Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Prevention, management, and complications

INTRODUCTION — Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) remains an important cause of brain injury in preterm infants. The negative impact of GMH-IVH on neurodevelopmental outcome is due not only to the direct consequences of GMH-IVH, but also to associated lesions including posthemorrhagic ventricular dilatation (PHVD) and white matter injury (WMI).

The prevention, management, complications, and outcome of GMH-IVH in the preterm infant are discussed in this topic review. The epidemiology, pathogenesis, clinical manifestations, and diagnosis of GMH-IVH are discussed separately. (See "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis".)

SEVERITY AND GRADING — The severity of GMH-IVH is based on the presence of blood in the germinal matrix and amount of blood in the lateral ventricles, evidence of white matter injury, and ventricular dilatation as demonstrated by cranial ultrasound (table 1):

●Grade I – Bleeding is confined to the germinal matrix (image 1) or GMH plus IVH <10 percent of the lateral ventricular area

●Grade II – GMH-IVH occupies 10 to <50 percent of the lateral ventricle volume (image 2)

●Grade III – GMH-IVH occupies more than 50 percent of the lateral ventricle volume and is associated with acute ventricular dilatation (image 3)

●Periventricular hemorrhagic infarction (PVHI; previously referred to as Grade IV IVH) – Hemorrhagic infarction in periventricular white matter ipsilateral to large IVH (image 4 and image 5)

Grades I and II are defined as low-grade GMH-IVH and grades III and PVHI (grade IV) as severe. Newborn infants with severe GMH-IVH have poorer neurodevelopmental outcome than infants with milder GMH-IVH. (See 'Outcome' below.)

PREVENTION — The most effective strategy for prevention of GMH-IVH is prevention of preterm birth, which is difficult to accomplish. However, there are other prenatal and postnatal interventions that may reduce the risk of GMH-IVH.

Prenatal and delivery room interventions — The following routine interventions have been reported to impact on the risk of GMH-IVH in preterm infants.

●Maternal transport – Although data are inconsistent regarding the benefit of maternal transport for reducing GMH-IVH for births of very preterm (VPT) infants (gestational age [GA] <32 weeks), maternal transport with adequate staffing and equipment should be considered for VPT infants as outcomes are poorer for infants born outside and transferred to level III neonatal care units compared with inborn infants. (See "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Neonatal transport' and "Inter-facility maternal transport", section on 'Background'.)

●Treatment of chorioamnionitis – Detection and treatment of maternal intrauterine infection is associated with a decrease in the risk of GMH-IVH.

●Antenatal corticosteroids – Antenatal corticosteroids should be given to mothers at risk for preterm delivery within the next seven days to reduce the risk of neonatal respiratory disease. In addition, antenatal corticosteroids reduce the risk of GMH-IVH as illustrated by a systematic meta-analysis of clinical trials showing antenatal corticosteroids given before preterm birth to reduce the risk of IVH (RR 0.58, 95% CI 0.45-0.75) [1]. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery" and "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Antenatal glucocorticoid therapy'.)

●Tocolytic therapy – Observational findings from the 2011 EPIPAGE-2 preterm cohort suggest that the use of tocolytic therapy, which delays preterm delivery resulting in reduced mortality and the risk of severe IVH (grade III and periventricular hemorrhagic infarction (PVHI) [grade IV]) [2].

General neonatal care — Although evidence based on clinical trials is limited, the following general measures based on expert opinion and clinical experience are used in the care of VPT infants to reduce the risk of GMH-IVH after birth [3,4]. (See "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Pathogenesis' and "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Possible risk factors'.)

●Delayed cord clamping – It is uncertain whether delayed clamping of the umbilical cord reduces GMH-IVH based on the available literature. Nevertheless, delayed clamping (30 to 60 seconds) is suggested for vigorous preterm infants who do not need resuscitation, because of potential associated benefits including reduced risk of GMH-IVH. In contrast, cord milking may be associated with increased risk for severe GMH-IVH and should be avoided [5,6]. Delayed cord clamping and milking are discussed separately. (See "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Delivery' and "Labor and delivery: Management of the normal third stage after vaginal birth", section on 'Early versus delayed cord clamping' and "Labor and delivery: Management of the normal third stage after vaginal birth", section on 'Cord milking'.)

●Neonatal resuscitation – Prompt and appropriate resuscitation in the delivery room, including efforts to avoid hemodynamic instability or conditions that impair cerebrovascular autoregulation, such as hypoxia, hypercarbia, hyperoxia, and hypocarbia. High tidal volumes during mask ventilation have been associated with GMH-IVH and should be avoided [7]. (See "Neonatal resuscitation in the delivery room" and "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Additional risk factors'.)

●Blood pressure (BP) management – Hypotension and hypertension should be avoided, and if present, corrected judiciously. Acute changes in BP, which are associated with GMH-IVH, should be avoided if possible. Interventions for low BP should be reserved for preterm infants who also have evidence of poor perfusion (neonatal shock). Neonatal shock and the e management of low BP in extremely preterm infants (GA <28 weeks) is challenging and is discussed in greater detail separately (See "Neonatal shock: Management" and "Assessment and management of low blood pressure in extremely preterm infants", section on 'Management approach'.)

●Handling and minimizing elevations of BP – Care should be provided that maintains the infant's head in the midline and tilted upwards with avoidance of sudden elevation of the legs and flushing/rapid withdrawal of blood [8-10]. However, in a large retrospective study, no protective effect for midline position was noted [11]. But, the study result must be interpreted carefully as the number of infants who were outborn and transferred was disproportionally higher in the midline position group, and there was no mention on whether an ultrasound was performed on admission after transport to screen for IVH development prior to practice application.

