Phenobarbital-Induced Bullous Lesions in a Non-Comatose Patient

Abstract

The occurrence of dermatologic bullous lesions is well documented after barbiturate-induced intoxication. However, this rare clinicopathological entity is only reported in the setting of a coma or lethal toxicity. This case report presents a case of a noncomatose, nonimmobilized patient who developed systemic bullous lesions after an acute ingestion of oral phenobarbital. This case illustrates that bullous lesions can occur as a result of nontoxic phenobarbital doses, in the setting of subtherapeutic barbiturate blood levels. All previous case reports suggest phenobarbital produces either systemic bullae only after coma-inducing doses or a localized bullous reaction after a high-dose intravenous injection. This case challenges this previously held association. Although the exact mechanism of the bullous lesions remains unknown, the clinical course of the lesions is self-limiting and does not require additional care beyond the treatment of a superficial burn. Resolution of the bullae occured within two weeks. Since phenobarbital continues to be prescribed orally, physicians should recognize bullous lesions as a potential adverse dermatological reaction from non-lethal oral doses of phenobarbital.

Introduction

Bullous lesions have been well documented to occur after barbiturate-induced intoxication leading to a coma.^1,2 In fact, these lesions have been referred to as “coma blisters.”³ Coma blisters may occur as a result of various medications, but they more frequently appear in the setting of barbiturate-induced coma. It has been shown that these skin lesions, including barbiturate-induced bullae, can be distinguished by the specific histologic finding of sweat gland necrosis. They commonly occur in both pressure and non-pressure areas on the extremities and trunk.⁴ The bullae may form from one hour to several days post-drug ingestion.

This case report is of a non-comatose patient who developed systemic bullous lesions after an acute ingestion of oral phenobarbital over the course of a few days. The patient did not receive a toxic dose of the barbiturate and had a sub-therapeutic barbiturate blood level upon presentation. After an extensive review of the literature, to our knowledge there have been no reports of subtherapeutic phenobarbital blood levels inducing systemic bullous lesions in a non-comatose, nonimmobilized patient. All prior published case reports describe either intravenous injection, which reacted with bullae only near the injection site, or massive oral overdoses, after which the bullae were distributed over a variety of locations on the body.

Case Presentation

This patient is a 29-year-old Caucasian woman with an 11-year history of schizoaffective disorder of bipolar type, cannabis abuse, and borderline personality disorder. She had multiple prior psychiatric hospitalizations. Following a six-month inpatient admission at the state psychiatric hospital, she was discharged to a group home because of poor adherence to medications, persistent use of cannabis, and impulsive, sometimes aggressive, behavior. At the group home, she was allowed visitors, but she was under 24-hour supervision and was not allowed to go outside. She is currently managed by an intensive community wrap-around psychiatric service program. Her active medications included lorazepam (Ativan^®) 1mg t.i.d., ziprasidone (Geodon^®) 100mg b.i.d., and sertraline (Zoloft^®) 100mg q.d.

She initially presented to the emergency department for evaluation of multiple bullous lesions covering her hands bilaterally with a few lesions on her mid-back and right heel. When asked about her activities of the past week, the patient reported nothing out of the ordinary except for the ingestion of another resident's phenobarbital prescription medication. She claims to have orally ingested 2 to 3 pills of 100mg phenobarbital per day during the past three days. No other non-prescribed medications were ingested. The patient also admitted to using alcohol and cannabis recently, and she smokes one pack of cigarettes a day. She has a past substance abuse history, including crack cocaine, ecstasy, LSD, and methamphetamine. The development of the presenting bullous lesions occurred acutely during the third night of phenobarbital ingestion. The patient denied any previous history of skin lesions or rashes.

Upon presentation to the emergency department, 24 hours after her last phenobarbital dose, the patient was breathing spontaneously, but was quite lethargic and unable to follow simple commands. Vital signs in the emergency department were temperature 98.8 degrees Fahrenheit, heart rate 100/min, respiratory rate 18/min, blood pressure 121/75mmHg, and pulse oximetry 90 percent on room air; these normal vital signs suggested a lack of systemic involvement to explain the patient's presenting lethargy. Within an hour she became agitated, made violent verbal threats to the medical staff, and drew attention to herself by screaming and threatening to break her bullous lesions open against the emergency department's cubicle window. She was promptly given lorazepam 2mg intramuscularly and was admitted involuntarily to the inpatient psychiatric unit, where she required intermittent chemical restraint and was secluded once for safety due to her agitation.

