Neonatal sepsis is one amongst the three commonest illnesses affecting newborns and ranks as the topmost illness responsible for neonatal mortality especially amongst the preterm and low birth weight babies (<xref ref-type="bibr" rid="bib001">1</xref>).It has the potential to involve all organ systems of the body as a consequence of immaturity of the hematopoietic, body defence and immunological systems on the one hand and extensive handling of babies, skin & mucosal breaches and invasive interventions on the other hand, leading to higher incidence of sepsis in NICUs (<xref ref-type="bibr" rid="bib002">2</xref>). There is no doubt that untreated neonatal sepsis is uniformly fatal but even with the appropriate treatment, mortality is 25–30%. It is a nightmare for the pediatrician treating the baby with sepsis!Novel adjunctive therapies tried for sepsis beyond antibiotics and supportive care, including IVIG administration, granulocyte transfusion, G-CSF, GM-CSF etc. have had their share of the limelight at some point or the other in recent times. But, still none of them have really stood the test of time with the rigors of evidence based medicine and continue to be considered either experimental or given up as worthless therapy. Most of their status is still to be adjudged through ‘large multi-centric randomized double blind placebo controlled trials with reasonable follow up till 18–24 months of age’. What remains to be seen is whether there are any more promising modalities on the horizon for treating neonatal sepsis or for the time being one has to juggle around with antibiotics, aggressive supportive therapy and concentrate more on strengthening the preventive aspects.One such modality is exchange transfusion. It has been tried by various workers but as yet there is no consensus regarding the utility of this modality in treating neonatal sepsis. Questions which need to be answered are, does exchange transfusion carried out as an adjunctive therapy for sepsis have any biological plausibility & rationale; if yes, has it been tested systematically and if so what is the current status? Moreover, is the therapy safe and is it cost effective?<sec id="sec002"><title>Biologic plausibility

Journal of Neonatology

0971-2179

National Neonatology Forum

jn-20-4-004

REVIEW ARTICLE

Exchange transfusion in neonatal sepsis-Does it help?

Nangia

Sushma

Associate ProfessorSaili

Arvind

Professor

Division of Neonatology Deptt of Pediatrics, Lady Hardinge Medical College, New Delhi 110 001.

10–122006

204333337

Neonatal sepsis is one amongst the three commonest illnesses affecting newborns and ranks as the topmost illness responsible for neonatal mortality especially amongst the preterm and low birth weight babies (<xref ref-type="bibr" rid="bib001">1</xref>).It has the potential to involve all organ systems of the body as a consequence of immaturity of the hematopoietic, body defence and immunological systems on the one hand and extensive handling of babies, skin & mucosal breaches and invasive interventions on the other hand, leading to higher incidence of sepsis in NICUs (<xref ref-type="bibr" rid="bib002">2</xref>). There is no doubt that untreated neonatal sepsis is uniformly fatal but even with the appropriate treatment, mortality is 25–30%. It is a nightmare for the pediatrician treating the baby with sepsis!Novel adjunctive therapies tried for sepsis beyond antibiotics and supportive care, including IVIG administration, granulocyte transfusion, G-CSF, GM-CSF etc. have had their share of the limelight at some point or the other in recent times. But, still none of them have really stood the test of time with the rigors of evidence based medicine and continue to be considered either experimental or given up as worthless therapy. Most of their status is still to be adjudged through ‘large multi-centric randomized double blind placebo controlled trials with reasonable follow up till 18–24 months of age’. What remains to be seen is whether there are any more promising modalities on the horizon for treating neonatal sepsis or for the time being one has to juggle around with antibiotics, aggressive supportive therapy and concentrate more on strengthening the preventive aspects.One such modality is exchange transfusion. It has been tried by various workers but as yet there is no consensus regarding the utility of this modality in treating neonatal sepsis. Questions which need to be answered are, does exchange transfusion carried out as an adjunctive therapy for sepsis have any biological plausibility & rationale; if yes, has it been tested systematically and if so what is the current status? Moreover, is the therapy safe and is it cost effective?<sec id="sec002"><title>Biologic plausibility

A neonate is immunologically immature though not immuno-deficient at birth. The newborn has incompletely developed robust immune competence at birth for the simple reason that it had to survive and grow in the maternal environment without any adverse effects of maternal-fetal antigenic incompatibility (3). The newborn thus is handicapped in terms of neutrophil numbers and its adequate functioning. It is known that neutrophil adhesion & chemotaxis are diminished at birth even in term babies, although phagocytosis and microbicidal activities are normal but they too diminish when a neonate is stressed as with sepsis, acidosis, shock or asphyxia (4). The monocyte function as well as its flux in response to infection is similarly diminished. In addition, the opsonic activity is quantitatively diminished and thus the alternate complement pathway functions suboptimally.(5) The cellular component of immune system has T cells which have poor cytokine production, poor CTL response and the humoral components also have low IgG and IgM levels as compared to adults and reach adult values only by 4–6 and 1–2 years respectively (6,7).

