Incidence of rhabdomyolysis occurrence in psychoactive substances intoxication: a systematic review and meta-analysis

Rhabdomyolysis is a potentially life-threatening condition induced by diverse mechanisms including drugs and toxins. We aimed to investigate the incidence of rhabdomyolysis occurrence in intoxicated patients with psychoactive substances. In this review, three databases (PubMed, Scopus, Web of Science) and search engine (Google Scholar) were searched by various keywords. After the screening of retrieved documents, related data of included studies were extracted and analyzed with weighted mean difference (WMD) in random effect model. The highest incidence of rhabdomyolysis was observed in intoxication with heroin (57.2 [95% CI 22.6–91.8]), amphetamines (30.5 [95% CI 22.6–38.5]), and cocaine (26.6 [95% CI 11.1–42.1]). The pooled effect size for blood urea nitrogen (WMD = 8.78, p = 0.002), creatinine (WMD = 0.44, p < 0.001), and creatinine phosphokinase (WMD = 2590.9, p < 0.001) was high in patients with rhabdomyolysis compared to patients without rhabdomyolysis. Our results showed a high incidence of rhabdomyolysis induced by psychoactive substance intoxication in ICU patients when compared to total wards. Also, the incidence of rhabdomyolysis occurrence was high in ICU patients with heroin and amphetamine intoxication. Therefore, clinicians should anticipate this complication, monitor for rhabdomyolysis, and institute appropriate treatment protocols early in the patient’s clinical course.


Search strategy
To find related publications, a combination of related keywords was used in databases and search engine including PubMed, Scopus, Web of Science, and Google Scholar.In addition, we used a manual search to develop literature search, references of selected studies, and citations of studies.The final search was updated on July 15, 2022 before data analysis.The keywords used included a combination of suggested words by Medical Subject Heading (MeSH) and other related words, as represented in details in Table 1.Finally, three limitations in research including human studies, publication date (2000-2022), and English language studies were applied.All processes related to the literature search were done independently by two researchers (AA and ES).

Eligibility criteria
Eligibility criteria were elaborated based on the PICO structure (Population, Intervention\ Exposure, Comparator, and Outcomes), and the studied population under meta-analysis were intoxicated patients hospitalized in the intensive care unit (ICU) or general wards due to poisoning, overdose, and abuse and were either conscious or unconscious.Intoxicated patients following enteral, parenteral, and inhalational use of methadone, cocaine, heroin, tramadol, amphetamine, methamphetamine, ecstasy, MDMA (3,4-methylenedioxyamphetamine), opioid, synthetic cannabinoids, methanol, and ethanol were included in the study.Mono-intoxication means patients who were intoxicated with a single substance and multi-intoxication means patients who had co-exposure and concomitant substances were detected in their drug screen.The research question of the study was occurrence of rhabdomyolysis among hospitalized intoxicated patients.Exclusion criteria were case reports and case series with less than 5 samples, review or editorial articles, none English language manuscripts, and studies on children.

Study selection, data items, data collection
Retrieved observational studies from selected databases with relevant exposures were imported into EndNote citation management software.After removing duplicate studies, title and abstract of remained studies were screened and data extraction was done by two independent researchers (TB and SA).Data extraction forms contained the author's name, year, age, gender, country, continent, study design, type of psychoactive drug, sample size, type of hospitalization wards (ICU/total wards), multi-intoxication (yes/no), dose of substance, and patient's medical history.Also, subgroup analyses for hospitalization ward, multi-intoxication, and geographical Table 1.The search strategy in all databases/search engine.

Google Scholar
This search strategy was repeated for all psychoactives and 300 initial results were reviewed; ((poisoning OR intoxication) AND (opioid) AND (rhabdomyolysis OR creatine kinase)) area (based on World Health Organization regions) was performed.In subgroup analysis, ICU means studies that included only ICU patients and total wards mean studies reporting total patients hospitalized in ICU and general wards.Also, poisoning with multiple drugs means studies reporting overall occurrence of rhabdomyolysis for all intoxicated patients but did not determine the incidence of rhabdomyolysis separately for each substance.Rhabdomyolysis occurrence was defined if included patients had CPK > 1000 IU/L/ or CPK > 5 × ULN 14 .Any disagreement at each stage was checked by a third researcher (AA).

Risk of bias of included studies
The risk of bias in studies was assessed independently by two researchers (MB and MS) and disagreements were discussed and checked by the third researcher (KE).For this purpose, Newcastle-Ottawa scale (NOS) was used to assess quality of nonrandomized studies in meta-analyses and the number of stars indicated methodological quality of articles.

