Short-Term Outcome of Pulmonary Hemorrhage in Very-Low-Birth-Weight Preterm Infants

Pediatrics and Neonatology (2013) 54, 330e334

Available online at www.sciencedirect.com

journal homepage: http://www.pediatr-neonatol.com

ORIGINAL ARTICLE Short-term Outcome of Pulmonary Hemorrhage in Very-Low-Birth-Weight Preterm Infants

Ting-An Yen, Ching-Chia Wang, Wu-Shiun Hsieh, Hung-Chieh Chou, Chien-Yi Chen, Po-Nien Tsao*

Department of Pediatrics, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei, Taiwan

Received Nov 6, 2012; received in revised form Feb 7, 2013; accepted Apr 22, 2013

Key Words Background: Severe pulmonary hemorrhage is a serious complication with a high mortality rate high frequency in preterm infants with respiratory distress syndrome (RDS) after surfactant therapy. The aim oscillation of this study is to evaluate the efﬁcacy of our current management strategy for neonates with ventilator; severe pulmonary hemorrhage. preterm; Materials and methods: Very-low-birth-weight (VLBW) infants who developed severe pulmo- pulmonary nary hemorrhage were studied from January 2006 to August 2011. Treatment for severe pulmo- hemorrhage; nary hemorrhage in our neonatal intensive care unit (NICU) included intratracheal epinephrine surfactant; spraying/irrigation, blood component therapy, and as necessary, surfactant supplement ther- very-low-birth-weight apy was administered in cases that secondary RDS was diagnosed. High frequency oscillatory ventilation (HFOV) was utilized when hypoxia or respiratory acidosis persisted under conventional mechanical ventilation (CMV). We then described the clinical courses of severe pulmonary hemorrhage following our management. Results: A total of 18 (3.2%) out of 469 VLBW infants developed severe pulmonary hemorrhage. The mean gestational age was 27 weeks, the mean birth weight was 822 g, and the onset age was 2.5 days after birth. There was no severe pulmonary hemorrhage-associated mortality during this period with the exception of one case, in which an infant died after the parents refused to do further therapy. Sixteen (88.8%) neonates had RDS and 13 received surfactant therapy. Twelve (66.6%) cases developed secondary RDS following the onset of severe pulmonary hemorrhage, and four cases received surfactant supplement therapy. In the surfactant

supplement group, alveolar-arterial oxygen difference (AaDO2) and oxygenation index (OI) during the 2e4 hours postpulmonary hemorrhage period showed statistically signiﬁcant

* Corresponding author. Department of Pediatrics, National Taiwan University Hospital, National Taiwan University College of Medicine, Number 7, Chung-Shan South Road, Taipei 100, Taiwan. E-mail address: [email protected] (P.-N. Tsao).

1875-9572/$36 Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved. http://dx.doi.org/10.1016/j.pedneo.2013.04.005 Pulmonary hemorrhage in preterm infants 331

improvement, whereas the other group only showed a tendency toward improvement without reaching statistical signiﬁcance when compared to the baseline data. Duration of high oxygen

requirement [deﬁned as fraction of inspired oxygen (FiO2) > 40%] was also less in the surfactant supplement group. Conclusion: This data suggests that our current strategy is effective for treating severe pulmonary hemorrhage in VLBW infants. Surfactant therapy for severe pulmonary hemorrhage may also be beneﬁcial for improving lung function and may shorten the duration of high oxygen requirement. Copyright ª 2013, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.

