Pastore, S.*; Londero, M.*; Barbieri, F.*; Di Leo, G.†; Paparazzo, R.‡; Ventura, A.*
Parenteral nutrition (PN) is considered a lifesaving treatment in children with intestinal failure, mainly because of short bowel syndrome and congenital gut dysfunction, such as epithelium defects and motility anomalies. Several complications have been, however, recognized. In particular, it has been documented that children receiving long-term PN frequently develop metabolic bone disease. The development of bone disease associated with long-term PN was attributed to several factors related to PN preparation: changes in the balance of calcium and phosphorus intake, excess or low administration of vitamin D, deficiency of vitamin K, fluoride or copper, and aluminium toxicity (1). Nevertheless, it has not yet been well established whether PN alone or also the disease responsible for intestinal failure could determine the alterations of bone mineral status.
As bone mineral density (BMD) correlates well with fracture risk, all such patients should undergo BMD evaluation during PN treatment. Presently the method of choice for bone assessment is the dual-energy x-ray absorptiometry (DXA) (1). In children, BMD measures must be related to age-specific and sex-specific normative values (z-score).
Various attempts have been made to treat intestinal failure–associated bone disease, both by modifying PN composition and by administering substances such as calcitonin, parathyroid hormone, and sexual hormones. Bisphosphonates are now well established as the major drug used in adults for the treatment of bone disease associated with excessive reabsorption. Nishikawa et al (2) showed that intravenous pamidronate improved BMD in the lumbar spine after pamidronate infusion in patients receiving home PN for short bowel syndrome. Haderslev et al (3) studied the efficacy of intravenous clodronate therapy compared with placebo in patients receiving long-term home PN. The study included 20 patients, 10 in the intervention arm and 10 receiving placebo. Patients in the clodronate arm received 1500 mg of drug every 3 months for 1 year. The mean BMD of the lumbar spine and hip increased compared with placebo; however, the increase was not significant.
In 2006, Raman et al (4) assessed the effect of intravenous pamidronate on BMD prospectively in 11 patients receiving home PN. The authors observed that these patients had a trend toward improvement in the T score of lumbar spine and hip after a minimum of 18 months of therapy, which became significant for the spine when those patients receiving steroids were excluded.
The use of bisphosphonates has become a widely accepted treatment in children and adolescents with various other bone disorders, including osteogenesis imperfecta, Paget disease, hypercalcemia of malignancy, and myositis ossificans (5). Nevertheless, there have been no paediatric studies on bisphosphonates use for intestinal failure–associated bone disease.
We describe the long-term outcome of 6 paediatric patients with intestinal failure–associated bone disease treated with cyclical intravenous pamidronate.
In this observational study we considered retrospectively all of the children receiving cyclic home PN for at least 3 years. All of the patients underwent a DXA scan to evaluate bone mineral status. Exclusion criteria were intermittent PN, corticosteroid treatment, and metabolic imbalance between calcium and vitamin D.
Calcium and phosphorus administrations were approximately 30 and 25 mg/kg daily, respectively, with a daily maximum dose of 500 to 600 mg of calcium and 450 to 500 mg of phosphorus as solution. Vitamin D dose was between 200 and 400 IU depending on oral intake and 25-OH vitamin D blood levels. The calorie-to-nitrogen ratio was higher than 150 as recommended (5,6).
Individual calcium, phosphorus, calorie to-nitrogen ratio, and vitamin D intake at the date of first DXA are indicated in Table 1. These intakes were not changed subsequently. Energy intake was mainly by PN (Table 1).
BMD of the lumbar spine was evaluated using DXA scan; all BMD measurements were given as grams per square centimetre and correlated to a z-score. All of the patients with a z-score lower than −2.5, or with a z-score −1 to −2.5, and clinical features of bone disease (bone pain or spontaneous fractures) were treated with intravenous pamidronate, 30 mg/m2 once per month for 6 months consecutively.
We obtained informed consent from the parents of the children. The DXA scans were obtained after each cycle of pamidronate administration. We followed up the patients for a mean of 108 months. Calcium homeostasis was assessed before and after each pamidronate administration.
Descriptive analysis was performed and, given the skewness of the data, medians and interquartile ranges (IQRs) were reported as measures of central tendency. Groups of z-score values were compared for significant differences using Wilcoxon rank sum test.
Six paediatric patients (5 boys and 1 girl) on extremely long PN (at least 3 years) were considered: 2 were affected by short bowel disease (SBS) and 4 by mucosal anomalies (MAs). A total 2 of 4 with MA had autoimmune enteropathy, 1 had tufting enteropathy, and 1 had microvillus atrophy. In these patients osteoporosis has been documented by DXA with a z-score lower than −2.5 SD (standard deviation) (Table 2). None of them experienced fractures. All of the patients with bone disease had started PN before 2000.