●Respiratory support – Respiratory support as needed to ensure adequate oxygenation and ventilation, and avoid hypoxemia. (See "Neonatal target oxygen levels for preterm infants", section on 'Oxygen target levels' and "Respiratory support, oxygen delivery, and oxygen monitoring in the newborn".)

●Management of metabolic derangements – Metabolic abnormalities, especially those that cause fluid shifts across cell membranes (eg, hyperosmolality, hyperglycemia, and hypoglycemia), should be prevented. Metabolic acidosis or alkalosis should be corrected carefully. Bicarbonate therapy should be avoided in preterm infants because it is associated with an increased risk of GMH-IVH. (See "Neonatal acute kidney injury: Pathogenesis, etiology, clinical presentation, and diagnosis", section on 'Presentation due to other laboratory abnormalities'.)

●Limiting transfusion – Avoidance of transfusion of packed red blood cells (RBCs) during the first days of life may decrease the incidence of severe IVH [12]. However, further investigation is required to determine whether the observation is due to the clinical status of the infant (ie, more severely ill infants requiring transfusion are more likely at risk of severe IVH), or if transfusion is an independent risk factor. (See "Red blood cell transfusions in the newborn".)

●Correction of coagulopathy – Administration of blood products (eg, fresh frozen plasma [FFP] and recombinant-activated Factor VII) is generally provided to correct coagulopathy. However, data is insufficient and inconsistent to determine if coagulation abnormalities are associated with an increased risk of GMH-IVH [13-15] and whether correction of coagulopathy reduces the incidence of GMH-IVH. There is no convincing evidence to support prophylactic administration of FFP to prevent GMH-IVH [16]. If a GMH-IVH is present on the admission cranial ultrasound, we would recommend testing prothrombin time (PT), activated partial thromboplastin time (aPTT), and international normalized ratio (INR). If any of these tests are prolonged based on neonatal thresholds, we would consider administration of FFP to prevent extension of the bleed (table 2 and table 3) [17].

●Platelet transfusion – Platelet transfusions are given prophylactically to neonates with a platelet count of <20,000/microL. A higher threshold for transfusion of <50,000/microL is used for extremely preterm infants (GA <28 weeks). (See "Neonatal thrombocytopenia: Clinical manifestations, evaluation, and management", section on 'Platelet transfusion'.)

●Patent ductus arteriosus – The incidence of GMH-IVH is greater in preterm infants with patent ductus arteriosus (PDA) compared with those without PDA [18]. However, the optimal management of PDA remains uncertain, but early treatment of a clinically significant PDA, particularly with indomethacin, may reduce IVH severity [19]. (See "Patent ductus arteriosus in preterm infants: Pathophysiology, clinical manifestations, and diagnosis", section on 'Systemic and cerebral blood flow effects' and "Patent ductus arteriosus in preterm infants: Management", section on 'Overview'.)

Unproven or ineffective interventions

Antenatal interventions — The following antenatal interventions do not appear to reduce the risk of GMH-IVH:

●Phenobarbital – Antenatal administration of phenobarbital to women in preterm labor does not reduce the risk of neonatal intracranial hemorrhage or early death [20]. Postnatal administration of phenobarbital also does not appear to prevent GMH-IVH in preterm infants. In a systematic review, postnatal administration of phenobarbital did not reduce the rate of GMH-IVH, severe GMH-IVH, neurologic impairment, ventricular dilatation, or death, and was associated with increased need for mechanical ventilation [21].

●Vitamin K – Administration of vitamin K to the mother prior to very preterm birth does not appear to significantly prevent GMH-IVH. In a meta-analysis, antenatal vitamin K was associated with a reduction of severe GMH-IVH, but when low quality studies were excluded from the analysis, the effect was no longer statistically significant [22].

●Magnesium sulfate (MgSO₄) – Although antenatal magnesium sulfate reduces the risk of cerebral palsy [23], it does not appear to reduce the risk of GMH-IVH in preterm infants between 24 and 32 weeks gestation. In a meta-analysis the effects of antenatal MgSO₄ for the prevention of cerebral palsy, there was no difference in the rate of GMH-IVH on serial cranial ultrasounds for the group receiving MgSO₄ and controls (21.5 versus 21.9 percent; relative risk [RR] 0.98, 95% CI 0.87-1.09) and for severe GMH-IVH (grades III and IV) (4.5 versus 5.2 percent; RR 0.63, 95% CI 0.63-1.09) [24]. (See "Neuroprotective effects of in utero exposure to magnesium sulfate".)

Delivery — The available data suggest that the mode of delivery does not affect the risk of GMH-IVH. As a result, cesarean delivery should be performed according to standard obstetric indications. (See "Germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) in the newborn: Pathogenesis, clinical presentation, and diagnosis", section on 'Delivery' and "Delivery of the low birth weight singleton fetus" and "Delivery of the low birth weight singleton fetus", section on 'Choosing the route of delivery'.)

Postnatal interventions — The following postnatal interventions have not been shown to safely reduce the long-term effects of GMH-IVH:

●Postnatal nonsteroidal antiinflammatory agents – Prophylactic cyclooxygenase (COX) inhibitors have been proposed as an intervention to reduce the incidence of PDA and improve neonatal outcomes. However, this approach does not appear to be more effective than early treatment of a symptomatic PDA and it would unnecessarily expose many infants who would not develop significant PDAs to COX inhibitors with potentially serious adverse effects. As a result, we generally recommend not administering prophylactic treatment with COX inhibitors in preterm infants. This issue is discussed in detail separately. (See "Patent ductus arteriosus in preterm infants: Management", section on 'Prophylactic/early therapy'.)

