The association between cocaine use detected on drug screening and rhabdomyolysis

Abstract

Rhabdomyolysis is associated with patients presenting to the emergency department who test positive for cocaine on rapid drug screening, including those neither agitated nor restrained. The aim was to determine patient characteristics, frequency, and severity of cocaine-associated rhabdomyolysis. We studied cocaine-positive patients also tested for rhabdomyolysis with creatine kinase (CK) over a six-year period and recorded demographics, vital signs, disposition, diagnoses, and routine labs. We then compared patients with and without rhabdomyolysis. There were 215 patients. The majority were male, black, middle-aged, and smoked tobacco. Psychiatric, neurological, and cardiovascular final diagnostic groups were most common. Other substances detected on drug screening included ethanol (18%) and amphetamines (36%). Rhabdomyolysis (CK ≥ 1000 U/L) was diagnosed in 54 (25%) patients, with median (IQR) CK of 2052 (1305–4959) and range of 1009–40,647 U/L. There were significant differences in gender, race, disposition, and renal/cardiac laboratory test results between patients with and without rhabdomyolysis. We performed multiple logistic regression analysis to reveal elevated BUN, blunt trauma, male gender, and amphetamines detected on drug screen as associated factors for developing rhabdomyolysis. The frequency of rhabdomyolysis in patients who used cocaine was 25%, with certain associated patient characteristics.

Keywords:

• Cocaine
• rhabdomyolysis
• methamphetamine
• amphetamine
• trauma
• cardiac

Introduction

Patients who use cocaine and other stimulants, such as methamphetamine, may present to the emergency department (ED) for acute care of a wide spectrum of illnesses, injuries, and mental health issues. Based on the 2019 United Nations World Drug Report, cultivation of the coca plant species Erythroxylum coca and cocaine production has recently reached the highest recorded level in history [1]. An estimated 18 million adults worldwide used cocaine in the year 2017, with the highest prevalence in North America and Europe. Based on the most recent 2018 National Survey on Drug Use and Health, both cocaine use and overdose deaths have been steadily rising since 2010 [2,3].

Rhabdomyolysis is a syndrome due to direct or indirect muscle injury, myocyte death, and release of intracellular contents into the bloodstream leading to complications such as acute renal failure and death [4]. Cocaine use has been associated with the risk of developing rhabdomyolysis in case reports and studies dating back to the 1970s [5–11]. A commonly held belief is that cocaine-associated rhabdomyolysis results from agitation or resistance to physical restraint. However, it has been reported that some non-agitated, non-restrained patients developed rhabdomyolysis [5,8–10]. This suggests other factors resulting in direct and indirect myocyte injury were responsible, such as blunt trauma, direct toxicity of cocaine on myocytes, or vasoconstriction causing further ischemic injury and free radical formation [5,12,13].

Patients presenting to the ED may not be forthcoming about illicit drug use. Clinicians must often decide whether or not to screen for drug use using rapid qualitative urine testing, and how to interpret positive test results within the context of a patient’s signs and symptoms. This includes screening for rhabdomyolysis, especially in non-agitated, non-restrained patients. Furthermore, a positive test result does not necessarily correlate with drug use immediately before presenting to the ED and may be subject to false positives. The frequency and severity of rhabdomyolysis associated with recent cocaine use detected on drug screening have not been well-defined, and there is potential for missing this important diagnosis. Early identification and treatment of rhabdomyolysis can prevent or mitigate acute and chronic renal impairment. To further understand this association, we conducted this study to compare the characteristics of cocaine-positive patients presenting to the ED who were and were not diagnosed with rhabdomyolysis, and to determine which factors correlated with developing rhabdomyolysis.

Methods

We conducted a retrospective review of the electronic medical records of a convenience sample of patients presenting to a university-affiliated ED from July 1, 2012 to July 1, 2017. This ED is the level I trauma center for the surrounding metropolitan area of two million residents and represents a tertiary care center for northern and central California. With approximately 80,000 patient-visits per year, the ED also serves as the de facto safety net for underserved, undocumented, and uninsured patients, especially those requiring acute psychiatric care. This ED also contractually accommodates victims and perpetrators brought by law enforcement.

