Epidemiology of giant cell arteritis in Waikato, Aotearoa New Zealand

Giant cell arteritis (GCA) is the most common primary vasculitis in adults over 50 years of age. It is a granulomatous vasculitis and classically involves the aorta and its branches, the distribution of which helps to classify it as either cranial or extra-cranial.¹ Symptoms include temporal headaches, visual loss, scalp sensitivity, jaw claudication and limb claudication. A feared complication of untreated disease is permanent vision loss.

Ethnicity may be a risk factor for GCA, with Scandinavian countries reporting the highest incidence rates.² It appears to be much less common in Asian and African American patients.^3–7 Age also appears to be a risk factor, with peaking incidence in the 70–79-year age group with a female predominance.⁷

Investigation of GCA has historically been with temporal artery biopsy (TAB), which has been the gold standard. Colour Doppler ultrasound of the temporal and axillary arteries (CDUS) has been used with increasing frequency over the last decade depending on the availability and expertise in local centres. Ultrasound features of GCA include the halo sign (a hypoechoic ring around the lumen of the temporal or axillary artery) or vessel wall thickening that can lead to stenosis or occlusion of a blood vessel.^8,9 Computed tomography (CT), magnetic resonance imaging (MRI) and F-fluorodeoxyglucose positron emission tomography with computed tomography (FDG-PET) supplement these investigations, particularly in extra-cranial large vessel vasculitis.

The American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR) released new classification criteria in 2022.¹⁰ Patients must be 50 years or older with additional clinical criteria of polymyalgic symptoms, vision loss, jaw or tongue claudication, scalp tenderness and abnormal examination of the temporal artery. Investigations with inflammatory markers, TAB, CDUS and FDG-PET are used in the criteria. Patients with six or more points reach a classification for GCA.

The foundation of treatment for GCA is corticosteroids, and while untreated disease is associated with significant morbidity, its treatment is also associated with morbidity for patients.¹¹ Newer steroid-sparing treatments include tocilizumab, an interleukin-6 inhibitor, or janus kinase (JAK) inhibitors. In resource-limited countries like Aotearoa New Zealand, appreciation of the epidemiology of a condition is crucial when advocating for government support for these evolving therapies, which are currently not routinely available for our patients.

Aotearoa New Zealand is a country of 4.6 million people with 16 regions. As of the 2018 Census, there are 70.2% Europeans (made up mostly of New Zealand Europeans [NZE] with a smaller group of Other Europeans), 16.5% Māori, 15.1% Asian and 8.1% Pacific peoples, along with smaller proportions of other nationalities. Waikato is the fourth largest region in Aotearoa New Zealand, making up 9.5% of the population.¹² Our primary objective was to examine the incidence of GCA in Waikato, which, given its large population, is likely to be reflective of the epidemiology of GCA nationally.

Methods

Beginning in 2013, a new process was developed for handling suspected cases of GCA in the Waikato Region. General practitioners, physicians and those likely to encounter suspected cases were asked to discuss cases with the on-call rheumatologist. Cases were therefore identified prospectively. In addition, a search of ultrasonography lists in the radiology department and a histology search of all temporal artery biopsies performed from January 2014 to December 2022 were screened for patients suspected of having GCA.

The electronic records were reviewed, including clinic letters, primary care referrals, ultrasound request forms, laboratory results, discharge summaries and electronic prescriptions. Data were collected regarding patient demographics, clinical and laboratory features of their GCA, investigations, treatment and mortality. Ethnicity was reported in conjunction with the Statistics New Zealand method of reporting and if patients identified as two ethnicities, both were noted.

Cases of GCA were defined as the final clinical diagnosis recorded by the rheumatologist or ophthalmologist. Where a patient had been evaluated more than once over the time period, only the original evaluation was retained for analysis.

CDUS involved ultrasound of both temporal and axillary arteries. Positive CDUS was classified as the halo sign and was indeterminate if only vessel wall thickening was noted. TAB was classified as positive if there was evidence of active inflammation and the pathology report supported a diagnosis of GCA.

