Parapelvic cysts, a distinguishing feature of renal Fabry disease

Abstract

Background

Fabry’s disease (FD) is a rare, multi-organ lysosomal disease, caused by the deficiency of the enzyme α-galactosidase A, and is difficult to diagnose. Although parapelvic cysts (PC) were previously associated with FD, their prevalence and significance are unclear.

Methods

The present study aimed to: (i) evaluate, by renal ultrasound, the real prevalence of PC and of their determinants in a multicentre, nationwide cohort of FD patients (n = 173, Study 1) and (ii) ascertain whether a greater accuracy of PC detection improved their identification, in FD patients from a single centre (n = 67, Study 2). In both studies, for each FD patient, an age- and renal function-matched subject was selected for comparison (1:1).

Results

In Study 1, PC were detected in 28.9% of FD subjects and in only 1.1% of control subjects (P < 0.001). The presence of other renal abnormalities did not differ between the groups, nor differences exist in the main demographic and laboratory parameters between the groups. In Study 2, the greater accuracy of ultrasound increased PC prevalence from 29.8% to 43.3% in the same subjects (P < 0.05). In both studies, no correlation was detected between PC and the main demographic, clinical and biochemical parameters, including use of enzyme replacement therapy (P < 0.1, minimum value). Finally, no difference existed between FD patients with and without PC.

Conclusions

The present study suggests that the presence of PC in renal patients should alert physicians to consider the diagnosis of FD, primarily in subjects with an unclear family history of renal disease and in the presence of other stigmata of the disease.

biomarkers, chronic renal failure, Fabry disease, parapelvic cyst, ultrasonography

INTRODUCTION

Renal involvement is a significant cause of morbidity and reduced life expectancy in patients with Fabry’s disease (FD), an X-linked lysosomal storage disorder characterized by absent or reduced activity of the enzyme α-galactosidase A (GLA), which results in progressive accumulation of glycosphingolipids, mainly globotriaosylceramide (Gb3), in a variety of tissues and organs and determines renal failure, cardiomyopathy and cerebrovascular disease [1].

Kidney involvement in FD follows Gb3 deposition in all types of renal cells and is characterized by a progressive disease that leads, in untreated patients, to end-stage renal failure from the third to fifth decade of life [2]. Although some clinical signs of FD are already present in childhood, patients are often diagnosed relatively late in the course of the disease due to the absence of specific clinical markers, while a timely diagnosis and the prompt start of enzyme replacement therapy (ERT) may be beneficial in stabilizing renal function or slowing its decline [3, 4].

The presence of parapelvic cysts (PC) has already been linked to FD and reported in the literature as a possible feature of its renal involvement [5–10]. Nevertheless, the prevalence of this specific finding as well as the nature of its association with FD remains undefined to date.

Therefore, the first aim of our study was to evaluate the real prevalence of PC and of their determinants in a large cohort of patients with FD, using renal ultrasound (US, Study 1). Furthermore, we also investigated whether a more detailed renal US, focused on PC detection, may have improved their identification in FD patients (Study 2).

MATERIALS AND METHODS

Study design and procedures

The present study is divided into two branches, namely Study 1 and Study 2.

Study 1 represents a multicentre retrospective study, conducted on a cohort of 173 FD patients, all with regular follow-ups in nine Specialized Care Hospital Units in Italy, aimed to detect the prevalence of PC and factors potentially associated with their presence using the renal US routinely performed during patients’ follow-up. All the participating centres received a specific form to fill out from the coordinating centre, including data on renal function, renal US and cardiovascular (CV) involvement. Inclusion criteria in the study were: genetically proven FD (age ≥18 years at the time of US), availability of kidney US examinations and signed informed consent for participation in the study. FD patients on renal replacement therapy or affected by any renal anomaly affecting renal structure were excluded from the study. In each centre, clinical and laboratory data relevant to FD were collected by trained physicians or obtained by direct interviews.