●Vitamin E – Although vitamin E, a strong antioxidant, reduces the risk of IVH, especially severe GMH-IVH, it is associated with an increased risk of sepsis, particularly when given intravenously to preterm infants [25]. As a result, it is not recommended for routine use to prevent GMH-IVH.

●Ethamsylate – Ethamsylate (also referred to as etamsylate), a hemostatic agent, has been proposed to reduce the risk of GMH-IVH. A systematic meta-analysis reported that treatment with ethamsylate compared with placebo decreased the risk of GMH-IVH in preterm infants less than 31 weeks gestation (relative risk [RR] 0.63, 95% CI 0.47-0.86), but there was no difference in mortality or developmental outcome at two years of age [26]. As the available data do not demonstrate long-term benefits in infants less than 32 weeks gestation (most at-risk group), we do not suggest the routine prophylactic use of ethamsylate. In addition, ethamsylate is not available in the United States.

●Erythropoietin – A multicenter trial reported that high-dose erythropoietin (EPO) administered within 24 hours after birth did not reduce the risk of GMH-IVH [27]. As a result, it should not be given as a preventive measure to reduce the risk of GMH-IVH.

A single-center trial reported that repeated low-dose EPO in 316 preterm infants (<32 weeks GA) diagnosed with IVH within 72 hours of delivery did not affect mortality or the risk of neurologic disability at 18 months of corrected age [28]. However, multivariate analysis showed that the combined outcome of death and disability at 18 months of age was lower in the EPO group compared with the placebo group (14.9 versus 26.4 percent, odds ratio 0.40, 95% 0.20-0.80). Nevertheless, further studies are needed to determine the neuroprotective effect of EPO in infants with IVH.

Our approach — Our approach to reduce the risk of GMH-IVH encompasses interventions that are used as best practices in general for the prenatal and postnatal management of very preterm infants (GA <32 weeks) and includes the following:

●Maternal, intrauterine transport.

●Antenatal administration of glucocorticoid steroids to pregnant women who are at 23 to 34 weeks gestation and at increased risk of preterm delivery within the next seven days.

●Delayed cord clamping for vigorous preterm infants who do not need resuscitation.

●Prompt resuscitative efforts for preterm infants who are not vigorous.

●Respiratory support as needed to ensure adequate oxygenation and ventilation.

●Correction of metabolic disorders.

●Fluid resuscitation and inotropic support reserved for infants with low blood pressure and inadequate perfusion.

●Coagulation correction.

●PDA intervention only for symptomatic infants.

●A bundle of nursing interventions including reduced and more gentle handling, optimal positioning, and avoidance of flushing/rapid blood withdrawal and sudden leg elevation) [8].

MANAGEMENT — No specific therapy exists to limit the extent of GMH-IVH after it has occurred. Treatment of GMH-IVH is supportive and directed towards preservation of cerebral perfusion, minimization of any further brain injury, and early detection of complications [3].

●Supportive care

•Maintenance of arterial perfusion to avoid hypotension or hypertension and preserve cerebral blood flow without significant or sudden perturbations. (See "Assessment and management of low blood pressure in extremely preterm infants", section on 'Management approach'.)

•Adequate oxygenation and ventilation with specific avoidance of hypocarbia, hypercarbia, and acidosis. (See "Neonatal target oxygen levels for preterm infants".)

•Provision of appropriate fluid, metabolic, and nutritional support. (See "Fluid and electrolyte therapy in newborns".)

•Correction of coagulopathy

●Seizure management – Seizures should be treated to avoid any associated impairment of cerebral oxygenation and cerebral perfusion, or elevations of systemic blood pressure. (See "Clinical features, evaluation, and diagnosis of neonatal seizures" and "Treatment of neonatal seizures".)

●Serial monitoring – For neonates with an established diagnosis of severe GMH-IVH, we suggest monitoring with twice weekly cranial ultrasound for four weeks after its onset to detect posthemorrhagic ventricular dilatation (PHVD), the major complication of severe IVH, prior to the development of signs and symptoms due to PHVD. However, ultrasound monitoring does not obviate the need for clinical monitoring and all affected neonates should be monitored for increasing head circumference and/or signs and symptoms of increased intracranial pressure (ICP), which are late findings of PHVD. Signs of elevated ICP in preterm neonates include apnea, bradycardia, irritability, sunsetting, and feeding difficulties. The management of PHVD is discussed below. (See 'Management' below.)

●Unproven therapies [29-31] such as recombinant-activated Factor VII, which is being tested as a potential treatment for newborns with GMH-IVH, but large randomized trials have not been conducted yet to demonstrate efficacy or safety [32,33].

COMPLICATIONS

Posthemorrhagic ventricular dilatation (PHVD) — Posthemorrhagic ventricular dilatation (PHVD), also referred to as posthemorrhagic hydrocephalus (PHH), is a major complication of severe GMH-IVH (ie, Grade III and periventricular hemorrhagic infarction [PVHI, previously referred to as grade IV IVH]). PHVD is associated with increased risk of mortality and neurodevelopmental impairment (NDI) [34-36]. Of note, it is especially important to differentiate grade III GMH-IVH with acute dilatation due to the clot filling more than 50 percent of the lateral ventricle from ventricular dilatation due to PHVD with a grade II GMH-IVH, as confusion between the two conditions commonly occurs in the clinical setting.

PHVD is thought to be caused by impaired reabsorption of the cerebrospinal fluid (CSF) due to inflammation of the subarachnoid villi by blood [3,35,37]. Transforming growth factor beta one (TGF-B1), one of the inflammatory factors, stimulates the production of extracellular matrix, which causes scarring and obstruction of arachnoid villi [38-40]. This results in communicating hydrocephalus, in which the entire ventricular system is dilated. Less frequently, infants can have noncommunicating hydrocephalus due to obstruction by a clot or scarring within the ventricular system, most often at the level of the aqueduct. (See "Hydrocephalus in children: Physiology, pathogenesis, and etiology", section on 'Pathogenesis'.)

