Nephrocalcinosis and Renal Dysfunction in Pediatric Intestinal Failure

Abstract

Background:

Outcomes of pediatric intestinal failure (PIF) have improved recently, with other comorbidities, such as increased echogenicity/nephrocalcinosis on ultrasound (US) in long-term survivors now evident. We evaluated the significance of nephrocalcinosis over time in PIF and its impact on renal function.

Methods:

Retrospective analysis on a cohort of PIF patients was performed. Presence of nephrocalcinosis and/or increased renal echogenicity (identified on US), estimated glomerular filtration rate (eGFR; ml · min ⁻¹ · 1.73m⁻2), renal tubular function, PN volume (ml · kg⁻¹ · day⁻¹) and PN exposure time (hours/day) were reviewed annually over a follow-up period of 2 years. Outcomes in the nephrocalcinosis versus normal US groups were compared.

Results:

Forty patients (28 boys, median age 2.7 years) were followed for 2 years. Fifteen (38%) had either increased echogenicity or nephrocalcinosis (group 1) at initial US. US were normal in the remaining 25 (62%) on initial assessment (group 2). eGFR did not differ between group 1 and group 2 at baseline (118 vs 133, P=0.51) and year 2 (130 vs 131, P=1.00). The percentage of patients with abnormal markers of tubular function was similar in both groups at year 2 (high urine calcium: creatinine 33 versus 30, P=0.83; high urine calcium: citrate 39 versus 42, P=0.87; low urine citrate: creatinine 15 versus 17, P=1.00; high urine oxalate: creatinine 39 versus 25, P=0.77).

Conclusions:

A large proportion of PIF patients with a history of parenteral nutrition (PN) exposure have nephrocalcinosis and/or increased echogenicity on US. Over a 2-year follow-up period, however, these abnormalities had no impact on eGFR or renal tubular function.

Intestinal failure (IF) is defined as the reduction in gut function below the minimum required for the absorption of macronutrients and/or water and electrolytes, such that intravenous supplementation is required to maintain hydration and growth (1,2). The most common causes of pediatric IF are protracted postinfectious diarrhoea and short bowel syndrome (SBS) in developing and developed countries, respectively (3). The overall incidence of SBS is 24.5 per 100,000 live births with a significantly higher incidence in preterm infants (4).

Parenteral nutrition (PN) remains the cornerstone of management of irreversible IF (1,5). It is not, however, devoid of complications. Over the last couple of decades, numerous advances in the management of pediatric intestinal failure (PIF) have resulted in improved long-term survival for infants and children receiving PN (6). These include advances in surgical techniques, lipid minimization strategies or the use of fish oil-based lipid emulsions and the establishment of multidisciplinary teams specialising in intestinal rehabilitation (IR). Despite these advances, patients with IF remain at risk of progressive liver disease, infectious complications, vascular thromboses, electrolyte imbalances, metabolic bone disease, and significant psychosocial and developmental sequelae associated with the use of daily PN and IF complications (3-5,7,8). Renal dysfunction has been reported in both adult and pediatric patients on long-term PN, characterised by a progressive decline in estimated glomerular filtration rates (eGFR) and sonographic abnormalities, such as increased echogenicity, nephrocalcinosis [mineral precipitation within the renal parenchyma (9)], and nephrolithiasis (10-13). The mechanism for this decline in renal function is yet to be identified (10,11,14).

We recently reported on the prevalence of renal abnormalities in PIF patients with a history of prolonged PN use followed by our intestinal rehabilitation programme (IRP) from 2013 to 2014 (15). Twenty-two out of 54 (41%) patients had nephrocalcinosis and/or increased echogenicity detected on ultrasound (US). The group of patients with sonographic abnormalities had shorter colonic remnants, a history of longer PN exposure, and a higher percentage of PN calories than their normal counterparts. Although these findings appeared significant, their long-term implications remain unknown.

We evaluated the significance of nephrocalcinosis over time in PIF and its impact on renal function.