We created a Clarity database (Epic Systems Corporation, Verona, Wisconsin, USA) to identify those patients who underwent qualitative urine toxicology screening and were positive for cocaine and/or its main metabolite benzoylecgonine. Of this group, we selected those further screened for rhabdomyolysis with creatine kinase (CK). The qualitative urine toxicology screen was performed using a UniCel DxC 800 Synchron (Beckman Coulter Inc., Brea, California) to detect cocaine and benzoylecgonine, at a cutoff level of 300 ng/mL. This antibody-specific assay has been tested for 26 different drugs which may represent false positives and only detects cocaine and benzoylecgonine [14]. This test also specifically detects d-methamphetamine, d-amphetamine, MDMA, and methylenedioxyamphetamine (MDA), collectively referred to as “amphetamines” in this study.

We recorded gender, age, race, and tobacco use. All ages were included. We abstracted initial vital signs, disposition from the ED, discharge diagnoses, and pertinent laboratory results such as blood urea nitrogen (BUN), creatinine, ethanol level, troponin I, and B-type natriuretic peptide (BNP), as these represent commonly ordered laboratory tests in the ED and are important indicators of renal and cardiac function. Only the first set of laboratory results were chosen, as the setting of this study was the initial workup and consideration for rhabdomyolysis in the ED. Peak laboratory results from serial measurements after admission to the hospital were thus not included in the analysis. Normal ranges for these laboratory results as defined by our hospital’s criteria were as follows: CK (0–250 U/L); BUN (8–22 mg/dL); creatinine (0.44–1.27 mg/dL); ethanol (0–80 mg/dL); troponin I (0–0.04 ng/mL); and BNP (1–100 pg/mL). Rhabdomyolysis was defined as four times the upper limit of normal, which was greater than or equal to 1000 U/L.

Discharge diagnoses were grouped into related categories based on the 10^th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) to enable proportion analysis, and included (with examples): blunt trauma (motor vehicle accidents and assaults); penetrating trauma (gunshot and stab wounds); cardiovascular (chest pain, acute coronary syndrome, congestive heart failure, dysrhythmia, etc.); pulmonary (pneumonia, chronic obstructive pulmonary disease, pulmonary embolus, upper respiratory infection, etc.); integumentary (cellulitis, abscess, burn, etc.); psychiatric (overdose, psychosis, depression, agitation, anxiety, etc.); neurological (stroke, headache, etc.); musculoskeletal (myalgia, rhabdomyolysis, back pain, etc.); gastrointestinal/genitourinary (urinary tract infection, abdominal pain, etc.); gynecological/obstetric (vaginal bleeding, pelvic inflammatory disease, pregnancy, etc.); endocrine/hematological/oncological (diabetic ketoacidosis, thyrotoxicosis, sickle cell disease, cancer, anemia, sepsis, renal failure, etc.) [15]. We analyzed the differences between cocaine-positive patients with and without rhabdomyolysis using Student’s t-test, Mann-Whitney U-test for non-parametric data, χ² test, and multiple logistic regression (MedCalc version 19.1, Ostend, Belgium). Results are reported as mean ± standard deviation (SD), median with interquartile ranges (IQR), odds ratio (OR), and 95% confidence intervals (CI) unless otherwise stated. Statistical significance was assumed at a level of p ≤ 0.05. This study was approved by the medical center’s Institutional Review Board.

