PEDIATRICS Vol. 106 No. 4 October 2000, pp. 849-851

EXPERIENCE AND REASON:
Successful Long-Term Peritoneal Dialysis in a Very Low Birth Weight Infant with Renal Failure Secondary to Feto-Fetal Transfusion Syndrome

The frequency of acute renal failure in the very low birth weight (VLBW) infant is estimated to be between 6% and 8%.^1-3 Although peritoneal dialysis (PD) is considered the treatment of choice for most newborns and older infants with end-stage renal failure, PD has been used only on a limited basis in the VLBW infant. Many newborns who require PD do not have primary renal failure, but require dialysis secondary to prenatal, perinatal, and/or postnatal conditions resulting in poor renal perfusion and subsequent renal failure including placental insufficiency, congenital cardiac anomalies, respiratory failure, and/or sepsis.³ The decision to offer dialysis usually occurs after failure of conservative management. Indications for dialysis are based on evidence of intractable acidosis, hyperkalemia, and/or progressive azotemia. Premature newborns, especially VLBW infants, historically have not been considered good candidates for PD because of technical feasibility and high morbidity from infection, poor weight gain, and complications from inadequate dialysis.^1-3 Although there are several published reports describing successful short-term PD in VLBW premature infants for acute transient renal failure, there is no literature describing the use of PD to treat chronic renal failure in this patient population.^2-6 We report what we believe is the first case of a surviving VLBW twin with renal failure secondary to feto-fetal transfusion syndrome who was successfully managed with long-term PD for 14 weeks.

	CASE REPORT

The patient, twin B, was an inborn, 930-g male product of a monozygotic twin gestation complicated by twin-to-twin transfusion. Twin B was the smaller donor twin whose prenatal course was complicated by significant oligohydramnios, while the larger 1230-g recipient twin A developed polyhydramnios, poor cardiac function, and mild hydrops resulting from volume overload. The twins were delivered via emergency cesarean section at 29 and 5/7 weeks' gestation because of worsening cardiac function in the recipient twin. Apgars for twin B were 7 and 8, at 1 and 5 minutes, respectively. Shortly after delivery, both twins were intubated for worsening respiratory distress. Twin B, because of worsening respiratory failure, was converted to high-frequency oscillatory ventilation within the first 24 hours. His hospital course was complicated by anuria that persisted from birth despite initial volume and diuretic therapy. He additionally developed progressively worsening hypotension over the first 48 hours of life requiring pressor support with dopamine, dobutamine, and epinephrine, up to a maximum of 20, 15 and .3 µg/kg/minute, respectively. Echocardiac evaluation on day 2 revealed a hemodynamically significant patent ductus arteriosus that was surgically ligated (medical therapy with indomethacin was not attempted because of anuria). Twin B remained unstable postoperatively requiring continued pressor therapy that gradually improved through the first week of life allowing weaning of support. Renal ultrasound on day 2 demonstrated increased renal parenchymal echogenicity, but otherwise normal appearing kidney and bladder architecture, and documented renal arterial flow bilaterally by Doppler. Renal size dimensions included a longitudinal length of 2.9 cm on the right, and 3.0 cm on the left, within the normal range for gestational age.

Persistent anuria through the first week of life required management with fluid restriction down to 45 to 50 mL/kg/day limiting our ability to provide adequate nutrition. Over the course of this period, twin B developed worsening electrolyte abnormalities including a mild metabolic acidosis and hyperkalemia, in addition to a rising serum blood urea nitrogen (BUN) to 47 mg/dL and serum creatinine to 5.2 mg/dL. Fluid overload further complicated his respiratory distress contributing to the development of pulmonary edema and resulting in significant increased support on high-frequency ventilation. Because of the above concerns, coupled with continued anuria, informed consent was obtained from the parents for PD. A Tenckhoff PD catheter was placed in the right lower quadrant on day 7, and dialysis was started the next day.