1. Introduction 2.2. Management protocol

Pulmonary hemorrhage occurs mainly in ventilated preterm All cases received 0.5 cc epinephrine (1:10000) spraying/ infants with severe respiratory distress syndrome (RDS), who irrigation via ET each time after ET suction until pulmonary have received surfactant therapy. The incidence of pulmo- hemorrhage was resolved. Patients with severe pulmonary nary hemorrhage was around 0.5e11% among a population of hemorrhage received ET suction only when they had dys- very-low-birth-weight (VLBW) infants.1e6 Pulmonary hem- pnea, hypercapnea, or hypoxemia. The epinephrine spray- orrhage is associated with significant morbidity and mortal- ing was conducted as fast as possible to create a splash via a ity, and the mortality rate is high, ranging from 50% to force injection of a mixture of 0.5 cc diluted epinephrine 82%.7e10 Risk factors for pulmonary hemorrhage included with 1 cc air in a 5 cc syringe connected to an oral-gastric prematurity, intrauterine growth restriction, respiratory tube. If respiratory failure was reflected as FiO2 > 60% with problems, patent ductus arteriosus (PDA), bleeding prob- a mean airway pressure (MAP) > 12 cm H2O, and/or per- lems, ventilator usage, and surfactant treatment.4,11,12 sisted respiratory acidosis (pH < 7.25) in spite of conven- Pulmonary hemorrhage was thought to be a complica- tional mechanical ventilator (CMV) support, we switched tion of surfactant therapy. Thus, clinicians might hesitate out the ventilator to a high frequency oscillation ventilator to administer surfactant again in neonates presenting with [HFOV; SensoreMedics model 3100 (Critical Care Corp., pulmonary hemorrhage. However, the presence of hemo- Yorda Linda, CA, USA)] to avoid barotrauma while main- globin, red cell membrane lipids, and serum proteins can taining optimal lung volume. Ventilator setting, arterial increase the surface tension resulting from surfactant blood gases, oxygenation index (OI), and alveolar-arterial 13 dysfunction and cause secondary RDS. oxygen difference (AaDO2) were recorded at specific The purpose of this study is to analyze the incidence and intervals for at least 8 hours. The decision to administer short-term outcome of severe pulmonary hemorrhage in surfactant following the pulmonary hemorrhage-induced VLBW infants following our current treatment, including RDS was made by each attending neonatologist. Blood intratracheal epinephrine spraying/irrigation; blood component therapy was given if coagulopathy or thrombo- component therapy; and in some cases, surfactant sup- cytopenia was evident. plement in our NICU from January 2006 to August 2011. 2.3. Outcomes 2. Materials and Methods Resolution of pulmonary hemorrhage in our cases was defined as cessation of fresh blood aspirate from the ET and 2.1. Patients achieving clinical stabilization by decreasing oxygen de- mand and ventilator setting. The outcome was a change in This 5-year retrospective study was conducted in the respiratory status following therapy as reflected by change neonatal intensive care unit at National Taiwan University in OI, AaDO2, and the duration of high oxygen requirement Hospital between 2006 and 2011. All infants weighing (defined as time to wean FiO2 < 40%) between different <1500 g and who had developed severe pulmonary hem- therapy groups. Statistical analysis was performed using the orrhage were eligible for the study. The definition of severe student t test. A p value < 0.05 was considered statistically pulmonary hemorrhage was based on previously established significant. criteria; namely the presence of blood-stained aspirates from endotracheal tube (ET) with either of the followings: multilobular infiltrates on chest radiograph, increased 3. Results ventilation support (fraction of inspired oxygen (FiO2) increase of >0.3 from the baseline), or acute drop of hem- A total of 469 VLBW preterm neonates were admitted to our atocirt (>10%).8 Secondary RDS was defined as deteriorated NICU during this period, of which 26 (5.5%) cases developed respiratory symptoms in clinical course and white-out pulmonary hemorrhage; however, only 18 (3.8%) patients pattern from the chest radiograph. fulfilled the criteria of severe pulmonary hemorrhage. The 332 T.