- Email: [email protected]
Incidence of outcomes potentially associated with corticosteroid therapy in patients with giant cell arteritis
Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Incidence of outcomes potentially associated with corticosteroid therapy in patients with giant cell arteritis☆ Jessica C. Wilson, PhDa, Khaled Sarsour, PhD, MPHb, Neil Collinson, PhDc, Katie Tuckwell, PhDc, David Musselman, MDb, Micki Klearman, MDb, Pavel Napalkov, MDb, Susan S. Jick, DScd, John H. Stone, MD, MPHe, Christoph R. Meier, PhD, MSa,d,f,n a
Basel Pharmacoepidemiology Unit, University of Basel, Basel, Switzerland Genentech, South San Francisco, CA, USA Roche Products Ltd., Welwyn Garden City, UK d Boston Collaborative Drug Surveillance Program, Boston University School of Public Health, Lexington, MA, USA e Massachusetts General Hospital Rheumatology Unit, Harvard Medical School, Boston, MA, USA f Hospital Pharmacy, University Hospital Basel, Basel, Switzerland b c
a r t i c l e in fo
Keywords: Corticosteroids Giant cell arteritis Health services research Outcomes research Serious adverse events
a b s t r a c t Objective: Giant cell arteritis (GCA) is an inflammatory disorder of blood vessels that preferentially affects large- and medium-sized arteries. High-dose oral corticosteroids (CS) are the mainstay of GCA therapy. Using data from the UK Clinical Practice Research Datalink, we quantified and compared the incidence of selected potentially CS-associated adverse outcomes in patients with and without GCA. Methods: We conducted a retrospective follow-up study of GCA and non-GCA patients to examine the incidence of adverse outcomes attributable to CS use. Eligibility criteria for the GCA group included a first-time diagnosis of GCA at age 50 years or older with receipt of Z1 prescription(s) for prednisolone. GCA patients were matched to a GCA-free comparison group of equal size on age, sex, general practice, and calendar time. We estimated incidence rates and incidence rate ratios (IRRs) for diabetes, osteoporosis, glaucoma, fractures, serious infection requiring hospitalization, and death for GCA and non-GCA patients and compared all-cause hospitalizations between the two groups. Results: The cohort consisted of 5011 GCA and 5011 matched non-GCA patients. Approximately 74% were women, and mean age at GCA diagnosis was 72.9 years. The IR for all outcomes was greater in the GCA group than the non-GCA group. IRRs [95% confidence intervals (CIs)] were as follows: diabetes 1.4 (1.2– 1.7), osteoporosis 2.4 (2.1–2.8), fractures 1.4 (1.2–1.6), glaucoma 2.0 (1.6–2.5), serious infection requiring hospitalization 1.5 (1.3–1.7), and death 1.2 (1.0–1.3). Conclusion: Compared with age- and sex-matched non-GCA patients, patients with GCA were at increased risk for diabetes, osteoporosis, fracture, and glaucoma and at a marginally increased risk for death. & 2016 Elsevier Inc. All rights reserved.
Introduction
☆
Sarsour, Musselman, Klearman, and Napalkov are employees of Genentech, a member of the Roche Group. Collinson and Tuckwell are employees of Roche Products Ltd. Stone are Principal Investigator in a Roche-funded clinical trial of a potential steroid-sparing agent in GCA. Wilson, Jick, and Meier have nothing to disclose. F. Hoffmann-La Roche Ltd. developed the initial concept for the study and provided funding for the research. n Corresponding author at: Basel Pharmacoepidemiology Unit Hospital Pharmacy, University Hospital Basel University of Basel, Spitalstrasse 26, CH-4031 Basel, Switzerland. E-mail address: [email protected] (C.R. Meier). http://dx.doi.org/10.1016/j.semarthrit.2016.10.001 0049-0172/& 2016 Elsevier Inc. All rights reserved.
