Final Protocol Approved

PROTOCOL

Comparison of Clonidine versus Phenobarbital as an Adjunct Therapy for Neonatal Abstinence Syndrome.

Hypothesis:

Infants being treated for neonatal abstinence syndrome with a combination of Clonidine and neonatal morphine sulfate will have a shorter length of treatment in comparison to a combination of phenobarbital and neonatal morphine sulfate.

Background and Aims:

Neonatal abstinence syndrome (NAS) is a symptom complex experienced by 55 to 94% of neonates who are exposed to intrauterine opioids (1-6). Clinical features of NAS involve multiple systems with predominance of neurological symptoms which include hyperexcitability, insomnia, irritability, hypertonia, hyperreflexia, tremors, and seizures. Other features include enteric symptoms (vomiting, diarrhea, feeding disturbances), and sympathetic and parasympathetic dysregulation including sweating, fever, tachypnea, and congestion. (7)

A greater awareness of maternal opioid addiction (heroin and cocaine) has lead to substitute therapies with methadone and buprenorphine (Subutex) ( 8-14). Currently NAS is seen secondary to maternal polydrug usage and hence the pharmacologic therapy for NAS needs greater attention. Pharmacotherapy in the neonatal period is based upon the Finnegan scoring system and usually initiated when consecutive scores are ≥ 8. Recent studies have shown that combination therapies are superior to monotherapy with neonatal morphine sulfate (NMS). Phenobarbital, a barbiturate and anticonvulsant/sedative, has been shown to reduce the length of hospitalization, decrease severity of withdrawal, as well as decrease hospital costs and care giver demands (15-17). Adjunct therapy with clonidine, an α2-adrenergic receptor agonist, is safe, effective and reduces length of treatment when compared to treatment with opioid and placebo (18-22).

Phenobarbital is an antiepileptic with action on GABA(A) receptors and has been shown in animal models to inhibit neuronal cell proliferation, survival, and neurogenesis (23). In human infants long term treatment with Phenobarbital may result in neuro-developmental compromise (24) Due to these potential harmful effects of phenobarbital alternative therapies should be explored more thoroughly including clonidine.

Specific Aims: To examine whether the duration of neonatal morphine sulfate (NMS) treatment is reduced when accompanied by clonidine (NMS_C, experimental group) vs Phenobarbital (NMS-P, standard). Secondary outcomes include length of stay for NAS, treatment failures, and adverse events (bradycardia, hypo- and hypertension).

We hypothesize that clonidine when compared to phenobarbital as an adjunct therapy, will have shorter length of stay, with fewer treatment failures and side effects.

Methods:

This study will be a prospective, randomized, non-blinded clinical trial of neonatal morphine sulfate and clonidine (NMS/C) versus NMS and phenobarbital (NMS/P). Infants will be recruited from the Baystate Children’s Hospital Neonatal Intensive Care Unit (NICU) and Neonatal Continuing Care Nursery (NCCN) after IRB approval has been obtained for the study. A written informed consent will be obtained from the parents. Since Department of Children and Family Services (DCF) is always notified secondary to maternal drug usage, if DCF has custody of the infant then the consent will be obtained from the DCF case worker assigned to the case.

Infants meeting the following criteria will be eligible for this study:

Inclusion criteria:

  • 0 to 15 days of age
  • Prenatal exposure to opioids with development of moderate to severe NAS (2 consecutive abstinence scores of ≥ 8)
  • Medically stable

Exclusion criteria:

  • Gestational age < 35 weeks
  • Infants withdrawing from maternal use of benzodiazepines
  • Intrauterine growth retardation (birth weight below the 5th percentile)
  • Congenital heart disease
  • Congenital anomalies
  • Medically unstable

Study Groups:

Two groups will be compared in this trial. The experimental therapeutic approach – NMS with clonidine—will be compared to the standard therapeutic approach—NMS with phenobarbital.

a)New treatment group (NMS-C):

Initiation of therapy for patients randomized to the NMS-C group will follow the schedule below:

Finnegan Score / NMS
mg/kg/day divided q3hr / Clonidine
mcg/kg/day divided q6hr
8-10 / 0.32 / 6
11-13 / 0.48 / 8
14-16 / 0.64 / 10
≥ 17 / 0.80 / 12

* If needing NMS > 0.80 mg/k/day then increase dose in increments of 0.16 mg/kg/day till Finnegan score <8

+ Clonidine escalation may be limited by hypotension or bradycardia

b)Standard treatment group (NMS-P):

Treatment management of patients receiving standard therapy is described below.

