Statistics from Altmetric.com
Editor,—Webb et al justify their policy of performing nephrectomy in healthy infants with prenatally detected multicystic dysplastic kidney (MCDK) on the basis of their findings in three children and their interpretation of the literature.1 Paediatricians may wish to scrutinise the methodology and conclusions of this paper more closely before referring asymptomatic infants for ‘prophylactic’ nephrectomy. On the evidence provided the diagnosis of hypertension in their first two patients has not been conclusively proved. Furthermore, it is difficult to see how the renal pathology in their third patient can be attributed to MCDK in the absence of cysts and, more importantly, ureteric atresia.
Making a diagnosis of hypertension in children is problematic—indeed obtaining reliable reproducible blood pressure readings in fractious infants is very difficult. The measurement of peripheral plasma renin activity, which was performed in these children, is a relatively poor guide to the aetiology of hypertension. A credible diagnosis requires selective renal vein renin sampling.
The role of echocardiography in the diagnosis and assessment of mild or moderate hypertension in childhood has yet to be validated. The authors have not identified the source of the normal ranges quoted—an important consideration since different studies have yielded different ‘normal’ values for the parameters quoted and the ‘normal’ range has often been derived from a limited number of individuals in different age groups.
In their third patient, a hypertensive girl of 14, retrograde pyelography is said to have revealed ‘remnant renal tissue in the right loin’. Atresia of the pelviureteric junction or proximal ureter is a characteristic feature of MCDK which, on retrograde pyelography, is associated with the distinctive finding of a blind ending ureter. In this case the flow of contrast across a patent ureter into a remnant of the collecting system, effectively excludes the diagnosis of MCDK.
It is disappointing that the authors felt it necessary to quote the outdated findings of an informal postal survey of American paediatric urologists2 to support their arguments for nephrectomy. Published in 1978 and spanning an undefined period, this survey predates the era of prenatal ultrasound and meets none of the criteria demanded of a scientific study. In addition to misquoting the number of respondents to the survey (136 not 48), the authors have omitted to inform their readers that 10% of the paediatric urologists surveyed reported serious operative complications resulting from nephrectomy for MCDK including deaths, renal failure, intestinal obstruction, and removal of the wrong kidney.
The experience of other centres in relation to the risk of hypertension in MCDKs is different to that reported by Webb et al. Data on 441 children with MCDK was submitted to the Multicystic Kidney Registry3 from 49 centres in the United States and Canada. Of the 441, 260 were managed conservatively with follow up of up to five years in some cases. No cases of hypertension were reported in this series.
Rickwood et al reported a series of 44 children with prenatally detected MCDKs.4 Blood pressure was measured routinely over a mean follow up of three years in 38 of these children and none was found to be hypertensive. These findings mirror our experience in Leeds where we have not encountered any case of hypertension associated with prenatally detected MCDK nor has an MCDK been identified in a child presenting with hypertension.5
Analysing 454 cases of hypertension admitted to the renal unit at the Hospital for Sick Children in Great Ormond Street, from 1975–85, Dealet al 6 documented a wide spectrum of underlying renal pathology without encountering a single MCDK (J E Deal, personal communication).
A retrospective study of 42 children undergoing nephrectomy for renal hypertension in Glasgow did not include a single case of MCDK.7 Similarly, a detailed review of 22 cases of surgically treated hypertension of renal origin in childhood, undertaken in Boston, did not document a single MCDK.8
It is important to distinguish clearly between other forms of renal dysplasia and MCDK, a relatively common and distinctive anomaly characterised by tense, non-communicating cysts and absent or non-functioning renal parenchyma, which is almost invariably associated with complete ureteric atresia or severe distal ureteric obstruction. The authors’ description of the radiological and pathological features of their third case clearly indicate that the renal pathology belongs in one of the former categories of dysplasia rather than representing a genuine MCDK.
Cases of hypertension conclusively linked to MCDKs have very rarely been reported but published evidence indicates that the magnitude of risk is extremely low. Menster et al, reviewing the published literature on the risks associated with prenatally detected MCDKs, concluded ‘It no longer seems advisable to routinely remove MCDK in young patients for either diagnostic or prophylactic reasons’.9
Paediatricians working in centres in which conservative management is undertaken can be reassured that the overwhelming weight of published evidence continues to favour a non-operative approach to the management of prenatally detected MCDKs in healthy infants.
Dr Webb and colleagues comment:
Mr Thomas and Dr Fitzpatrick raise some important questions and we are grateful for the opportunity to respond to them in detail.