Rationale

The rationale for performing an exchange transfusion in newborns with sepsis is related to the mechanisms that result in reduced morbidity and improved survival. Exchange transfusions may help by removing the circulating bacteria, the toxins, their inflammatory mediators and provide trophic factors and replete neutrophils. Neutropenia as well as poor neutrophil function has been reported in sepsis. Exchange transfusion improves oxygen carrying capacity and also oxygen delivery by providing higher amount of adult hemoglobin. It increases opsonic and granulocytic activity by providing heat labile & heat stable plasma proteins, antibodies and components of complement system along with fibronectin. It improves perfusion and tissue oxygenation, decreases hemorrhagic complications, and enhances the humoral and cellular inflammatory response (8,9).

Clinical evidence

The literature is replete with anecdotal case reports, case series and even some clinical trials reporting survival in the babies subjected to exchange transfusions when all other available modalities had been tried and had failed. The literature search reveals only a handful of clinical trials most of which are controlled trials and only two randomized controlled trials. The first report of exchange transfusion for treatment of neonatal sepsis was published in 1975 in which it was reported that 85% of the sclerematous neonates (10) survived compared to none in the control group. Most of the trials reported in literature have demonstrated better survival and other better outcomes in exchanged versus not exchanged especially when fresh blood was used for this purpose. The improved survival has been documented in both term & preterm septicemic babies.

A few studies have shown improvements in neutrophil numbers after exchange transfusion (ET) in neutropenic neonates and have advocated use of ET instead of Granulocyte transfusion as the latter may be difficult technically and time consuming (11). Christensen et al. (12) used ET in 6 patients and reported three fold increase in neutrophil numbers and improved outcome. Similarly Hall et al. (13) evaluated serum opsonic activity and peripheral neutrophil counts before and after exchange transfusion in 41 infants with early onset group B streptococcal septicemia. They reported significant increase in immature and mature neutrophils in the peripheral blood post-exchange transfusion among survivors although the increases in opsonic antibody levels did not reach statistical significance. It is possible that exchange transfusion with fresh whole blood may enhance mobilization of neutrophils in infants with adequate bone marrow storage pools. De Curtis et al. (14) reported improvement of phagocytosis and nitroblue tetrazolium reduction after exchange transfusions in two preterm infants with severe septicemia. However, they cautioned that if exchange transfusion is used in the management of neonatal sepsis, special attention should be given to the immunological properties of the donor blood (15). There are few clinical reports in sclerematous neonates who underwent repeated exchange transfusions and could be salvaged. Xanthou et al. (10) reported on two babies with severe sepsis who survived following ET, Vain et al. (16) reported on survival of 7 babies out of 10 after repeated ETs in babies with severe sepsis who developed sclerema on antibiotic treatment and supportive therapy. Lemos et al. (17) & Kavada et al. (18) similarly reported on post operative cardiac surgery patients who survived after ET used for therapy of sepsis. Dalvi et (19) reported on 53 neonates with severe or unresponsive sepsis subjected to therapeutic exchange transfusions (ET) using 170 ml/kg of citrated blood less than 24 hours old. The procedure was repeated up to a maximum of 4 times. The success of therapy was adjudged by resolution of sclerema and/or improvement in clinical features. The mean time for recovery following ET was 19.6 + 12.4 h (range: 1–48 h). The overall survival was 77.4%. Kokk T et al. (20) reported on markers of oxidative stress before and after exchange transfusion for neonatal sepsis in 8 neonates and found that there was short-term elevation of conjugated dienes (CDs) and thiobarbituric acid reactive substances (TBARSs) (p < 0.05) and increased serum antioxidant capacity (AOC) (p = 0.06). Chromatographic migration of the red blood cells (CM RBCs), a measure of RBC deformability, also improved after ET. They concluded that ET for neonatal sepsis does not cause significant oxidative stress; rather, it may help to restore the antioxidant depots and improve RBC deformability and microcirculation. Murakami et al. (21), studied changes of cerebral hemodynamics and oxygenation in 3 unstable septic newborns using near infrared spectroscopy (NIRS) during exchange transfusion and reported that two-way exchange transfusion increases cerebral blood volume and improves cerebral oxygen delivery. Furthermore NIRS was useful in monitoring changes in cerebral hemodynamics and oxygenation in unstable septic neonates.