Finnegan Score / NMS
mg/kg/day divided q3hr / Phenobarbital
mg/kg/day divided q8hr
8-10 / 0.32 / 6
11-13 / 0.48 / 8
14-16 / 0.64 / 10
≥ 17 / 0.80 / 12

* If needing NMS > 0.80 mg/k/day then increase dose in increments of 0.16 mg/kg/day till Finnegan score <8

Clinical Care of Study Participants

Once enrolled in the study infants will have their daily care as per the pre-existing NICU and NCCN protocols for vital signs and other clinical parameters required for daily care. Infants in the Clonidine group will be initially admitted to the NICU for 24hrs to ensure stable blood pressures before transitioning to the NCCN. Since blood pressure monitoring is not a routine in the NCCN, infants in the study will have their blood pressures monitored every 6hrs for the first 48hrs after starting and stopping the protocol, in between they will get twice daily blood pressure monitoring at 12h intervals, in addition to the other routine standard vital signs.

In addition, all infants will have Finnegan scores documented q3h. Since enrollment in the study will mean that the infant is undergoing NAS, based on their randomization , the infants will be started on NMS and additional therapy with clonidine 6 -12 mcg/kg/day divided every 6 hours or Phenobarbital 6 -12 mg/kg/day divided every 8 hours, with starting dosage based on initial abstinence score.

The NMS solution used will be 0.4 mg/ml as per the current protocol used in the NICU based on the Finnegan scores to guide escalation and de-escalation of morphine.

The Clonidine solution used will be 10 mcg/ml and the protocol used will be based on Finnegan scores and studies published by Holder et al and Agthe et al. If after initiation of Clonidine, bradycardia and/or hypotension are noted the dose will be decreased to the next lower dose until the vital signs stabilize.

Study withdrawal criteria: Infants with persistent side effects of bardycardia and/or hypotension not responding to dose adjustments will be eligible for withdrawal from the protocol. Most adverse effects are mild and tend to diminish with continued therapy. The most frequent side effects reported in pediatric and adult population (which appear to be dose-related) are dry mouth, occurring in about 40 of 100 patients; drowsiness, about 33 in 100; dizziness, about 16 in 100; constipation and sedation, each about 10 in 100. The cardiovascular side effects are bradycardia and hypotension are usually mild responding to dose adjustments. Severe side effects such as AV block are rare (i.e.1 in 100) and seen when subjects received concomitant therapies with Clonidine such as beta blockers or methylphenidate.

Weaning Protocol. Infants will be maintained on original pharmacotherapy for 48 hours following initiation. Each infant will have morphine sulfate reduced by 10% of absolute dose if abstinence scores remain ≤ 8 for 24 hours. Each infant will be eligible to be weaned in this step-wise fashion once in a 24 hour period. If an infant develops recurrence of withdrawal symptoms with abstinence scores > 8 x 2 or more over a 24 hour period following a wean will be on hold, if scores significantly higher persistently (>10 x 2) then the morphine dosage will be escalated to the last effective dose.

Infants will be continued on clonidine at the last effective dose until the morphine sulfate is weaned off in order to prevent recurrence of symptoms of NAS. After morphine wean has been completed, clonidine will be weaned by 50% in a two step reduction (every 24 hours) until off. The stepwise wean of clonidine will be undertaken in order to prevent the occurrence of rebound hypertension and sympathetic overactivity; two adverse effects that have been observed with abrupt discontinuation of clonidine (18)

Starting Clonidine Daily Dose / First clonidine weaning dose / Second clonidine weaning dose / Third weaning step
6 mcg/kg/day / 3 mcg/kg/day x 24 hours / 1.5 mcg/kg/dose x 24 hours / Discontinue clonidine
8 mcg/kg/day / 4 mcg/kg/day x 24 hours / 2 mcg/kg/day x 24 hours / Discontinue clonidine
10 mcg/kg/day / 5 mcg/kg/day x 24 hours / 2.5 mcg/kg/day x 24 hours / Discontinue clonidine
12 mcg/kg/day / 6 mcg/kg/day x 24 hours / 3 mcg/kg/day x 24 hours / Discontinue clonidine

Phenobarbital is not weaned during the hospital stay and infants are discharged home on it per current unit protocol.

Infants will be observed for 36-48 hours off Clonidine prior to discharge home.