Dialysis was initiated with small 10 mL (10 mL/kg) dwells over 45 minutes with 24 cycles per day using a 1.5% glucose dialysate. Because of poor ultrafiltrate removal, the dialysate glucose concentration was increased to 2.5%, and dwell volumes were gradually increased to 25 mL/cycle (25 mL/kg) over the next week as tolerated. Twin B responded well to this regimen with significant fluid removal after the first week, decreasing his weight from 1.56 kg to 1.2 kg, as well as his serum BUN and creatinine levels to 21 mg/dL and 2.1 mg/dL, respectively. After 1 month of clinically adequate PD, the mean ultrafiltrate volume was 121 mL/d and the calculated Kt/V equaled 2.1. On day 25, twin B began producing small quantities of urine (<.50 mL/kg/hour). Urinalysis at this time revealed a pH of 7.0, 1 to 4 red blood cells per high-power field, sodium 29 meq/L, potassium 13 meq/L, chloride 26 meq/L, and creatinine of 4.2 mg/dL. Despite his new onset urine output, twin B remained oliguric and unable to come off PD. After 8 weeks of dialysis, PD was temporarily discontinued after an initial revision of the PD catheter because of intermittent obstruction. While off PD, twin B remained oliguric, and his serum creatinine increased from 1.8 to 3.7 mg/dL. In addition, he suffered an increased intolerance of feedings and worsening respiratory distress because of fluid overload, resulting in a reinstitution of dialysis at <72 hours.

Over the course of the next several weeks, an increased PD cycle volume to 60 mL (25 mL/kg) was obtained, an adequate volume to allow for conversion to continuous cyclic peritoneal dialysis (CCPD), which twin B tolerated well. Follow-up renal studies at 3 months of life included an ultrasound demonstrating worsening bilateral corticomedullary differentiation, persistent increased echogenicity but good interval growth to a length of 3.7 cm on the right and 3.9 cm on the left. Normal forward blood flow by Doppler to both kidneys was observed. Additional studies included a voiding cystourethrogram noting grade II vesicoureteral reflux on the left, and a dimercaptosuccinic acid scan showing markedly diminished uptake and excretion bilaterally consistent with parenchymal disease.

After 14 weeks of dialysis (31/2 months chronological age), twin B underwent another peritoneal catheter revision and bilateral inguinal hernia repair. Dialysis was again discontinued at this time to allow for surgical recovery. Postoperatively, however, urine output gradually improved over the next several days to >1 to 2 mL/kg/hour while maintaining stable hemodynamics and serum creatinine levels (1.8 mg/dL). Over the course of the next week, adequate urine output with stable electrolytes and renal function was maintained. Creatinine clearance at this time was measured at 9 mL/minute/1.73 m² after a 24-hour urine collection (estimated glomerular filtration rate (GFR) 11 mL/minute/1.73 m² using the Schwartz formula). Twin B was able to tolerate good enteral intake with an appropriate weight gain up to 3.7 kg, and he was successfully discharged from the hospital off PD therapy.

While on PD, serum electrolytes, including calcium and phosphorus, were maintained within normal ranges without the need for supplemental therapy, except for sodium citrate which was begun to treat a mildly persistent metabolic acidosis. Significant problems with hypotension during PD were not experienced. Enteral feeds were begun on day 18 and slowly advanced up to 175 mL/kg/day over the course of the next 2 months (150 kcal/kg/day, 6 g of protein/kg/day), with a resulting slow increase in body weight, length and head circumference. Enteral nutrition consisted of expressed breast milk supplemented with human milk fortifier, ProMod (Ross Products, Columbus, OH) for added protein calories, and Nephrocaps (Fleming and Co, Fenton, MO) as a daily vitamin supplement. After discontinuation of PD, caloric intake was decreased to 130 kcal/kg/day and protein to 4 g/kg/day. A gastrostomy tube was placed at 12 months to supplement oral feedings and improve weight gain. At 15 months, feedings were converted to PM 60/40, and with solid foods, caloric intake was targeted to 3 g/kg/day of protein and 100 to 110 kcal/kg/day.

