In the respiratory management of preterm infants of less than 33 weeks gestational age and very low birth weight infants, it is essential to understand changes in chest X-ray and blood gas findings in addition to observing their clinical symptoms. Further, to evaluate the severity of lung injury, measurement of inflammatory markers and lung injury markers in infants is very useful. Factors that have been reported to be measured relatively easily at the bedside are peripheral blood leukocyte count and its fraction, urinary β₂ microglobulin (u-BMG) resulting from hypercytokinemia [1-3], and serum levels of KL-6 [4], SP-A, and SP-D [5] as markers of lung injury. In Japan, chronic lung disease (CLD) in preterm infants is classified into three groups based on its etiology: a group that develops CLD after respiratory distress syndrome, a group that develops CLD with intrauterine infection, and a group that combines neither. This group is divided into a severe group, which shows a bubbly shadow on radiography, and a moderate group, which shows a hazy shadow. The other two groups are classified into seven types to clarify the management method for postnatal infants [6,7]. We experienced a preterm, female, extremely low birth weight infant (ELBWI) with highly intense fetal inflammatory response syndrome due to intrauterine infection. Her early postnatal u-BMG was below the cutoff value [8] for CLD but was associated with leukemoid reactions and markedly elevated serum KL-6 levels. As a result, the onset of CLD was unavoidable, but to reduce injury to her immature lungs as much as possible after birth and to avoid tracheal intubation, we continued treatment with bi-level nasal continuous positive airway pressure (CPAP) from the acute stage, which resulted in the avoidance of home oxygen therapy (HOT) and her successful discharge from hospital. The infant’s subsequent growth and development are going well. Takami et al. already reported CLD type III (Wilson-Mikity syndrome) in which serum KL-6 is markedly elevated from the early postnatal period [9]. However, their report is a study of lung injury-repair markers such as KL-6 and SP-D, and there are no reports of hypercytokinemia occurring at the same time. The present ELBWI did not require mechanical ventilation management or HOT, despite both the degree of CLD and the markedly high value of KL-6 in the early postnatal period. We focused on the slow change in the peak of u-BMG, which was observed to occur after the peak of KL-6. It has already been reported that an increase in u-BMG indicates a similar increase in interferon gamma (IFN-γ), which is an inflammatory cytokine [10-13], and is useful for predicting the onset of CLD and determining prognosis [1-3]. In this case, in addition to the inflammatory effect of IFN-γ, another effect was strongly suggested, that repair of damaged lungs may be a self-defense effect that reduces intense lung fibrosis [14-16]. In addition to the evaluation of various markers at individual time points in the early postnatal period, we report on the importance of evaluating their transitions over the entire time course.

The mother was hospitalized from the 26^th week of pregnancy for management of genital bleeding and increased abdominal tension. At 28 weeks and 1 day of gestation, an emergency cesarean section was performed due to exacerbation of signs of intrauterine infection and fetal asphyxia. The umbilical cord was stained yellow, and some ulcers were found (10 days later, placental pathology revealed chorioamnionitis [CAM] of Blanc stage III and funisitis of stage II). The female infant’s birth weight was 866 g. After adequate oral suctioning and resuscitation were performed, she was transferred to the NICU on mask CPAP. Because spontaneous breathing was not completely established, aminophylline was administered intravenously for apnea due to prematurity, and SiPAP^TM was selected as the ventilator for delivery of bi-level nasal CPAP. Ventilation settings were pressure-high: 8 cmH₂O, pressure-low: 4 cmH₂O, frequency: 20/min, inspiratory time: 1 sec, and F_IO₂ = 0.35. Venous blood examination on admission revealed a white blood cell count of 33,400/µL, CRP 0.4 mg/dL, IgM 198.0 mg/dL, pH 7.21, PCO₂ 56.8 mmHg, BE -4.3 mmol/L, and BS 50 mg/dL. No signs of respiratory distress syndrome were present on chest X-ray. We anticipated the onset of CLD with intrauterine infection (CLD III or III′) and wanted to avoid tracheal intubation as much as possible to minimize postnatal lung barotrauma and volutrauma and to ensure future quality of life (QOL) for the infant. Thus, we planned to continue respiratory management with bi-level nasal CPAP. The target blood gas pH value at 72 hours after birth, which is the critical time, was set to 7.25 or higher, and thereafter, the pH value was maintained at 7.20 or higher except in transient instances. An increase in PCO₂ up to 65 mmHg was allowed. The nurses adjusted the administered oxygen concentration appropriately to maintain SpO₂ between 90% and 95%. As a result, the initial F_IO₂ was 0.35, and it was increased once to 0.4 from 2 to 4 days of age but then gradually decreased to 0.21 by 76 days of age. Bi-level nasal CPAP was continued from 0 to 13 days, and monophasic nasal CPAP was performed from 13 to 56 days (Figures 1 and 2).

