bs_bs_banner
Pediatrics International (2014) 56, 566–570
doi: 10.1111/ped.12321
Original Article
Predicting onset of chronic lung disease using cord blood cytokines Daishi Takao,1 Satoshi Ibara,2 Takuya Tokuhisa,3 Chie Ishihara,2 Yoshinobu Maede,2 Takako Matsui,2 Hironobu Tokumasu,4 Kyoko Sato,2 Eiji Hirakawa,2 Chika Kabayama2 and Masakatu Yamamoto2 1 Division of Pediatrics, Maternal and Perinatal Care Center, Aizenbashi Hospital, Osaka, 2Division of Neonatology, Maternal and Perinatal Care Center, Kagoshima City Hospital, Kagoshima, 3Division of Neonatology, Fukuda Hospital, Kumamoto and 4Division of Pediatrics, Kurashiki Central Hospital, Kurashiki, Japan Abstract
Background: Applicability of cord blood interleukin-6 (IL-6) and interleukin-8 (IL-8) as markers for early prediction of the onset of chronic lung disease (CLD) due to intrauterine infection was investigated in the present study. Methods: Eighty very low-birthweight infants with chorioamnionitis were divided into two groups: the CLD group (42 patients) and the non-CLD group (38 patients), according to the presence or absence of CLD, and the clinical background and cord blood IL-6 and IL-8 levels in each group were compared and investigated. Results: The CLD group had significantly longer duration of mechanical ventilation and hospitalization (P < 0.05) and significantly higher IL-6 and IL-8 (P < 0.01) than the non-CLD group. Using the receiver operating characteristic curves of CLD onset for both IL-6 and IL-8, the cut-off value of IL-6 for predicting onset of CLD was 48.0 pg/mL, and its sensitivity and specificity were 76% and 96%, respectively. The cut-off value for IL-8 was 66.0 pg/mL, and its sensitivity and specificity were 71% and 82%, respectively. Conclusion: The cord blood levels of both IL-6 and IL-8 were significantly higher in the CLD group, indicating that both IL-6 and IL-8 are useful predictors of onset of CLD.
Key words chorioamnionitis, chronic lung disease, cytokines, inflammation, neonate.
Gomez et al. reported that in utero hypercytokinemia induced inflammatory organ damage in fetuses in a manner similar to that in systemic inflammatory response syndrome (SIRS) in adults and increased the morbidity of postnatal chronic lung disease (CLD) and periventricular leukomalacia. This condition has become commonly referred to as fetal inflammatory response syndrome (FIRS).1 Although the survival rate of premature newborns has markedly improved with progress in perinatal care, the incidence of CLD has still not been reduced.2,3 The present study investigated and evaluated clinical applicability of the measurement of cord blood cytokines, that is, interleukin-6 (IL-6) and interleukin-8 (IL-8), as indices for predicting the onset of CLD from an early stage of life.
Methods The subjects were 80 very low-birthweight (VLBW) infants in whom chorioamnionitis (CAM) was diagnosed based on the results of placental pathology. They were admitted to the Maternal and Perinatal Care Center, Kagoshima City Hospital during a 9 year period from January 2003 to December 2011, and their cord blood IL-6 and IL-8 levels were measured. Forty-two infants who developed CLD were assigned to the CLD group and Correspondence: Daishi Takao, MD, Division of Pediatrics, Maternal and Perinatal Care Center, Aizenbashi Hospital, 5-16-15 Naniwaku, Nihonbasi, Osaka 556-0005, Japan. Email: [email protected] Received 5 September 2013; revised 24 December 2013; accepted 8 January 2014.
© 2014 Japan Pediatric Society
38 infants without CLD were assigned to the non-CLD group. Perinatal background and cord blood IL-6 and IL-8 levels were compared and investigated between the groups. CLD was defined as a sustained respiratory disorder that required oxygenation beyond 28 days after birth. Cord blood was collected at birth by an attending physician from the Division of Neonatology and centrifuged at 15 000 g for 5 min. The serum was stored frozen at −70°C until measurement. IL-6 and IL-8 levels were measured using the ELISA method. The severity of CAM was determined using Blanc’s classification.4 Written consent was obtained from the parents/family when samples were collected. The study design was approved by an ethics review board. Dr. SPSS IIJ for Windows (SPSS, Chicago, IL, USA) was used for statistical analysis and Mann–Whitney U-test and analysis of covariance were applied. Differences were considered significant for P < 0.05.