Renal US evaluations were performed for each individual in the single centre according to their practice and equipment. The requested data included: determination of renal diameters, presence of PC, cortical cysts and/or other renal abnormalities (such as urinary tract obstruction, ampullary pelvis, ectasia of calyces, sinus lipomatosis, renal atrophy and decreased cortical thickness, increased echogenicity). All the gathered data were then entered a central database at the coordinating centre (Chair of Nephrology, University Federico II of Naples, Italy) for evaluation.

To compare the prevalence of PC between FD patients and patients with chronic renal failure (CRF) of different aetiologies, we identified the US of a control subject with comparable age and renal function (1:1 case–control study) for each recruited FD patient, with the obvious exclusion of cystic diseases.

Study 2 represents a cross-sectional study on 67 FD patients followed at the coordinating centre of University Federico II of Naples, Italy, and previously included in Study 1. Its aim was to demonstrate whether more accurate ultrasounds, carried out by a single expert radiologist and focused to reveal the presence of PC in FD patients, influenced the prevalence data reported in Study 1.

Accordingly, these 67 FD patients (along with their respective controls) were re-examined with a new US, and their data were compared with those reported in the last US of their follow-up (performed by different radiologists) and included in Study 1.

In Study 2, all US examinations were carried out using a HDI 5000 Sono-CT color Doppler system (Philips Healthcare, Bothell, WA, USA), equipped with a 5–2 MHz curved array transducer. The operator was unaware of the participants’ clinical details and laboratory findings. All participants were scanned in the supine, lateral and prone positions after an overnight fast, to minimize the presence of bowel gas. A complete B-mode evaluation of renal morphology was performed. Similarly, a new ultrasound was performed also on the 67 patients selected for the control group.

In both studies, cysts were classified according to their echogenicity and their location. In particular, they were defined as ‘parapelvic’ if originating in the renal parenchyma and extending into, and primarily expanding within, the renal sinus. PC are typically oval, sharply delineated, echo-free/hypoechoic, usually multiple in number, not interconnected and oriented in a radial direction toward the hilus [11, 12].

Renal function was expressed as estimated glomerular filtration rate (eGFR), calculated with the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [13]. Other laboratory values were measured using standard hospital laboratory techniques. GLA enzyme activity was classified as ‘absent’ (<1% of normal values) or ‘residual’ (1–5% of normal values) because of the different measurement methods used in the recruiting centres.

The study conformed to the principles of the Declaration of Helsinki. The institutional Review Board or Ethics Committee at each site approved the protocol, and written informed consent was obtained from all participants.

Statistics

Variables were summarized as mean ± standard deviation (SD) when using numerical variables, and as frequencies (%) when using qualitative factors. Comparison between groups was performed using the Student’s t-test for independent samples and the Chi-square test. The association between the presence of PC and FD disease in both studies was further quantified by using crude and age/gender-adjusted odds ratio with the corresponding 95% confidence intervals (CIs). All tests were two-sided and significance was set at an alpha level of 0.05.

RESULTS

Study 1 (retrospective study)

The characteristics of the 173 FD patients and of their control group enrolled in the retrospective study are reported in Table 1. There were no differences between the groups in demographic and clinical data nor in other laboratory parameters, with the exception of a slight difference in gender distribution.

Table 1

Baseline characteristics of FD patients versus controls of Study 1

Parameter	FD (n = 173)	Controls (n = 173)	P
Age (years)	44.2 ± 14.7	47.4 ± 18.3	NS
Male gender	70 (40.4)	85 (49.1)	NS
eGFR (mL/min/1.73 m2)	87.5 ± 34.7	83.4 ± 27.5	NS
Proteinuria (mg/day)	404.2 ± 807.3	512.8 ± 619.3	NS

Data are means ± SD or n (%). NS, not significant.