-A. Yen et al average gestational age (GA) and birth weight (BW) of these Table 1 Characteristics of patients with severe pulmo- patients were 27 2 weeks and 820 246 g, respectively. nary hemorrhage. Among the 18 VLBW neonates with severe pulmonary hemorrhage, 16 (88.9%) patients had RDS and 13 required Characteristics Patients with surfactant therapy. PDA was documented by cardiac echog- severe pulmonary raphy in 8 patients (44.4%). Among the 13 cases with RDS hemorrhage requiring surfactant therapy, severe pulmonary hemorrhage (n Z 18) always appeared within 72 hours after surfactant treatment. Birth weight (g) 820 246 There were 5 patients who did not survive to the discharge, Gestational age (wk) 27 2 and one was associated with severe pulmonary hemorrhage Sex M/F 9/9 for which his parents refused further therapy, and this case Age (d) 2.4 1.3 was excluded from the following analysis. The other four RDS requiring surfactant therapy 13/18 (72.2) patients survived from pulmonary hemorrhage but later died Hemorrhage onset within 13/13 (100) from other causes: One patient developed pneumothorax 72 h of surfactant and pneumopericardium incidentally and died 3 days later Epinephrine use (irrigation/spraying) 18/18 (100) following the resolution of pulmonary hemorrhage. Two pa- HFOV use 14/17 (82.3) tients had refractory hypotension: in one patient, it was Pulmonary hemorrhage-related RDS 12/17 (70.5) caused by cardiac tamponade at the age of 44 days, and the Secondary surfactant therapy 4/12 (33.3) other one developed severe intraventricular hemorrhage, for severe pulmonary hemorrhage complicated with hyperkalemia at 3 days of age. One infant OI prior to hemorrhage 8 1 died at 1 month of age from systemic fungal infection. The AaDO2 prior to hemorrhage 146 95 majority of pulmonary hemorrhage cases were resolved after Death due to severe pulmonary 1/18 (5.5)* 3e5 treatments of epinephrine spraying, even when the hemorrhage neonates were still under a high setting of ventilation sup- Survived to discharge 13/18 (72.2) port. Twelve (70.5%) of the 17 patients developed secondary RDS because of surfactant dysfunction, and four of them Data are presented as n (%) or mean SD unless otherwise received an additional dose of surfactant therapy. HFOV was indicated. OI: [(FiO2 Mean Airway Pressure)/PaO2] 100, AaDO : [(713 FiO2) (PaCO /0.8)] PaO . utilized in 14 neonates (Table 1). 2 2 2 * One patient died after his parents refused further therapy. To analyze the effect of surfactant therapy for those AaDO2 Z alveolar-arterial oxygen difference; HFOV Z high infants with severe pulmonary hemorrhage-induced sec- frequency oscillatory ventilation; OI Z oxygenation index; ondary RDS, we compared the blood gases, OI, and AaDO2 RDS Z respiratory distress syndrome. between treatment with and without the surfactant group. The basic profile of both groups (gestational age and birth weight), initial values of blood gases, OI, and AaDO2 after hemorrhage is not a good indication for HFOV therapy. The OI severe pulmonary hemorrhage induced RDS were not and AaO2 both decreased dramatically during HFOV support. significantly different between these two groups. However, All patients survived from severe pulmonary hemorrhage the average values of OI and AaDO2 were higher in the without HFOV-related complications. surfactant treatment group (Table 2). The average time of surfactant therapy was 1.2 hours following the development of severe pulmonary hemor- Table 2 Characteristics of patients who developed sec- rhage, with a range of 0.5 hour to 1.5 hours. In the sur- ond RDS following severe pulmonary hemorrhage. factant treatment group, there was a statistically Surfactant Without surfactant p* Z significant improvement in AaDO2 (p 0.029) and OI, group (n Z 4) group (n Z 8) (p Z 0.012) when comparing the values between the initial Gestational 27 1.4 27.1 1.45 0.93 posthemorrhage period (0e1 hour) and the 2- to 4-hour age (wk) posthemorrhage period, while whereas the group without Birth weight 902.5 2521.7 858.6 252.9 0.74 surfactant treatment only showed a tendency toward Sex (M/F) 3/1 5/3 improvement (AaDO , p Z 0.457; OI, p Z 0.606; Table 3). 2 Values at hemorrhage In addition, half (4/8) of the neonates without surfactant MAP 15.0 2.2 12.3 3.3 0.19 treatment therapy had increased OI in the second hour FiO 1 0 0.9 0.1 0.26 compared with the first hour after the onset of severe 2 pH 7.13 0.17 7.14 0.16 0.85 pulmonary hemorrhage (data not shown). By contrast, all PaCO 66.4 54.3 60.7 15.9 0.35 neonates in the surfactant treatment groups had substan- 2 OI 37.8 9.3 29 19.7 0.45 tially decreased OI in the second hour following the onset of AaDO 563.6 122.2 535.7 71.6 0.35 severe pulmonary hemorrhage when compared to the first 2 postpulmonary hemorrhage hour. In the surfactant treat- Data are presented as mean standard deviation. Z ment group, the duration to wean FiO2 < 40% after pul- AaDO2 alveolar-arterial oxygen difference; Z Z monary hemorrhage was also shorter compared to the other FiO2 fractionated inspired oxygen; MAP mean airway Z Z group (p Z 0.05). pressure; OI oxygenation index; PaCO2 partial arterial Z After the development of severe pulmonary hemorrhage, carbon dioxide; RDS respiratory distress syndrome. * unpaired two-tailed t test, and p < 0.05 is considered HFOV was provided to 12 patients who originally used CMV. significant. HFOV worked well in all cases even though pulmonary Pulmonary hemorrhage in preterm infants 333

