Prolonged intravenous immunoglobulin treatment in very low birth weight infants with late onset sepsis

1 Introduction

Sepsis in the newborn represents a frequent cause of morbi-mortality, despite advances in new antimicrobials and neonatal intensive care [1]. Very low birth weight (VLBW) infants have the highest risk of developing sepsis [1]. The greater susceptibility to severe infections of newborns, especially preterm infants, is due mainly to immunological immaturity [2]. VLBW infants have antibody deficiency, complement deficiency, low fibronectin concentrations, as well as neutrophil dysfunction [2]. Preterm neonates, especially in the VLBW infant, also have physiologic hypogammaglobulinemia due to an incomplete acquisition of maternal antibodies [3]. Administration of intravenous immunoglobulin (IVIG) in such neonates could improve increase level of immunoglobulins and improve immunity function [4]. Based on these findings, different immunotherapeutic methods have been used as an adjuvant treatment for septicemia in newborns such as IVIG [5, 6], granulocyte transfusions, fibronectin administration and exchange transfusion [7]. Several studies in the last decades have tried to determine the effectiveness of IVIG as an adjuvant treatment for neonatal sepsis. Many authors [5, 9] have addressed this issue with different outcomes. However, there are differences in study designs such as sample size, recruitment criteria, dosage and type of immunoglobulins. Most publications include suspected and proven sepsis [5, 8], and a few of them include only bacteriologically confirmed sepsis [9, 10]. This makes it difficult to compare among different studies. According to several studies, the duration of the administration of IVIG has been different, with doses between 250 and 500 mg/kg [6, 10]. The hypothesis for the present investigation is that a greater dosage and longer period of infusion of IVIG could improve the outcome of sepsis in VLBW infants. The objective of the present study was to evaluate the effectiveness of prolonged IVIG use as an adjuvant treatment for sepsis in very low birth weight infants.

2 Methods

This was a prospective, randomized, controlled, non-blind study. Very low birth weight infants (under 1500 g) admitted to the neonatal intensive care unit of the Guillermo Grant Benavente Hospital in Concepción, Chile, were studied during a period of two years. VLBW infants with suspected late onset sepsis were selected. The criteria for sepsis were: clinical infection with positive blood cultures, and/or signs of infection with laboratory findings suggestive of bacterial infection. Late onset sepsis was considered 72 hours after birth [11]. Clinical sepsis was considered when blood cultures were negative but the complete blood count showed three or more of the following characteristics: alterations of the immature/total neutrophil ratio≥0.2; immature/mature neutrophil ratio≥0.3, increased immature neutrophils; leukopenia (< 5000/mm³) or leukocytosis (> 20000/mm³) as to postnatal age; degenerative changes of neutrophils and thrombocytopenia under 150.000/mm³ [12]. For the diagnosis of negative coagulase sepsis, a positive blood culture is required in the presence of clinical signs and symptoms of the infection: fever greater than 38°C rectal or hypothermia less than 37°C, bradycardia or apnea, in addition to one or more alteration of the laboratory tests mentioned above [12, 13]. The preterm neonates were randomly selected in two groups. Group I received regular antibiotic treatment (vancomycin or cloxacilin plus cefotaxime). The remaining newborns (Group II) received the same antimicrobials plus standard polyvalent IVIG (Sandoglobulin®, CSL Behring), at a dose of 0.5 g/kg/day during 7 days. Patients were randomized by a computer-generated sequence. The clinician selected the treatment option by closed envelope method. The IVIG was infused during 3 hours, with simultaneous monitoring of heart rate, blood pressure, pulse oxymetry and temperature. The antibacterial treatment, general supportive measures and clinical characteristics of both groups of infants were similar. In group II, serum IgG was measured at initiation of therapy, and once daily during the treatment period. IgG measurement was made by radioimmunoassay (RIA). Possible side effects from IVIG were monitored by serial hemograms, aminotransferases, and creatininemia during treatment.