Results

There were 1940 patients evaluated in the ED with a positive toxicology screen for cocaine during the study period. A total of 215 cocaine-positive patients were tested for rhabdomyolysis with total CK and were included in the study (Table 1). The majority of patients were male (n = 149, 69%), black (n = 127, 59%), aged 46 ± 15 years, and smoked tobacco (n = 134, 62%). Concomitant substances detected on drug screening included ethanol (n = 39, 18%) and amphetamines (n = 77, 36%). Psychiatric (n = 54, 25%), neurological (n = 35, 16%), and cardiovascular (n = 25, 12%) final diagnostic groups were the most common. The majority (n = 135, 63%) were admitted to the hospital from the ED, and twelve (6%) were discharged to an inpatient psychiatric facility. There was no significant difference in initial vital signs between the two groups (Table 2).

Table 1. Comparison of cocaine-positive patients with and without rhabdomyolysis.

	CK ≥ 1000 U/L	CK < 1000 U/L	p	Χ^2
Total	54 (25%)	161 (75%)
Female	9 (17%)	57 (35%)
Male	45 (83%)	104 (65%)	0.02	6.7
Age	47 ± 15	46 ± 15	0.8*
Range	16–73	16–78
Race			0.02	9.5
Asian	0	2 (1%)
Black	34 (63%)	93 (58%)
Hispanic	3 (6%)	17 (11%)
Native American	0	3 (2%)
White	7 (13%)	36 (22%)
Other	10 (18%)	10 (6%)
Amphetamines positive	22 (41%)	55 (34%)	0.4	0.8
Ethanol positive	11 (20%)	28 (17%)	0.6	0.2
Tobacco use	33 (61%)	101 (63%)	0.8	0.1
Diagnostic group			0.1	12
Blunt trauma	7 (13%)	9 (6%)
Penetrating trauma	3 (5%)	15 (9%)
Cardiovascular	10 (19%)	15 (9%)
Pulmonary	3 (5%)	11 (7%)
Integumentary	1 (2%)	8 (5%)
Psychiatric	9 (17%)	45 (28%)
Neurological	7 (13%)	28 (17%)
Musculoskeletal	0	3 (2%)
GI/GU	7 (13%)	11 (7%)
OB/GYN	0	0
Endocrine/Heme/Onc	7 (13%)	16 (10%)
Disposition			0.006	10
Home	7 (13%)	56 (35%)
Admit	44 (81%)	91 (57%)
AMA/Eloped	0	4 (2%)
Jail	0	1 (1%)
Psychiatric	3 (6%)	9 (5%)
Expired	0	0

Abbreviations: AMA = against medical advice; CK = creatine kinase; GI = gastrointestinal; GU = genitourinary; GYN = gynecology; Heme = hematology; OB = obstetrics; Onc = oncology.

* = Student’s t-test.

Table 2. Initial vital signs and laboratory results.

Initial vital signs [mean ± SD]	CK ≥ 1000 U/L	CK < 1000 U/L	p*
Pulse (beats/min)	83 ± 16	81 ± 14	0.3
Range	34–136	55–121
SBP (mmHg)	126 ± 21	127 ± 20	0.7
Range	67–180	60–192
DBP (mmHg)	73 ± 15	76 ± 14	0.1
Range	23–109	24–113
Respirations (breath/min)	17 ± 4	17 ± 3	1
Range	6–29	6–24
Temperature (F)	98.1 ± 1.5	98.1 ± 0.7	1
Range	90.3–104.9	96.8–102
Laboratory results [median (IQR)]	CK ≥ 1000 U/L	CK < 1000 U/L	p^†	Z-score
CK (U/L)	2052 (1305–4959)	242 (116–418)	<0.0001	−10.9
Range	1009–40,647	11–989
BUN (mg/dL)	21 (11–40)	13 (9–20)	0.0003	−3.6
Range	5–174	3–100
Creatinine (mg/dL)	1.6 (1.2–2.9)	1.2 (0.9–1.7)	<0.0001	−4.4
Range	0.63–25.48	0.45–13.12
Ethanol (mg/dL)	154 (48–214)	131 (57–230)	0.7	0.4
Range	0–239	0–370
Troponin I (ng/mL)	0.06 (0.02–0.2)	0.02 (0.01–0.06)	0.004	−2.9
Range	0.01–3.04	0.01–3.25
BNP (pg/mL)	601 (41–1069)	246 (67–813)	0.4	−0.9
Range	17–5000	16–1433

Abbreviations: BNP = B-type natriuretic peptide; BUN = blood urea nitrogen; CK = creatine kinase; DBP = diastolic blood pressure; IQR = interquartile range; SD = standard deviation; SBP = systolic blood pressure.