Physical exam was unremarkable except for the external integument examination. It revealed several variously sized bullae located on both hands. All bullae were tense, fluid-filled, with minimal clear, nonodorous drainage, which were extremely tender to light touch. The right hand was remarkable for a large 6cm in diameter, intact lesion on the palmar surface (Figure 1). Along the dorsum of each finger, there were three or four blisters (Figure 2). None of the bullae were surrounded by erythema or induration. Only a few of these lesions were open and draining serous fluid. The underlying tissue was light pink in color without ulceration. Additionally, there were two bullae on her mid-back (Figure 3) and one bulla on the posterior surface of the right heel over the Achilles tendon (Figure 4). There were no ocular, oral, or other remarkable cutaneous findings. At no time was the patient short of breath or edematous in other tissues.

Figure 1.

Right hand, palmar surface, 6cm bulla in the center of the palm, Day 2 of hospitalization

Figure 2.

Right hand, dorsal surface, Day 2 of hospitalization

Figure 3.

Posterior thorax, bullae lesions opened by patient, Day 2 of hospitalization

Figure 4.

Right achilles tendon bulla, measuring 2cm in diameter, Day 2 of hospitalization

Initial laboratory results revealed a normal complete blood count (except for a mildly elevated WBC count at 14.8 K/µL with 85 percent neutrophils), complete metabolic panel was within normal limits (except for AST at 114IU/L), and normal urinalysis. Serum beta-HCG was negative. The urine drug screen was positive for cannabinoids, barbiturates, and benzodiazepines, and negative for amphetamines/methamphetamines, cocaine, opiates, phencyclidine hydrochloride, and propoxyphene. Follow-up GC/MS identification and quantification of the positive benzodiazepine level was not done. Blood alcohol levels were not obtained and are usually not obtained in the emergency department for patients who do not present with evidence of recent alcohol consumption, such as the odor of alcohol on the breath. Her quantitative Delta-9-tetrahydrocannabinol (THC) and THC/creatinine ratio values were 286ng/dL and 2577ng/g, respectively, indicating a substantial and recent THC ingestion. The presenting phenobarbital level was 6.1µg/mL (therapeutic range: 12.0–39.0µg/mL). An electrocardiogram showed normal sinus rhythm without abnormalities.

The patient was continued on lorazepam 1mg t.i.d., ziprasidone 100mg b.i.d., sertraline 100mg q.d., and acetaminophen/hydrocodone (Lortab^®) 500mg/5mg every six hours p.r.n. pain. Maintenance intravenous fluids were initiated, along with intravenous vancomyocin 1.5g q.d. and piperacillin/tazobactam (Zosyn^®) 3.375g q.i.d. The lesion dressings were changed twice daily with silver sulfadiazine topical cream (Silvadene^®). The patient remained agitated throughout the initial days of admission. On the third day, the patient refused all psychotropic medications. After two days of intravenous antibiotics, she was transitioned to oral cephalexin (Keflex^®) 500mg t.i.d. for two additional days, upon the patient's request. General surgery, plastic surgery, internal medicine, wound care, and dermatology were consulted for the bullous lesions. The dermatology consultation stated that a skin biopsy would most likely be nonspecific at the time of biopsy and endorsed the diagnosis of phenobarbital-induced bullous disorder. The bullae remained tense, clearly defined with a moderately thick skin wall. When the roof of a bulla was opened, pale yellow, fibrinous fluid freely flowed out, leaving a raw, erythematous base. The opened bulla had a similar appearance to a superficial second-degree burn (Figures 5 and 6). All the bullae slowly resolved without scarring over the next few weeks. No further lesions arose during the subsequent weeks of the hospitalization.

Figure 5.

Right hand, dorsal surface, Day 11 of hospitalization

Figure 6.

Right hand, palmar surface, Day 11 of hospitalization

Discussion

Phenobarbital belongs to the barbiturate drug class, formerly a more pervasive and popular prescription drug of abuse. Currently, barbiturates are mainly used intravenously as anesthetics or orally as anticonvulsants. They primarily act on the gamma-aminobutyric acid (GABA) channels in the central nervous system (CNS) to suppress the activity of excitable tissue.^5,6 Phenobarbital is a long-acting barbiturate, which has the least abuse potential and exerts its effect up to 12 hours post-ingestion. Common neuropsychiatric effects from excessive doses of phenobarbital, other than CNS depression, include acute cognitive impairment, usually causing delirium, irritability, combativeness, and paranoia.⁵ Dermatological findings are much less common. In fact, they occur approximately six percent of the time in barbiturate poisonings. But, 50 percent of patients with dermatological findings after lethal phenobarbital toxicities will manifest tense, clear bullous skin lesions of unclear etiology.⁶ Additionally, 50 percent of Steven-Johnson syndrome and toxic epidermal necrolysis patients who reported anticonvulsant medication use stated phenobarbital was the drug of use.⁷

One of the earliest case reports stated that severe barbiturate poisoning is characterized by the appearance of multiple bullous lesions on the ocular conjunctivae and skin.⁸ A 1965 observational study revealed that 19 (6.5%) out of 290 patients suffered from definite bullous lesions after barbiturate intoxication.¹ This same study found an association between bullae location and sites of pressure. The onset of lesions does not correlate with coma depth or cardiovascular or respiratory side effects.