The infant’s serum KL-6 was remarkably elevated at over 3000 U/mL, which is the main factor reflecting lung tissue damage, her white blood cell count exceeded 40,000/µL, NLR was confirmed, and her u-BMG level, which is an indicator of CLD onset, also rose to 20×10⁴ µg/gCr (Figure 3). Urinary N-acetyl-β-d glucosaminidase activity (NAG) was slightly increased and decreased to clinical reference value between 24.9 and 68.4 IU/gCr which was reported by Sakata et al. [17].

Her renal tubular damage was estimated to have been mild. Therefore, although only temporarily, we felt that the infant’s prognosis was poor because of increased revels of u-BMG and KL-6. However, from the nursing perspective, efforts were made to stabilize the infant’s respiratory condition by adjusting the optimal oxygen concentration and changing the infant’s position according to her condition at the time. Antibiotics were used for only 5 days after birth, and steroids were not used at all during hospitalization. During the hospital course, the infant had no intracranial hemorrhage or periventricular leukomalacia, did not require laser photocoagulation for retinopathy of prematurity, and was not indicated for HOT. At 96 days after birth, the energetic infant was discharged weighing 2956 g.

For clinicians in neonatal fields, the ability to predict comorbid pathology of an infant is useful in establishing preventative measures to minimize its onset. In preterm infants following intrauterine infection, and especially in very low birth weight infants, the fact that the development of CLD after birth can be predicted early in life is essential so that it remains as a major complication that minimally affects the prognosis of the infants. The mechanism by which CLD develops in preterm infants is the incomplete repair of damage to lung epithelial cells during maturity [18,19]. There are many causes of lung injury. Of these, Wilson-Mikity syndrome, which is caused by intrauterine infection, originates from CAM and funisitis and results in hypercytokinemia in the fetal period [7,20]. Neutrophils accumulate in the fetal lung and infiltrate into the alveolar space to release digestive enzymes such as elastase, resulting in the destruction of immature lung tissue [21,22]. Without complete repair, the infant is born and is forced to breathe spontaneously with an adverse lung condition. As a result, hyperinflation of some alveoli and the collapse of other different alveoli occur, resulting in unique emphysematous lesions.

Recently, Romero et al. proposed the concept of fetal inflammatory response syndrome, which causes cytokine storms that spread non-infectious inflammatory responses to infants in addition to intrauterine infections [23].

Nishimaki et al. reported that there was a significant difference in u-BMG of preterm infants up to 2 days after birth with and without CAM and with or without the subsequent onset of CLD [1-3]. Elevated u-BMG was due to increased production of IFN-γ, supporting the presence of hypercytokinemia in the infant and emphasizing the convenience of utilizing the non-invasive collection of urine rather than blood sampling in infants. Furthermore, they examined the weekly transition of the u-BMG level after birth and reported that it decreased once from the 1-week-old value to 2 weeks after birth and then gradually increased toward the last 4 weeks, especially in severe cases of CLD. They emphasized the fact that the u-BMG value continues to rise.

Hagiwara et al. reported the usefulness of serum KL-6 as a predictor of the onset of CLD [24,25]. Furthermore, Oue et al. of the same group reported a significant difference in both the onset of severe CLD depending on the presence or absence of microorganisms in the stomach of preterm infants and in the KL-6 level within 4 weeks after birth [26].