There are a total of 6 trials reported in the literature of which two are randomized controlled trials. The earliest trial reported was in 1977 by Tollner et al. with a sample size of 33. This trial used historic controls and reported a case fatality of 50% in the ET group and 100% in the control group (22). Bossi et al. (23) on 35 babies with sepsis reported 55% survival in ET group versus only 38% in the controls. Narayanan et al. (24) studied 60 babies with sepsis who were treated with steroids alone, steroids with ET and standard therapy and reported improved survival in the ET group. Gunes et al. (25) subjected 88 infants with gestation age ranging from 32 to <37 weeks with sepsis to either ET with fresh whole blood or IVIG and studied the effect of these on immunoglobulin G and M levels over a 24-hour period. ET was performed on 33 infants, 33 infants received IVIG and 22 infants served as controls. There were nine deaths (27%) in the IVIG group, seven (21%) in the ET group and nine (41%) in the control group (p>0.05). There were no differences between the initial and 24-hour IgG levels in the IVIG group, ET and control group. IgM levels rose significantly 12 hours after ET and elevated IgM levels persisted for over 24 hours. The mortality was least in the ET group. The two randomized controlled trials reported so far have been by Mathur (8) et al. and Sadana (9) et al. who studied neutropenic and sclerematous septic neonates respectively. Mathur et al. performed ET in neutropenic septic neonates and reported increased ANC, decreased band forms, improved neutrophil function and reduction of mortality from 70% in the controls to 35% in the exchanged group. Sadana et al. reported increased immunoglobulin levels but no difference in C3 levels in septic sclerematous neonates treated with ET versus no ET. The mortality decreased from 95% in the control group to 50% in the ET group. The same are depicted in a tabulated form in Table no. 1. Table 1

Summary of studies comparing usefulness of Exchange Transfusion in Neonatal Sepsis

Author	Year	Sample size	Trial type	Results

Tollner (22)	1977	33	Controlled (Historic controls)	Case Fatality 50% (ET)100% (control)
Bossi (23)	1981	35	Controlled	Survival 55% (ET) 38% (control)
Narayanan (25)	1982	60	Controlled	Improved survival after ET
Mathur (8)	1993	30	RCT	Survival 65% (ET) Control 30%
Sadana (9)	1997	40	RCT	Survival 50% ET5% control
Gunes (24)	2006	88	Controlled (IVIG 33, ET 33, Control 22)	IgM levels higher with ET, No change in IgG/IgM levels with IVIG

Even though all the above reports have generally reported favourable outcome with ET, the problem with extrapolation of these to routine clinical practice are manifold. Firstly, all the reported clinical trials are done on small number of subjects ranging from 35–45. More importantly the study designs have been of poor quality and have methodological flaws. There is lack of uniform inclusion criteria viz. not all cases are culture positive sepsis, some have included clinical sepsis, others neutropenic neonates with sepsis and yet others have included sclerematous neonates. Control groups again have varied from historical controls to concurrent controls with ratios of 1:1 to 2:1 and at times just an arbitrary number. There has been no mention of sample size calculation or its consideration in design of the study. Some studies have stratified the groups by gestation and others by weight. A few of the studies have used other adjunctive therapies in addition to exchange transfusion such as IVIG, Steroids and online plasma exchange and have reported improved outcomes which may well be the result of additional therapies. Lastly, the outcomes studied have been equally variable ranging from increase in neutrophil numbers, increase in immunoglobulin levels, improvement in coagulation disturbance parameters, resolution of sclerema with clinical improvement and reduction of mortality. Since, there is so much diversity in the scant reported literature, any kind of a systematic review or recommendations for clinical practice is not possible and has not been attempted.

Safety

Exchange transfusion is not a procedure without risks especially in incompetent hands. Complications include over transfusion, perforation of major vessels, hypocalcaemia, citrate toxicity, hypothermia, hypoglycaemia, thrombocytopenia, necrotizing enterocolitis, GVHD, bacterial & viral infections. In addition with repeated ETs there is a risk of production of leucocyte and platelet antibodies.

Current Status

At the present time exchange transfusion in neonatal sepsis can only be considered experimental. Hence, the situations where one could resort to ET are sclerematous neonates and neonates with DIC unresponsive to appropriate antibiotics and supportive care.

Recommendation

Currently, as ET is not a standard therapy. A large multi-centric randomized controlled trial with appropriate methodology must be carried out to prove or refute the role of ET in neonatal septicemia. This modality may however be tried in a situation where a neonate is in DIC as a consequence of sepsis or has developed sclerema or is continuing to be neutropenic despite appropriate antibiotics and supportive therapy.

