Jump to comment:

• Almost certainly a straw that could break the camel's back.
  Richard G Fiddian-Green
  Published on: 20 April 2005
• Necrotising Enterocolitis following blood transfusion
  Andrew Short
  Published on: 14 April 2005

• Published on: 20 April 2005

  Almost certainly a straw that could break the camel's back.

  • Richard G Fiddian-Green, FRCS, FACS

  Dear Editor,

  Andrew Short et al. draw attention to Marik and Sibbld's observation that, "change in intraluminal [in reality intramucosal or intramural :pHi) as distinct from intraluminal] pH, as measured by gastric tonometry, was more marked if the red cells had been stored for more than 15 days". Their study showed that if there was a benefical effect on tissue oxugentation it was confined to fresh blood. More im...

  Show More

  Dear Editor,

  Andrew Short et al. draw attention to Marik and Sibbld's observation that, "change in intraluminal [in reality intramucosal or intramural :pHi) as distinct from intraluminal] pH, as measured by gastric tonometry, was more marked if the red cells had been stored for more than 15 days". Their study showed that if there was a benefical effect on tissue oxugentation it was confined to fresh blood. More importantly they found an inverse association between the change in pHi and the age of the transfused blood (r = -.71; P <.001). In those patients receiving blood that had been stored for more than 15 days, the gastric intramucosal pH consistently decreased following the red blood cell transfusion increasing the severity of their reductive stress. Their findings were a refinement of earlier findings by Silverman and Tuma [1]. A more recent study has failed to show either benefit [improvement in pHi] or harm [decrease in pHi] even when age of the blood was considered [2]. The issues have been addressed more broadly but in a different context [3].

  There are so many factors that could precipitate an acute decline in energy charge in gut mucosa and hypothetically cause NEC, including premature feeding [4], that it must be very difficult if not impossible to manage neonates objectively without knowing the pHi [5]. Instituting feeds or giving RBCs in a neonate that has an abnormally low pHi would seem, for example, most unwise unless the interventions were shown to promptly return the pHi to normality.

  References

  1. Silverman HJ, Tuma P. Gastric tonometry in patients with sepsis. Effects of dobutamine infusions and packed red blood cell transfusions. Chest. 1992 Jul;102(1):184-8.

  2. Walsh TS, McArdle F, McLellan SA, Maciver C, Maginnis M, Prescott RJ, McClelland DB. Does the storage time of transfused red blood cells influence regional or global indexes of tissue oxygenation in anemic critically ill patients? Crit Care Med. 2004 Feb;32(2):364-71.

  3. Fiddian-Green RG. Open versus laparoscopy assisted colectomy. Lancet. 2003 Jan 4;361(9351):74; author reply 75-6.

  4. Schloerb PR, Wood JG, Casillan AJ, Tawfik O, Udobi K. Bowel necrosis caused by water in jejunal feeding. JPEN J Parenter Enteral Nutr. 2004 Jan-Feb;28(1):27-9.

  5. Hatherill M, Tibby SM, Denver L, Marsh MJ, Murdoch IA. Early detection of necrotizing enterocolitis by gastrointestinal tonometry. Acta Paediatr. 1998 Mar;87(3):344-5.

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 14 April 2005

  Necrotising Enterocolitis following blood transfusion

  • Andrew Short, Consultant Paediatrician
  • Other Contributors:
    • Andrew Gallagher, Munir Ahmed

  Dear Editor

  We read with interest the article questioning whether blood transfusion increases the risk of necrotising enterocolitis (NEC). We, like many paediatricians, are convinced of an association and would like to add to the anecdotal evidence so that in future greater thought will be given to the role of "top-up" transfusions in preterm infants.

  Case 1

  A male infant was born by caesarean...

  Show More

  Dear Editor

  We read with interest the article questioning whether blood transfusion increases the risk of necrotising enterocolitis (NEC). We, like many paediatricians, are convinced of an association and would like to add to the anecdotal evidence so that in future greater thought will be given to the role of "top-up" transfusions in preterm infants.

  Case 1

  A male infant was born by caesarean section for pre-eclampsia at 29 weeks gestation. There had been absent end diastolic flow on antenatal doppler scans and the baby was growth retarded with a birth weight of 990g (2nd centile).[2] doses of betamethasone were given antenatally. He developed respiratory distress, requiring ventilation for 1 day followed by nasal continuous positive airway pressure (CPAP) for a further 6 days. He started on parenteral nutrition (TPN) on day 4 and minimal enteral feeds with expressed breast milk began on day 5. He was on full oral feeds by day 11, when clinical signs of a patent ductus arteriosus (PDA) were detected. The duct closed clinically with a 5-day course of Indomethacin. By this stage he had become anaemic (Hb 8.4 g/dl) and he received a transfusion of 10 ml/kg packed cells on day 13 without complication.