Results The severity of CAM was stage 1 in one infant, stage 2 in seven infants, and stage 3 in 34 infants in the CLD group. The severity was stage 1 in 11 infants, stage 2 in 10 infants, and stage 3 in 17 infants in the non-CLD group, indicating that the number of stage 3 and stage 1 infants tended to be high in the CLD group and the non-CLD group, respectively (Table 1). Median gestational duration was 25.0 weeks in the CLD group and 28.2 weeks in the non-CLD group, and median birthweight was 760 g in the CLD group and 1162 g in the
Predicting onset of CLD Table 1 CAM stage vs presence of CLD CAM Stage 1 Stage 2 Stage 3
CLD group (n = 42) 1 7 34
Non-CLD group (n = 38) 11 10 17
CAM, chorioamnionitis; CLD, chronic lung disease.
non-CLD group, showing that both gestational duration and birthweight were significantly shorter and lower in the CLD group. White blood cell count and IgM level did not differ significantly between the two groups, but median C-reactive protein (CRP) was 0.12 mg/dL in the CLD group and 0.05 mg/dL in the non-CLD group, indicating that CRP was significantly higher in the CLD group. Median days of mechanical ventilation was 58 days in the CLD group and 23 days in the non-CLD group, indicating that the duration of mechanical ventilation was significantly longer in the CLD group. Median days of hospitalization was 171 days in the CLD group and 126 days in the non-CLD group, indicating that the duration of hospitalization was significantly longer in the CLD group. Certainly, gestational age and birthweight were significantly lower in the CLD group, which may have affected the results. After adjustment for gestational age, however, mean days of mechanical ventilation remained longer in the CLD group (67.9 days) compared with the non-CLD group (17.8 days). Mean days of hospitalization also remained longer in the CLD group (203.0 days) compared with the non-CLD group (125.1 days) after adjustment. Moreover, there was no correlation between the levels of cord blood cytokines and gestational age or birthweight.
567
The incidence of severe intraventricular hemorrhage ≥grade 3 was 20% in the CLD group and 11% in the non-CLD group, indicating no significant difference. The score on the Revised Kyoto Scale of Psychological Development did not differ significantly between the two groups either at corrected age of 1 year 6 months, or 3 years (Tables 2,3). Cord blood IL-6 and IL-8 levels were 111.0 pg/mL and 92.0 pg/mL, respectively, in the CLD group and 17.0 pg/mL and 33 pg/mL, respectively, in the non-CLD group, indicating that both IL-6 and IL-8 were significantly higher in the CLD group (both P < 0.01; Fig. 1). Using receiver operating characteristic (ROC) curves of CLD onset for both IL-6 and IL-8, the cut-off of IL-6 for predicting onset of CLD was 48.0 pg/mL, the sensitivity was 76%, specificity was 96%, and the area under the curve (AUC) was 0.848. The cut-off for IL-8 was 66.0 pg/mL, the sensitivity was 71%, specificity was 82%, and AUC was 0.817 (Fig. 2).
Discussion The risk factors for onset of CLD include congenital factors associated with immaturity such as early gestational age and low birthweight, and acquired factors such as barotrauma due to mechanical ventilation after birth and oxygen toxicity due to high oxygen concentration.5 The incidence of CLD due to acquired factors such as mechanical ventilation and oxygen toxicity has decreased in recent years as a result of improved neonatal management such as artificial lung surfactant replacement therapy and more advanced mechanical ventilators. There has, however, been an increasing number of CLD cases among more premature infants.6,7 Inhibition of fetal alveolar formation by exposure to inflammatory cytokines in utero as a consequence of intrauterine infection and CAM is considered to be a key factor in the onset of CLD,
Table 2 Subject clinical characteristics vs presence of CLD CLD group Median (range) 25.0 (22–29) 760 (435–1388) 4.2 (0–108) 0.12 (0.01–1.31) 58 (13–112) 171 (111–421) 20 72 (31–89) 77 (10–96)
Gestational duration (weeks) Birthweight (g) IgM (mg/dL) CRP (mg/dL) Mechanical ventilation (days) Hospitalization (days) Severe IVH (%) KSPD DQ (1.5 years) KSPD DQ (3 years)
Non-CLD group Median (range) 28.2 (23–30) 1162 (552–1434) 6.0 (0–90.7) 0.05 (0.01–1.19) 23 (3–35) 126 (81–240) 11 78 (56–103) 88 (54–102)
P
Pediatrics International (2014) 56, 566–570
