LIVERPOOL, ENGLAND – Cardiovascular disease was the predominant cause of death of patients with granulomatosis with polyangiitis 1-5 years after a diagnosis in a study by U.K. researchers, suggesting that this could be a target for future intervention.

While active disease was the No. 1 cause of death within the first year of diagnosis in 40% of patients with granulomatosis with polyangiitis (GPA), it was overtaken by cardiovascular disease (CVD) as the main cause of death in 37.5% of patients in the next 4 years from diagnosis.

Nevertheless, active disease remained an important cause of death, accounting for 10% of deaths at 1-5 years, 18.2% at 5-10 years, and 16.7% at 10-15 years.

“The idea for this study came from patients with vasculitis who were polled by Vasculitis UK,” Fiona A. Pearce, MBBS, explained at the British Society for Rheumatology annual conference.

“Further research into mortality was one of their top priorities as patients want to know the honest truth about what is going to happen to them,” added Dr. Pearce, of the division of epidemiology and public health at the University of Nottingham (England).

GPA is a rare type of vasculitis that is estimated to occur in around 1,350 people every year in the United Kingdom. Mortality is known to be high, with around 11%-14% of patients dying in the first year of diagnosis, but there are few data on what happens over a longer term.

The aim of the study was therefore to examine patient survival in the long term – what were the mortality rates several years post diagnosis? Did the risk of death remain high throughout this time and did the causes of death change?

“Of course, the important clinical question was, “Can we then improve things?’ ” Dr. Pearce said.

Two U.K. databases – Clinical Practice Research Datalink and Hospital Episode Statistics – were used to identify patients with GPA diagnosed between 1998 and 2014 and match each case to 10 controls based on age, gender, and family practice. Data on the cause and date of any deaths were then obtained from the Office of National Statistics.

Overall, 465 cases of GPA were matched to 4,610 controls. The median age of participants was 61 years and 57% were male. There were 50 cases with more than 10 years of follow-up data available and data on 139 deaths could be analyzed.

A survival analysis showed that there was a significant reduction in cases versus controls, “but it’s not a constant,” Dr. Pearce noted.

“In the time immediately after diagnosis, the risk of death in people with GPA is very high, and over the first 6 months it tails off.” Then the mortalities are very similar, albeit much lower, to those of controls, but with another dip around 8 years.

The number at risk of death in the second year was 310 for GPA vs. 3,543 for controls and 230 vs. 2,622 at 4 years, 138 vs. 1,704 at 6 years, 93 vs. 1,136 at 8 years, and 49 vs. 658 at 10 years.

Looking at the data another way, the hazard ratios for death comparing GPA cases to controls was 24.5 in the first month, 14.6 in months 1-2, 7.5 by 2-3 months, 4.3 at 3-6 months, 1.6 at 6 months to 8 years, and 3.2 at 8-10 years.

Mortality seems to have improved with time. Splitting the cohort into two time periods based on their diagnosis, those diagnosed between 2008 and 2014 had a lower risk of death than did those diagnosed between 1998 and 2007, although it was still more than four times higher than the background population.

The leading cause of death in patients with GPA 5-10 years after diagnosis was cancer (30.3% of cases), but when this was compared against the general population, the risk was no greater (hazard ratio, 1.0).

GPA cases were also 2.9 times more likely than controls to die as a result of an infection, suggesting that this together with CVD could be a target for mortality reduction strategies.

“We can’t get away from the fact that although this is a large study, there are still small numbers of patients because this is rare disease,” Dr. Pearce observed. “We also don’t have detailed clinical information on each patient, so we can’t look for associations or clinical phenotypes at diagnosis, and there are no biomarkers.”

From a clinical perspective, she noted, it is important to remember that deaths in the first year are mainly from active disease and to continue to try to diagnose and treat the condition as early as possible.

Arthritis Research UK funded the trial. Dr. Pearce had nothing to disclose.

SOURCE: Pearce F et al. Rheumatology. 2018;57(Suppl. 3):key075.204.

Heart disease in GPA exacts high toll in year 2 and beyond