Risk factors — The risk of PHVD increases with the severity of GMH-IVH, decreasing gestational age (GA), and greater severity of illness [3,41,42]. It commonly occurs in infants with grade III GMH-IVH or PVHI (grade IV GMH-IVH) [42-44]. In a study based on data from two large patient databases including 147,823 preterm infants with GMH-IVH, the risk of PHVD increased with the severity of GMH-IVH; 1, 4, 25 and 28 percent for grades I, II, III and PVHI, respectively [44].

Clinical presentation and course — PHVD usually begins within one to three weeks after the onset of severe IVH. The early stages of PHVD are detected by screening cranial ultrasounds. The clinical presentation of increasing head circumference and signs of increased intracranial pressure (ICP) occurs late in the course and usually presents several weeks after ventricular dilatation is detected by brain imaging studies (image 6 and image 7 and image 8) [45].

There are three different clinical courses of PHVD [3,41]:

●Spontaneous arrest without a need for intervention (40 percent)

●Rapid progression (10 percent)

●Persistent slow progression (50 percent), in which 20 percent of total infants do not progress after intervention with temporizing CSF drainage (from serial lumbar punctures [LPs] or reservoir but no shunting), and 30 percent eventually require permanent shunting.

An additional 5 percent of infants with spontaneous or treatment-associated arrest of PHVD, will develop late progression of their disease usually during the newborn period or, rarely, after discharge from the neonatal intensive care unit (NICU) in the first year of life.

Both mortality and impaired neurodevelopmental outcome are greater in preterm infants who develop PHVD compared with those without PHVD, especially extremely preterm infants [34,36]. The deleterious effects of PHVD are thought to be caused by injury to the periventricular white matter [46]. In one small study, magnetic resonance imaging [MRI] around term equivalent age (TEA) showed decreased volumes of deep gray matter and cerebellum in infants with PHVD compared with controls without PHVD, implying a greater amount of white matter injury (WMI) in infants with PHVD [47]. Diffusion tensor imaging (DTI) studies also showed injury and developmental disruptions of the periventricular WMI in preterm infants related to PHVD [48-50]. (See 'White matter injury' below and 'Outcome' below.)

Diagnosis — The diagnosis is made by serial cranial ultrasounds that demonstrate increasing ventricular dilatation. In our center, ultrasounds are performed at least twice weekly for infants who have severe IVH (eg, grade III and PVHI). (See 'Early intervention' below.)

Serial measurements of the following indices are recorded and plotted to determine if there is progressive dilatation, the rate of change, and if and when intervention should be performed (image 9) [51-55].

●The ventricular index (VI) is defined as the distance between the falx (double-fold of dura mater that descends through the interhemispheric fissure in the midline of the brain to separate the cerebral hemispheres) and the lateral wall of the anterior horn based on imaging through the coronal plane at the level of the foramen of Monro. It is the main determinant for intervention for infants with PHVD.

●Anterior horn width (AHW) is defined as the diagonal width of the anterior horn measured at its widest point in the coronal plane at the level of the foramen of Monro.

●Thalamo-occipital distance (TOD) is defined as the distance between the outermost point of the thalamus at its junction with the choroid plexus and the outermost part of the occipital horn in the parasagittal plane.

●Frontal temporal horn ratio (FTHR) and frontal occipital horn ratio (FOHR) are obtained by measuring the widest distance of the frontal and occipital horns, respectively, and temporal horns and dividing the average of these measurements by twice the largest biparietal distance.

●Resistance index of cerebral artery blood flow is defined as the difference between peak systolic flow velocity and end diastolic flow velocity/peak systolic flow velocity.

Differential diagnosis — PHVD is distinguished from the main differential diagnosis of nonprogressive ventricular dilatation (also referred to as stable ventriculomegaly or ex-vacuo ventricular dilatation) by the progressive dilatation seen on serial cranial ultrasounds. Nonprogressive ventricular dilatation occurs in 25 percent of preterm infants with GMH-IVH and is caused by cerebral atrophy. Some degree of cerebral atrophy may also be present in infants with PHVD [41]. It is important to distinguish between the two conditions as the approach to management differs. In those with ex-vacuo dilatation, the ventricular shape may be irregular and dilatation is especially pronounced posteriorly. The subarachnoid space and interhemispheric fissure are widened as well.

No effective preventative measures — Based on the assumption that PHVD is primarily due to inflammatory response of the subarachnoid villi to the presence of blood, early LP and fibrinolytic agents were suggested as preventive measures by removing inflammatory blood products from the CSF. However, these interventions have not been shown to be effective in preventing PHVD and as a result, they are not recommended.

●Early LP or ventricular taps – A meta-analysis reported no difference in the outcomes of shunt placement, death, disability, and multiple disabilities between repeated LPs or ventricular taps and supportive measures alone in neonates at risk for developing PHVD [56]. Ventricular taps should be avoided, as these will result in multiple needle tracks within the brain.

●Fibrinolytic agents – Injection of fibrinolytic agents into the ventricular system of infants with IVH has been suggested as an intervention to prevent PHVD. However, there are mixed results in regards to overall benefit, and there appear to be significant adverse events associated with fibrinolytic therapy [57-59]. As a result, fibrinolytic therapy for PHVD is not recommended for routine administration and should continue to be viewed as experimental.

●Combination of drainage and fibrinolytic agents ‒ In a small observational study, external ventricular drainage was combined with daily urokinase intraventricular injections in 21 of 43 infants [60]. There were no reported serious complications but it remains unclear whether the combination therapy was more effective than drainage alone.

