Background
Rhabdomyolysis is common in the HIV-positive population, particularly in those with advanced disease [
1]. A cohort study from Kaiser Permanente reported a 10-fold higher incidence of rhabdomyolysis among the HIV-positive compared with HIV-negative individuals (265 events/100,000 py versus 26 events/100,000 py; 95% CI: 8.5–12.0;
p < 0.001) [
1]. The spectrum of the illness varies from asymptomatic elevation in muscle breakdown biomarkers to a life-threatening disease causing severe acute kidney injury (AKI), major electrolyte disturbances and death in some cases. Rhabdomyolysis is responsible for 5–25% of all cases of AKI [
2] and is associated with death in up to 10% of these cases [
3].
Risk factors unique to this population include HIV itself [
4,
5], co-infection with HCV [
6,
7], medication-related adverse effects, drug-drug interactions [
8], and alcohol and illicit drug abuse [
9]. Despite the notably higher incidence of rhabdomyolysis among HIV-positive persons, the literature in this patient population is sparse. Studies to date have been relatively small and dependent on diagnostic codes (e.g. ICD-9 codes) [
1,
10‐
16]. To more comprehensively evaluate the risk factors for rhabdomyolysis and its impact on clinical outcomes in the context of HIV infection, we aimed to investigate the incidence, patient characteristics, etiologies and outcomes of rhabdomyolysis in a large cohort of HIV-positive patients cared for at a tertiary care center in Maryland, USA between June 1992 and April 2014.
Discussion
To the best of our knowledge, this study represents the largest to date evaluating causes and outcomes of rhabdomyolysis in an HIV-positive population. Our study showed a relatively high incidence of rhabdomyolysis as defined by CK ≥1000 IU/L. Infection, compression injury and drug/alcohol use were the most common etiologies present in these patients. In contrast to other studies, multiple concurrent potential etiologies were frequently identified in this population.
These processes lead to muscle necrosis through direct sarcolemmic injury or by interfering with the oxidative or glycolytic energy pathway, ultimately leading to lethal intra-sarcoplasmic calcium overload, ATP depletion and subsequently muscle cell death. The incidence of AKI was very high in this patient population at 46%, though in large part a result of an infection associated with the rhabdomyolysis.
The incidence of rhabdomyolysis in our HIV population was 943 per 100,000 person-years. Reported data from Kaiser Permanente on HIV-positive individuals showed an incidence rate of 265 per 100,000 person-years [
1]. Studies assessing incidence rates of rhabdomyolysis in the general population have specified those based on particular etiologies. The incidence rate of rhabdomyolysis secondary to lipid-lowering drugs ranged between 2.46 to 4.4 cases/100,000 person-years with statin monotherapy and up to 37–60 cases/100,000 person-years with statin/fibrate combination [
24,
25]. The latter is comparable to the incidence rate of rhabdomyolysis associated with statins in our population which is around 55 per 100,000 person years. The incidence rate was 22.2 cases per 100,000 person-years in military trainees with exertional rhabdomyolysis [
26]. Therefore, compared to the prior studies, the incidence rate in our population was significantly higher, suggesting higher risk of rhabdomyolysis in the HIV-positive population. This is likely related to the higher risk of infection in this population, higher rates of drug exposures (illicit and non-illicit), and possibly higher risk of drug-drug interactions. In addition, differences among studies with regards to incidence could also be due to differing thresholds used to define rhabdomyolysis. Similar to other studies of rhabdomyolysis, the majority of patients were men. This likely reflects their increased muscle mass and possible increased frequency of exposure to risk factors.
More than one potential contributory factor was present in 33% of patients. Multiple risk factors resulted in more severe injury and higher CK levels. This is in contrast to the study by Melli et al. which did not show a correlation between increasing number of risk factors and higher CK levels [
27].
Infection was the most frequently encountered etiology among the HIV-positive population, but it was associated with milder CK elevations. Seventy-five percent of patients with infection had a CK level between 1000 and 5000 IU/L. Sepsis as a major cause for rhabdomyolysis in the HIV-positive population was identified in six out of a series of seven patients by Joshi and Liu [
16] and in seven out of a series of 20 by Chariot et al. [
15]. Pneumonia was the most common etiology identified in our population; consistent with what has been demonstrated in a study of 52 patients by Blanco et al. [
28] who assessed infectious versus non-infectious causes of rhabdomyolysis. Respiratory infections constituted 38% of the infectious etiologies. This could reflect the fact that pneumonia accounts for most infectious admissions to the hospital [
29], but also raises the question of whether hypoxia secondary to pneumonia further potentiates muscle cell necrosis. Two or more concomitant infections – viral, bacterial or fungal - were seen in 27.5% of our patients. This likely speaks to the immunocompromised status of this particular set of patients and their vulnerability to infectious complications.