Twin B had several PD-related complications during his hospital course to include acute infectious peritonitis. On day 18, a cloudy peritoneal dialysate was noted (4600 white blood cells/mL) with a subsequent positive culture for Pseudomonas aeruginosa. After initial broad-spectrum antibiotic coverage, therapy was narrowed to amikacin and cefipime given both intravenously and intraperitoneally via the dialysate for a total 14 days, successfully establishing negative cultures without the need for discontinuation of dialysis or catheter removal. Additional significant complications included the development of bilateral scrotal edema, inguinal hernias, and several episodes of dialysis catheter obstruction as previously discussed. All of these complications were successfully managed either medically or surgically allowing for the resumption of effective dialysis.

Twin B had serial cranial sonograms performed through the first month of life, all of which were normal without evidence of intraventricular hemorrhage or ischemic injury. His respiratory course was complicated, as he remained on high-frequency ventilation until day 40 when he was successfully converted to conventional ventilation with extubation to nasal cannula 4 days later. He remained on oxygen with underlying mild chronic lung disease at the time of discharge, however, was able to successfully wean to room air several months later. He additionally was discharged from the hospital on erythropoietin and iron therapy (begun shortly after going on PD) for management of his underlying anemia.

Twin B is currently 15 months old and has a urine output of 2 to 3 mL/kg/hour off PD, and is maintaining stable serum BUN (70 mg/dL) and creatinine levels (1.7 mg/dL). His estimated GFR has modestly improved since discharge up to near 20% of normal. He has achieved a weight of 5450 g (<5th percentile corrected), and a length of 72 cm and head circumference of 45 cm (both at 5th percentile corrected). He remains on erythropoietin, iron, and supplemental sodium citrate, but otherwise is not on other routine medications. Serum parathyroid has consistently remained in the normal range without evidence of secondary hyperparathyroidism. He presently is receiving scheduled occupational and physical therapy for mild to moderate gross and fine motor developmental delays, but otherwise is developmentally appropriate with normal vision and hearing.

	DISCUSSION

Top Introduction Discussion Conclusion References

We believe this article describes the first successful use of long-term PD on a VLBW with renal failure. Although our patient currently has recovered adequate renal function to sustain him off dialysis, his present poor GFR virtually guarantees that he will ultimately require renal replacement in the form of long-term dialysis or transplantation as he grows. However, given his relatively good clinical outcome, we feel that this case supports the consideration of PD in the VLBW after individual evaluation of nonrenal comorbidity and appropriate counseling of families.

Previous reports in the literature involving PD in the preterm VLBW infant have been limited to cases describing only brief transient use of dialysis. In 1980, Kanarek et al² described a successful case of 30 hours of PD in a VLBW infant instituted for metabolic acidosis and hyperkalemia. Sizun et al⁶ in 1993 reported 3 cases of PD in VLBW infants. All 3 infants were suffering from acute renal failure secondary to either hypoxia or sepsis, and underwent PD over a range of 31 to 70 hours. Although 1 case of PD was ineffective, the remaining 2 cases successfully corrected fluid and electrolyte balance with eventual restoration of urine output. Unfortunately, all 3 infants subsequently died of extra-renal causes after the cessation of dialysis. A retrospective review of PD in neonates by Matthews et al¹ in 1990 demonstrated efficacy of PD in neonates <60 days old. Only 8/31 of the infants were premature (<38 weeks' gestational age) and the average weight was 3.67 kg at the time of catheter insertion, with the smallest being 1.9 kg. The mean duration of dialysis was 16 days, with the longest lasting 73 days. Overall, 19/31 infants died at intervals ranging from 1 to 90 days postcatheter insertion. Five of the 12 survivors remained on chronic dialysis awaiting renal transplantation at the time of publication, although specific gestational age and weights of these survivors were not discussed. Matthews' review showed that infants <60 days old whose renal failure was isolated (without other significant end-organ injury), were more likely to undergo successful dialysis and have better survival than infants with multi-organ involvement. Additional reviews have shown mortality rates ranging from 10% to 35% in the population of infants <1 year old undergoing PD.⁷ Infants who require PD and are anuric or oliguric have an even higher mortality rate (64%) in comparison to infants with adequate urine output (20%).⁷