METHODS

We performed a retrospective longitudinal analysis on this previously reported cohort of 44 patients (15) identified between 2013 and 2014 and followed by our multi-disciplinary IR program at The Hospital for Sick Children in Toronto. All patients who had received annual abdominal/renal US scans (which is routine for all patients on PN in our program) for a minimum of 3 scans over 2 years (baseline, year 1, and year 2) were included in the study. The primary outcome measure was eGFR. Secondary outcome measures included the presence of increased echogenicity on annual US and renal tubular function based on urine analysis. Patients who did not have at least 3 annual US scans and patients with a history of previous liver or intestinal transplant were excluded from the analysis.

Data were collected from the electronic patient chart. Patient demographics and intestinal anatomy were collected at baseline. Demographic data obtained included date of birth, gender, gestational age, and aetiology of IF. Intestinal failure was defined as the need for PN for more than 42 days or less than 75% of predicted length of small intestine (16). Residual intestinal anatomy including presence of an ostomy and both absolute length and percentage of expected length based on established norms (17) of both small and large bowel were collected. Radiology results for abdominal/renal US were collected to determine the presence of nephrocalcinosis or increased echogenicity.

Patients were stratified into 2 groups at baseline; group 1, abnormal US—patients with increased echogenicity and/or nephrocalcinosis identified on abdominal US, and Group 2, normal US—patients with no renal abnormalities on US. The following outcome measures were collected at baseline and then at yearly intervals for 2 years for both groups; estimated glomerular filtration rate (eGFR, ml · min⁻¹ · 1.73m⁻²), PN volume (ml · kg⁻¹ · day⁻¹), PN exposure time (hours per day), and renal tubular function on urine test (measured only at year 2). Ultrasound changes over time were also documented at each yearly interval for both groups.

eGFR was calculated using the modified Schwartz formula (18,19). Renal tubular function was assessed using the following parameters: urine calcium to creatinine ratio, urine calcium to citrate ratio, urine citrate to creatinine ratio, and urine oxalate to creatinine ratio. Age-appropriate reference ranges for these renal values were obtained from several sources to provide parameters and age ranges (20-23).

Data were collected by an MD and an experienced clinical nurse specialist. Experienced pediatric sonographers at our institution performed renal scans. Data were analyzed using medians and interquartile ranges (IQR) for continuous variables and proportions for categorical variables. The Mann-Whitney U and Fisher exact tests were used for continuous and categorical variables, respectively. An alpha-value of 0.05 was considered statistically significant. IBM SPSS Statistics 22 (2013) was used for the analyses.

RESULTS

Forty out of 54 patients (28 boys, median age 2.7 years) from the original cohort (15) were followed for a period of at least 2 years. The remaining patients were excluded from the analysis because of follow-up period below 2 years.

Patient demographics and characteristics at baseline are outlined in Table 1. Fifteen out of 40 patients (37.5%) had an abnormal renal US at baseline (Group 1) characterised by increased renal echogenicity or nephrocalcinosis. The groups had no differences in age at time of US (group 1=2.5 years vs group 2=3 years, P=1.00) or gestational age (group 1, 37 weeks vs group 2, 35 weeks, P=0.739).

TABLE 1.

Patient demographics at baseline ultrasound

	Abnormal US (group 1) N=15	Normal US (group 2) N=25	P value
Sex (male)	12 (80%)	16 (64%)	0.477
Median gestational age; weeks (IQR)	37 (32.5–38.5)	35 (33–38)	0.739
Median age at baseline US; years (IQR)	2.5 (0.5–4.4)	3 (0.9–4.6)	1
Etiology
Gastroschisis	4 (26)	8 (32)	0.973
Small bowel atresia	3 (20)	4 (16)
Necrotizing Enterocolitis	2 (13)	4 (16)
Volvulus	2 (13)	3 (12)
Hirschsprungs disease	2 (13)	1 (4)
Tufting enteropathy	1 (7)	2 (8)
Other	1 (7)	3 (12)
Anatomy
Median SB % (IQR)	30% (10.3–79.5)	25% (12–62)	1
Median LB % (IQR)	65% (47.5–100)	85% (50–100)	1
Intact colon	5 (33%)	12 (48%)	0.512
Stoma present	8 (53%)	14 (56%)	1
TPN dependent at baseline	14 (93%)	20 (80%)	0.381
PN exposure at baseline
Median days on PN (IQR)	840 (183.5–1423)	695 (341–928)	0.84
Median hours of PN/day (IQR)	13 (0–15.5)	12 (8–14)	1
Median % PN calories (IQR)	37 (0–93)	35 (0–77)	0.954
Median ionized calcium levels; mmol/L (IQR)	1.26 (1.24–1.32)	1.29 (1.25–1.33)	0.97