The diagnosis of bone disease was made after a median period of 127.5 PN months. They all showed a significant improvement in BMD evaluated after 6 and 12 months of pamidronate treatment.
The median z-score of the patients before starting pamidronate treatment was −3.66 SD (IQR 1) compared with −3.21 SD (IQR 0.80) after 6-month treatment and −2.2 SD (IQR 0.94) after 12-month treatment (P = 0.03). The results were confirmed at the last DXA after a mean 108-month follow-up.
Levels of serum calcium, phosphorus, and vitamin D were normal in all of the patients. In 5 of 6 patients, calciuria had already increased at the base line (ratio of urinary calcium/urinary creatinin >0.3), and in 2 of these cases there was a further increase subsequent to pamidronate infusion. Alkaline phosphatase level was not increased; almost all of our patients had high levels of osteocalcin, a marker of bone turnover, also at the beginning of treatment (Table 3). Furthermore, the level of osteocalcin tended to increase following pamidronate infusion. Kidney ultrasound before and after treatment with pamidronate showed no sign of nephrocalcinosis.
A total of 5 of 6 patients had hypocalcaemia in the 2 days following treatment in both cycles that was corrected by increasing calcium intake in the 3 days following pamidronate infusion. One patient experienced cramps of lower limbs. Three patients complained of bone pain during infusion that lasted until the following day, which was treated by administration of antalgic therapy. One patient had a pseudoinflammatory reaction during one of the infusions (increased erythrocyte sedimentation rate, polymerase chain reaction, and fever); the episode was self-limiting and the patient fully recovered after administration of 1 dose of steroid. All of these adverse effects were mild and did not require treatment termination.
Osteoporosis is a serious adverse effect of PN in children as well as in adults (7–9). Many factors seem to be involved in the development of intestinal failure–associated bone disease. In the past, great importance was attributed to the role of aluminium contamination of total parenteral nutrition ingredients, namely the amino acids. Excess doses of vitamin D may also play a role in metabolic bone disease. In the absence of the need for enteral calcium absorption, the vitamin D contained in the multivitamin preparations used in PN solutions must not exceed normal daily intake levels because it can be toxic to bone. Doses of 400 IU/day in paediatric patients have not been associated with complications (10); however, the single most important contributing factor to bone disease is a negative calcium balance (11,12). Besides these single factors, the main problems with PN are the possible precipitation of calcium and phosphorus in PN solutions and the possibility of increased urinary mineral losses. Another potential cause of poor bone mineralisation in patients is vitamin K deficiency. Some vitamin K–dependent proteins (such as osteocalcin and protein S) are involved in bone formation, and so vitamin K supplementation (1 mg daily) may reduce calciuria, although the real effects are not well known (13).
In our institute composition of PN has been the same since 1975; at that time mixtures with glucose, aminoacids, lipids, electrolytes, and vitamins were introduced. We have always used pure crystalline aminoacids, Trophamine in newborns and then Freamine plus sodium chloride, sodium lactate, gluconate calcium, potassium chloride and potassium aspartate, magnesium sulfate and Addamel (1 vial/day). We have always used Vitalipid and Soluvit for vitamin intake. Therefore, we think that in our series PN is not the main cause of bone disease; probably the disease responsible for intestinal failure can determine alterations of bone mineral status.
The study by Diamanti et al (14) did not find significant correlations between bone mineral status, PN duration, and nutrient intake, but the authors observed that patients with congenital gut dysfunctions, such as epithelium defects and motility anomalies, were at major risk for developing metabolic bone disease. We unfortunately considered a small cohort of patients to be able to find a correlation between metabolic bone disease and any underlying disease responsible for intestinal failure.
Some studies demonstrated the utility of bisphosphonates for improving bone mass density in adults (2–4); nevertheless, there have been no paediatric studies about bisphosphonates effect on intestinal failure–associated bone disease.
Bisphosphonates are structural analogues of inorganic pyrophosphate, but they are not subject to enzymatic and chemical breakdown. As the bisphosphonate of choice, we have selected pamidronate, one of the most used compounds, owing to its proven efficacy in treating metabolic bone disease in adults receiving long-term PN (2,4) and in children affected by osteoporosis because of other causes (15,16).
For pamidronate, the classic administration scheme includes a 3-day infusion every 3 months. Steelman and Zeitler (17) evidenced that a single-day pamidronate infusion is equally efficacious. This regimen is more convenient in paediatric patients because it requires shorter hospitalisation time and it is less stressful for children. Because all of our patients receiving PN have intestinal failure, intravenous administration of pamidronate is necessary for obtaining optimal drug levels.
Diamanti et al (14) documented that there is a high prevalence of metabolic bone disease in patients with intestinal failure undergoing long-term treatment with PN, but bone mineral status increases regularly under PN treatment too. Moreover, they showed that there is a linear correlation between BMD and weight and height in patients.