GLA activity was absent in 65 patients (37.6%), residual in 94 patients (54.3%) and not reported in the remaining 14 individuals. A total of 56 distinct GLA gene mutations were represented, accounting for the different clinical manifestations and organ involvement. In particular, 117 patients (67.6%) showed CV manifestations or symptoms of CV diseases (left ventricular hypertrophy, arrhythmia, diastolic dysfunction, valvular abnormalities or coronary artery disease).

Table 2 shows the frequency of renal abnormalities detected by US in the two study groups. Renal cysts, either parapelvic or cortical, were detected in 76 FD patients (43.9%), with a size ranging between a few millimeters and 8 cm. Fifty patients had PC (28.9%) and 26 cortical cysts (15.0%). Moreover, 12 of these patients showed both parapelvic and cortical cysts (6.9%). Among the 50 patients with PC, 33 of them proved to have multiple cysts (66%), of whom 19 showed a bilateral involvement (38%). Other renal abnormalities were present in only four patients (2.3%).

Table 2

Frequency of renal abnormalities detected by US in the FD patients versus controls in Study 1

Abnormality	FD (n = 173)	Controls (n = 173)	P
None	93 (57.8)	144 (83.2)	0.001
Cortical cysts	26 (15)	21 (12.1)	NS
Parapelvic cysts	50 (28.9)	2 (1.1)	<0.001
Other	4 (2.3)	6 (3.5)	NS

Data are reported as numbers and (%). NS, not significant.

In control subjects, the prevalence of renal cysts was much lower (13.2%), with the presence of PC in only two subjects (1.1%, P < 0.001 versus FD patients, also in gender-adjusted analysis), while no difference existed in frequency of cortical cysts, in other renal abnormalities or in kidney longitudinal diameters between FD and control subjects (data not reported).

No correlation was found between the presence of PC and any of the selected demographic, clinical and biochemical parameters.

Finally, no significant difference in age, gender, GLA activity, type of gene mutations, CV involvement, kidney function or in protein excretion was detected between the FD patients with PC compared with those without (data not shown).

Study 2 (cross-sectional study)

Demographic and clinic characteristics of patients included in Study 2 (Table 3) were comparable to those of the overall population of the retrospective study (data not shown). Mean GLA activity in patients of Study 2 was 2.5 ± 2.0 nmol/mL/h, with 42 patients (62.7%) showing the typical CV manifestations of FD (left ventricular hypertrophy, arrhythmia, diastolic dysfunction, valvular abnormalities or coronary artery disease). Moreover, 29 distinct GLA gene mutations were represented: 48 patients had missense mutation (71.6%), 7 showed insertions (10.4%), 6 had deletions (8.9%) and 6 had splicing mutations (8.9%). Finally, 52 patients (77.6%) were under ERT (mean duration: 54 ± 22 months): 57% treated with agalsidase beta and 43% with agalsidase alfa.

Table 3

Main parameters of FD patients of the cross-sectional study compared with their matched controls

Parameter	Follow-up patients (n = 67)	Controls (n = 67)	P
Age (years)	41.6 ± 14.2	44.2 ± 14.7	NS
Male gender	28 (41.8)	70 (40.4)	NS
eGFR (mL/min/1.73 m2)	94 ± 31.9	87.5 ± 34.7	NS
Proteinuria (mg/day)	408.8 ± 918.3	404.2 ± 807.3	NS

Data are means ± SD or n (%). NS, not significant.

Table 4 shows the frequency of renal abnormalities detected after the new US in our FD patients, compared with the data reported retrospectively in the same patients in Study 1. The overall frequency of renal cysts was slightly higher than in Study 1 (65.7% versus 49.2%, not significant), due to the increased detection of PC, which was significantly higher than in Study 1 (43.3% versus 29.8% in Study 1, P < 0.05). On the other hand, the number of cortical cyst remained basically unchanged. Similar to Study 1, 62% of FD patients of Study 2 showed multiple cysts (n = 18), and 44% had bilateral cysts (n = 13).