* = Student’s t-test.

^† = Mann-Whitney U test.

Of the 215 cocaine-positive patients, 54 (25%) had rhabdomyolysis, with a median (IQR) CK of 2052 (1305–4959) U/L and a range of 1009 to 40,647 U/L. Our univariate analysis of cocaine-positive patients with and without rhabdomyolysis revealed significant differences in gender (p = 0.02), race (p = 0.02), and disposition (p = 0.006). There were also differences in initial renal function laboratory results between the rhabdomyolysis and non-rhabdomyolysis patients: BUN 21 (11–40) mg/dL versus 13 (9–20) mg/dL, p = 0.0003) and creatinine 1.6 (1.2–2.9) mg/dL versus 1.2 (0.9–1.7) mg/dL, p < 0.0001). For the cardiospecific laboratory tests, troponin I was higher between the two groups: 0.06 (0.02–0.2) ng/mL versus 0.02 (0.01–0.06) ng/mL (p = 0.004).

We performed subgroup analysis between solely cocaine-positive patients and cocaine- and amphetamines-positive patients. For the cocaine-only subgroup (n = 138, 64%), patients with blunt trauma (p = 0.03, χ² = 4.2) and cardiovascular (p = 0.007, χ² = 7.2) diagnoses were more likely to have rhabdomyolysis, whereas psychiatric diagnoses were less likely (p = 0.009, χ² = 6.7). For the cocaine/amphetamines subgroup (n = 77, 36%), we detected no significant differences for gender, race, disposition, or initial vital signs. Only differences in BUN (p < 0.0001) and creatinine (p = 0.01) were significant.

We performed a multiple logistic regression analysis of demographics, substance co-positivity, initial vital signs, laboratory results, and diagnoses. This revealed blunt trauma (OR = 4.9, 95% CI: 1.1–21.8; p = 0.04), elevated BUN (OR = 4.6, 95% CI: 1.6–12.9; p = 0.004), male gender (OR = 3.3, 95% CI: 1.3–8.3; p = 0.01), and positive screen for amphetamines (OR = 2.4, 95% CI: 1.0–5.4; p = 0.04) to be significant factors associated with rhabdomyolysis in cocaine-positive patients.

Discussion

In this study, rhabdomyolysis occurred in cocaine-positive patients for a wide range of illness and injury patterns in the ED that were not characterized by agitation, seizure, immobilization, or physical restraint. In the existing literature, the syndrome of rhabdomyolysis in patients who use cocaine and other stimulants had commonly been attributed to extreme agitation or seizures, with frequent or prolonged isometric muscle contraction and hyperthermia from excessive monoamine concentration within the central and peripheral nervous systems [4]. However, cocaine-associated rhabdomyolysis also occurs in patients who are neither agitated nor restrained, which may be explained by direct mechanisms of cocaine toxicity [8–10]. In vitro studies of murine skeletal muscle incubated with cocaine demonstrated rhabdomyolysis occurred less than one hour after exposure, suggesting a direct toxic effect [16,17]. Authors of an in vivo study highlighted yet another mechanism of cocaine-induced rhabdomyolysis by measuring significantly increased levels of glutathione, a free radical scavenger, and thiobarbituric acid reactive substances, a measurement of lipid peroxidation, in mice treated with a single dose of intravenous cocaine [17]. Another in vivo study of the direct effect of cocaine on skeletal eel muscle showed increased activity of cytochrome oxidase, which catalyzes the final step in the mitochondrial electron transfer chain with reduction of oxygen to water [18]. Cocaine-induced vasoconstriction may also be responsible for critical changes in cytochrome oxidase hyperactivity leading to an ischemic phase, a reperfusion phase with increased reactive oxygen species, and a mitochondrial dysfunction phase then apoptosis [19]. Cocaine activates the apoptotic caspase and p38 pathways from pathological release of cytochrome C from the mitochondria into the cytosol, which irreversibly leads to cell death [20–24].