The primary outcome was mortality at hospital discharge. The study was approved by the Scientific Ethics Committee of our center, and explicit consent was obtained from the newborn’s parents prior to initiation.

Fisher’s exact test was applied for the statistical analysis of mortality in each group, and ANOVA test for the study of IgG levels between days. A p-value under 0.05 was considered statistically significant.

3 Results

Forty VLBW infants were included. Mean birth weights±standard deviation were 1132±189 g and 1075±178 g for groups I and II respectively. No predominance of either sex was observed (Table 1). Blood culture positivity was 70.0% in group I and 75.5% in group II. Staphylococcus epidermidis was the most common isolated bacteria in both groups (Table 2). All neonates under study developed clinical signs of septicemia and had alterations in blood count suggesting bacterial infection. Average base serum IgG level in group II was 486 mg/dL, and increased to 852 mg/dL (p < 0.01) after the first IVIG dose. No significant IgG level change was observed in Group I (Fig. 1). IgG levels increased gradually after each day of infusion, reaching an average value of 1898 mg/dL at the seventh day of treatment (Fig. 1). A direct correlation (r = 0.94) was observed between serum IgG levels and days of treatment (Fig. 2). Mortality at hospital discharge was 25.0% in Group I, and 5.0% in group II, a statistically significant difference (p < 0.03). (Table 3). No cardiorespiratory, hemodynamic, anaphylactic or other IVIG-attributable adverse effects were observed. There were no changes in hemoglobin, creatinine or aminotransferase levels.

OPEN IN VIEWER

Fig.1

Serum Immunoglobulin G (IVIG) levels in very low birth weight infants with sepsis.

OPEN IN VIEWER

Fig.2

Correlation between serum IgG levels and days of IVIG infusion.

The infection of the newborn represents an important morbidity and mortality cause, especially in preterm and very low birth weight infants [1]. In spite of the current antimicrobial agents and various support therapies, sepsis-related lethality is still high [1]. The high incidence of severe infection in newborn babies can be partly explained from an incomplete development of humoral and cellular defense mechanisms [2]. The presence of maternal origin antibodies in the newborn plays a major roIe in preventing infection. However, transplacental IgG acquisition can only occur significantly after 32 weeks from gestation which explains the immunoglobulin deficiency in preterm infants [3]. In the last decades, IVIG use in neonatal sepsis has been a cause of major interest [4–10]. Some researchers have suggested that IVIG administration would generate a better immune response from the neonate to bacterial disease which can subsequently lead to decreased lethality [4–6]. Due to methodological heterogeneity and design deficiencies in some studies, this hypothesis has not been confirmed. It should be noted that a systematic review of IVIG administration in adults with severe sepsis showed a significant reduction in mortality [14].

In the present study, mortality in IVIG treated children was significantly lower than in the control group (Fisher’s exact test, p = 0.03). This outcome has also been observed in several other studies in the past [5, 9]. Recently the same result has been published, but with IgM- enriched intravenous immunoglobulins [15, 16]. It should be noted that different IVIG presentations, doses and administration schemes have been used by different authors with favorable outcomes. Sidiropoulos et al analyzed the effect of 0.5–1 g/kg/day of polyvalent IVIG on six consecutive days [5]. They enrolled 82 newborns with bacterial infections -confirmed and clinical sepsis- and observed a reduction in mortality rates in both groups [5]. The design and results from this study are very similar to the present investigation. Gökalp et al., [9] also used a similar treatment plan of 4 polyvalent IVIG doses of 0.5 g/kg during an 8 day period (days 1, 2, 3 and 8). They also found a lower rate of lethality, as well as lower rate of complications and recovery time in the IVIG group. Nevertheless, it is difficult to compare these results with other studies because they only included infants with a single infectious agent (Salmonella typhimurium), and a single antimicrobial (Cefoperazone) [9]. Other authors have observed similar results in late preterm [17] and preterm infants [18] with the use of polyvalent IVIG. In the latter [18] there was a trend for lower mortality rates in the IVIG group (33.3% vs 13.3%), but it was not statistically significant (p = 0.06). However, there was a lower hospital stay in the treated group (p < 0.05) [18].