Table 4

Comparison of renal abnormalities detected by US in 67 FD patients, included in Study 1 (standard, retrospective US) and then in Study 2 (new, ‘focused’ US)

Abnormality	Study 1 pts (n = 67)	Study 2 pts (n = 67)	Controls (n = 67)
None	23 (34.1)	21 (31.3)*	50 (74.6)**
Cortical cysts	13 (19.4)	15 (22.4)	14 (20.9)
Parapelvic cysts	20 (29.8)	29 (43.3)*	1 (1.5)**
Other	2 (3.0)	2 (3.0)	2 (3.0)

Data of control patients after the new US are also reported. Data are reported as numbers and (%).

*P < 0.05 versus Study 1.

**P < 0.001 versus both Study 1 and Study 2.

No difference was detected in renal abnormalities in control patients of Study 2 after the new US compared with those described in Study 1.

Both cortical cysts and PC were already evident at a young age, with no difference between genders (Figure 1).

FIGURE 1

Percentage of FD patients with kidney cysts by decade and gender.

Finally, no correlation was found between the presence of PC and any of the collected clinical and demographic parameters.

DISCUSSION

In this national, multicentre study, we demonstrated that the prevalence of PC, observed during a routine US study, is significantly higher in FD patients (29%) compared with patients with kidney disease of similar age and level of renal function (1.1%) and that such prevalence becomes even higher (43%) when the specific US was performed.

In this study we confirm the preliminary data reported several years ago by Ries et al. [5] in a small cohort of 24 male FD patients, selected for an ERT trial because of severe neuropathic pain. Indeed, these authors showed a 50% prevalence of PC, as detected by computed tomography (CT) and confirmed by magnetic resonance imaging (MRI) study. Our analysis, however, overcomes some limitations of Ries’ paper, since we have analysed a much larger, nationwide cohort of patients of both genders, selected only on the basis of the presence of an US in their clinical report. This approach allowed us to analyse data of heterogeneous FD patients with different age, organ involvement and/or genetic mutations, either treated or untreated with ERT.

The PC prevalence in our retrospective Study 1 (i.e. using standard US) is lower than that reported by Ries et al. (29% versus 50%). However, it should be noted that when US were performed to specifically detect PC in Study 2, such percentages greatly increased to 43.3%, approaching the one previously described in highly selected patients [5]. Interestingly, different from the general population, where the presence of cysts is rare before age of 50 years, and mostly affects males, in our FD patients cysts were already present at a young age, with no difference between genders (Figure 1). This result is in line with what has also been described in a previous study by Glass et al. [6] that examines a consistent number of patients with FD, although their prevalence of PC is much lower (14.5%) compared with our study.

Therefore, our study confirms the extraordinarily high prevalence of PC in FD patients compared with patients with renal failure or with the general population, in whom it ranges between 1.28% and 6% including autopsy studies [14–16]. Furthermore, it suggests that the presence of such findings should alert the nephrologist to consider a possible diagnosis of FD in patients with CRF of unknown aetiology, and to start in-depth laboratory and clinical examinations to confirm the presence of the disease. It is also noteworthy that the ability to report this finding simply requires an expert and motivated radiologist, without requiring a more expensive or complex second level imaging examination. In fact, renal US also allows the differential diagnosis between PC and hydronephrosis, since cysts are never interconnected, as commonly observed in calyceal groups in hydronephrosis, and their echolucency cannot be traced down to the ureter.

The second relevant result of our study is the lack of statistical association between PC and demographic, clinical or laboratory data of our patients. In particular, this analysis reasonably excludes that the presence of PC in FD is related with patients’ age, as previously suggested [5]. This further differentiates PC from solitary, simple renal cysts, which are rare below 30 years of age, and whose incidence commonly increases to approximately 50% during the sixth decade of life [17, 18].