IQR = inter-quartile ranges; LB = large bowel; PN = parenteral nutrition; SB = small bowel; TPN = total parenteral nutrition; US = ultrasound.

Gastroschisis was the most common diagnosis in both groups (26% and 32% in groups 1 and 2, respectively), closely followed by small bowel atresia (20% and 16%) and necrotising enterocolitis (13% and 16%). Bowel anatomy did not differ between the 2 groups; 33% and 48% of patients in group 1 and group 2, respectively had an intact colon. Patients in group 1 had a slightly longer but nonsignificant exposure to PN than those in group 2 (840 vs 695 days, P=0.84, and 37% vs 35% PN calories; P=0.954). Median serum ionized calcium levels were not statistically different between group 1 and group 2 either (1.26 vs 1.29 mmol/L, P=0.97).

eGFR did not differ between groups over the study period (Table 2). Calculated eGFR at baseline was 117.6 ml · min⁻¹ · 1.73m⁻² for group 1 and 133.4 for group 2 (P=0.514), 120.5 versus 126.9 ml · min⁻¹ · 1.73m⁻² for year 1 (P=1.00) and 129.6 versus 130.6 ml · min⁻¹ · 1.73m⁻², respectively, at year 2 (P=1.00) (Table 2).

TABLE 2.

Estimated glomerular filtration rate at baseline and follow-up period

	Abnormal US Group 1 (n=15)	Normal US Group 2 (n=25)	P-value
Median Baseline eGFR (IQR)	117.6 (98.0–146.8)	133.4 (115.2–155.6)	0.514
Median Year 1 eGFR (IQR)	120.5 (102.2–152.6)	126.9 (111.0–141.7)	1.000
Median Year 2 eGFR (IQR)	129.6 (103.0–138.6)	130.6 (108.1–139.4)	1.000

eGFR = estimated glomerular filtration rate (ml · min⁻1 · 1.73m⁻²); IQR = interquartile ranges; US = ultrasound.

Table 3 delineates the PN time (in hours/day) and PN volume (in ml · kg⁻¹ · day⁻¹) for both groups over the study period. PN exposure did not differ between group 1 and group 2 at baseline (15 vs 14 hours/day, P=1.00), year 1 (12 vs 12 hours/day, P=0.934) and year 2 (11 vs 12 hours/day, P=0.870) respectively. PN volume also did not differ between group 1 and group 2 at baseline (114 vs 85 ml · kg⁻¹ · day⁻¹, P=0.191), year 1 (65 vs 74 ml · kg⁻¹ · day⁻¹, P=1.00) and year 2 (82 vs 62.4 ml · kg⁻¹ · day⁻¹, P=0.899), respectively.

TABLE 3.

PN hours and volume (per day) at baseline and follow-up period

	Abnormal US group 1 (n=15)	Normal US group 2 (n=25)	P value
Median hours/day on PN (IQR) at:
Baseline	15.0 (12.0–24.0)	14.0 (10.0–20.0)	1.000
Year 1	12.0 (0–15.0)	12.0 (0–16.0)	0.934
Year 2	11.0 (0–14.0)	12.0 (0–15.0)	0.87
PN volume/day (IQR) at: (ml/kg)
Baseline	114 (78–145.0)	85 (39.0–125.0)	0.191
Year 1	65.0 (0–140.0)	74.0 (0–130.0)	1.000
Year 2	82.0 (0–120.0)	62.4 (0–107.3)	0.899

PN = parenteral nutrition; US = ultrasound.