Giant cell arteritis (GCA) is a form of primary systemic vasculitis that predominantly affects large and medium-sized arteries, with a predilection for the aorta and vessels to the head and neck. Vascular involvement in GCA, characterized by granulomatous inflammation of the aorta and its main branches, results in arterial narrowing and blood flow obstruction [1,2]. The disease almost exclusively affects persons older than 50 years of age, with incidence rates (IRs) peaking in the eighth decade of life [3–5]. Women are more likely to develop GCA than men. To date, the etiology of GCA is unknown. Genetic and environmental factors
2
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
are thought to be involved, and it is widely accepted that both the adaptive and the innate immune systems play important roles in the pathophysiology of this disease [3]. High-dose systemic corticosteroids (CS), predominantly prednisolone in the United Kingdom, are the cornerstone of GCA therapy. The duration of CS treatment varies by patient, but, because of disease relapses when patients are on or off treatment, the mean therapy duration is 5–6 years [6]. Long-term CS use is associated with a variety of adverse events (AEs), including weight gain, osteoporosis, fractures, hypertension, diabetes, cataracts, and myopathy [7,8]. Few studies, however, have examined the incidence of serious AEs specifically in GCA patients, a patient population often characterized by advanced age and many attendant comorbidities. A recent UK population-based study that examined the incidence and characteristics of GCA reported a variety of prevalent comorbidities associated with GCA, including osteoporosis, infections, cardiovascular diseases, diabetes, hypokalemia, and intraocular lens replacement. Many of these comorbidities are related to CS use [9]. We investigated the IRs of a series of adverse outcomes considered to be associated with prolonged CS therapy in GCA patients. We focused on diabetes, osteoporosis, fractures, glaucoma, and infections requiring hospitalization and compared the rates of these events with rates of the same events occurring in non-GCA patients. We also assessed mortality rates and the incidence of all-cause hospitalizations among GCA patients and a non-GCA comparison group.
Methods
their GCA diagnosis. We defined study entry as the date of this first prescription. For the comparison group, we randomly identified patients without GCA who were matched 1:1 to the GCA patients on age (year of birth), sex, general practice, study entry year, and years of history in the CPRD before study entry. All patients were required to have at least 3 years of recorded medical history before entry. Follow-up (i.e., study entry) began on the date of first prednisolone prescription for each GCA patient and the same date for the corresponding matched patient in the GCA-free comparison group. We excluded patients in both the GCA and the comparison groups if they had a diagnosis of any type of cancer, alcoholism, drug abuse, or HIV diagnosis before study entry. To further validate the identified GCA group, we conducted a sensitivity analysis using a more stringent, previously validated definition of GCA as described by Smeeth et al. [12]. Here GCA Table 1 Baseline cohort characteristics for GCA and non-GCA comparison patients Characteristics
GCA group, n ¼ 5011
Non-GCA group, n ¼ 5011
Sex Male Female
1298 (25.9) 3713 (74.1)
1298 (25.9) 3713 (74.1)
Age group, years 50–59 60–69 70–79 80–89 90þ
470 1205 2071 1158 108
466 1208 20 1161 107
(9.4) (24.0) (41.3) (23.1) (2.2)
(9.3) (24.1) (41.3) (23.2) (2.1)
χ2
–
–
Data sources
Mean age (SD)
The Clinical Practice Research Datalink (CPRD) provides anonymized health care information on approximately 8 million patients in the United Kingdom, with data extending back to 1987. Information on demographics (age, sex, weight, and height), consultation outcomes and diagnoses, specialist referrals, details on all prescribed medications, and lifestyle factors (e.g., tobacco and alcohol use) is recorded by specially trained general practitioners in the CPRD. READ codes, a disease coding system utilized in the UK National Health Service (NHS), are used to classify medical diagnoses. Recorded information on diagnoses and drug exposure has been validated repeatedly and has proven to be of high quality [10,11]. The study protocol was reviewed and approved by the Independent Scientific Advisory Committee for Medicines and Healthcare Products Regulatory Agency Database Research. Hospital Episode Statistics (HES) data contain details of all patient admissions to NHS hospitals in England only (not Northern Ireland, Scotland, or Wales), constituting approximately 55% of patients in the CPRD. The data collected include patient demographics and clinical and administrative details coded using the International Classification of Diseases, 10th Revision (ICD-10) system. The most up-to-date HES data available were used for this study, covering the years 1997–2012.
