney disease. Clinical sporadic mutations that can give rise of ADPKD suggest genetics 94 (1):125-131. doi:https​://doi.org/10.1111/cge.13249​ 8. McHugh K, Stringer DA, Hebert D, Babiak CA (1991) Sim- that it is important to be aware of risk factors for ADPKD in ple renal cysts in children: diagnosis and follow-up with US. the absence of known family history [6, 7]. Radiology 178 (2):383-385. doi:https​://doi.org/10.1148/radio​ In conclusion, incidentally discovered renal cysts in pedi- logy.178.2.19875​97 atric patients carry a low risk of malignancy and a small but 9. Wallis MC, Lorenzo AJ, Farhat WA, Bagli DJ, Khoury AE, Pippi Salle JL (2008) Risk assessment of incidentally detected complex non-zero association with ADPKD. Among cyst character- renal cysts in children: potential role for a modifcation of the istics, bilaterality, multiplicity, large size, and increased size Bosniak classifcation. The Journal of 180 (1):317-321. on follow-up imaging have a signifcant association with doi:https​://doi.org/10.1016/j.juro.2008.03.063 ADPKD in pediatric patients. Further defnitive ADPKD 10. Tada S, Yamagishi J, Kobayashi H, Hata Y, Kobari T (1983) The incidence of simple renal cyst by computed tomography. Clinical testing is recommended for patients without a known family radiology 34 (4):437-439 history of ADKPD who demonstrate cyst size > 2.5 cm and/ 11. Israel GM, Silverman SG (2011) The incidental renal mass. Radi- or bilateral or multiple cysts on initial imaging. Additionally, ologic clinics of North America 49 (2):369-383. doi:https​://doi. a one-year follow-up US is recommended in patients with a org/10.1016/j.rcl.2010.10.007 12. Silverman SG, Israel GM, Herts BR, Richie JP (2008) Manage- single cyst < 2.5 cm, with cyst size >2.5 cm or increase in ment of the incidental renal mass. Radiology 249 (1):16-31. cyst size on follow-up imaging being a potential indication doi:https​://doi.org/10.1148/radio​l.24910​70783​ for more defnitive ADPKD evaluation. 13. Karmazyn B, Tawadros A, Delaney LR, Marine MB, Cain MP, Rink RC, Jennings SG, Kaefer M (2015) Ultrasound classifca- tion of solitary renal cysts in children. Journal of pediatric urol- ogy 11 (3):149.e141-146. doi:https​://doi.org/10.1016/j.jpuro​ Compliance with ethical standards l.2015.03.001 14. Bosniak MA (2012) The Bosniak renal cyst classifcation: 25 years Conflict of interest The authors declare that they have no confict of later. Radiology 262 (3):781-785. doi:https​://doi.org/10.1148/ interest. radio​l.11111​595 15. Koutlidis N, Joyeux L, Mejean N, Sapin E (2015) Management of simple renal cyst in children: French multicenter experience of 36 cases and review of the literature. Journal of pediatric urology References 11 (3):113-117. doi:https​://doi.org/10.1016/j.jpuro​l.2015.03.003

1. Marietti S, Woldrich J, Durbin J, Sparks S, Kaplan G, Chiang Publisher’s Note Springer Nature remains neutral with regard to G (2013) Urologic fndings on computed tomography of the jurisdictional claims in published maps and institutional afliations. abdomen and pelvis in a pediatric population. Journal of pedi- atric urology 9 (5):609-612. doi:https​://doi.org/10.1016/j.jpuro​ l.2012.05.014 2. Torres VE, Harris PC, Pirson Y (2007) Autosomal dominant polycystic kidney disease. Lancet (London,

1 3