Tubular function, a potentially more sensitive marker of nephrocalcinosis-related morbidity than eGFR, was assessed at year 2 (Table 4). No difference was found between the 2 groups in any of the parameters measured to assess renal tubular function.

TABLE 4.

Tubular function at year 2

At year 2	Percentage of patients with abnormal US (n=13)	Percentage of patients with normal US (n=27)	P value
High urine calcium: creatinine	33.3	30	0.833
High urine calcium: citrate	38.5	41.7	0.870
Low urine citrate: creatinine	15.4	16.7	1.000
High urine oxalate: creatinine	38.5	25	0.770

US = ultrasound [age-specific reference ranges reproduced from (20)].

Ultrasound findings noted at baseline did change over time for patients in both groups (Fig. 1). The increased echogenicity and/or nephrocalcinosis identified in 9 out of 15 patients in group 1 at baseline had resolved on follow-up imaging at year 2; all of these 9 patients were on PN at baseline, and only 3 of them were off PN by year 2. Seven patients from group 2 with normal ultrasounds at baseline developed sonographic renal abnormalities by year 2; 6 of these 7 patients were on PN at baseline, with an additional patient becoming PN-dependent at year 2. Thus, there did not appear to be any relation to PN use and a change in the appearance or resolution of renal sonographic abnormalities in either group over the 2-year period.

FIGURE 1.

Ultrasound changes over time from baseline to year 2 for group 1 and group 2.

DISCUSSION

Increased echogenicity on US is a nonspecific marker of renal parenchymal damage, outside the neonatal period where the kidneys are usually echogenic (15,24,25). Several known causes of increased echogenicity include nephrocalcinosis, interstitial inflammation, tubular disorders, and glomerular disease. Nephrocalcinosis is defined as increased mineral precipitates in the renal parenchyma (9,25). Premature babies, particularly those born <32 weeks, are at increased risk of nephrocalcinosis; however, in most cases the nephrocalcinosis resolves within the first few years of life (9). Other risk factors for nephrocalcinosis include electrolyte derangements, acid base disturbance, frequent episodes of dehydration, frequent episodes of sepsis, and metabolic bone disease (20,26,27), all of which are also seen in patients with IF (15,24,28).

The importance of chronic kidney disease (CKD) in long-term PN patients (10-14) and in intestinal transplant recipients is becoming more widely recognized (14,29-34). Most of the data in these reports, however, are in the adult IF population. A study on adult IF patients published in 1993 showed a 3.5% decline in creatinine clearance per year in 33 patients who received PN for an average of 8 years (11). Just over half of these patients also had impaired renal tubular function, as measured by tubular resorption of phosphate. A more recent study from 2014 in 33 adult home PN and 22 intestinal transplant patients, showed an annual decline in eGFR of 2.8% and 14.5% over 7 and 6 years, respectively (14). Another retrospective study on 70 Finnish patients with pediatric onset IF, published in 2018 (35), showed that 20 (29%) patients had decreased eGFR (defined as <89 ml · min⁻¹ · 1.73m⁻2 or <62 ml · min⁻¹ · 1.73m⁻2 for those ages 12–19 months) at a median age of 5.7 years following a median PN duration of 3.2 years.

Interestingly, none of these 20 patients with decreased renal function had a history of renal disease, stones, or any evidence of nephrocalcinosis on most recent US, compared with the other 50 patients with normal renal function. The 20 patients with decreased renal function had significantly longer duration of PN (3.2 vs 0.9 years) and shorter percentage of age-adjusted small bowel length remaining (22% vs 32%) compared with the 50 patients with normal renal function. The authors, however, did not use a single consistent method for calculating eGFR; various methods including the Schwartz equation, the Bedside Schwartz equation, as well as 51Cr-EDTA measurements were used.

In our study, renal function was assessed by measuring eGFR calculated by the modified Schwartz formula. It is well recognized that creatinine is a poor marker of renal function, especially in the early stages of renal damage. It is also now known that PIF patients receiving long-term PN have abnormal body composition characterized by low fat free mass and high fat mass (36). It would, therefore, be interesting to note if formal assessments of renal function, such as 51Cr-EDTA measurements would have made a difference to our findings. Given the retrospective nature of the data collection, this was not possible to assess, and it remains as one of the limitations of our study.