BMI 12–18.4 18.5–24.9 25–29.9 30–60 Missing
99 1624 1653 891 744
(2.0) (32.3) (33.0) (17.8) (14.9)
92 1596 1575 876 872
(1.8) (31.9) (31.4) (17.5) (17.4)
0.0131
Smoking status Never Current Former Missing
2415 778 1479 339
(48.2) (15.5) (29.5) (6.8)
2517 665 1362 467
(50.2) (13.3) (27.2) (9.3)
o0.0001
Alcohol use Never Current Former Missing
1244 3102 81 584
(24.8) (61.9) (1.6) (11.7)
1058 3158 93 702
(21.1) (63.0) (1.9) (14.0)
o0.0001
Comorbidity Rheumatologic disease Polymyalgia rheumatica Renal disease Peripheral vascular disease Peptic ulcer disease Myocardial infarction Mild liver disease Moderate liver disease Hemiplegia Diabetes Diabetes with complications Dementia Congestive heart disease Chronic pulmonary disease Cerebrovascular disease
1090 911 446 296 299 305 1 32 25 516 119 32 300 1249 521
(21.8) (18.2) (8.9) (5.9) (6.0) (6.1) (0.02) (0.64) (0.5) (10.3) (2.4) (0.6) (6.0) (24.9) (10.4)
274 127 358 224 245 255 2 12 21 484 105 58 228 952 376
(5.5) (2.5) (7.1) (4.5) (4.9) (5.1) (0.04) (0.24) (0.4) (9.7) (2.1) (1.4) (4.6) (19.0) (7.5)
o0.0001 o0.0001 0.0012 0.0012 0.0173 0.0297 0.5636 0.0025 0.5544 0.2862 0.3441 0.0003 0.0013 o0.0001 o0.0001
Average number comorbidities
1.03
Study population The study population was composed of an initial cohort of all patients in the CPRD 50 years of age and older with a first-time READ code for GCA between January 1, 1995, and August 31, 2013. In order to increase the proportion of validated and confirmed GCA cases in the cohort, patients were required to have at least one recorded prednisolone prescription at or within the 6 months after
72.9 (9.1)
72.9 (9.1)
0.70
–
–
BMI, body mass index; GCA, giant cell arteritis. Values in bold are statistically significant.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
patients were required to have received at least two prednisolone prescriptions, one prescription at or within the 6 months following their GCA diagnosis, with the second prescription being within 6 months of the first. As with the initial GCA group, a comparison group was matched 1:1 to the “twice criteria” GCA group.
3
We calculated crude IRs with 95% CIs for all outcomes of interest among the GCA group and the non-GCA comparison group, stratified by age and sex. We also estimated crude incidence rate ratios (IRRs) with 95% CIs.
Results Follow-up analysis of outcomes of interest Study population and characteristics We followed up all patients in the study population and accumulated person-time from study entry until the first event in the following list of events: diagnosis of an outcome of interest, death, end of the patient's medical record, or end of the study (August 31, 2013). Outcomes of interest were incident diabetes or a new prescription for a drug to treat diabetes, incident osteoporosis, incident glaucoma, or a newly prescribed therapy to lower intraocular pressure, bone fracture, serious infection requiring hospitalization, all-cause hospitalizations, and death (Supplementary Table S1). We examined each outcome separately; thus, a patient could have been included in more than one analysis. For the subset of patients with HES linkage, we combined CPRD and HES data in order to assess “all-cause hospitalizations” and “infections requiring hospitalization.” We excluded patients from this analysis if they had an entry date outside the dates of available HES data (and the corresponding date in the matched non-GCA group). To be included in the final HES-linked cohort, both GCA and matched non-GCA comparison patients were required to have available HES data linkage, thus ensuring equal access to outcomes information for both groups. For the outcomes diabetes, osteoporosis, and glaucoma, we established separate sub-cohorts from which we excluded GCA and non-GCA patients if they had a previous diagnosis of the outcome of interest before study entry. Definitions of all outcomes are provided in Supplementary Table S1.