Compression injury was the second most common etiology in our population. We attribute this to the increased prevalence of poly-substance abuse in our study population, with associated increased risk for immobilization and consequent pressure on muscle groups. It might also reflect the socioeconomic background of this population that puts them at increased risk for traumatic injury [
30,
31].
HIV-associated inflammatory myositis was present in 11 cases (3%). All of these were in the post-HAART era. Melli et al. described two cases of HIV associated myositis over a 9-year period [
27]. The increased incidence in our study might imply an increased recognition of this disease entity or, possibly, poor adherence of our study cohort with their antiretroviral therapy, since 10 of 11 had viral load >400 copies per ml.
Statins were responsible for >50% of medication-induced myotoxicity, whereas in other reports, antipsychotic medications were the leading etiology of medication-induced rhabdomyolysis [
27], and the cause of recurrent rhabdomyolysis in 10% of patients [
28]. Our results are likely explained by the increased prevalence of dyslipidemia and atherosclerotic disease associated with cART and thus the increased frequency of statin use. In addition, drug-drug interactions between statins (those metabolized through the cytochrome P450 3A4 [CYP3A4] enzyme system: simvastatin, lovastatin, and atorvastatin) and protease inhibitors, particularly the powerful CYP3A4 inhibitors ritonavir or cobicistat, are well known [
8]. Previous studies have reported myotoxic medications to be responsible for 11% of cases [
27]. The lower percentage in our study likely reflects disproportionate rates of infection, drug and alcohol use in this cohort, and possibly reflects medication non-adherence in our patient population.
AKI developed in 46% of our patients. Comparison of the incidence of rhabdomyolysis associated AKI in the HIV-positive population to the general population is difficult due to varying definitions of AKI used and different cutoffs for CK level to define rhabdomyolysis; the reported incidence in the general population ranges from 13 to 46%, so that rhabdomyolysis-associated AKI in this particular HIV-positive population is at least at the higher end of the spectrum.
Higher CK levels were associated with increased incidence of AKI. This is consistent with findings from several other studies [
3,
32]. Not surprisingly, the need for RRT was also higher in the higher CK groups, though the difference did not reach statistical significance between the three different CK categories. This is likely secondary to the small proportion that needed dialysis in each group.
The mortality rate was higher in the intermediate and high CK groups at 21% and 18% respectively, compared with 11% in the low CK group. We attributed this high mortality to the increased proportion of patients with infection in those groups (55% and 38% respectively|), where CK elevations were likely related to the infection itself. CK elevation with infection reflects muscle ischemia from a prolonged hypo-perfusion which also results in ischemic renal injury, and intracellular acidosis leads to ATP depletion and increased risk of muscle necrosis. In addition, mild elevations of CK have been reported in febrile patients [
33].
As was shown previously by Ward et al. [
18], the mortality of hospitalized patients with a CK level > 1000 IU was higher in those who developed AKI versus those who did not develop AKI. This was true in our population as well. In addition, rhabdomyolysis in this HIV population was associated with a 1.5–4 fold increase in the rate of death than the general population (13% in this study compared to 3.4–10% in the general population reported in prior literature) [
27].
Our study has several limitations. It is a retrospective study where data collection was based on chart reviews. Nonetheless, we did not solely rely on reported diagnosis in discharge summaries. Follow up clinic notes, response to interventions/treatments, laboratory values, toxicology screens, electromyographies and biopsies were reviewed on each patient when available. Also, this is a single urban center study with a predominantly African American population and many of those studied had virologically uncontrolled HIV disease. As noted above the study also spanned the different treatment eras of HIV care. These factors limit its generalizability to other cohorts of different races and ethnicities and less severe HIV disease.
On the other hand, this is the largest HIV cohort studied regarding the epidemiology, causes and outcomes of rhabdomyolysis. Patients were selected based on their lab results rather than on ICD 9 codes, which may miss some mild cases of rhabdomyolysis. To our knowledge, there are 2 other studies that assessed rhabdomyolysis in the context of HIV infection. In a series of 20 hospitalized HIV-positive patients in France [
15], rhabdomyolysis was responsible for 10% of cases of biopsy-proven AKI. In another series of seven patients, three patients developed AKI, and one of those patients died [
15,
16].