Statistical analysis

Descriptive data are presented as frequencies for categorical variables. Continuous variables are presented as mean with standard deviation (SD). Where symptoms were not reported, this was reported as missing data; thus, percentage calculations are of valid percent rather than total percent. T-tests were used to compare mean age at diagnosis. Chi-squared tests were used to compare deaths between ethnicities. Incidence was calculated by examining the number of new instances of GCA in proportion to the population of the Waikato Region for each year over the duration of the study. Population counts for the Waikato Region were provided by age groups from Statistics New Zealand, a centralised government agency that collects and manages official statistical data for Aotearoa New Zealand. Yearly population projections for the Waikato Region are mostly based around 2018 Census data.¹³ After monitoring incoming patients for GCA over a span of 9 years and meticulously tracking occurrences of deaths, prevalence was calculated by determining the total number of individuals with GCA at the end of the observation period and expressing this count as a proportion of the total population under surveillance (i.e., Waikato Region and people over 50 years old). The mean annual mortality rate was calculated from the yearly death count among the changing total population of total known cases of GCA in the Waikato Region. Moreover, a standardised mortality ratio (SMR) was derived comparing the observed deaths among cases of GCA versus what would be expected using the age-specific rates of the surrounding population. Analyses were conducted in SPSS and R software. All significance tests were two-tailed and values of p<0.05 were considered significant.

Ethics statement

National ethics approval was granted by HDEC (reference: 2023 EXP 15448) and there was a local assessment through the Waikato Hospital who also approved the project (RD023025), which included review by Te Puna Oranga Māori Consultation Research Review Committee.

Results

Cases of GCA and demographics

From January 2014 to December 2022, a total of 761 evaluations took place for patients with suspected GCA and were referred for either CDUS, TAB or both. Of these, 214 individuals had a final diagnosis of GCA. The majority of the patients with GCA were managed by rheumatology (65.9%), followed by ophthalmology (17.8%), a combination of both specialities (8.5%) and other specialities (usually general medicine or neurology; 7.9%). Most patients were female (68.2%) with a mean age of 74.0 (SD 8.3) years.

Patients with GCA were most commonly of European background in 93.9% (201/214), Māori 7.9% (17/214) and Asian 0.5% (1/214). The remaining patients (2.8%) included other minorities or patients who did not state an ethnicity. There were no cases of GCA in patients of Pacific peoples descent. Seventeen patients identified as having both European and Māori ethnicity; thus, the total percentage is greater than 100%.

The mean age of European patients with GCA (74.2 [SD 8.3] years) was significantly older than those identifying as Māori (70.1 [SD 8.9] years) (p=0.047).

Table 1 identifies the baseline clinical characteristics of patients with GCA. There was a mean duration of symptoms of 26.4 (SD 29.5) days at the time of referral or initial review.

Analysis of moving averages for annual numbers of cases of GCA showed a statistically significant peak in 2019 and a smaller one in 2021.

View Table 1, Figure 1.

Diagnostic tests

In patients diagnosed with GCA, CDUS was performed in 167/214 (78.0%). Of these ultrasounds performed, 17 (10.2%) were positive with a halo sign, 88 (52.7%) were negative and 62 (37.1%) were indeterminate with vessel wall thickening noted.

TAB was performed in 168 (78.5%) of the patients with GCA and 67 (39.9%) had positive histology reported. The ACR/EULAR 2022 classification criteria for GCA were positive in 83.6% of patients.

The sensitivity and specificity of CDUS compared to clinical diagnosis was 10.2% (95% confidence interval [CI] 6.0–15.8%) and 99.8% (95% CI 98.9–99.9%) respectively. The sensitivity and specificity of a non-negative CDUS (i.e., either a positive or indeterminate result) was 47.3% (95% CI 39.5–55.2%) and 91.2% (95% CI 88.3–93.5%) respectively.

For TAB, the sensitivity and specificity compared to clinical diagnosis was 39.9% (95% CI 32.4–47.7%) and 100.0% (95% CI 97.5–100.0%) respectively. For the ACR/EULAR 2022 classification criteria, the sensitivity was 83.6% (95% CI 78.0–88.3) and specificity was 53.1% (95% CI 48.7–57.5%).

Eleven patients with GCA had an FDG-PET scan performed looking for large vessel vasculitis (LVV). There was evidence of LVV in five patients (45.5%) with none (0/5) of these patients having a positive CDUS and one (1/3) having a positive TAB.

Two patients with GCA had neither a CDUS nor a TAB. Their data had been collected prospectively as part of the CDUS referral process. Both had significant referral delays where the treating rheumatologist decided against further tests but the case was clinically consistent with GCA.