We identified 5011 patients 50 years of age and older who had an incident diagnosis of GCA during the study period. All GCA patients received at least one prednisolone prescription within the 6 months after the GCA diagnosis. We also identified at random a matched comparison group of 5011 patients with no GCA diagnosis. Table 1 displays the basic characteristics of patients included in the study population. Most (74%) GCA patients were women, and the mean age at disease onset was 72.9 years. The median number of prednisolone prescriptions prescribed to GCA patients during follow-up was 17 (Q1 ¼ 6, Q3 ¼ 38). Less than 18% of nonGCA patients had a prednisolone prescription, the median number during follow-up was 3 (Q1 ¼ 1, Q3 ¼ 12). Figure 1 displays the final numbers for each of the separate cohorts following the exclusion of patients with a prevalent outcome. Incidence of the clinical outcomes of interest During follow-up, 594 patients (6.6%) in the GCA and the nonGCA groups combined developed diabetes, 778 (8.4%) received an osteoporosis diagnosis, 387 (4.2%) had an incident glaucoma diagnosis, 733 (7.3 %) sustained a fracture, 822 (8.2%) were hospitalized for a severe infection, and 1234 (12.3%) died. In the GCA group, the median time (in years) for the occurrence of diabetes, glaucoma, osteoporosis, infection, fractures, and death
Initial Cohort 5,011 GCA group 5,011 Non-GCA group
Diabetes Cohort 4,479 GCA group 4,517 Non-GCA group 532 GCA patients and 494 non-GCA patients with a history of diabetes were excluded Osteoporosis Cohort 4,567 GCA group 4,681 Non-GCA group 444 GCA patients and 330 non-GCA patients with a history of osteoporosis were excluded
Glaucoma Cohort 4,600 GCA group 4,662 Non-GCA group 411 GCA patients and 349 non-GCA patients with a history of glaucoma were excluded Fractures and Death Cohorts 5,011 GCA group 5,011 Non-GCA group No previous exclusions
HES-Linked Cohort 2,491 GCA group 2,491 Non-GCA group No previous exclusions
Fig. 1. Exclusions before follow-up.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
4
Table 2 IRs (per 1000 person-years) and IRR with 95% CIs for all outcomes of interest in GCA and non-GCA patients GCA group Outcomes, n
Non-GCA group PT, years
IR (95% CI)
Outcomes, n
PT, years
IR
IRR (95% CI)
Diabetes Male Female
340 85 255
24,017.8 5646.3 18,371.5
14.2 (12.7–15.7) 15.1 (12.2–18.6) 13.9 (12.3–15.7)
254 75 179
25,395.7 6055.9 19,339.8
10.0 (8.9–1.3) 12.4 (9.9–15.5) 9.3 (8.0–10.7)
1.4 (1.2–1.7) 1.2 (0.9–1.7) 1.5 (1.2–1.8)
Osteoporosis Male Female
532 63 469
23,203.1 6409.5 16,793.6
22.9 (21.1–24.9) 9.8 (7.7–12.6) 27.9 (25.5–30.5)
246 28 218
26,124.4 6932.0 19,192.4
9.4 (8.3–10.7) 4.0 (2.8–5.8) 11.4 (10.0–13.0)
2.4 (2.1–2.8) 2.4 (1.6–3.8) 2.5 (2.1–2.9)
Fractures Male Female
433 57 376
26,227.3 6605.4 19,621.9
16.5 (15.0–18.1) 8.6 (6.7–11.2) 19.2 (17.3–21.2)
340 29 311
27,707.9 7064.5 20,643.4
12.3 (11.0–13.6) 4.1 (2.9–5.9) 15.1 (13.5–16.8)
1.4 (1.2–1.6) 2.1 (1.3–3.3) 1.3 (1.1–1.5)
Glaucoma Male Female
253 60 193
24,837.9 6,000.9 18,837.0
10.2 (9.0–11.5) 10.0 (7.8–12.9) 10.3 (8.9–11.8)
134 37 97
26,451.0 6491.4 19,959.6
5.1 (4.3–6.0) 5.7 (4.1–7.9) 4.9 (4.0–5.9)
2.0 (1.6–2.5) 1.8 (1.2–2.6) 2.1 (1.7–2.7)
Serious infection Male Female
476 120 356
12,559.8 3099.4 9460.5
37.9 (34.7–41.4) 38.7 (32.5–46.1) 37.6 (34.0–41.7)
346 103 243
13,454.1 3331.0 10,123.1
25.7 (23.2–28.5) 30.9 (25.6–37.4) 24.0 (21.2–27.2)
1.5 (1.3–1.7) 1.3 (1.0–1.6) 1.6 (1.3–1.9)
Mortality Male Female
653 197 456
27,932.0 6826.6 21105.3
23.4 (21.7–25.2) 28.9 (25.1–33.1) 21.6 (19.7–23.7)
581 196 385
28,724.6 7117.8 21,606.8
20.2 (18.7–21.9) 27.5 (24.0–31.6) 17.8 (16.1–19.7)
1.2 (1.0–1.3) 1.1 (0.9–1.3) 1.2 (1.1–1.4)
2101 548 1553
4587.0 1114.4 3472.6
1876 492 1384
6758.5 1663.1 5095.4
Hospitalization Male Female
458.0 (443.6–472.5) 491.7 (462.5–521.01 447.2 (430.7–463.8)
277.6 (267.0–288.4) 295.8 (274.4–318.2) 271.6 (259.6–284.0)
1.7 (1.6–1.8) 1.7 (1.5–1.9) 1.7 (1.5–1.8)
CI, confidence interval; GCA, giant cell arteritis; IR, incidence rate; IRR, incidence rate ratio; PT, person-time. Values in bold are statistically significant.