Synthesis of results
The number of intoxicated patients in the studies was considered as denominator of the fraction and the number of samples containing occurrence of rhabdomyolysis was placed in the numerator.The effect size for rhabdomyolysis incidence in each subgroup was determined as pooled effect size with 95% confidence interval (CI).
Weighted mean difference (WMD) was used to compare values of renal function indexes including blood urea nitrogen (BUN), creatinine (Cr), and CPK between intoxicated patients with and without rhabdomyolysis.The pooled effects size was estimated using random effect model by considering disparities between studies.Heterogeneity between studies was estimated by Cochrane Q test and I 2 index.The type of hospitalization wards, geographical area, and multi-intoxication were considered to find the source of heterogeneity in the subgroup analysis.Publication bias was determined by Egger's regression and Begg's test.Sensitivity analysis was performed to assess the impact of a single study on the results.All statistical analyses for meta-analysis were done in Stata software (version 16.0; Stata Corporation).

Study selection and characteristics
The process of study selection is depicted in the PRISMA flow diagram (Fig. 1).In the systematic search of electronic databases, 2493 nonredundant studies were found, of which 62 articles were potentially relevant.After www.nature.com/scientificreports/reading their full texts, 5 articles were excluded due to lack of required data and 57 articles met the inclusion criteria for final analysis.As described in Table 2, the percentage of hospitalized men in all included studies was high compared to women.The minimum and maximum mean age of patients was 19.4 and 46.0 years, respectively.The reported median age ranged from 19 to 50 years.A summary of included article characteristics is described in Table 2. Based on the types of psychoactive substances, a total of 57 articles containing 3,122,944 intoxicated patients were screened.Included articles were 11 studies for opioids, 6 for methadone, 11 for synthetic cannabinoids, 7 for cocaine, 17 for amphetamines, 5 for methanol, 4 for ethanol, 4 for heroin, 2 for tramadol, and 15 for multiple drugs.Most of the studies were done in the United States (n = 24), Iran (n = 13), Canada (n = 3), and Australia (n = 3).

Hospitalization ward and multi-intoxication
The pooled effect size in the subgroup of ICU patients was higher than in the total wards (Table 3).Incidence of rhabdomyolysis in ICU patients and total wards was respectively 13.6% vs 8.1% for opioids, 19.3% vs 16.5% for methadone, 25% vs 8.2% for synthetic cannabinoids, 40.0%vs 25.4% for cocaine, 71.5% vs 22.7% for amphetamines, 21% vs 1.75% for methanol, 100% vs 40.6% for heroin, and 30.5% vs 26.7% for multiple drug poisoning.In addition, pooled effect size in the subgroup of total wards was influenced by the severity of intoxication, which was different in the included studies.In comparison with multi-intoxication subgroup, pooled effect size of rhabdomyolysis occurrence was high in mono-intoxication subgroup except for synthetic cannabinoids (6.1% vs 13.2%) and amphetamines (24.1% vs 31.6%).

Geographic area
As it was showed in Table 3, according to the subgroup analysis based on the geographic area, the highest incidence of rhabdomyolysis occurrence was related to amphetamines in American region (65.6% [35.6-90.5])and synthetic cannabinoids in European region (30.7% [18.9-43.8]),whereas the highest incidence of rhabdomyolysis occurrence in Eastern Mediterranean was related to other psychoactive substances.Subgrouping by geographic region reduced heterogeneity between studies.

Pooled mean effect size of renal function indexes
Table 4 shows the comparison of mean effect size of renal function indexes in patients with and without rhabdomyolysis.Accordingly, the value of BUN (WMD = 8.78, p = 0.002), Cr (WMD = 0.44, p < 0.001), and CPK (WMD = 2590.9,p < 0.001) was significantly high in patients with rhabdomyolysis compared to those patients without rhabdomyolysis.

Publication bias and sensitivity analysis
Table 3 represent specified p-values related to the publication bias in each type of psychoactive substances with Begg's and Egger's test, indicating that there was no publication bias in most of the intoxications.Sensitivity analysis was performed for all tests applied for meta-analysis and the results showed that none of the pooled effect size was influenced by a single study.