Incidence and prevalence of GCA in Waikato

The incidence of GCA was calculated for cases identified with TAB and also for cases identified clinically (i.e., on the experience of the rheumatologist with or without supporting imaging or histological data). The mean annual incidence (MAI) of biopsy positive GCA was 4.6 per 100,000 people over 50 years (95% CI 3.5–5.7). The MAI of clinical GCA was 14.7 per 100,000 people over 50 years (95% CI 12.7–16.6). The annual incidence for clinical GCA is reflected in Figure 1. The confidence intervals are overlapping and thus there does not appear to be any change in the annual incidence rate over the 9-year period. The estimate of the lower bound of prevalence of clinically positive GCA was 94.6 per 100,000 people over 50 years.

COVID-19

Aotearoa New Zealand had an initial lockdown period due to the COVID-19 pandemic in March–May 2020 and a subsequent lockdown in August–September 2021, when there was community transmission. During these periods, the service remained in place where patients and physicians had access to CDUS and TAB. Reviews occurred via telephone consultation unless the patient was unwell, in which case a face-to-face review was organised.

Mortality

There were 47 deaths in the cohort over the study period in patients with GCA. The mean age at time of death was 81.1 (SD 7.5) years with a range of 64–95 years of age. Females made up 59.6% of deaths. Mean time to death from referral was 29.1 (SD 23.6) months. The mean annual mortality rate was 38.4 per 1,000 cases over 50 years. The SMR was 1.18 (95% CI 0.83–1.52). Most deceased patients were of European descent (42/47, 89.4%) with four deaths in Māori patients (8.5%) and one not stated. The difference between deaths between Europeans and Māori was not statistically significant (p=0.27).

Infection was the most common cause of death, accounting for 16.7% (8/48) of deaths, followed by malignancy in 12.5% (6/48), cerebrovascular disease in 12.5% (6/48) and cardiovascular disease in 10.4% (5/48). Half of these deaths due to infection (4/8) occurred prior to the onset of the COVID-19 pandemic and of the four infection-related deaths in patients occurring in 2020 onwards, none were positive for COVID-19 at the time of their death. In addition to causes of mortality already described, there was one patient who died from an aortic dissection, one patient from venous thromboembolism (VTE) and two from peptic ulcer disease.

Discussion

This is the largest epidemiological study on GCA in Aotearoa New Zealand published to date and certainly the first to document the epidemiology of GCA in Waikato.

We report a MAI for clinically diagnosed GCA of 14.7 per 100,000 people over 50 years and a lower incidence of TAB-positive GCA of 4.6 per 100,000 people over 50 years. This latter finding is likely due to only 78.5% of GCA patients having a TAB performed, and only 39.9% being positive. There are three studies to date looking at the incidence of GCA in Aotearoa New Zealand and they cover Otago,¹⁴ Canterbury¹⁵ and Counties Manukau.¹⁶ Collectively, these report a MAI of biopsy-positive GCA of 10.5–12.73 per 100,000 people over 50 years^14,15 and a MAI of clinically positive GCA of 11.4–15 per 100,000 people over 50 years old.^15,16

Data on sensitivity and specificity for TAB and CDUS in Aotearoa New Zealand are only available in one of the three studies (Counties Manukau¹⁶), with a slightly higher sensitivity for TAB of 57% and CDUS 26%. Specificity is similar with 100% for TAB and 97% for CDUS.

This unity between four epidemiology studies in Aotearoa New Zealand supports the notion that cases of GCA are largely stable over the last decade. Certainly, over our study period in Waikato, cases of GCA did not appear to be rising.

World-wide, pooled incidences of GCA are 10.00 (95% CI 9.22–10.78) cases per 100,000 over 50 years, which is lower than our study but likely reflects the combination of high-risk countries (Scandinavian), moderate-risk countries (European, American) with low-risk countries (Asian) in different proportions to what we see in Aotearoa New Zealand.² It is likely, given Aotearoa New Zealand is made up predominantly of Europeans, that the epidemiology of GCA in our country will reflect other European countries.

The prevalence of GCA in our study was 94.6 per 100,000 over 50 years. This is also higher than the pooled prevalence of 51.7 per 100,000 people over 50 years from the meta-analysis by Li et al.,² yet very similar to the prevalence of 87.9 per 100,000 people over 50 years reported by a European study.¹⁷ It is likely, given Aotearoa New Zealand is made up predominantly of Europeans, that the epidemiology of GCA in our country will reflect other European countries. While the methods of the study were not dedicated to prevalence calculations, we believe that the duration of the study, life expectancy and average age of admittance means it is possible to make an estimate on the lower bound for prevalence.