Management

Early intervention — Our approach to management for PVHD is based on early detection of asymptomatic cases identified by findings of dilatation on serial ultrasound examinations (algorithm 1) [35]. We suggest initiating treatment if ultrasound measurements of lateral ventricles demonstrate signs of progressive dilatation rather than waiting until clinical signs are evident.

We recommend that all infants with severe IVH (grades III and PVHI) are monitored twice a week with cranial ultrasound for evidence of ventricular dilatation [35].

●For infants who are four weeks after the diagnosis of severe IVH and have no signs of dilatation on serial ultrasounds, no further treatment or monitoring is required as they have not progressed to PHVD.

●If ventricular dilatation is observed, signs of progressive dilatation are monitored by cranial ultrasound performed twice a week with serial measurements of ventricular index (VI), anterior horn width (AHW), thalamo-occipital distance (TOD), and resistance index (RI) of cerebral artery blood flow velocity. Tools for plotting ventricular measurements are available based on postmenstrual age 24 to 42 weeks and for postmenstrual age between 24 and 29 weeks, and used to categorize infants for PVHD risk [35,53].

●Infants with significant and persistent lateral ventricular dilatation with a VI >4 mm above the 97^th percentile and AHW >10 mm or TOD >25 mm (high risk) for postmenstrual age (PMA) are categorized as having ventricular dilation of severe risk [35,51,52,61,62].

•Serial LPs to drain CSF are initially performed to maintain a VI <4 mm above the 97^th percentile for PMA. A cranial ultrasound is performed the next day, and the decision to perform a subsequent LP is based on the findings of persistent dilatation on the follow-up study. LPs are performed not more often than once a day and restricted to 10 mL/kg per LP. If there is no further dilatation, monitoring with cranial ultrasound is continued twice a week up to 4 weeks. If there is no sign of progressive dilatation during this time period, no further treatment or monitoring is required.

•Temporary ventricular access device (VAD) is placed when more than three to five serial LPs are performed and VI and AHW remains above normal values or if LPs failed to decrease VI or if LPs are not well tolerated. Over the course of 7 to 10 days following VAD placement, drainage is performed once or twice a day with an initial aliquot of 10 mL/kg, with a goal of keeping the VI well below 4 mm above the <97^th percentile for PMA. The frequency and volume of the drainage is modified based on daily ultrasound measurements.

•A permanent ventriculoperitoneal shunt (VPS) is placed in the following clinical settings when the infant's weight is greater than 2 kg, the CSF protein is below 1.5 g/L, the erythrocyte count is less than 100/mm³, and:

-Dilatation continues to progress despite drainage using VAD

or

-Persistent need for VAD drainage after 4 weeks

●Infants with a VI between the 97^th percentile and 4 mm above the 97^th percentile are categorized with ventricular dilatation of moderate risk, and are monitored with daily physical examination and cranial ultrasounds performed twice weekly:

•If dilatation stops or stabilizes, no further treatment or monitoring is necessary.

•If there is progressive dilatation, intervention is started with initial serial LPs followed by drainage using VAD and, if needed, permanent shunting.

Support for this approach is based on the following:

●A multicenter observational study of 127 preterm infants with PVHI demonstrated 18- to 24-month neurodevelopmental outcome was better for survivors receiving early intervention (n = 78) identified with progressive dilatation on screening ultrasound compared with those (n = 78) with late intervention initiated only when there were clinical signs of rapid increase in head circumference and raised ICP (n = 49) [63]. In the early intervention group, initial intervention mostly consisted of temporizing measures, with only 20 percent of infants subsequently needing a VPS, while in the late intervention group, VPS was the most common initial intervention used. Most of the survivors in the early intervention group (55 of 62 evaluated infants) had cognitive and motor scores within the normal range, whereas one-half of the survivors in the late intervention group (14 of 27 infants) had moderate to severe NDI.

●Although a multicenter clinical trial, DRIFT (ie, drainage, irrigation, and fibrinolytic therapy). was stopped early because of excess secondary hemorrhage in the intervention group (35 percent), a follow-up study reported the intervention group (DRIFT) versus the control group (ventricular taps) were likely to survive without severe cognitive disability (66 versus 35 percent) at 10 year follow-up [58,59]. These results provide support for the hypothesis that washing out harmful substances from the ventricles reduces overall brain injury after PHVD, however, DRIFT is a complicated procedure requiring highly skilled clinical monitoring and intervention over 72 hours.

●Multicenter clinical trial of 126 preterm infants that compared timing of intervention (serial lumbar punctures to drain CSF) based on moderate versus high risk criteria established by cranial ultrasound measurements of VI and AHW [64]. The composite adverse outcome (defined as death or cerebral palsy or Bayley composite cognitive/motor scores <-2 SDs at 24 months corrected age) was less often seen with earlier intervention (35 versus 51 percent). In addition, MRI at term equivalent showed less brain injury (based on the Kidokoro Global Brain Abnormality Score) and smaller ventricular volumes for the group treated at a lower threshold [65].

●Indirect evidence

•In a meta-analysis that included 2533 infants with PHVD who underwent a temporizing neurosurgical procedure, a meta-regression analysis showed older age for TNS was associated with placement of a VPS for permanent drainage and moderate to severe neurodevelopmental impairment [66]. TNS included LPs, ventricular tap, VAD, external ventricular drain, ventricular subgaleal shunt, and/or neuroendoscopic lavage. Limitation of this analysis is due to the significant heterogeneity of the included studies for all outcomes due to differences in inclusion criteria, definition of PHVD and PHH, and threshold used for treatment (clinical signs of ICP, ultrasound findings of ventricular dilatation).