  On day 24, at a corrected age of 33 +5 days, he was in 28% 02 with occasional brief desaturations. He was tolerating expressed breast milk feeds. His haemoglobin was 8.2 g/dl and a transfusion of 20 ml/kg packed cells was prescribed for anaemia that was judged to be symptomatic (increased oxygen requirement, desaturations, apnoea and poor weight gain). The blood was cross matched against maternal serum, screened for irregular antibodies and drawn up through a 40 micron filter before infusion over 4 hours. Frusemide (0.5 mg/kg) was given intravenously half way through the transfusion. Feeds were not stopped prior to the transfusion. There had been no clinical or haematological suspicion of sepsis and no suggestion of feed intolerance. One day before the transfusion the C Reactive Protein was less than 5 and total white cell count 8.0.

  Two hours into the transfusion the nursing staff noted that he was unsettled, and he developed a tachycardia (200 bpm) at the end of the transfusion. He passed a normal stool, but within 2 hours began to vomit with a tense distended abdomen. An abdominal x ray confirmed extensive radiological changes of NEC. He required intubation and transfer to a Surgical Unit, where laparotomy confirmed extensive small bowel involvement with 40% of the gut affected. He died 5 weeks after bowel resection.

  Case 2

  Following spontaneous rupture of membranes at 25 weeks gestation and treatment for chorio amnionitis, a female infant was delivered at 25 +2 weeks gestation. She required resuscitation at birth, including adrenaline, sodium bicarbonate and fluid boluses, but responded well and was ventilated for 2 days for respiratory distress syndrome. Her birth weight was 840g (50th centile). She required one top-up transfusion on day 3, which was given without complication.

  At 3 days of age she was weaned from ventilation to nasal CPAP, and then into air the following day. TPN was commenced on day 3, and continued until expressed breast milk feeds were fully established on day 12. She had episodes of desaturation related to feeds, which were treated with sodium alginate (Gaviscon) and Ranitidine. Breast milk fortifier was added because of poor weight gain at 20 days of age.

  On day 25 she was weaning off nasal CPAP, required 23% oxygen and had very occasional desaturations and bradycardias. Her haemoglobin was 8.9g/dl, and there was no indication of sepsis (C Reactive Protein of 8 and white cell count 15.3). She was prescribed a top-up transfusion of 20 mls/kg packed cells over 4 hours. No Frusemide was given. Apart from a transient rise in resting heart rate from 160 to 190 bpm, her observations remained stable throughout the transfusion, and her feeds were not discontinued. Five hours after completing the transfusion she developed severe apnoeas and bradycardias associated with abdominal distension. She deteriorated rapidly with extensive radiological changes of NEC with perforation. Despite resuscitation and transfer to a Neonatal Surgical Unit she died from disseminated intravascular coagulation and multi-organ failure. Blood cultures taken after her deterioration grew E.Coli.

  Discussion

  We found one reference linking cases of NEC to blood transfusions.[3] During an outbreak of NEC affecting 33 babies over a 6-month period in a single neonatal unit in Oklahoma a risk factor analysis was performed and compared with 37 controls. If a baby had received at least one blood transfusion the odds ratio for developing NEC was 15.5. Using a logistic regression model including other risk factors (treatment with frusemide, caffeine or theophylline) the relative risk for infants who had received a packed cell transfusion was 8.98 compared to controls. No temporal association was suggested, and the precise mechanism for such a causal link was not identified.

  The authors suggested that there may be regional alterations in oxygen delivery caused by the altered oxygen carrying characteristics of stored blood, by phthalate esters leached from polyvinyl-chloride containers or by transient polycythaemia and hyperviscosity. Standard transfusion volumes of 15-20 ml/kg will not alter the haematocrit by more than 0.2, but local variation in blood flow in the mesenteric circulation has been postulated in association with hyperviscous packed cell transfusion.[4] Marik and Sibbald found that in a group of septic adult patients on an intensive care unit gut oxygen availability paradoxically decreased after red cell transfusion. The change in intraluminal pH, as measured by gastric tonometry, was more marked if the red cells had been stored for more than 15 days. It is possible that transfusion of poorly deformable red cells causes areas of localised ischaemia in the gut microcirculation, disrupting the intestinal mucosal barrier and facilitating bacterial spread. Localised areas of endothelial dysfunction caused by circulatory changes may adversely affect nitric oxide production, further exacerbating intestinal ischaemia.[5] There is also evidence that red cells develop reduced levels of intracellular ATP and 2,3 diphosphoglycerate with time and that this affects both the red cell deformability and oxygen delivery.