Our study has similar baseline demographics compared to other New Zealand studies that have noted a female predominance of 65.5–71%^14–16 and mean age of 72.8–74.2 years.^14,15

Despite Māori and Pacific peoples making up 23.9% and 4.5% of the Waikato population respectively,¹² only 7.0% of our patients with GCA were Māori and none were of Pacific peoples origin. Other Aotearoa New Zealand studies have also demonstrated that GCA is uncommon in Māori and Pacific peoples.^15,16 One factor contributing to this may be the lower life expectancy in these groups¹² and the known risk of age and GCA.⁷ Asians make up 9.5% of the population of Waikato, yet we noted very few Asians with GCA, a finding consistent with the very low incidence in this group noted by other studies.⁴

Our study documented lower sensitivities of TAB and CDUS compared to those in Counties Manukau¹⁶ and other world-wide studies.^18–22 TAB is a difficult investigation to acquire in Waikato Hospital due to local referral issues and an under-resourced vascular department, where biopsies commonly occur outside of the optimal window. This may account for the lower positive biopsy rates. The lower sensitivity of CDUS was also seen in the Counties Manukau study and the cause remains unclear.¹⁶ A separate study analysing the efficacy of a fast-track pathway to investigate GCA in Waikato has noted that early commencement of corticosteroids results in fewer positive CDUS²³ scans, which has been noted in other studies as well.⁶ However, early corticosteroids are mandatory to prevent complications of disease. We also suspect that ultrasonographers have under-reported some positive CDUS scans, describing findings as “suggestive of GCA” rather than noting the presence of a halo sign. This has been noted on a separate retrospective review of several CDUS scans used in the data collection; however, the significance remains unclear as static ultrasound appears inferior to real-time ultrasound, though not formally evaluated in GCA.²⁴

The ACR/EULAR 2022 classification criteria also performed at lower sensitivity and specificity in our study compared to a recent review of their performance in the United States,²² which noted a sensitivity of 92.6% and specificity of 71.8%.

The mortality rate was 38.4 per 1,000 people over 50 years old. This compares to a pooled annual mortality rate of 20.4 (95% CI 17.8–23.0) cases per 1,000 people over 50 years.² While higher than the pooled mortality rate, it is similar to some of the European studies included in this meta-analysis and thus is probably consistent given our predominantly European population.

Our study noted no increase in the SMR for patients with GCA, which is concordant with literature. Several meta-analyses looking at all-cause mortality in GCA compared to the general population have found that the SMR is not increased with GCA.^25,26 However, on subgroup analysis in these studies, there appears to be an increase in mortality if patients are recruited through a hospital setting.²⁶ An increase in mortality is also noted during the first 2 years of treatment, which disappears after 10 years.^27,28

From literature, the leading cause of death in GCA is cardiovascular disease followed by cerebrovascular disease, infection and malignancy.²⁶ Our study showed causes of death were due to infection, followed by malignancy and cerebrovascular disease, and lastly by cardiovascular disease. This discrepancy may be due to the low numbers of deaths (n=47) and shorter duration of follow-up for patients recruited in the latter half of the study period, which may not reflect causes of death for patients who died later than 2023. Despite the study period crossing the COVID-19 pandemic, death due to COVID-19 was uncommon and unlikely to explain the higher number of infection-related deaths. Over half of deceased patients had a time from referral to death of under 2 years (53.2%), consistent with the known increased mortality risk noted in the first 2 years.^27,28

We acknowledge that patients being investigated in the private healthcare system were not captured in this study and that the private health community in Aotearoa New Zealand does account for a significant portion of the rheumatology workforce.²⁹ The access to CDUS in the Waikato private sector is limited and unlikely to occur at short notice except via the Waikato fast-track pathway, which was set up in 2013 to enable rapid access (i.e., same day or next working day) to CDUS for patients. Thus, most patients in the private sector are referred to the public system and would be captured in this dataset. It is unlikely that patients living out of Waikato would be investigated by the Waikato Hospital because access to healthcare was usually strictly limited to District Health Board zones. This may not be the case for private cases of GCA, which may have crossed boundaries. Overall, it is likely that we have recorded most cases of GCA over a 9-year period. Other minor limitations of this study include its retrospective nature along with some missing data due to inadequate documentation of clinical symptoms.

Conclusion

The epidemiology of GCA in Waikato is comparable to other epidemiological studies in Aotearoa New Zealand, reflecting a stability in the incidence of GCA nationally over the past decade. GCA appears to be less common in Māori, Pacific peoples and Asian patient groups and occurs at a younger age in Māori patients compared to European. We noted no increase in mortality, with common causes of mortality being infection followed by malignancy and cerebrovascular disease.