Randomization Scheme

Infants who meet the inclusion criteria, after informed consent from the parents or guardians will be randomly assigned to one of the study groups- either NMS-C or NMS-P. Patients will be randomized to groups under equal allocation according to a computer generated randomization procedure. Patients will be randomized in blocks sizes of 4 and 6. Block randomization will insure that study groups will remain balanced at the completion of each block and that unbalanced study groups is minimized if the study is terminated during a block interval.

To insure that study groups are balanced on important patient characteristics, randomization will be stratified by type of maternal drug use: Methadone/ buprenorphine (Subutex) versus Polydrug use.

Primary Outcome Measures

The primary outcome will be length of treatment with NMS in the two study arms. It will be defined as the total number of days the infants were on NMS before successfully weaning off of it completely. Phenobarbital is not weaned during the hospital stay and infants are discharged home on it hence time to wean off of NMS is a more appropriate measure of treatment efficacy.

Secondary Outcome Measures

Our secondary outcomes will be

a)Total length of hospital stay for NAS treatment. It will be defined as total number of days the infant was in the hospital while on pharmacotherpeutic agents for NAS treatment. Occasionally infants will stay hospitalized for non-medical reasons increasing the hospital stay.

b)The mean total treatment dose and mean daily dose of NMS required in the two

study arms will be recorded

c)Total number of treatment failures: Will be defined as the number of times that the infant failed a wean and had to be escalated to the previous dose

d)Adverse effects:

1)Bradycardia will be defined as sustained HR <80 beats/minute

2)Hypotension will be defined as mean blood pressure <2 SD for the infants gestational age (Mean BP = Mean GA)

3)Hypertension will be defined as mean blood pressure > 2 SD for the infants gestational age (Mean BP = Mean GA)

Data Collection:

Maternal data would include detailed drug history during the pregnancy (drug and dosages), urine toxicology, concomitant alcohol and smoking history.

Infant data will include gestational age, birth weight, Apgar score at 5 minutes, baseline heart rate and blood pressure at admission, then every 6h for 48h followed by every 12 hours after initiation of the protocol, and 24 and 48hours after discontinuation of therapy, type of feedings (formula versus breast milk). If formula is the preferred by mothers than all infants in the study will be on a soy formula. All infants will have daily vital signs monitoring as per the current NICU/NCCN policy. In addition infants in the NCCN will have blood pressures recorded daily at 12h interval.

Specific data will also be collected on each infant including:

  • Amount of NMS necessary to control symptoms
  • Duration of NMS therapy
  • Duration of hospitalization for NAS treatment
  • Treatment failure

Sample Size Estimation

Sample size and statistical power were estimated using PASS (J. Hintze, PASS ver. 8.0.7, 2008, Kaysville, UT). The current mean duration of treatment for NAS at the Baystate Children’s hospital is 25.8 days. A 20% reduction in the duration of treatment (i.e., approximately 5 days) is considered clinically meaningful. Assuming a common standard deviation of approximately 8.5 days (based on our clinical data), a sample size of 55 infants per group (total of 110 patients) would provide more than 80% power for a 2-tailed test of significance at a critical value of 5%.

In the event that survival analysis is used, a sample size of 55 infants per group would provide about 80% power to detect a 40% reduction in time-to-weaning from NMS, assuming a two-sided test of significance at a critical level of 5%. This hazard ratio is equivalent to a reduction in the median time-to-weaning from 26 days to 15 days. This analysis assumes that approximately 5% of infants will be of infants will be censored by the end of follow-up. Our clinical experience indicates that there will be little or no censoring.

Data analysis

Treatment and control groups will be compared on baseline characteristics to identify patient or clinical factors that may not have been adequately balanced by the randomization procedure. Continuous variables will be compared using 2-sample t-tests or its non-parametric equivalent (Wilcoxon rank-sum test), while categorical variables will be compared using Pearson chi-square or Fisher’s exact test. Differences on baseline characteristics achieving a p-value of .15 or less will be included in subsequent multivariable analysis. For analyses using the survival approach, bivariable analyses will be based on the Kaplan-Meier product-moment estimator. Statistical testing will used the log-rank test.

Multivariable analyses will be conducted using two approaches. First, if data on time-to-weaning are not skewed and there is no censoring, linear regression will be used. The primary outcome will be the length of treatment with NMS in the two study arms, defined as the mean number of days the infants were on NMS before successfully weaning off of NMS completely. Independent variables will include the dichotomous variable representing study group, as well as baseline characteristics requiring adjustment. If time data are highly skewed, the time data will be log transformed and linear procedures will be applied.