References1Lawn

Cousens

Zupan

. 4 million neonataldeaths: When? Where? Why?. Neonatal survival 1, Neonatal Survival Series. Lancet Mar, 2005; 9–18.2Polin

Parravicini

Regan

Taeusch

. Bacterial sepsis and meningitis. In ‘Avery's Diseases of the Newborn’ 8^th Edn. Eds Taeusch

Ballard

Gleason

, 2005; 551–577.3Beer

Billingham

,: The immunobiology of mammalian reproduction, Eanglewood Cliffs, NJ, Prentice-Hall, 1976.4Berg

Pillingerv

. The Neutrophil. Function and regulation in innate and humoral immunity. Clin Immunol. 2001; 99: 7–17.5Winkelstein

. The complement system of the fetus and neonate. In ‘Fetal and Neonatal Physiology’. Vol 2, 3^rd Ed. Eds. Polin

Fox

Abman

, Saunders; 2004: 1549–1554.6La Pine

Hill

. Host defence mechanismsagainst bacteria. In ‘Fetal and Neonatal Physiology‘. Vol 2, 3^rd Ed. Eds. Polin

Fox

Abman

, Saunders; 2004: 1475–1486.7Insel

Looney

. The B Lymphocyte system: Fundamental immunobiology. In Steihm

(eds); Immunological disorders in Infants & Children, 5^th Ed. Philadelphia, Elsevier saunders, 2004: 53–57.8Mathur

Subramanian

Sharma

Puri

. Exchange transfusion in neutropenic septicemic neonates: effect on granulocyte functions. Acta Paediatr. 1993; 82 (11): 939–943.9Sadana

Mathur

Thakur

. Exchange transfusion in septic neonates with sclerema: effect on immunoglobulin and complement levels. Indian Pediatr. 1997; 34 (1): 20–25.10Xanthou

Xypolyta

Anagnostakis

Economou-Mavrou

Matsaniotis

. Exchange transfusion in severe neonatal infection with sclerema. Arch Dis Child. 1975; 50 (11): 901–902.11Christensen

Hill

Anstall

Rothstein

. Exchange transfusion as an alternative to granulocyte concentrate administration in neonates with bacterial sepsis and profound neutropenia. J Clin Apher. 1984; 2 (2): 177–183.12Christensen

Anstall

Rothstein

. Use of whole blood exchange transfusion to supply neutrophils to septic, neutropenic neonates. Transfusion. 1982; 22 (6): 504–506.13Hall

Shigeoka

Hill

. Serum opsonic activity and peripheral neutrophil counts before and after exchange transfusion in infants with early onset group B streptococcal septicemia. Pediatr Infect Dis. 1983; 2 (5): 356–358.14De Curtis

Vetrano

Romano

D'Antonio

Paludetto

Ciccimarra

. Improvement of phagocytosis and nitroblue tetrazolium reduction after exchange transfusions in two preterm infants with severe septicemia. Pediatrics. 1982; 70 (5): 829–830.15Pelet

. Exchange transfusion in newborninfants: effects on granulocyte function. Arch Dis Child. 1979; 54 (9): 687–690.16Vain

Mazlumian

Swarner

Cha

. Role of exchange transfusion in the treatment of severe septicemia. Pediatrics. 1980; 66 (5): 693–697.17Lemos

. Exchange transfusion in treatment ofsepsis. Pediatrics. 1981; 68 (3): 471–472.18Kawada

Nakamura

Yamagishi

Nomura

Kanazawa

Kagimoto

Okazaki

Ooishi

. [Exchange transfusions in sepsis after the pediatric open heart surgery] [Article in Japanese]. Kyobu Geka. 2001; 54 (12): 999–1002.19Dalvi

Rao

Rangnekar

Fernandez

. Exchange transfusions in neonatal sepsis. Indian Pediatr. 1991; 28 (8): 956–957.20Kokk

Talvik

Zilmer

Kutt

Talvik

. Markers of oxidative stress before and after exchange transfusion for neonatal sepsis. Acta Paediatr. 1996; 85 (10): 1244-1246.21Murakami

Yamashita

Nishimi

Inoue

Matsuishi

Kato

. Changes of cerebral hemodynamics and oxygenation in unstable septic newborns during exchange transfusion. Kurume Med J. 1998; 45 (4): 321–325.22Tollner

Pohlandt

Heinze

Henrichs

. Treatment of septicemia in the newborn infant: Choice of initial antimicrobial drugs & the role of exchange transfusion. Acta Pediatr Scand, 1977, 66: 605–610.23Bossi

Meister

Pfenninger

. Exchangetransfusion for severe neonatal septicemia. Pediatrics. 1981; 67 (6): 941.24Narayanan

Mitter

Gujral

. A comparative study on the value of exchange and blood transfusion in the management of severe neonatal septicemia with sclerema. Indian J Pediatr. 1982; 49 (399): 519–523.25Gunes

Koklu

Buyukkayhan

Kurtoglu

Karakukcu

Patiroglu

. Exchange transfusion or intravenous immunoglobulin therapy as an adjunct to antibiotics for neonatal sepsis in developing countries: a pilot study. Ann Trop Paediatr. 2006; 26 (1): 39–42.