The pathogenesis of the skin changes from normal to bullous lesions is unknown. Various studies propose mechanisms of barbiturate-induced bullae formation. Biopsy specimens of the bullae show necrosis of the eccrine sweat glands whereas adjacent normal skin fails to reveal any pathologic changes.^1,3,9,10 This argues against the hypothesis that the growth of bullae is a consequence of direct spread of cutaneous drug toxicity.⁹ The appearance of several bullae in different locations, unconnected by abnormal skin, is also inconsistent with that hypothesis. Another theory suggests that bullae formation results from uninterrupted pressure (presumably due to the immobile state in coma) exacerbated by low blood flow states due to shock or vasoactive drugs.^1,3,9,10 Moreover, local anoxic states produce a necrotic bulla and further destroy the most metabolically active cells of the eccrine sweat glands before any of the other skin cells. More recent studies have proposed various immune mechanisms leading to bullae formation.^11,12

This case demonstrates similar features to other case reports of patients with coma blisters^1,3 and those with bullae at the site of localized barbiturate administration.^2,10 Our patient's distribution of bullae and time to onset of the lesions were also fairly comparable to comatose patients in a past British study.¹ The striking difference is that this patient did not receive a toxic, coma-inducing dose nor localized, intravenous administration of phenobarbital. After oral-only ingestion of phenobarbital, all the bullous lesions erupted simultaneously, and no new lesions were noted after admission. The locations of these lesions were not localized to one extremity or site of ingestion. Some of her lesions were localized to areas likely to be exposed to increased stretch, such as the Achilles tendon area. But, other lesions, similar to those described in previous studies, were found on the dorsum or contiguous surfaces of the fingers and hand, which tend to have less pressure.^2,8,13

The fact that no new lesions appeared while the patient was under direct observation by the medical team suggests a previous, recent topical chemical exposure. It is possible that the patient was topically exposed to the active ingredients of phenobarbital by manually handling the phenobarbital tablets. Along the same line, the patient could have been topically exposed to some other chemical or toxin besides phenobarbital. However, this would not explain the distribution of bullae on her back and Achilles tendon skin. Exposure to poison ivy-like toxins is unlikely since she had not gone outside.

It is to be noted that this patient is impulsive and potentially manipulative, capable of ingesting or contacting substances despite her group home's “24-hour supervision” as evidenced by her use of another resident's phenobarbital and significant amounts of marijuana and perhaps alcohol as well. Therefore, any conclusions as to the etiology of her clinical presentation must be preliminary and partially speculative.

Her initial lethargic and uncooperative presentation and subsequent agitation is suggestive of delirium despite sub-therapeutic phenobarbital blood level. However, this patient is known to have multiple episodes of agitation and emotional outbursts in all her previous hospitalizations without obvious evidence of delirium. Possible explanations include hypersensitivity to barbiturates and/or other psychoactive compounds, cumulative effect of multiple ingested compounds as evidenced by positive urine drug screen, or ingestion of other unknown or undetected compounds. Additionally, she could have ingested additional lorazepam or other benzodiazepines or alcohol. Contributing effects of marijuana ingestion must also be considered. Any of the above could also be contributing factors to her dermatologic findings. Nonetheless, no reports have identified cases of cannabis leading to dermatologic bullous lesions nor cannabis and phenobarbital interacting synergistically to cause these adverse effects. Alcohol ingestion is a possibility, but unlikely since her liver function enzymes were only mildly elevated. Phenobarbital is also known to be metabolized by the liver, but it is uncertain if it would cause elevation of the liver function enzymes. At this time, there is no clear explanation of why this patient's AST level is solely elevated.

Assuming that her dermatologic lesions were in fact wholly or partially caused by phenobarbital oral ingestion, the clinical significance of this case is that bullous lesions can occur as a result of oral ingestion of doses of phenobarbital similar to those that are prescribed. All previous case reports suggest phenobarbital produces bullae only after large enough doses to induce coma, or when injected, only in the location surrounding the injection site. This suggests either subcutaneous leakage or higher concentrations in the veins near the injection site before dilution systemically. This case challenges those previously held assertions. Fortunately, the course of the lesions is self-limiting and does not require additional care beyond the treatment of a superficial burn. A wide-spectrum antibiotic ointment, such as silver sulfadiazine topical cream, is sufficient to prevent secondary infection of ruptured bullae. The majority of the bullae resolve within two weeks, in which it is replaced by clinically intact normal skin. No conclusion can be made regarding the mechanism of bullous lesion formation from this single case report. Nevertheless, as phenobarbital continues to be prescribed orally, physicians should be aware of this potential adverse dermatological effect.