Secondly, we will use survival analytic approaches ( Cox proportional hazard regression) to compare the study groups, particularly in the event that there is censoring (e.g., due to transfers or withdrawals from the study). Proportional hazards regression will be used to control for baseline variables requiring adjustment despite randomization. Survival models will be developed and model assumptions will be verified using methods described by Hosmer and Lemeshow (1999).

All subjects will be analyzed on an intention-to-treat basis.

Early Termination of Study

To determine whether early termination of the trial is necessary due to larger than expect benefit or larger than expected harm, we plan one interim look at the data when 50% of the patients have completed follow-up. We will use the Obrien-Fleming (1979) spending function to adjust the alpha comparison at the interim analysis and the final analysis in order to maintain an overall critical level of 5% (overall p-value is 5%). Thus, the trial will be terminated early if at the interim analysis the absolute difference in the proportion of A-V block between the study groups achieves a significance level of p ≤ 0.003. For the final analysis at study completion, significance testing will be conducted at p ≤ 0.047. This adjustment ensures that the comparisons will be conducted at an overall significance level of 5%.

Data Storage and Confidentiality:

  1. Certificate of Confidentiality: All maternal and infant data will be recorded from the infant’s medical record and de-identified, hence the information collected will not pose any adverse consequences for subjects or damage their financial standing, employability, insurability, or reputation. A certificate of confidentiality is thus not needed.
  2. Each infant will be given a study number allocated to them, which will be linked to the infant’s medical record number. The key to the study number will be stored by the PI on her personal office computer. No one other than the PI will have access to the key, ensuring maintenance of confidentiality.
  3. The PI will be the only one with access to the key to the identity code. The de-identified infant data will be accessible to Dr. Brooke Surran, Dr. Bhavesh Shah, Dr. Paul Visintainer, Ms. Susan Chamberlain and Ms. Kathleen Kopcza.
  4. No data will be moved/shared outside of BaystateMedicalCenter. If presenting the analyzed data for research meetings or publication, then only the de-identified data will be presented.

Risks and benefits:

Multiple studies have shown clonidine used in the above mentioned doses to have relatively few side effects with the main one being hypotension at the initiation and rebound hypertension with abrupt weaning.

If at any time any study modification to therapy is not tolerated by the infant (bradycardia, seizures, hypoxia, hypo or hypertension for age and condition), the infant will be assessed and returned to previous dosages for stabilization.

This comparison study is potentially of great significance. If clonidine is proven to be equally effective in treatment of NAS many of the detrimental effects of phenobarbital therapy may be avoided for infants on long term pharmacotherapy for treatment of withdrawal with shorter length of hospital stay.

References:

1. Khalsa JH, Gfroerer J. Epidemiology and health consequences of drug abuse among pregnant women. Semin Perinatol. 1991;15:265–270

2. Dicker M, Leighton EA. Trends in diagnosed drug problems among newborns: United States, 1979–1987. Drug Alcohol Depend. 1991;28:151–165

3. Fricker HS, Segal S. Narcotic addiction, pregnancy, and the newborn. Am J Dis Child 1978;132(4):360–366.

4. Harper RG, Solish GI, Purow HM, Sang E, Panepinto WC. The effect of a methadone treatment program upon pregnant heroin addicts and their newborn infants. Pediatrics 1974;54(3):300–305.

5. Madden JD, Chappel JN, Zuspan F, Gumpel J, Mejia A, Davis R. Observation and treatment of neonatal narcotic withdrawal. Am J Obstet Gynecol 1977;127(2):199–201.

6. Ostrea EM, Chavez CJ, Strauss ME. A study of factors that influence the severity of neonatal narcotic withdrawal. J Pediatr 1976;88(4pt1):642–645.

7. Finnegan, L.; Connaughton, JJ.; Kron, R. A scoring system for evaluation and treatment of the neonatal abstinence syndrome: a new clinical and research tool. In: Marselli, P.; Garanttini, S.; Sereni,F., editors. Basic and Therapeutic Aspects of Perinatal Pharmacology. New York, NY: Raven; 1995.p. 139-152.

8. Doberczak TM, Kandall SR, Friedmann P. Relationships between maternal methadone dosage, maternal-neonatal methadone levels, and neonatal withdrawal. Obstet Gynecol. 1993;81:936–940