The rate of infective endocarditis (IE) in people who inject drugs (PWID) has been rapidly increasing in recent years, explains Michael Silverman, MD. Dr. Silverman and colleagues hypothesized that the type of opioid being injected could be an integral part of the increased rate of IE. “We postulated that exposure to controlled-release hydromorphone could be associated with an increased risk of IE among PWID,” he adds. To determine the effect that injecting controlled-release hydromorphone could have on PWID, Dr. Silverman and colleagues conducted a retrospective cohort study.

Comparing IE Frequency

For a study published in The Lancet Infectious Disease, the researchers used a large database to analyze 60,529 PWID. Previous research from the study team suggested that sugars in controlled-release opioids can stabilize viral survival within injection equipment, leading to increased risk for HIV and Hepatitis C transmission. Additionally, controlled release formulations require more handling than regular release formulations, which can lead to greater risk for Staphylococcus aureus contamination, and the same sugars within the controlled release formulation can prolong S. aureus survival in injection equipment, putting patients at risk for bacteremia.

The study team estimated controlled-release hydromorphone exposure and risk of IE by first analyzing patient data in a population-level analysis, identifying whether they lived in regions with high (≥25%) or low (≤15%) hydromorphone prescription rates. Patients were then matched 1:1 by additional baseline characteristics to compare IE frequency. Patients with prescription data were analyzed to identify those who filled prescriptions for controlled-release or immediate-release hydromorphone and matched with patients who filled prescriptions for other opioids. Again, the two groups were compared for IE frequency.

Controlled-Release Hydromorphone & IE

Between April 2006 and September 2015, 733 (1.2%) patients had IE. “Patients living in high hydromorphone use areas were matched with patients in low hydromorphone use areas,” explains Dr. Silverman. “With many patients obtaining their opiates from local diversion of prescriptions, this allowed us to look at local availability for injection. We identified 254 (1.6%) hospital admissions in PWID to be for IE in regions with high hydromorphone use, while 113 (0.7%) were among PWID in regions with low use. This approach controlled for total opioid prescribing within the local communities, and therefore, suggests that local community availability of this particular opioid was associated with IE incidence.”

Through the patient-level analysis, 3,884 patients were matched (Table) to find frequency of IE. Patients who filled prescriptions for hydromorphone had a higher rate of IE than those who filled prescriptions for other opioids (2.8% vs 1.1%). This association was seen in patients who had taken controlled-release hydromorphone (3.9% vs 1.1% (adjusted Odds Ratio, 3.3) but was not significant with immediate-release hydromorphone. “This confirms that the drug excipients and specific practices needed to inject the controlled release preparation predispose patients to IE,” adds Dr. Silverman.

Protecting PWID

When considering next steps, Dr. Silverman believes avoiding prescribing controlled-release opioids could reduce the burden of infectious complications, such as HIV, Hepatitis C, and IE. “Other controlled-release opioids also need to be studied to determine their ability to support bacterial and viral contamination in injection drug equipment,” adds Dr. Silverman. “In light of ongoing diversion and misuse, the specific rates of infectious complications of various drugs when injected need to be clarified so that only those at lowest risk are prescribed. The details of patient practices associated with injection of drugs needs further study so that we can intervene more effectively in protecting our community.”

References

Hydromorphone and the risk of infective endocarditis among people who inject drugs: a population-based, retrospective cohort study

https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(19)30705-4/fulltext