•Although a multicenter clinical trial DRIFT, (ie, drainage, irrigation, and fibrinolytic therapy) was stopped early because of excess secondary hemorrhage in the intervention group (35 percent), a follow-up study reported the intervention group (DRIFT) versus the control group (ventricular taps) were likely to survive without severe cognitive disability (66 versus 35 percent) at 10 year follow-up [58,59]. These results provide support for the hypothesis that washing out harmful substances from the ventricles reduces overall brain injury after PHVD; however, DRIFT is a complicated procedure requiring highly skilled clinical monitoring and intervention over 72 hours.

Management of symptomatic newborns — In our center, for the rare infant with PHVD who presents with clinical signs or symptoms of increased ICP (eg, bulging fontanelle, splayed cranial sutures, increasing head circumference by >2 cm/week, apneas, feeding problems), we suggest initiating central spinal fluid (CSF) drainage. Our preferred intervention is insertion and tapping from a temporary VADs rather than serial lumbar punctures. CSF drainage may be increased to 15 mL/kg day, but care should be taken not to reduce the VI too rapidly because of rapid hemodynamic changes and risk of rebleeding [67].

Interventions for CSF drainage — Management of progressive PHVD include the following interventions for cerebral spinal fluid (CSF) drainage [3,68]:

●Serial LPs to drain cerebral spinal fluid (CSF) – Available data suggest that serial LPs do not reduce the need for subsequent shunting and do not delay the progression of hydrocephalus [56,68,69]. However, serial LPs may alleviate increasing intracranial pressure (ICP).

●Ventricular access devices (VADs) are often used as a temporizing means to drain CSF to avoid or delay the need for a permanent shunt [70-72]. The rationale for deferring permanent ventricular shunt placement is to avoid VPS obstruction, which can occur in the acute setting immediately after a severe IVH due to high levels of protein in CSF. In addition, older infants are usually better surgical candidates and sometimes ventricular dilatation will stop progressing such that placement of a permanent VPS can be avoided.

Temporizing drainage procedures include:

•Tunneled subcutaneous ventricular drain to a subcutaneous reservoir, which can be tapped for CSF removal (usually through the ventricular reservoir), which is the procedure used in our center. In variations of this system, the tunneled catheter drains into a surgically prepared pouch in the supraclavicular region or subgaleal space [73-75]. A subgaleal shunt may function as a continuous drain and not require tapping unless drainage is inadequate or shunt obstructs distally. Subgaleal shunts, as compared with ventricular reservoirs, reduce the need for daily CSF aspiration [69] and prolong the time period before permanent shunt placement [76]. There appears to be no difference in eventual need for permanent shunting when comparing ventricular reservoir patients and subgaleal shunt patients [72,77].

•Tunneled subcutaneous ventricular drain to an external drip chamber.

However, there are no definitive data to show which form of temporizing drainage of CSF in PHVD is best [78]. A meta-analysis comparing the different approaches showed no significant differences in rates of obstruction, infection, death, or neurodevelopmental outcome [79].

●Neuroendoscopic lavage (NEL) irrigates turbid intraventricular fluid aspirate remnants of hematomas. Irrigation is continued until anatomic landmarks are visualized. A repeat NEL may be needed and around 40 percent of the infants required a ventriculoperitoneal shunt [80].

●Permanent ventricular shunt for continuous CSF drainage cannot be performed if there is excessive blood in the CSF, because blood may block the shunt. The most common shunt placed in preterm infants is VPS. VPS, considered to be the definitive treatment for PHVD, can be associated with significant morbidity, especially in extremely preterm and low birth weight infants. Complications with shunts include infection and shunt malfunction. There are data that suggest that repeated revisions are associated with reduced cognitive function [81]. Endoscopic third ventriculostomy, with or without choroid plexus cauterization, may be an effective treatment alternative to a VPS in some cases. These procedures are discussed in greater detail separately. (See "Hydrocephalus in children: Management and prognosis", section on 'CSF diversion procedures'.)

●Although medications to reduce CSF production, such as acetazolamide and furosemide, have been used, there is no evidence that these medications are either effective or safe in preterm infants with PHVD. In particular, available data suggest that these agents do not decrease the need for shunting or decrease mortality [68,69]. Also, a meta-analysis of diuretic intervention studies showed poorer outcome, including increased risk of motor impairment and nephrocalcinosis, in treated infants [82].

Research efforts — Ongoing research is focused on determining whether noninvasive neurophysiological assessment, such as flash visual evoked potentials and amplitude integrated electroencephalography, near-infrared spectroscopy (NIRS), and possible biomarkers of IVH and PHVD (eg, activin or S100B, chemokines and cytokines, amyloid precursor protein [APP] in the CSF) may be useful measures in the future to improve the timing for intervention in the management of PHVD [83-87].

White matter injury — Infants with severe IVH (grade III and PVHI/grade IV) are also at risk for having WMI. This can initially be recognized with cranial ultrasound as increased echogenicity in the periventricular white matter, or on MRI as either increased signal intensity on a T2-weighted sequence or as punctate lesions in the white matter which can be hemorrhagic or more often ischemic in nature. On a repeat MRI in infancy, sequelae in the white matter can be seen as abnormal signal in the periventricular white matter suggestive of gliosis.

Cystic periventricular leukomalacia (c-PVL) characterized by periventricular focal necrosis with subsequent cystic formation has become a less common manifestation of WMI [88,89]. The observed decline may be a reflection of over-diagnosis of c-PVL in earlier studies, as cases of PVHI that remain separate from the lateral ventricles and evolve into multiple cysts may have been misclassified as c-PVL. However, it is more likely to be due to a real decline in c-PVL [88]. c-PVL presents two to three weeks after injury with a typical distribution dorsolateral to the external angles of the lateral ventricles and involves the region adjacent to the trigones and to the frontal horn and body of the lateral ventricles. There are data to suggest that GMH-IVH may exacerbate c-PVL, due to the presence of non-protein-bound iron in the CSF [46].