PD has a theoretical advantage in the premature neonate secondary to the large peritoneal surface area to body ratio that provides improved dialysis efficiency.³ PD is contraindicated in certain circumstances such as necrotizing enterocolitis or recent abdominal surgery. PD is preferred over other techniques such as hemodialysis in small infants due to vascular access problems and coagulation control. There are case reports of continuous arteriovenous hemofiltration and veno-venous syringe driven ultrafiltration being used in infants weighing <1000 g, however, these techniques have only been demonstrated to be successful over the course of hours as opposed to days.³ Reported complications with PD in infants include peritonitis, exit site infections, leaks around the exit site, catheter obstruction requiring revision or replacement, abdominal wall hernias, and several case reports of bowel perforation secondary to erosion from the dialysis catheter tip.^1,4,6,8

A noteworthy aspect of our case, in addition to our patient's gestation and size, was the cause of our patient's renal failure. Transient renal failure in the donor twin of feto-fetal transfusion syndrome has been previously reported by Christensen et al.⁹ They report the case of transient renal insufficiency occurring in the first week of life in the smaller 805-g donor twin from a 28-week gestation. Acute renal failure resulted in a maximum serum creatinine of 2.8 mg/dL, followed by subsequent recovery of renal function, and a decrease of serum creatinine down to .5 mg/dL without medical intervention or the need for dialysis. Renal sonography demonstrated normal parenchyma and size with adequate blood flow bilaterally. They hypothesized that the renal injury was probably secondary to chronic renal hypoperfusion in utero secondary to feto-fetal transfusion syndrome.⁹ Our patient's renal failure was most likely similar to the pathophysiology reported in this case. Additional postnatal factors occurring in the first several days of life, however, including severe respiratory failure, systemic hypotension, and a patent ductus arteriosus requiring surgical closure, complicated our patient's course and very likely may have contributed to his renal injury. Other reports have described severe renal injury and dysfunction postnatally in feto-fetal transfusion syndrome resulting in renal cortical necrosis secondary to release of thromboplastin and/or necrotic emboli from a demised fetus.¹¹

We feel an important factor in our patient's successful outcome was the ability to provide good enteral nutrition after the initiation of PD. Our patient was able to eventually consume his entire caloric intake orally after he reached appropriate gestation, but subsequently required gastrostomy placement to supplement his diet. Dietary goals should include at least 100% of the estimated daily requirement for age as well as excess protein supplementation while undergoing PD. For infants who are unable to or do not tolerate oral feeds, early gastrostomy placement should be considered. Our patient was able to achieve linear growth by the establishment of early complete enteral feedings, and thereby avoid the morbidity and limitations associated with parenteral nutrition and inadequate caloric intake, including the development of renal osteodystrophy. Although twin B has not been treated with growth hormone to date, he presently is undergoing evaluation to begin therapy in an attempt to augment his present growth velocity.

Weaning of PD support may be considered when endogenous urine output increases and/or the serum creatinine significantly declines from its established baseline in the absence of a concurrent change in the dialysis regimen. It is possible to contribute to ongoing oliguria by maintaining a hypovolemic state while on dialysis. Although our patient did have an eventual increase in urine output after the cessation of PD postsurgery allowing the discontinuation of therapy, we feel that the duration of his therapy was appropriate. He did fail an initial trial off PD 2 months into treatment. Serial renal Doppler studies documented good blood flow bilaterally to both kidneys, and his serum creatinine, predicted GFR, and endogenous urine output remained unchanged before discontinuation of PD. Despite an absence of changes in these variables in our patient, we feel it is likely that we would have begun to see improvement in these variables and been able to begin to gradually wean support had PD treatment continued. However, the appropriate timing of weaning of PD can be difficult to determine in the face of ongoing therapy. Given this, periodic trials of discontinuation of PD, even in the absence of improvement in predictive variables for renal function, should be considered once maintenance therapy has been established and the patient is stable.