In mild-to-moderate rhabdomyolysis, CK peaks between 24 and 72 h and typically returns to normal by five days [5]. The short half-life of cocaine of approximately one hour compared to the development of rhabdomyolysis overs hours and days further infers these effects are caused by cocaine metabolites and/or secondary to events in other organs such as liver and kidney. Cocaine (benzoylmethylecgonine) is metabolized by plasma and hepatic cholinesterase enzymes to its major metabolite benzoylecgonine and other lesser metabolites, such as ecgonine methyl ester, ecgonine, and norcocaine, which have much longer half-lives than cocaine [25]. These metabolites are also implicated in indirect cardiovascular toxicity and direct intracellular toxicity [26]. An in vitro study of the toxicity of cocaine and its metabolites on isolated mitochondria demonstrated norcocaine and its subsequent metabolites had the greatest inhibitory effect on mitochondrial respiration [27].

In this study we found cocaine-positive patients with rhabdomyolysis had higher levels of BUN and creatinine than those without rhabdomyolysis. In our regression analysis, BUN, which reflects hepatic urea cycle activity from protein digestion, is a test of renal function and had the highest odds of all laboratory studies for association with rhabdomyolysis. Creatinine, which is the breakdown product of phosphocreatine in muscle, is another important test of renal function. A 1990 study of 29 cocaine-positive patients who were diagnosed with rhabdomyolysis stratified them into three separate severity subgroups: mild, characterized by anxiety, tachycardia, diaphoresis, dyspnea, or chest pain; moderate, characterized by delirium, agitation, fever, leukocytosis, or an elevated serum creatinine; and severe, characterized by seizure, coma, hypotension, arrhythmia, or cardiac arrest [9]. The authors determined the patients at highest risk for complications of rhabdomyolysis were those in the moderate or severe groups. Average initial creatinine was 2.9 ± 2.9 mg/dL for the moderate subgroup and 3.3 ± 1.9 mg/dL for the severe subgroup, which was similar to our findings of elevated creatinine.

We found troponin I to be elevated in cocaine-positive patients with rhabdomyolysis compared to those without rhabdomyolysis. Troponins I and T are considered high-sensitivity tests for cardiac myocyte injury. In a study of 109 ED patients with rhabdomyolysis, Li and associates reported a 17% frequency of false-positive cardiac troponin I levels, with poor correlation between peak CK and troponin I levels [28]. The second cardio-specific laboratory test BNP is a hormone secreted by ventricular cardiomyocytes in response to stretching caused by increased ventricular blood volume. The finding of elevated BNP has been shown to be an early predictor of cardiac complications in cocaine users [29]. Our results showed abnormal median BNP for both rhabdomyolysis and non-rhabdomyolysis subgroups, but the difference did not reach statistical significance (Table 2). Based on these findings, we believe abnormal elevation of these commonly ordered renal- and cardio-specific tests should heighten clinical suspicion for rhabdomyolysis with further screening with CK.

The proportion of cocaine-positive males with rhabdomyolysis was significantly higher than females in our study, and male gender was also associated with rhabdomyolysis on regression analysis. The majority of cocaine-positive patients who developed rhabdomyolysis were male (83% versus 65% of those who did not develop rhabdomyolysis), and this trend is consistent with other studies [7,9–11]. In general, males tend to use stimulants at a higher rate than females [30]. Differences in gonadal hormones, such as estrogen and progesterone, between genders are also important determinants [31,32]. Several in vivo studies have demonstrated gender and hormonal differences in response to cocaine use, with females being less susceptible to cocaine-associated cardiovascular toxicity and having different psychomotor responses to equivalent doses of cocaine than males [33–35]. Males also have been shown to have lower activity levels of plasma cholinesterase than females, which leads to slower peripheral metabolism of cocaine [36,37].