Data published from our IRP showed that a large proportion of PIF patients have increased echogenicity and/or nephrocalcinosis on surveillance abdominal US (15). These sonographic abnormalities were associated with prolonged PN exposure, the presence of a stoma, and a shorter colonic remnant. Interestingly in our present study, no differences were found in PN exposure, the presence of a stoma or a shorter colonic remnant between the 2 groups, although a slight, nonsignificant trend towards a shorter colonic remnant was noted in patients with an abnormal renal US (Group 1). This may have been because of smaller study numbers in each group.

The cause of the sonographic abnormalities in PIF patients still remains unclear, but is likely to be multifactorial. Although the findings from the original study by Kosar et al indicated an association between prolonged PN use and radiological renal abnormalities, their medium and long-term clinical significance was unknown. Our results now indicate that PN has no impact on eGFR and renal tubular function for up to 2 years of follow-up. This is one of the first studies examining the longer term impact of increased echogenicity and/or nephrocalcinosis in PIF patients.

Hyperoxaluria, hypercalciuria, and hypocitraturia are known risk factors for the development of renal calculi (10,20-22,26,37) and are associated with nephrocalcinosis. Renal tubular function in our group of patients was assessed at 2 years using urine calcium to creatinine ratio, urine calcium to citrate ratio, urine citrate to creatinine ratio and urine oxalate to creatinine ratio. Although a significant number of patients had abnormal urinary tests as can be seen in Table 4, our results show an equal distribution of these markers of tubular dysfunction in patients with and without increased echogenicity/nephrocalcinosis on US. Given the retrospective nature of this study and the lack of sufficient data, we were unable to assess renal tubular function using beta-2 microglobulin or tubular resorption of phosphate. It would be interesting to see if these markers or the ones already used for this study show significant abnormalities or changes over time in this cohort of patients in future prospective analyses.

Cyclical dehydration may also be a risk factor for the development of long-term CKD in pediatric IF patients on PN. Most of our patients had urine collected in the morning, a time when they are at their most hydrated state and getting off PN. This may have skewed the results of the urinary markers, and different results could have potentially been obtained in the later hours of the day. In future prospective studies, it would be interesting to see if any differences in these ratios would be observed if urine collection was timed just before PN hook-up in the evening, a time when they would theoretically be at their most dehydrated state.

Our results also show changes in the sonographic renal abnormalities over time in both groups of patients. Seven out of 25 patients with normal US images at baseline developed abnormalities at 2-year follow-up, whereas 9 out of 15 patients with abnormal findings at the outset saw normalization of these findings by the 2-year mark.

There was no association between PN use at the time and these sonographic changes. It is, however, important to note that the diagnosis of increased renal parenchymal echogenicity can be subjective (38,39), with 1 study showing a mean variation of 3% between measurements obtained by 2 different individuals on identical sonographic images (38), and another study showing a coefficient of variation of 0.83 on repeated measurements made by 2 different individuals (40). Interoperator differences could, therefore, have affected these results to some degree.

CONCLUSION

In summary, a large proportion of PIF patients with a history of prolonged PN exposure have increased echogenicity and/or nephrocalcinosis on US. In the short-term over a 2-year follow-up, however, the presence of these abnormalities has no impact on eGFR or renal tubular function. Further long-term studies are required to evaluate the significance of these sonographic abnormalities over time.

What Is Known

• A gradual decline in renal function is a known complication for adult patients with prolonged parenteral nutrition exposure. Data on pediatric patients are limited.

• Sonographic abnormalities, such as nephrocalcinosis and increased renal echogenicity have been reported in pediatric intestinal failure patients with prolonged parenteral nutrition exposure.

What Is New

• In the short-term, sonographic abnormalities, such as nephrocalcinosis or increased renal echogenicity have no impact on estimated glomerular filtration rate or renal tubular function in pediatric intestinal failure patients.