The course of c-PVL varies as demonstrated by an observational study of 7063 infants from the National Institute of Child Health and Human Development Neonatal Research Network that reported development of c-PVL in 433 infants (6.1 percent) [90]. Of the affected infants, c-PVL was diagnosed based on ultrasound findings in 40 percent within the first 28 days of age, and c-PVL disappeared prior to 36 weeks postmenstrual age (PMA) in one-half of these preterm infants and persisted to 36 weeks PMA in the other half. In the remaining 60 percent of the infants with c-PVL, cysts were first observed at 36 weeks PMA. c-PVL was associated with subsequent development of cerebral palsy (CP), intellectual impairment, and cerebral visual disturbances and it appears that the risk of adverse outcome was similar for those with cysts at 36 weeks PMA and those in whom cysts had disappeared.

OUTCOME

Mortality and short-term morbidity — Mortality increases with the severity of GMH-IVH.

●In a study based on information from two large ultrasound databases of preterm infants in the United States with GMH-IVH, mortality rates increased with increasing severity of GMH-IVH, with rates of 4, 10, 18 and 40 percent for grades I, II, III, and periventricular hemorrhagic infarction (PVHI, grade IV), respectively [44].

●In a study from the Netherlands in neonates with severe GMH-IVH (grade III and PVHI), the mortality was almost 30 percent for neonates with a grade III and 40 percent for those with a PVHI [91]. Two-thirds of the survivors developed posthemorrhagic ventricular dilatation (PHVD).

●In a study combining data from the Netherlands and Canada, the mortality was 40 percent for preterm infants with PVHI. One-third of the survivors developed posthemorrhagic ventricular dilatation (PHVD) [36].

Severe IVH also increases the risk of associated white matter injury [92]. (See 'White matter injury' above.)

Long-term outcome — The long-term outcome of infants who survive with GMH-IVH worsens with increasing severity of GMH-IVH, decreasing gestational age (GA), and the requirement of a permanent shunt for progressive PHVD. However, it remains unclear if infants with low-grade GMH-IVH have long-term sequelae, as data are inconsistent.

Severity of GMH-IVH — The following studies demonstrate that the risk of neurologic impairment (cerebral palsy [CP] and neurodevelopmental impairment [NDI]) increases with GMH-IVH severity (table 4):

●In a follow-up study of 1812 infants (GA <33 weeks) born in 1997, CP was diagnosed at five years of age in 8, 11,19, and 50 percent of survivors with grades I, II, III and PVHI, respectively [93].

●Outcome at 2 years corrected age was assessed with Bayley Scales of Infant and Toddler Development, 3^rd Edition (BSID-III) for 883 infants enrolled in the Preterm Erythropoietin Neuroprotection Trial. The risk of CP increased with GMH-IVH grade: 6, 16, 33, and 64 percent for those with Grade I, bilateral Grade II, Grade III, and PVHI, respectively [94].

●In a regional study of 1472 infants (GA <28 weeks) born between 1998 and 2004, neurodevelopmental delay was diagnosed in 3.4, 7.8, and 17.5 percent of preterm infants with no GMH-IVH, grades I to II, and grade III and PVHI at two to three years corrected age, respectively [95]. In addition, the risk of CP (6.5, 10.4, and 30 percent), deafness (2.3, 6, and 8.6 percent), and blindness (0, 0, 2 percent) varied for infants with no GMH-IVH, grades I to II, and grade III and PVHI, respectively. Notably, even survivors with low-grade GMH-IVH were more likely to exhibit neurodevelopmental delay, CP, and deafness when compared with children with no GMH-IVH.

●In a follow-up study of 499 extremely preterm (EPT) infants (GA <28 weeks) infants, CP was diagnosed at 8 years of age in 8, 15, 18, 26, and 75 percent of infants with no GMH-IVH, grades I, II, III, and PVHI, respectively [96]. It should, however, be noted that c-PVL was found in 6 and 4 percent of preterm born children with grades I and II IVH, respectively, and 13 and 25 percent of preterm born children with grades III and PVHI, respectively. Children who had more severe grades of IVH (grades III and PVHI) had an increased risk of impaired intellectual ability and academic skills compared with other children but not those with grades I and II.

Although data including large population studies suggest that preterm infants with low-grade (grade I and II) GMH-IVH are not at increased risk for long-term NDI compared with controls without GMH-IVH [97-99], there are other studies that indicate that this population may have more long-term morbidity compared with survivors without GMH-IVH.

●Studies that found similar NDI for survivors with mild GMH-IVH compared with controls without GMH-IVH include:

•A study based on data from the Infant Health and Development Program (IHDP), a multisite trial of education intervention for low birth weight infants, found no difference in intellectual function, academic achievement, or behavior between children with evidence of mild GMH-IVH and those without GMH-IVH at age 3, 8, and 18 years [97].

•A report from the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network report of extremely low birth weight (ELBW) infants (BW <1000 g) born between 2006 and 2008 showed similar neurodevelopmental outcomes between infants with grade I and II GMH-IVH and those without GMH-IVH at 18 and 22 months postmenstrual age (PMA) [98]. After adjusting for confounding variables, there were no differences in the rate of CP, gross motor limitations, cognitive and language scores on the Bayley Infant Development Indexes (3^rd edition), or composite measures of neurologic impairment.

●In contrast, the following studies suggest that survivors with low-grade GHM-IVH are at increased risk for NDI:

•The Multicenter Randomised Indomethacin Intraventricular Haemorrhage Prevention Trial, a small study of 44 ex-preterm adolescents with isolated grade II IVH, found that individuals with low-grade GMH-IVH were more likely to have cognitive and executive function impairment compared with preterm survivors without GMH-IVH and term controls [100].