We also have reported a very high incidence of CLD in preterm infants who develop a leukemoid reaction that causes abnormally high peripheral leukocyte counts in the early postnatal period and a high leukocyte count in the tracheal airway aspirate in the early stage of life [27]. We found that the onset of neonatal leukemoid reaction can be predicted, and others reported that the onset of neonatal leukemoid reaction is caused by the synergistic effect of IL-6 and G-CSF among various cytokines [28].

In this case as well, the 28-week-ELBWI was judged to have inevitable CLD onset because of the markedly high cord blood IgM level, the presence of CAM, and the presence of leukemoid reaction in the early neonatal period. Of note is the chronological transition of the serum KL-6 and u-BMG levels. The u-BMG value within 2 days after birth was 52,014.1 µg/gCr, which was clearly lower than the cutoff value of 100,000 µg/gCr for the onset of CLD [1-3]. But on day 3, the serum KL-6 level was already very high at 1233 U/mL and then surged within a short period of time to reach a peak value of 3016 U/mL on day 9. The transition of the u-BMG value increased thereafter, reached a peak value of 199,857.7 µg/gCr at the age of 19 days, and then decreased. An increase in BMG is substitute for an increase in IFN-γ, which resulted in the presence of hypercytokinemia in the infant [1-3]. This provided evidence of an inflammatory response occurring in the lungs and is why the high u-BMG levels were predictors of CLD onset. With this in mind, an increase in u-BMG occurs, which results in lung damage, and is followed by its subsequent repair. The rise in KL-6 should reflect this in that there should be a change in KL-6 following the change in u-BMG. In the present infant, however, the peak of KL-6 occurred 10 days earlier than that of u-BMG, and this mechanism does not apply.

IFN-γ has antifibrotic effects other than an inflammatory reaction [14-16]. It has been reported that INFγ in adult patients with idiopathic interstitial pneumonia is decreased [29]. IFN-γ induces and activates Smad7, a transforming growth factor (TGF)-β signal inhibitor, through activation of the JAK (Janus kinase) 1/STAT (signal transducer and activator of transcription) 1 signaling pathway, and as a result, finally through inhibition of phosphorylation, and the association of the TGF-β signal promoters Smad3/Smad4 suppresses the TGF-β1 signaling pathway [16]. IFN-γ also suppresses the proliferation of PDGF-induced lung fibroblasts [30]. It is further suggested that IFN-γ suppresses Th2 cytokines such as IL-4 and IL-13, which correct the Th1/Th2 balance and promote fibrosis, and brings about an inhibitory effect on fibrosis [31].

As reported in this case and by Takami et al. [9], a markedly high value of KL-6 in the 3000–4000 U/mL range has been reported to result in a poor prognosis in adults. The present patient avoided tracheal intubation and HOT and had a good course, not because of further hypercytokine storms after birth and the promotion of lung damage, but because the increase in IFN-γ was excessive. We suggest that this was the result of a biological defense reaction to suppress fibrosis.

In our study of the onset of CLD, we already measured various chemical mediators in the airway aspirates of the group that develops CLD following intrauterine infection. Among them, it was confirmed that IL-8 showed a high value early after birth [32], but thereafter, it was discovered that anti-IL-8-IgM autoantibodies also increased at the same time [33]. This suggests the possibility of a self-defense reaction to excessive inflammation in the lung fields.

In preterm infants with neonatal toxic shock syndrome such as exanthematous disease reported by Takahashi et al. [34,35], the phenomenon that an anti-inflammatory reaction occurring due to a marked increase in IL-10 coexists as one mechanism to avoid aggravation is also similar in the lungs, i.e., a self-defense reaction that protects against lung damage.

Measurements at given points in time are evidently useful in the assessment of various biomarkers. In addition, assessment of the time course of these markers is also important. Thus, observing both the individual time points and the time course of these markers is important to clearly and accurately grasp the pathological condition of neonates. In the near future, it will be necessary to accumulate additional cases of Wilson-Mikity syndrome and examine them from the aspect of repair mechanisms in premature lung injuries.

The authors have no conflicts of interest to declare.