	CONCLUSION

Top Introduction Discussion Conclusion References

In feto-fetal transfusion syndrome, anuria in the donor twin, even when quite prolonged, can be followed by recovery of adequate renal function and potential discontinuation of dialysis support. The use of PD in the VLBW infant is technically feasible and can provide effective long-term therapy, and should be considered for supportive management of VLBW infants with prolonged renal failure after obtaining the appropriate consent. Institution of PD in the tiny infant remains technically difficult and not without both acute and chronic risks, and a decision to perform PD on the VLBW infant must be carefully examined on an individual case basis. Recovery of long-term renal function is difficult to predict, and if PD is acutely used, the potential need for continued long-term support in the face of survival, including dialysis as a bridge for renal transplantation, must be carefully explained to the family.

Karin E. Rainey, MD
Robert J. DiGeronimo, MD
Jose Pascual-Baralt, MD
Wilford Hall USAF Medical Center
Lackland AFB
San Antonio, TX 78236

	FOOTNOTES

The opinions and assertions contained herein are the private views of the authors and are not construed as official or as reflecting the views of the Department of Defense (and/or the Department of the Air Force).

Received for publication Oct 5, 1999; accepted Apr 27, 2000.

Reprint requests to (R.J.D.) Department of Pediatrics, Wilford Hall, USAF Medical Center, 59th MDW/MMNP, 2200 Berquist Dr, Suite 1, Lackland AFB, San Antonio, TX 78236. E-mail: robert.digeronimo{at}59mdw.whmc.af.mil

	ABBREVIATIONS

VLBW, very low birth weight; PD, peritoneal dialysis; BUN, blood urea nitrogen; GFR, glomerular filtration rate.

	REFERENCES

Top Introduction Discussion Conclusion References

1. Matthews D, West K, Frederick J, Peritoneal dialysis in the first 60 days of life. J Pediatr Surg. 1990; 25:110-116 [CrossRef][Medline]
2. Kanarek K, Root E, Sidebottom R, Williams P Successful peritoneal dialysis in an infant weighing less than 800 grams. Clin Pediatr. 1982; 21:166-169
3. Coulthard MG, Vernon B Managing acute renal failure in very low birthweight infants. Arch Dis Child. 1995; 73:187-192 [Free Full Text]
4. Huber R, Fuchshuber A, Huber P Acute peritoneal dialysis in preterm newborns and small infants: surgical management. J Pediatr Surg. 1994; 29:400-402 [CrossRef][Medline]
5. Coulthard M, Sharp J Haemodialysis and ultrafiltration in babies weighing under 1000 g. Arch Dis Child. 1995; 73:162-165 [Abstract/Free Full Text]
6. Sizun J, Giroux J, Rubio S, Guillois B, Alix D, De Parscau L Peritoneal dialysis in the very low-birth-weight neonate (less than 1000 g ). Acta Paediatr. 1993; 82:488-489 [CrossRef][Medline]
7. Ellis E, Pearson D, Champion B, Wood E Outcomes of infants on chronic peritoneal dialysis. Adv Perit Dial. 1995; 11:266-269 [Medline]
8. Verrina E, Zacchello G, Perfumo F, Clinical experience in the treatment of infants with chronic peritoneal dialysis. Adv Perit Dial. 1995; 11:281-284 [Medline]
9. Christensen A, Daouk G, Norling L, Postnatal transient renal insufficiency in the feto-fetal transfusion syndrome. Pediatr Nephrol. 1999; 13:117-120 [CrossRef][Medline]
10. Tapper D, Watkins S, Burns M, Hickman R, Avner E Comprehensive management of renal failure in infants. Arch Surg. 1990; 125:1276-1281 [Abstract/Free Full Text]
11. Heijst AF, Nijhuis JG, Bult P, Renal failure in the surviving monochorionic twin after death of the co-twin in utero. Pediatr Nephrol. 1996; 10:51-54 [CrossRef][Medline]