•A study performing MRI using tract-based spatial statistics (TBSS) in 24 infants with low-grade IVH found microstructural impairment in the periventricular white matter. These microstructural white matter changes were associated with poorer outcome at 24 months [101].

Permanent shunt placement — Permanent shunt placement is also a predictor for poor outcome as noted by the following studies:

●A study from the NICHD Neonatal Research Network of infants born between 1993 and 2002 showed a negative effect of shunt placement on neurodevelopmental outcome [102].

•The mean Bayley Mental Developmental Indexes (MDI) (2^nd edition) at 18 to 22 months PMA were 74, 66, 71.5, and 60 for grade III without shunt, grade III with shunt, PVHI without shunt, and PVHI with shunt, respectively.

•The mean Bayley Psychomotor Developmental Indexes (PDI) were 77, 64, 73, and 55 for grade III without shunt, grade III with shunt, PVHI without shunt, and PVHI with shunt, respectively.

•The risk for CP was 23, 57, 37, and 80 percent for grade III without shunt, grade III with shunt, PVHI without shunt, and PVHI with shunt, respectively.

●In an analysis from the NICHD Neonatal Research Network of 353 preterm infants born with a GA <27 weeks and severe IVH who survived to 36 weeks PMA, shunt placement was the most significant factor associated with either late death or neurodevelopmental impairment (NDI), while bilateral GMH-IVH and PVHI had a comparatively smaller impact on outcome. One-third of the cohort (n = 117) survived without any NDI.

However, in the previously mentioned observational study, timing of intervention for PHVD appeared to be more important for long-term neurodevelopmental outcome than whether or not a shunt was placed [63]. In this cohort, the majority of survivors in the early intervention group had no evidence of NDI, and there was no difference in cognitive and motor testing scores between patients managed only with a temporizing procedure and those having a subsequent VPS placement. In contrast, only 2 of the 16 infants in the late intervention group needing intervention, mostly insertion of a permanent shunt, had a normal outcome. Eleven infants in the late intervention group did not have any intervention, and 8 (73 percent) of these 11 infants had a normal outcome. Of interest, ventricular size in this group of infants was comparable to all infants in the early intervention group.

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Intraventricular hemorrhage in newborns (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Introduction ‒ Although germinal matrix hemorrhage and intraventricular hemorrhage (GMH-IVH) is an important cause of brain injury in preterm infants, currently there is no specific therapy to limit the extent of hemorrhage or injury after it has occurred or to prevent the complications posthemorrhagic ventricular dilatation (PHVD) and white matter injury (WMI). Management is focused on prevention of GMH-IVH. (See 'Management' above.)

●Prevention ‒ The most effective strategy to prevent GMH-IVH is to reduce the risk of preterm birth. When preterm birth cannot be avoided, appropriate prenatal and delivery room care should be provided to the mother and neonate. This includes antenatal glucocorticoid administration, delayed cord clamping, prompt resuscitative efforts including appropriate fluid therapy and respiratory support as needed, correction of metabolic acidosis and coagulopathy, and efforts to minimize handling and provide optimal positioning of the infant, as described in separate topics. (See 'Our approach' above and "Preterm birth: Risk factors, interventions for risk reduction, and maternal prognosis" and "Labor and delivery: Management of the normal first stage" and "Neonatal resuscitation in the delivery room".)

●Management ‒ In neonates with GMH-IVH, management is supportive and focused on reducing further brain injury through preservation of cerebral perfusion and oxygenation by maintaining adequate mean arterial perfusion, oxygenation and ventilation, and providing appropriate fluid, metabolic, and nutritional support. In addition, seizures are treated in a timely manner to avoid hypoxia or hypotension. Ongoing surveillance that includes twice weekly cranial ultrasounds are used for early detection of posthemorrhagic ventricular dilatation (PHVD). (See 'Management' above.)

●Complications ‒ Complications of GMH-IVH include PHVD and white matter injury (WMI). (See 'Complications' above.)

●Posthemorrhagic ventricular dilatation ‒ PHVD, the most common complication, occurs in approximately 25 percent of infants with GMH-IVH. The incidence of PHVD increases with the severity of IVH (table 1). If untreated, PHVD may lead to progressive ventricular dilatation and increased intracerebral pressure (ICP), which increases mortality and long-term neurodevelopmental impairment. Our management approach is focused on early identification and intervention to avoid ICP and its potential adverse effects (algorithm 1). (See 'Management' above.)

•For infants with severe IVH (grades III and IV), we suggest ongoing monitoring with cranial ultrasound twice weekly to detect early signs of dilatation prior to the onset of symptoms (Grade 2C).

•For most infants with PHVD, we suggest intervening based on ultrasound evidence of progressive dilatation rather than waiting until clinical signs are evident (Grade 2C). We typically intervene if the ventricular index (VI) is >4 mm above the 97th percentile for postmenstrual age (PMA). Intervention consists of temporizing procedures such as serial lumbar punctures (LPs) or ventricular access device (VAD) to remove cerebral spinal fluid (CSF) (See 'Early intervention' above.)

•For infants with signs of increased ICP, we suggest neurosurgical intervention to remove CSF and decrease ICP (Grade 2C). (See 'Management of symptomatic newborns' above.)

•There are no known measures (LPs, ventricular taps, or fibrinolytic agents) that prevent PHVD. (See 'No effective preventative measures' above.)

●Outcome ‒ Increased severity of GMH-IVH (table 1) is associated with increased mortality and morbidity. (See 'Outcome' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Lisa M Adcock, MD, who contributed to an earlier version of this topic review.