Unfortunately, our work does not help clarify the pathogenesis of PC. Kidney cysts, in fact, as a distinctive feature of either acquired or inherited renal cystic pathologies, typically involve the renal parenchyma [19]. Conversely, PC originate from the sinus structures, presumably by lymphatic vessels enclosed in fibrous and fatty tissue with nerves and other vascular structures. Previous papers [20] have hypothesized that occlusion of such lymphatic vessels may result in subsequent dilatation, potentially responsible for ‘in loco’ cyst formation, but when and why this occurs remains unclear. We hypothesize, however, that a trigger condition should be present, since no other renal pathology is associated with the presence of cysts in the renal sinus except FD.

Recently, a possible role of glycosphingolipids has been postulated in the pathogenesis of human renal cystic diseases. In fact, it has been demonstrated that intra-renal levels of two glycosphingolipids, such as glucosylceramide (GlcCer) and ganglioside, were significantly higher in two mouse knockout models of polycystic kidney disease (PKD) and of nephronophthisis [21–23]. It has also been shown that the GlcCer synthase inhibitor Genz-123346 effectively inhibited the cystogenesis in these models orthologous to human PKD [24]. This evidence suggests that the modulation of glycosphingolipid metabolism could represent an effective approach for treatment of kidney cystic diseases, and allows us to hypothesize an indirect role of GB3 storage also in the pathogenesis of PC formation in FD, potentially associated with altered local metabolism or peculiar genetic regulatory processes.

Finally, some limitations of our study should be addressed, such as the absence of data of GB3 levels. Our results, however, indirectly suggest that the entity of GB3 storage is not crucial in determining the onset of PC, given the absence of relationship between cyst presence and the duration of FD or the use of ERT and the duration of such treatment. Indeed, the prevalence of cysts did not differ between FD patients under ERT and untreated subjects, although ERT is widely recognized as effectively reducing renal cell accumulation of GB3 [25–29]. A prospective study aimed to clarify whether either the number or the dimensions of PC are modified with time or after prolonged ERT treatment is warranted.

In conclusion, our study confirms the high prevalence of PC in classically affected FD males and females, and clearly demonstrates that a routinely US renal examination may detect the presence of such cysts, mostly when specifically researched, with CT or MRI scans as second level analyses in case of ambiguous finding. Although to date PC cannot be considered a pathognomonic sign of FD, its presence should alert both nephrologists and radiologists to consider the diagnosis of FD, especially in subjects with an unclear family history of renal disease and when other stigmata of the disease are evident. An early diagnosis of FD and the prompt start of ERT and of adjunctive therapies may prevent disease progression and the development of irreversible pathological changes [30–32].

AUTHORS’ CONTRIBUTIONS

A. Pisani had the original idea; E.R., L.P.A., A. Pellegrino, S.F., R. Mignani, M.V., Y.B., D.C., S.S., F.P., L.C., R. Manna, C.Z. and M. Santangelo took care of the patients; M.S., E.T., S.C., M.I. and D.D.R. performed the radiological studies; D.B. performed the statistical evaluation; A. Pisani, E.R. and M. Sabbatini wrote the paper. All the authors read and approved the final version of the manuscript.

FUNDING

Antonio Pisani, Sandro Feriozzi, Renzo Mignani and Federico Pieruzzi are consultant for Genzyme, a Sanofi company and for Shire; moreover, Antonio Pisani, Sandro Feriozzi, Renzo Mignani, Massimo Imbriaco and Federico Pieruzzi received travel grants by Genzyme, a Sanofi company and by Shire.

CONFLICT OF INTEREST STATEMENT

A. Pisani received reimbursement for attending symposiums, fees for speaking, funds for research and fees for consulting by Shire, Genzyme and Amicus Companies. A. Pisani warrants, on behalf of himself and his co-authors, that the paper is original, has not been formally published in any other journal, is not under consideration by any other journal and does not infringe any existing copyright or any other third party rights.

Author notes

Massimo Sabbatini and Eleonora Riccio authors share senior authorship.