doi: 10.1111/ped.12321
Original Article
Predicting onset of chronic lung disease using cord blood cytokines Daishi Takao,1 Satoshi Ibara,2 Takuya Tokuhisa,3 Chie Ishihara,2 Yoshinobu Maede,2 Takako Matsui,2 Hironobu Tokumasu,4 Kyoko Sato,2 Eiji Hirakawa,2 Chika Kabayama2 and Masakatu Yamamoto2 1 Division of Pediatrics, Maternal and Perinatal Care Center, Aizenbashi Hospital, Osaka, 2Division of Neonatology, Maternal and Perinatal Care Center, Kagoshima City Hospital, Kagoshima, 3Division of Neonatology, Fukuda Hospital, Kumamoto and 4Division of Pediatrics, Kurashiki Central Hospital, Kurashiki, Japan Abstract
Background: Applicability of cord blood interleukin-6 (IL-6) and interleukin-8 (IL-8) as markers for early prediction of the onset of chronic lung disease (CLD) due to intrauterine infection was investigated in the present study. Methods: Eighty very low-birthweight infants with chorioamnionitis were divided into two groups: the CLD group (42 patients) and the non-CLD group (38 patients), according to the presence or absence of CLD, and the clinical background and cord blood IL-6 and IL-8 levels in each group were compared and investigated. Results: The CLD group had significantly longer duration of mechanical ventilation and hospitalization (P < 0.05) and significantly higher IL-6 and IL-8 (P < 0.01) than the non-CLD group. Using the receiver operating characteristic curves of CLD onset for both IL-6 and IL-8, the cut-off value of IL-6 for predicting onset of CLD was 48.0 pg/mL, and its sensitivity and specificity were 76% and 96%, respectively. The cut-off value for IL-8 was 66.0 pg/mL, and its sensitivity and specificity were 71% and 82%, respectively. Conclusion: The cord blood levels of both IL-6 and IL-8 were significantly higher in the CLD group, indicating that both IL-6 and IL-8 are useful predictors of onset of CLD.
Key words chorioamnionitis, chronic lung disease, cytokines, inflammation, neonate.
Gomez et al. reported that in utero hypercytokinemia induced inflammatory organ damage in fetuses in a manner similar to that in systemic inflammatory response syndrome (SIRS) in adults and increased the morbidity of postnatal chronic lung disease (CLD) and periventricular leukomalacia. This condition has become commonly referred to as fetal inflammatory response syndrome (FIRS).1 Although the survival rate of premature newborns has markedly improved with progress in perinatal care, the incidence of CLD has still not been reduced.2,3 The present study investigated and evaluated clinical applicability of the measurement of cord blood cytokines, that is, interleukin-6 (IL-6) and interleukin-8 (IL-8), as indices for predicting the onset of CLD from an early stage of life.
Methods The subjects were 80 very low-birthweight (VLBW) infants in whom chorioamnionitis (CAM) was diagnosed based on the results of placental pathology. They were admitted to the Maternal and Perinatal Care Center, Kagoshima City Hospital during a 9 year period from January 2003 to December 2011, and their cord blood IL-6 and IL-8 levels were measured. Forty-two infants who developed CLD were assigned to the CLD group and Correspondence: Daishi Takao, MD, Division of Pediatrics, Maternal and Perinatal Care Center, Aizenbashi Hospital, 5-16-15 Naniwaku, Nihonbasi, Osaka 556-0005, Japan. Email: [email protected] Received 5 September 2013; revised 24 December 2013; accepted 8 January 2014.
© 2014 Japan Pediatric Society
38 infants without CLD were assigned to the non-CLD group. Perinatal background and cord blood IL-6 and IL-8 levels were compared and investigated between the groups. CLD was defined as a sustained respiratory disorder that required oxygenation beyond 28 days after birth. Cord blood was collected at birth by an attending physician from the Division of Neonatology and centrifuged at 15 000 g for 5 min. The serum was stored frozen at −70°C until measurement. IL-6 and IL-8 levels were measured using the ELISA method. The severity of CAM was determined using Blanc’s classification.4 Written consent was obtained from the parents/family when samples were collected. The study design was approved by an ethics review board. Dr. SPSS IIJ for Windows (SPSS, Chicago, IL, USA) was used for statistical analysis and Mann–Whitney U-test and analysis of covariance were applied. Differences were considered significant for P < 0.05.