Table 3 Values following severe pulmonary hemorrhage-induced RDS among neonates with (n Z 4) and without surfactant therapy (n Z 8). With surfactant Without surfactant

AaDO2 OI AaDO2 OI 0e1 h 563.6 122.2 37.8 9.3 524.5 71.4 27.5 21.1 1e2 h 465.4 131.0 28.6 17.0 488.6 114.8 26.5 14.8 2e4 h 334.7 144.4* 16.7 7.4* 471.2 136.2 20.6 14.9 y y Time taken to reach FiO2 < 40% (h) 13.4 (11e16) 41.1 (8e108)

AaDO2 Z alveolar-arterial oxygen difference; OI Z oxygenation index; RDS Z respiratory distress syndrome. * Compared to 0e1 hour data, paired two-tailed t test, and p < 0.05 is considered signiﬁcant. y Two groups comparison, unpaired two-tailed t test, and p Z 0.05.

4. Discussion following the onset of severe pulmonary hemorrhage. This indicates that secondary RDS after severe pulmonary hem- In this study, we demonstrated a good short-term outcome of orrhage is common in VLBW infants and should also be severe pulmonary hemorrhage in VLBW infants using our treated properly. As posited by Amizuka et al, lower birth current treatment policy. Our policy for managing severe weight and lower gestational age are significantly associated pulmonary hemorrhage included stopping pulmonary hem- with having a good response to exogenous surfactant after 20 orrhage by epinephrine spraying through an ET, thereby pulmonary hemorrhage. A shorter interval between the avoiding unnecessary endotracheal suction, and treating onset of pulmonary hemorrhage and surfactant administra- secondary RDS by HFOV with and without surfactant. Our tion showed an immediate response to the surfactant, fol- data showed that active pulmonary hemorrhage usually lowed by no increase in ventilator index. In our study, the stopped after 3e5 instances of epinephrine spraying. This average time of surfactant administration was 1.2 hours after suggests that using epinephrine spraying through an ET is an the development of severe pulmonary hemorrhage and e effective therapy to treat pulmonary hemorrhage.14 We improvement of OI and AaDO2 was most significant at 2 4 replaced conventional mechanical ventilation with HFOV in hours posthemorrhage in the surfactant group. No patients 14 out of the 18 patients as a result of hypoxemia and/or deteriorated or experienced a recurrent pulmonary hemor- persistent respiratory acidosis. Generally, HFOV is consid- rhage following surfactant administration. In another study ered less effective in an obstructive lung condition like a by Pandit et al, OI and a/A ratio showed significant pulmonary hemorrhage, and few preliminary studies aimed improvement during the 3- to 6-hour postsurfactant period. at analyzing the effect of HFOV in respiratory distress sec- Although no formal recommendation based on randomized ondary to pulmonary hemorrhage in VLBW infants have been study can be presented to evaluate the effect of surfactant 22 conducted.15e17 Ko et al16 reported that 13 (72%) of 18 of therapy in pulmonary hemorrhage, our data and previous their patients responded to HFOV and survived, and their OI reports suggest that surfactant therapy may be beneficial for 9,10,20,21 decreased significantly at 1 hour after HFOV therapy.16 cases with severe pulmonary hemorrhage. Alkharfy et al achieved a 59% of survival rate (10/17), and In conclusion, treatments including epinephrine spraying OI improved significantly at 3 hours after HFOV therapy.17 In through an ET, avoiding unnecessary endotracheal suction, our study, none of the patients died within 24 hours of severe blood component therapy, HFOV support, and surfactant pulmonary hemorrhage under HFOV therapy. Similarly, the therapy (in the case of secondary RDS) were found to be safe and effective for VLBW infants who developed severe FiO2, OI, and AaDO2 decreased significantly after HFOV use for 1 hour. These data suggest that HFOV could be used as an pulmonary hemorrhage. alternative ventilator therapy in cases where CMV fails to support VLBW infants with severe pulmonary hemorrhage. References Pulmonary hemorrhage appears to be a complication of surfactant therapy. Increased risk of pulmonary hemor- 1. Bhandari V, Gagnon C, Rosenkrantz T, Hussain N. Pulmonary rhage has been reported following surfactant therapy hemorrhage in neonates of early and late gestation. J Perinat 4,18 either as prophylaxis or treatment of RDS. Improved lung Med 1999;27:369e75. compliance after surfactant therapy with a left-to-right 2. Garland J, Buck R, Weinberg M. Pulmonary hemorrhage risk shunt at ductus arteriosus may contribute to pulmonary in infants with a clinically diagnosed patent ductus arte- hemorrhage. Surfactant therapy has also been demon- riosus: a retrospective cohort study. Pediatrics 1994;94: strated to cause in vitro cytotoxicity, leading to impaired 719e23. membrane integrity at the level of the alveolar capillary.19 3. Horbar JD, Wright LL, Soll RF, Wright EC, Fanaroff AA, The presence of hemoglobin, red cell membrane lipids, Korones SB, et al. A multicenter randomized trial comparing two surfactants for the treatment of neonatal respiratory and serum protein increased surface tension and inactivated 13 distress syndrome. National Institute of Child Health and the surfactant, resulting in secondary RDS. Therefore, Human Development Neonatal Research Network. J Pediatr there was a rationale for the use of surfactant in pulmonary 1993;123:757e66. 9,20,21 hemorrhage induced RDS. In our retrospective case 4. Raju TN, Langenberg P. Pulmonary hemorrhage and exogenous series, 13 neonates (72.2%) developed secondary RDS surfactant therapy: a metaanalysis. J Pediatr 1993;123:603e10. 334 T.-A. Yen et al