Discussion
The current study was a systematic review and meta-analysis of clinical data related to the incidence of rhabdomyolysis among intoxicated patients with psychoactive substances.To the best of our knowledge, this was the first systematic review conducted on rhabdomyolysis occurrence in psychoactive substance intoxication.Our results showed that pooled effect size for all categories of psychoactive substances was high in the subgroup of ICU patients compared with total wards.Also, intoxication with heroin (~ 100) and amphetamine (~ 71.5) showed the highest effect size for occurrence of rhabdomyolysis in ICU patients.In a study in Iran on 227 poisoned patients with refined opium extract, the majority of them (75.8%) were male.However, it has been documented that females have higher mitochondrial mass in skeletal muscle with greater oxidative phosphorylative capacities and therefore have greater protection against rhabdomyolysis 66,67 .
A broad range of neurological complications affecting both central nervous system (CNS) and peripheral nervous system (PNS) are encountered in heroin abusers 24 .CNS lesions include brain hypoxia and seizure, spongiform leukoencephalopathy, stroke, and myelopathy 68 while PNS involvement commonly manifests as compressive neuropathy or focal rhabdomyolysis 69 .The etiology of this acute PNS complication is unclear.Some studies found immunological causes in patients who developed rhabdomyolysis.Also, mechanical trauma is considered a potential mechanism of focal nerve injury and localized rhabdomyolysis in heroin abusers 70  effect size of rhabdomyolysis in ICU patients with heroin intoxication (100 [95% CI 39.7-100]) was the highest in our study.
The second highest effect size of rhabdomyolysis in ICU patients was observed in intoxication with amphetamines (71.5 [95% CI 12.8-100]).The etiology of amphetamine-induced rhabdomyolysis has traditionally been attributed to agitation and/or physical restrain with intense isometric muscle contraction.However, many patients who use amphetamines are not agitated or restrained but experience rhabdomyolysis.Some of the indirect mechanisms and cofactors are gender difference (males at higher risk), monoamine receptor polymorphisms, cocaine and sedative co-injection, seizure, sepsis, and hyperthermia 55 .Hyperthermia is a major toxic reaction to amphetamines that can lead to rhabdomyolysis, hypotension, disseminated intravascular coagulation (DIC), and AKI.Hyperthermia occurs as a result of complex interactions between serotonin, dopamine, norepinephrine, and environmental conditions 71 .In some studies that reported a high frequency of rhabdomyolysis, there are several potential explanations for poor clinical outcomes and rhabdomyolyses like hyperthermia, concertation of amphetamines, and prolonged hypoxia.Most of the severe morbidity and mortality in these cases can be attributed to hyperthermia effects 71 .
In this study we found a higher effect size for patients with mono-intoxication than those with multi-intoxication following use of psychoactive substances.In many studies, alcohol abuse has been identified as a main cause of rhabdomyolysis 8,38 and other studies regarding the etiology of rhabdomyolysis have reported opioid overdose as a significant contributing factor to rhabdomyolysis 8 .In this regard, Talaie et al. reported that opium poisoning is the most common cause of rhabdomyolysis (23.3%), followed by poisoning with benzodiazepines, phenobarbital, propranolol, aluminum phosphide, alcohol, and co-poisoning 8 .Also, Babak et al.in their study reported that rhabdomyolysis is mostly associated with methadone abuse, followed by opium abuse, and is more commonly correlated with poisoning in younger patients 8,38 .

Limitations
The final analysis in our study was substantially limited in the number and quality of studies available.Our review only included the published studies but we tried our best to contact researchers and obtain more information about their studies.Also, few studies in our review reported seizure induced by psychoactive intoxication that may contribute to rhabdomyolysis in intoxicated patients.The next limitation is that we had disparities in the distribution of intoxication severity in total wards subgroup affecting the effect size of the study.Furthermore, distribution of type of psychoactive substance in the subgroup of multiple poisoning was different between studies, which affects the effect size of each study.These limitations and lack of clarity in the studies caused high heterogeneity in our analysis.Another limitation of this study is that we searched only studies with English full text or at least English abstracts, and also subgroup analysis was not possible based on the dose and route of the substance used, blood levels of the drug, and other variables due to insufficient reported data in the studies.In addition, we did not perform any blinding process for all stages of study selection, quality assessment, and data extraction.Therefore, we propose running more comprehensive and original research in this regard to help make a better conclusion regarding the incidence of rhabdomyolysis in patients with psychoactive substance intoxication.

Conclusion
In conclusion, this systematic review and meta-analysis revealed high incidence of rhabdomyolysis occurrence in patients with heroin and amphetamine intoxication compared to other psychoactive substances.Clinicians should anticipate this complication, monitor for rhabdomyolysis particularly in the ICU, and institute appropriate treatment protocols early in the patient's clinical course.

Figure 1 .
Figure 1.Flow diagram of the literature search for studies included in meta-analysis.

Figure 2 .
Figure 2. Pooled incidence of rhabdomyolysis based on the types of amphetamines intoxication.

Table 3 .
Overall and subgroup incidence of rhabdomyolysis according to psychoactive substances intoxication.

Table 4 .
Weighted mean difference of renal indices according to rhabdomyolysis.BUN Blood Urea Nitrogen, Cr Creatinine, CPK Creatine Phosphokinase.