  Transfusion of blood products may cause a transfusion reaction, although pre-term infants are not thought to mount an antibody response and for this reason blood is not routinely cross-matched against the infant’s blood. There are several reports of blood transfusion causing acute haemolysis in babies with established NEC and T activation,[6] although the use of low anti-T titre plasma products has not been shown to reduce mortality,[7] and the current neonatal transfusion guidelines remain non-committal about the importance of T activation.[8] Standard neonatal transfusions consist of irradiated red cells reconstituted in sodium/adenine/glucose /mannitol (Sag-m) and will therefore be low titre products.

  It is possible that there is a chance association between the blood transfusion and the onset of symptoms in both these cases. Both infants suffered many of the known risk factors for NEC, including growth retardation, absent end diastolic flow, umbilical catheterisation and PDA for case 1, and pre term rupture of membranes, asphyxia, and umbilical catheterisation for case 2. Both were just outside the peak period of NEC incidence, and were fed on maternal breast milk, which confers protection against NEC.[9] The osmolarity of the feeds was increased by adding Gaviscon and breast milk fortifier, which could increase the risk of NEC, but both infants were tolerating full feeds with no prior signs of bowel sepsis. Blood transfusion may simply have unmasked latent NEC. Apnoeas and desaturations could potentially be an early symptom of NEC, but there were no associated clinical signs such as increased residual gastric volumes, vomiting, abdominal distension or altered bowel habit. Careful retrospective analysis of nursing records has not identified any change in physiological measurements prior to the administration of blood.

  Our unit has a low incidence of NEC, with these two cases representing the only confirmed cases in the last 18 months. We had no other infants with symptoms of NEC at the same time, and there was a three -month gap between the two cases. We do not routinely discontinue milk feeds prior to blood transfusion. However correspondence with colleagues suggests that many units do omit feeds for the duration of the blood transfusion. This correspondence has also highlighted that our two cases are not unique: an e-mail circulated to members of the British Association of Perinatal Medicine (BAPM) highlighting our concerns about blood transfusion has identified 16 units in the UK that have experienced similar cases. It is important to clarify whether certain transfusion practices make NEC more likely, and data should be collected prospectively on a national basis to try to address this. We would urge colleagues to ensure that any future cases are reported through the serious hazards of transfusion (SHOT) mechanism,[10] which is usually coordinated locally by the transfusion laboratory. Meanwhile the question should not be whether to stop feeds during a transfusion, but whether the transfusion is really necessary, particularly for infants with pre-existing risk factors for NEC.

  References

  1. Caplan MS, Jilling T New concepts in necrotising enterocolitis Current Opinion Pediatr 2001. 13 (2) 111-115

  2. Newell SJ Necrotising enterocolitis in Textbook of Neonatology : Ed Rennie JM, Roberton NRC. Churchill Livingstone 1999. 3rd edition pp 747-755.

  3. McGrady GA, Rettig PJ, Istre GR et al An outbreak of necrotising enterocolitis. Association with transfusion of packed red blood cells. American Journal of Epidemiology 126 (6) 1165-1172. 1987

  4.Marik PE, Sibbald WJ Effect of stored blood transfusion on oxygen delivery in patients with sepsis. J.A.M.A. 269 ; 63. 1993 : PP3024-3035

  5. Reber KM, Nankervis CA, Nowicki PT Newborn intestinal circulation. Physiology and pathophysiology Clin Perinatol 2002 Mar ; 29(1) : 23-29

  6. Grant HW, Hadley GR, Adhikari M, Fernandes-Costa F T-cryptantigen (TCA ) activation in surgical neonates : a hidden problem. Pediatr Surg Int 1998 ; 14 : 204-207

  7. Boralessa H, Modi N, Cockburn H et al RBC T activation and haemolysis in a neonatal intensive care population : implications for transfusion practice. Transfusion 42 Nov 2002 : pp1428-1434

  8. British Committee for Standards in Haematology Transfusion Task Force Transfusion Guidelines for Neonates and Older Children.

  9. Lucas A, Cole TJ Breast milk and neonatal necrotising enterocolitis Lancet 1990 Dec22-29 336 (8730 ) 1519-1523

  10. Serious hazards of transfusion website at www.shot.demon.co.uk

  11. Stockman JA Anemia of prematurity. Current concepts in the issue of when to transfuse Ped Clin N Amer . 1986; 33(1): pp111-128

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top