was 1, 1.5, 2, 2, 3, and 4 respectively. The incidence of all outcomes was higher in the GCA group than in the non-GCA comparison group (Table 2). All IRRs for diabetes, osteoporosis, glaucoma, fracture, infection requiring hospitalization, and death were between 1.2 and 2.0 and were significantly higher in the GCA group than the non-GCA cohort (Table 2). The sensitivity analysis revealed marginal difference in the reported IRs and IRRs for the outcomes in the more stringent “twice criteria” GCA group (Supplementary Table S2). After stratification of the analyses by sex, there was little evidence of effect modification for any outcome except fracture (Table 2). The IRR for fracture in patients with GCA compared with those without GCA was substantially higher for men (IRR ¼ 2.1, 95% CI: 1.3–3.3) than for women (IRR ¼ 1.3, 95% CI: 1.1–1.5). In the GCA and non-GCA group, a greater proportion of females (77% GCA:25% non-GCA) were prescribed osteoporosis medications compared to males (63% GCA:11% non-GCA). IRs for most outcomes across most age groups were higher for GCA patients than for the relevant non-GCA patients (Figs. 2 and 3). Incidence of all-cause hospitalization During follow-up, 2101 GCA patients and 1876 non-GCA patients each had at least one hospitalization. GCA patients were at greater risk for hospitalization than non-GCA patients (IRR ¼ 1.7, 95% CI: 1.6–1.8). The risk was similar in men and women (Table 2). Based on ICD-10 classifications, more patients in the GCA group than the non-GCA group (overall P o 0.001) were hospitalized for diseases of the circulatory system (ICD-10 codes I00I99) (26.0% vs 18.9%), digestive system (ICD-10 codes K00-K93) (24.9% vs 21.1%), eye and adnexa (ICD-10 codes H00-H59) (20.5% vs 17.0%), respiratory system (ICD-10 codes J00-J99) (15.3% vs 11.4%), endocrine, nutritional, and metabolic systems (ICD-10 codes E00E90) (3.7% vs 2.2%), and blood and blood-forming organs,
including disorders involving the immune mechanism (ICD-10 codes D50-D89) (4.8% vs 3.1%).
Discussion In this retrospective population-based follow-up study, the IRs of diabetes, glaucoma, osteoporosis, fractures, and serious infections were all higher in GCA patients than in a non-GCA comparison group. Increased risks for adverse outcomes such as diabetes, serious infections, and fractures have been reported to be between 1.3 and 3 times higher in GCA patients than in non-GCA patients [6,9], which is consistent with IRRs found in the current study. It is well recognized that CS therapy increases the risk for these outcomes and that these outcomes are often attributable to AEs resulting from long-term CS therapy. Increased risk for diabetes, fractures, osteoporosis, infection, and gastrointestinal effects has been reported in patients receiving prolonged CS treatment [12–16]. A small-scale, retrospective, US study in 120 GCA patients found that a substantial proportion of patients treated with CS (86%) developed AEs attributable to CS therapy [6]. Few studies have examined the occurrence of serious AEs of CS therapy by sex in GCA patients. The IRRs for diabetes, glaucoma, serious infections, and death in GCA patients compared with nonGCA patients did not differ substantially between women and men. In contrast to other studies that have suggested greater risk for osteoporosis and CS-related fragility fractures among women [17,18]; however, our study found the IRR for fracture was considerably higher among men. Given that men are generally considered to have fewer risk factors for insufficiency fracture, our finding is consistent with the concept that the presence of one major risk factor in GCA patients compared with non-GCA patients —namely, CS use—disproportionately increases the fracture risk for men. The United Kingdom has no recognized population screening
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
B
20 15 10 5 0
80 70 60 50 40 30 20 10 0
16
70
14
60
12
50
10 8 6 4 2 0
50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
40 30 20 10
Number of Glaucoma Cases
25
90
Incidence Rate (1,000 person-years)
30
Number of Diabetes Cases
Incidence Rate (1,000 person-years)
A
5
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
30
140
20
100
15
80
10
60 40
5 0
20 0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
35
140
30
120
25
100
20
80
15
60
10
40
5
20
0
Number of Fracture Cases
120
Number of Osteoporosis Cases
25
160
Incidence Rate (1,000 person-years)
D Incidence Rate (1,000 person-years)
C
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
Fig. 2. IRs (per 1000 person-years) and number of cases in GCA and non-GCA patients stratified by age at GCA diagnosis for (A) diabetes, (B) glaucoma, (C) osteoporosis, and (D) fractures. GCA, giant cell arteritis; IR, incidence rate.