5. Stevenson D, Walther F, Long W, Sell M, Pauly T, Gong A, 14. Chen YY, Wang HP, Lin SM, Chang JT, Hsieh KS, Huang FK, et al. et al. Controlled trial of a single dose of synthetic surfactant Pulmonary hemorrhage in very low birth weight infants: risk at birth in premature infants weighing 500 to 699 grams. The factors and management. Pediatr Int 2012;54:743e7. American Exosurf Neonatal Study Group I. J Pediatr 1992; 15. Pappas MD, Sarnaik AP, Meert KL, Hasan RA, Lieh-Lai MW. 120(2 pt 2):S3e12. Idiopathic pulmonary hemorrhage in infancy. Clinical features 6. Zola EM, Overbach AM, Gunkel JH, Mitchell BR, Nagle BT, and management with high frequency ventilation. Chest 1996; DeMarco NG, et al. Treatment Investigational New Drug 110:553e5. experience with Survanta (beractant). Pediatrics 1993;91: 16. Ko SY, Chang YS, Park WS. Massive pulmonary hemorrhage in 546e51. newborn infants successfully treated with high frequency 7. Dufourq N, Thomson M, Adhikari M, Moodley J. Massive pul- oscillatory ventilation. J Korean Med Sci 1998;13:495e9. monary haemorrhage as a cause of death in the neonate e a 17. AlKharfy TM. High-frequency ventilation in the management of retrospective review. S Afr Med J 2004;94:299e302. very-low-birth-weight infants with pulmonary hemorrhage. Am 8. Lin TW, Su BH, Lin HC, Hu PS, Peng CT, Tsai CH, et al. Risk J Perinatol 2004;21:19e26. factors of pulmonary hemorrhage in very-low-birth-weight in- 18. van Houten J, Long W, Mullett M, Finer N, Derleth D, fants: a two-year retrospective study. Acta Paediatr Taiwan McMurray B, et al. Pulmonary hemorrhage in premature infants 2000;41:255e8. after treatment with synthetic surfactant: an autopsy evalua- 9. Pandit PB, Dunn MS, Colucci EA. Surfactant therapy in neonates tion. The American Exosurf Neonatal Study Group I, and with respiratory deterioration due to pulmonary hemorrhage. the Canadian Exosurf Neonatal Study Group. J Pediatr 1992; Pediatrics 1995;95:32e6. 120(2 pt 2):S40e4. 10. Pandit PB, Dunn MS, Kelly EN, Perlman M. Surfactant replace- 19. Findlay RD, Taeusch HW, David-Cu R, Walther FJ. Lysis of red ment in neonates with early chronic lung disease. Pediatrics blood cells and alveolar epithelial toxicity by therapeutic 1995;95:851e4. pulmonary surfactants. Pediatr Res 1995;37:26e30. 11. Kluckow M, Evans N. Ductal shunting, high pulmonary blood 20. Amizuka T, Shimizu H, Niida Y, Ogawa Y. Surfactant therapy in ﬂow, and pulmonary hemorrhage. J Pediatr 2000;137:68e72. neonates with respiratory failure due to haemorrhagic pul- 12. Finlay ER, Subhedar NV. Pulmonary haemorrhage in preterm monary oedema. Eur J Pediatr 2003;162:697e702. infants. Eur J Pediatr 2000;159:870e1. 21. Wiswell TE. Expanded uses of surfactant therapy. Clin Peri- 13. Holm BA, Notter RH. Effects of hemoglobin and cell membrane natol 2001;28:695e711. lipids on pulmonary surfactant activity. J Appl Physiol 1987;63: 22. Aziz A, Ohlsson A. Surfactant for pulmonary haemorrhage in 1434e42. neonates. Cochrane Database Syst Rev 2012;7:CD005254.