Results The severity of CAM was stage 1 in one infant, stage 2 in seven infants, and stage 3 in 34 infants in the CLD group. The severity was stage 1 in 11 infants, stage 2 in 10 infants, and stage 3 in 17 infants in the non-CLD group, indicating that the number of stage 3 and stage 1 infants tended to be high in the CLD group and the non-CLD group, respectively (Table 1). Median gestational duration was 25.0 weeks in the CLD group and 28.2 weeks in the non-CLD group, and median birthweight was 760 g in the CLD group and 1162 g in the
Predicting onset of CLD Table 1 CAM stage vs presence of CLD CAM Stage 1 Stage 2 Stage 3
CLD group (n = 42) 1 7 34
Non-CLD group (n = 38) 11 10 17
CAM, chorioamnionitis; CLD, chronic lung disease.
non-CLD group, showing that both gestational duration and birthweight were significantly shorter and lower in the CLD group. White blood cell count and IgM level did not differ significantly between the two groups, but median C-reactive protein (CRP) was 0.12 mg/dL in the CLD group and 0.05 mg/dL in the non-CLD group, indicating that CRP was significantly higher in the CLD group. Median days of mechanical ventilation was 58 days in the CLD group and 23 days in the non-CLD group, indicating that the duration of mechanical ventilation was significantly longer in the CLD group. Median days of hospitalization was 171 days in the CLD group and 126 days in the non-CLD group, indicating that the duration of hospitalization was significantly longer in the CLD group. Certainly, gestational age and birthweight were significantly lower in the CLD group, which may have affected the results. After adjustment for gestational age, however, mean days of mechanical ventilation remained longer in the CLD group (67.9 days) compared with the non-CLD group (17.8 days). Mean days of hospitalization also remained longer in the CLD group (203.0 days) compared with the non-CLD group (125.1 days) after adjustment. Moreover, there was no correlation between the levels of cord blood cytokines and gestational age or birthweight.
567
The incidence of severe intraventricular hemorrhage ≥grade 3 was 20% in the CLD group and 11% in the non-CLD group, indicating no significant difference. The score on the Revised Kyoto Scale of Psychological Development did not differ significantly between the two groups either at corrected age of 1 year 6 months, or 3 years (Tables 2,3). Cord blood IL-6 and IL-8 levels were 111.0 pg/mL and 92.0 pg/mL, respectively, in the CLD group and 17.0 pg/mL and 33 pg/mL, respectively, in the non-CLD group, indicating that both IL-6 and IL-8 were significantly higher in the CLD group (both P < 0.01; Fig. 1). Using receiver operating characteristic (ROC) curves of CLD onset for both IL-6 and IL-8, the cut-off of IL-6 for predicting onset of CLD was 48.0 pg/mL, the sensitivity was 76%, specificity was 96%, and the area under the curve (AUC) was 0.848. The cut-off for IL-8 was 66.0 pg/mL, the sensitivity was 71%, specificity was 82%, and AUC was 0.817 (Fig. 2).
Discussion The risk factors for onset of CLD include congenital factors associated with immaturity such as early gestational age and low birthweight, and acquired factors such as barotrauma due to mechanical ventilation after birth and oxygen toxicity due to high oxygen concentration.5 The incidence of CLD due to acquired factors such as mechanical ventilation and oxygen toxicity has decreased in recent years as a result of improved neonatal management such as artificial lung surfactant replacement therapy and more advanced mechanical ventilators. There has, however, been an increasing number of CLD cases among more premature infants.6,7 Inhibition of fetal alveolar formation by exposure to inflammatory cytokines in utero as a consequence of intrauterine infection and CAM is considered to be a key factor in the onset of CLD,
Table 2 Subject clinical characteristics vs presence of CLD CLD group Median (range) 25.0 (22–29) 760 (435–1388) 4.2 (0–108) 0.12 (0.01–1.31) 58 (13–112) 171 (111–421) 20 72 (31–89) 77 (10–96)
Gestational duration (weeks) Birthweight (g) IgM (mg/dL) CRP (mg/dL) Mechanical ventilation (days) Hospitalization (days) Severe IVH (%) KSPD DQ (1.5 years) KSPD DQ (3 years)
Non-CLD group Median (range) 28.2 (23–30) 1162 (552–1434) 6.0 (0–90.7) 0.05 (0.01–1.19) 23 (3–35) 126 (81–240) 11 78 (56–103) 88 (54–102)
P