system in place to identify patients who have osteoporosis or who are at high risk for fracture. Instead, osteoporosis is identified opportunistically through a case-finding strategy based on history of fragility fracture or on the presence of clinical risk factors that include but are not limited to age, sex, and current glucocorticoid treatment for more than 3 months [19]. The high IRR observed in male GCA patients may highlight the need for additional clinical awareness of fracture risk in this patient group. An increased incidence of all-cause hospitalizations was observed in GCA patients compared with non-GCA patients. The most common causes of hospitalization in the GCA group were diseases of the cardiovascular system, closely followed by diseases of the digestive system and of the eyes. It appears likely that this observation reflects both steroid-related AEs and the pathophysiology of GCA, in which increased risks for cardiovascular disease [9,20] and vision impairment [4,21] have been observed independently of CS use in patients with GCA. We observed a modest increase in the risk for death in GCA patients compared with the matched non-GCA group. These results are consistent with most, but not all, studies in the literature that have generally demonstrated mortality rates comparable to or slightly higher than those of the general population [22–26] and that have reported increased mortality rates among GCA patients in the first years after GCA diagnosis [26–29]. Compared with the comparison group, GCA patients had a higher frequency of comorbidities before study entry, which might have contributed to the observed increased risk for mortality. Our study has a number of important strengths. The CPRD is a well-respected primary care database of high quality and completeness [10,11,30]. We were rigorous in our assessment of
incident cases of GCA to prevent misclassification, and strict exclusion criteria were applied to ensure that everyone included had a valid GCA diagnosis. The finding of little difference in results of the additional sensitivity analysis comprising patients with a stricter GCA definition further validates our GCA group and supports the presented findings. Moreover, this is one of few studies to examine the incidence of outcomes likely associated with prolonged CS therapy by sex. Finally, previous studies of GCA with regard to the development of serious AEs have been small, each involving fewer than 150 GCA patients [6,31–38]. The current study, which includes data on more than 5000 GCA patients, is one of the largest to date to examine the incidence of multiple, potentially CS-related effects in patients with this disease. The findings add further weight to the perceived potential clinical burden of prolonged CS treatment in GCA patients. Our study also has some potential limitations. There might have been a bias for osteoporosis diagnosis in the GCA group because of an increased likelihood of referring patients with GCA for fracture risk assessment. This potential bias might have been minimized, however, because the incidence of GCA peaks between 70 and 80 years of age, coinciding with the recommended age for fracture risk assessment: 65 years of age and older for women and 75 years of age and older for men [39]. Similar considerations apply to a potential bias in the detection of glaucoma. We did not exclude non-GCA patients who had a prednisolone prescription during follow-up; however, the percentage of users was low ( o 18%). In addition, the median number of prescriptions during follow-up was considerably lower than in the GCA exposed patients and was likely to have minimal influence on the reported IRs in the nonGCA group.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
6
Incidence Rate (1,000 person-years)
120
90 80
100
70
80
60 50
60
40
40
30 20
20
10
0
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+
Cases GCA
Age Group, years Cases Non-GCA Rate GCA
Number of Serious Hospital Infection Cases
A
Rate Non-GCA
90
180
80
160
70
140
60
120
50
100
40
80
30
60
20
40
10
20
Number of Death Cases
Incidence Rate (1,000 person-years)
B
0
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+
Cases GCA
Age Group, years Cases Non-GCA Rate GCA
Rate Non-GCA
Fig. 3. IRs (per 1000 person-years) and number of cases in GCA and non-GCA patients stratified by age at GCA diagnosis for (A) serious infections and (B) death outcomes of interest. GCA, giant cell arteritis; IR, incidence rate.