On the other hand, other studies of IVIG as adjuvant treatment for neonatal sepsis have not obtained favorable outcomes in terms of mortality [19–21]. Weisman et al. [19] studied 31 premature neonates with early onset sepsis which were randomly assigned to receive a single 0.5 g/kg dose of polyvalent IVIG or Albumin infusions. Short term mortality was significantly lower in the IVIG group (0/14 versus 7/17, p < 0.05). However, survival rate at 56 days was not affected [19]. VLBW infants may have many complications leading to mortality -not only sepsis- so long term mortality is unlikely to be influenced by a single treatment received during the first days since birth.

Other studies [10, 20] did not show a reduction in lethality either, but such research works differ significantly from the present study because they enrolled both term and preterm infants, and used a single 0.5 g/kg dose of IVIG. Term neonates possibly do not benefit as much from immunotherapy as VLBW infants and shorter treatment plans may not raise serum immunoglobulin to an effective level. The largest study to date is the INIS study [21], in which 3493 VLBW infants from 113 hospitals in nine countries were enrolled. They received two 0.5 g/kg doses of polyvalent IVIG in 48 hours. Newborns with suspected and confirmed sepsis were included, but only 39% of them had positive cultures. They found no difference in mortality or disability at 2 years of age. Secondary outcomes, such as in-hospital mortality, length of stay or a second episode of sepsis, also didn’t show any significant differences [21]. In contrast, the present study had a higher rate of infants with proven sepsis, and also, a longer infusion period and total dose of IVIG were used.

Polyvalent IVIG is mainly composed of IgG, but immunoglobulin M is responsible for the immediate response to infection [2]. This is why many researchers have focused on IgM enriched IVIG. The initial work by Haque et al., [6] showed very promising results with a significant reduction in lethality from sepsis. Most investigations regarding IgM-enriched IVIG have had favorable outcomes, with at least a trend for lower mortality [22], and others with a significant reduction in sepsis-related lethality [15, 23].

A recent Cochrane Library systematic review [24] included 3973 neonates from 9 studies and found no reduction in mortality during hospital stay, or death or major disability at two years of age in infants with suspected or proven infection. This outcome was largely influenced by the results of the INIS study [21], which accounted for 3493 of the 3973 patients. They considered these results to be undisputable, but there are many questions still unanswered as other authors have noted [25, 26]. In fact, if we only consider the other 8 studies included in the systematic review, the outcome would clearly favor the IVIG treatment (Fisher’s exact test, p < 0,02).

To our knowledge, no other study has used a total dosage of 0.5 g/kg of IVIG, and on 7 consecutive days of infusion. In our experience, a significant increase of average serum IgG levels was achieved from the first IVIG infusion. With the first IVIG dose an average IgG level of 800 mg/dL was achieved. Studies of IVIG pharmacokinetics in very low birth weight neonates have evidenced that doses between 250–750 mg/kg/per day determine significant increases of serum IgG, which helps to overcome the relative hypogammaglobulinemia in these children [4, 27]. This study showed a direct positive correlation (r = 0.94), leading to a significant cumulative effect over the period of administration. In the present research work, IVIG use was well tolerated by the newborns. No clinical or laboratory side effects were observed. This was also observed in most publications [5, 26]. Studies differ on the outcomes and reduction of mortality, but there is little doubt regarding its safety, even in VLBW infants. In the INIS study [21], only 12 adverse events were reported (including 2 deaths) in the 1759 infants receiving IVIG treatment. In the placebo group, 10 adverse events were reported (including 4 deaths). No explicit cause-effect relationship was reported [21].

In conclusion, we found that a seven day treatment plan of IVIG as an adjuvant therapy was well tolerated and could be effective in VLBW neonates with sepsis. There are still unanswered questions regarding this treatment option.

Disclosure statements

Authors declare no conflicts of interest.