We found blunt trauma and rhabdomyolysis to also be associated based on logistic regression. Blunt trauma may result in physical disruption of the sarcolemma, followed by release of sarcoplasmic contents into the vasculature and development of rhabdomyolysis [4]. Clinicians caring for blunt trauma patients in the ED should consider the potential for rhabdomyolysis [10]. Another finding was that concomitant drug screen positivity for amphetamines was significantly associated with rhabdomyolysis in cocaine-positive patients. There may be synergistic and additive effects between the two different stimulants, such as increasing the amount of reactive oxygen species present in cells. Cocaine and amphetamines both increase extracellular levels of monoamines from sympathetic nervous system-mediated adrenal medulla stimulation and through distinct molecular mechanisms within the central and peripheral nervous systems [38]. Cocaine inhibits monoamine re-uptake across the presynaptic membrane, whereas amphetamines enhance release of monoamines from synaptic vesicles. The addition of amphetamines results in even higher concentrations of monoamines within the synaptic cleft compared to cocaine alone, which leads to inflammation and apoptosis of surrounding cells [39]. These excess monoamines are eventually oxidized by monoamine oxidase-B to dihydroxyphenylacetic acid and hydrogen peroxide, which reacts with transition metal ions to produce highly toxic hydroxyl radicals via the Fenton reaction [40]. Other reactive oxygen species implicated in combined stimulant toxicity include hydroperoxides, alkylperoxides, hyponitrite, superoxide, alkoxyl, and alkylperoxyl radicals, which induce cell death from nucleic acid and protein damage [41]. Similar to cocaine, amphetamines have also been implicated in mitochondrial death and activation of caspase apoptotic pathways, and the combination of the two stimulants may also have an additive effect on toxicity [42].

In this study, 25% of cocaine-positive patients developed rhabdomyolysis. In 1991, Welch et al. conducted a prospective study of 68 cocaine-positive patients and found 24% developed rhabdomyolysis, which is similar to our finding [10]. In their study, myalgia was not useful in detecting rhabdomyolysis: only one patient with rhabdomyolysis had myalgia, and three patients without rhabdomyolysis had myalgia. A more recent study by O’Connor et al revealed a frequency of 33% in cocaine-positive patients [11].

Limitations

Limitations of this study include those inherent to a retrospective chart review. It also involves a convenience sample of patients whose treating clinicians ordered a CK, which may have resulted in a higher than actual frequency. The introduction of unintended bias is another potential limitation, as clinicians have varying opinions on the utility of drug screening in the ED with regard to impact on patient care, and which patients should undergo drug screening. Patients were selected on the basis of their drug screen results, which was done without confirmation drug testing. Confirmatory testing with gas chromatography-mass spectrometry does not reflect routine ED care, as it is expensive and results are not immediately available to the treating clinician. However, some patients may have been missed, as they did not undergo drug screening or gave a history of cocaine use during their ED evaluation. To model characteristic ED testing, we only considered initial laboratory values, and some patients who developed rhabdomyolysis or renal failure hours or days later may have been missed. Although the clinical relevance of an elevated CK may be questionable, we believe most ED clinicians would consider a CK greater than 1000 U/L to warrant further evaluation and treatment. Another limitation is our ED demographics may not be applicable to other regions.

Conclusion

One-quarter of cocaine-positive patients in this study had rhabdomyolysis. Characteristics associated with cocaine-positive patients with rhabdomyolysis included male gender, middle-age, blunt trauma, and positive drug screen for amphetamines. Laboratory tests associated with rhabdomyolysis included elevations in BUN, creatinine, troponin I, and BNP. Emergency clinicians should consider measuring CK in patients with any combination of these characteristics and laboratory results.

Disclosure statement

The authors have no conflict of interest to report.