We agree that the diagnosis of hypertension may be difficult in small infants and children, though having detected systolic blood pressures in excess of the 95th centile for both age and height on repeated occasions using an appropriately sized cuff with the patient in the resting state, we have met widely accepted criteria for the diagnosis of hypertension in all three cases.1-10 The presence of end organ damage (left ventricular hypertrophy diagnosed using accepted normal ranges1-11 ) in the first two patients confirms the longstanding duration and severity of this hypertension. It is generally agreed that raised peripheral plasma renin activity, while never truly diagnostic, is a useful indicator of renally mediated hypertension; renal vein renin sampling may be potentially technically difficult from the small vein which drains a MCDK and is generally used to predict response to surgery rather than to diagnose hypertension. The resolution of hypertension and echocardiographic changes after nephrectomy provide further supportive evidence for the hypertension in these three children being renally mediated.
MCDK represents the extreme of a spectrum of obstructive cystic dysplasia; while strict pathological criteria require the presence of an atretic ureter for a diagnosis of MCDK to be made, in practice, physicians and surgeons make a clinical diagnosis on the basis of DMSA and ultrasound findings. Many children with lesions such as those seen in our third case may therefore be under follow up with a clinical diagnosis of MCDK.
Our interpretation of page 213 of Bloom and Brosman’s paper1-2 was that 48 (35%) of the 136 urologists sent questionnaires responded to the section on clinical findings associated with MCDK. If, however, Thomas and Fitzpatrick’s interpretation is correct, 15% of 136 is a much larger number of urologists reporting hypertension than 15% of 48, and lends further weight to our argument supporting hypertension as a significant complication of the MCDK. Operative morbidity and mortality have fallen significantly since Bloom and Brosman’s publication; our previously reported absence of perioperative complications1-12 has continued over recent years.
The report of the Multicystic Kidney Registry reports on 441 children with MCDK in the United States and Canada of whom 181 underwent nephrectomy predominantly between 7 and 12 months of age,1-3confirming that nephrectomy continues to be considered a valid treatment option in many centres. Contrary to Thomas and Fitzpatrick’s statement, four children followed up conservatively developed hypertension, though this was ‘believed to be unrelated to the multicystic kidney’. Without an alternative explanation or details explaining why hypertension was unrelated, MCDK related hypertension cannot be excluded. Our paper reports three further cases of hypertension in children diagnosed as having MCDK in addition to those others previously reported in the literature.
The surgical versus conservative management of the MCDK will remain an area of great controversy. Only by collecting complete and accurate data from large, unselected series of patients treated both surgically and conservatively will the true incidence of complications of both approaches emerge, and we make a further plea for the setting up of a national registry to expedite this.
Editor,—Webb et al continue the debate about conservative versus surgical management of multicystic dysplastic kidney (MCDK) with case reports of three patients for whom surgery was undoubtedly indicated.2-1 We would, however, caution against the experience of tertiary referral centres being extrapolated to the general population.
Experience at Northwick Park Hospital, a large district general hospital, suggests that MCDK is more common than the published incidence figures of one in 4300 live births.2-2 We have reviewed all cases of MCDK diagnosed in infants born at Northwick Park Hospital in the 10 year period 1987–96 during which there were 33 537 live births. During that period, routine antenatal ultrasonography was carried out at 16–18 weeks’ gestation. Further scans were performed at a later stage in the pregnancy only if there was a clinical indication. Of the 14 cases of MCDK diagnosed on antenatal ultrasonography, 11 were detectable on the routine scan at 16–18 weeks’ gestation, including one case in which the pregnancy was terminated at 18 weeks because of contralateral renal agenesis and multiple congenital abnormalities. In three cases with a normal early scan, the abnormality was detected by ultrasonography at 25, 26, and 39 weeks’ gestation, respectively. One further MCDK was detected by palpation of a renal mass on the routine neonatal examination after birth. Thus the incidence of MCDK diagnosed in an unselected district general hospital population was one in 2400 live births. As the majority of mothers were scanned on only one occasion during their pregnancy, our figures may still represent an underestimate of the true incidence of this condition.
During the same 10 year period, only one infant with unilateral renal agenesis was identified on routine antenatal ultrasonography, compared with the published necropsy incidence in adults of one in 1070.2-3 Although the relatively large size of the fetal adrenal gland makes the diagnosis of unilateral renal agenesis difficult to identify on antenatal ultrasonography, our data may also be interpreted as suggesting that a large proportion of solitary kidneys identified in adult life may in fact be due to regression of MCDK. All of our infants with an antenatal diagnosis of MCDK had serial postnatal ultrasonography, a DMSA isotope scan, and micturating cystourethrogram performed postnatally. Associated renal tract abnormalities in the contralateral kidney were found in only three infants (two grade I and one grade IV vesicoureteric reflux), a similar incidence to that found in other series.2-4
Complication rates determined from case reports in the literature will, inevitably, be subject to reporting bias. They must be interpreted in the light of data on incidence and natural history in non-referral populations if the risks of conservative management are to be put in proper perspective. Meanwhile the dilemma of management of MCDK remains controversial.