In the present report, all of the patients treated with bisphosphonates had a normal nutritional state and kept following their own growth percentile for weight and for height, except 1. This patient (patient 2 in Table 2) presented a severe malnutrition before pamidronate infusion and had a significant nutritional improvement during the treatment, with a significant z-score increase compared with the other ones. Nevertheless, all our patients had a significant z-score improvement after 2 cycles of pamidronate treatment (Fig. 1). In our small sample anthropometrical variables (weight, height, and body mass index) are not related with the z-score trend (Table 4).
Our patients had normal levels of calcium, phosphorus, and vitamin D and proper nutrient intake. Osteocalcin levels were above upper limits of normality, which is a bone turnover marker, so its level can be correlated with bone growth. At the last follow-up, DXA scan showed that no patients had a z-score lower than −2.5 (Table 2); moreover, nobody developed bone fractures during the 108-month follow-up.
The study weakness is the absence of a control group, so it is difficult to assess the real role of bisphosphonates in BMD improvement and to establish whether there is a true biological benefit with the use of pamidronate. Moreover, we considered just a small patient population because the condition of interest is rare and we considered only patients on long-term PN (at least 3 years) and with no other risk factors for bone disease.
Finally, to the best of our knowledge, we are not aware of any study evaluating pamidronate effect on bone disease in children treated with long-term PN. In our experience, described in the present study, pamidronate is effective for improving BMD and safe in patients with intestinal failure–associated bone disease.
1. Kroger H, Kotaniemi A, Vainio P, et al. Bone densitometry of the spine and femur in children by dual energy x ray absorptiometry. Bone Miner 1992; 17:75–85.
2. Nishikawa RA, Siepler SE, Siepler JK, et al. Intravenous pamidronate improves bone mineral density in home parenteral nutrition patients. Clin Nutr 2003; 22:S88.
3. Haderslev KV, Tjellesen L, Sorensen HA, et al. Effect of cyclical intravenous clodronate therapy on bone mineral density and markers of bone turnover in patients receiving home parenteral nutrition. Am J Clin Nutr 2002; 76:482–488.
4. Raman M, Aghdassi E, Baun M, et al. Metabolic bone disease in patients receiving home parenteral nutrition: a Canadian study and review. J Parenter Enteral Nutr 2006; 30:492–496.
5. Peters C, Fischer JE. Studies on calorie to nitrogen ratio for total parenteral nutrition. Surg Gynecol Obstet 1980; 151:1–8.
6. Wesley JR, Coran AG. Intravenous nutrition for the pediatric patient. Semin Pediatr Surg 1992; 1:212–230.
7. Pironi L, Labate AM, Pertkiewicz M, et al. Prevalence of bone disease in patients on home parenteral nutrition. Clin Nutr 2002; 21:289–296.
8. Shike M, Harrison JE, Sturtridge WC, et al. Metabolic bone disease in patients receiving long term total parenteral nutrition. Ann Intern Med 1980; 92:343–350.
9. Harrison JE, Jeejeebhoy KN, Track NS. The effect of total parenteral nutrition (TPN) on bone mass. In: Coburn JW, Klein GL, eds. Metabolic Bone Disease in Bone Parenteral Nutrition. Baltimore, MD: Urban and Schwarzenberg; 1985:53.
10. Breaux CW Jr, Dukie D, Georgeson KE, et al. Calcium phosphate crystal occlusion of central venous catheters used for total parenteral nutrition in infants and children: prevention and treatment. J Pediatr Surg 1987; 22:829–832.
11. Buchman AL, Mourkarzel A. Metabolic bone disease associated with total parenteral nutrition. Clin Nutr 2000; 19:217–231.
12. Koo WW, Tsang RC, Succop P, et al. Minimal vitamin D and high calcium and phosphorus needs of preterm infants receiving parenteral nutrition. J Pediatr Gastroenterol Nutr 1989; 8:225–233.
13. Knapen MHJ, Hamulyak K, Vermeer C. The effect of vitamin K supplementation on circulating osteocalcin (bone Gla protein) and urinary calcium excretion. Ann Intern Med 1989; 111:1001–1005.
14. Diamanti A, Claudia Bizzarri C, Basso MS, et al. How does long-term parenteral nutrition impact the bone mineral status of children with intestinal failure? J Bone Miner Metab 2010; 28:351–358.
15. Falk MJ, Heeger S, Lynch KA, et al. Intravenous bisphosphonate therapy in children with osteogenesis imperfecta. Pediatrics 2003; 111:573–578.
16. Srivastava T, Alon US. The role of bisphosphonates in diseases of childhood. Eur J Pediatr 2003; 162:735–751.
17. Steelman J, Zeitler P. Treatment of symptomatic pediatric osteoporosis with cyclic single-day intravenous pamidronate infusions. J Pediatr 2003; 142:417–423.