Interaction between steroid injections and HIV protease inhibitors can disrupt cortisol levels

Corticosteroid injections can lead to dysfunctions in cortisol levels in people taking HIV protease inhibitors, according to research published in the online edition of the Journal of Acquired Immune Deficiency Syndromes. Investigators in the United States found that 11% of their HIV-positive patients developed hypothalamic-pituitary-adrenal (HPA) axis dysfunction after treatment with local corticosteroid injections. All were taking antiretroviral therapy based on a protease inhibitor (PI).The authors believe that a drug interaction was the reason.

“The fact that all cases of HPA-axis dysfunction in this cohort occurred in patients who were treated with a PI underscores the significant concern regarding drug-interactions in this population,” write the authors. “Considerable morbidity can result from either over- or under-dosing of essential medicines. Corticosteroid levels are known to increase when administered with ritonavir [Norvir] and other PIs, which are CYP34A inhibitors and thus result in reduced clearance of steroid…levels.”

Improvements in antiretroviral therapy mean that many HIV-positive people now have an excellent prognosis and are living well into old age. This means that the management of co-morbidities and the diseases of ageing are an increasingly important part of HIV care. Osteoarthritis and musculoskeletal complaints are common in older people and treatment with corticosteroid injections can offer relief.

Corticosteroids are metabolised by the liver using the CYP34A pathway. This is also utilised by protease inhibitors, meaning there is a significant risk of drug interactions, which could potentially lead to the development of HPA-axis dysfunction and Cushing’s syndrome (prolonged exposure to elevated levels of cortisol). Symptoms can include excessive tiredness, mood changes, increased appetite and weight gain, hypertension, weakness, bruising and fungal skin infections.

Investigators in Boston therefore examined the records of 170 HIV-positive people who had received therapy with corticosteroid injections between 2002 and 2011. They examined the prevalence of and for risk factors for HPA-axis dysfunction.

A total of 81 people (47%) were taking therapy based on a protease inhibitor, 56 (33%) were taking another form of HIV therapy and 29 (17%) were treatment naive.

The participants had a mean age of 52 years and 69% were male. Their mean CD4 cell count was 569 cells/mm³. The majority of participants (89%) had an undetectable viral load.

There were nine laboratory-confirmed cases (5%) of HPA-axis dysfunction. All the cases involved people whose HIV therapy was based on a protease inhibitor. This meant there was an 11% prevalence of HPA-axis dysfunction among people taking this class of antiretroviral drug.

Among those who developed HPA-axis dysfunction, three individuals received one corticosteroid injection, two received two injections and four people had received three injections in the previous six months. Eight of the participants received injections containing triamcinolone and one person was treated with methylprednisone. The median time between the first injection and the development of HPA-axis dysfunction was 31 days.

Four people (44%) required hospitalisation. There were no deaths and no cases of adrenal crisis. All nine patients were treated with 3 to 5 mg prednisone. The durations of steroid replacement therapy lasted between one week and more than three years.

The only risk factor for the development of HPA-axis dysfunction was receipt of two or more steroid injections in the previous six months (HR = 6.42; 95% CI, 1.20-34.33, p < 0.05).

“Our data show that corticosteroid injections pose a substantial risk of disrupting the endogenous HPA-axis dysfunction in HIV-infected patients on PIs leading to clinical presentation with signs and symptoms of steroid excess or adrenal insufficiency,” comment the authors. “The total number of steroid injections over a relatively short period of time is a significant predictor of subsequent development of HPA-axis dysfunction. It might be possible to reduce the risk of this adverse event by decreasing the amount of corticosteroid used in these injections for this at risk population. Pharmacokinetics studies could also be helpful in future investigations.”