In conclusion, the findings of this large population-based study indicate that patients with GCA are at increased risk for diabetes, osteoporosis, and glaucoma, are more likely to sustain a fracture, and are more likely to require all-cause hospitalization than are patients without GCA. The risk for death was found to be only marginally higher in patients with GCA than in an age- and sexmatched comparison group.
Author contributions All authors reviewed and revised the manuscript drafts and approved the final version for publication. CRM had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Study conception and design: All authors. Drafting the manuscript: Wilson, Jick, and Meier.Acquisition of data: Wilson, Jick, and Meier. Analysis and interpretation of data: All authors.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.semarthrit.2016. 10.001.
References [1] Seetharaman M, Paget S, Leibowitz E. Giant Cell Arteritis. 〈http://emedicine. medscape.com/article/332483-overview〉; 2012. [2] Gonzalez-Gay MA, Martinez-Dubois C, Agudo M, Pompei O, Blanco R, Llorca J. Giant cell arteritis: epidemiology, diagnosis, and management. Curr Rheumatol Rep 2010;12:436–42, http://dx.doi.org/10.1007/s11926-010-0135-9. [3] Chew SSL, Kerr NM, Danesh-Meyer HV. Giant cell arteritis. J Clin Neurosci 2009;16:1263–8, http://dx.doi.org/10.1016/j.jocn.2009.05.002. [4] Gonzalez-Gay MA, Vazquez-Rodriguez TR, Lopez-Diaz MJ, et al. Epidemiology of giant cell arteritis and polymyalgia rheumatica. Arthritis Rheum 2009;61:1454–61, http://dx.doi.org/10.1002/art.24459. [5] Villa-Forte A. Giant cell arteritis: suspect it, treat it promptly. Cleve Clin J Med 2011;78:265–70, http://dx.doi.org/10.3949/ccjm.78a.10131. [6] Proven A, Gabriel SE, Orces C, O'Fallon WM, Hunder GG. Glucocorticoid therapy in giant cell arteritis: duration and adverse outcomes. Arthritis Rheum 2003;49:703–8, http://dx.doi.org/10.1002/art.11388. [7] Kaushik P, Cote D. Giant cell arteritis. Indian J Rheumatol 2009;4:20–4, http: //dx.doi.org/10.1016/S0973-3698(10)60157-4. [8] Royal College of Physicians. Diagnosis and management of giant cell arteritis: concise guidance to good practice series; 2010. [9] Petri H, Nevitt A, Sarsour K, Napalkov P, Collinson N. Incidence of giant cell arteritis and characteristics of patients: data-driven analysis of comorbidities. Arthritis Care Res (Hoboken) 2014. http://dx.doi.org/10.1002/acr.22429. [10] Jick H, Jick SS, Derby LE. Validation of information recorded on general practitioner based computerised data resource in the United Kingdom. Br Med J 1991;302:766–8, http://dx.doi.org/10.1136/bmj.302.6779.766. [11] Khan NF, Harrison SE, Rose PW. Validity of diagnostic coding within the General Practice Research Database: a systematic review. Br J Gen Pract 2010;60:e128–36, http://dx.doi.org/10.3399/bjgp10X483562. [12] Smeeth L, Cook C, Hall AJ. Incidence of diagnosed polymyalgia rheumatica and temporal arteritis in the United Kingdom, 1990–2001. Ann Rheum Dis 2006;65:1093–8, http://dx.doi.org/10.1136/ard.2005.046912. [13] Saag KG, Koehnke R, Caldwell JR, et al. Low dose long-term corticosteroid therapy in rheumatoid arthritis: an analysis of serious adverse events. Am J Med 1994;96:115–23, http://dx.doi.org/10.1016/0002-9343(94)90131-7. [14] Blackburn D, Hux J, Mamdani M. Quantification of the Risk of Corticosteroidinduced Diabetes Mellitus Among the Elderly. J Gen Intern Med. 2002;17:717–20. [15] Van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C. Oral corticosteroids and fracture risk: relationship to daily and cumulative doses. Rheumatology (Oxford) 2000;39:1383–9. 〈http://www.ncbi.nlm.nih.gov/pubmed/11136882〉. [16] Van Staa TP, Cooper C, Leufkens HG, Bishop N. Children and the risk of fractures caused by oral corticosteroids. J Bone Miner Res 2003;18:913–8, http://dx.doi.org/10.1359/jbmr.2003.18.5.913. [17] Gabriel SE, Sunku J, Salvarani C, O'Fallon WM, Hunder GG. Adverse outcomes of antiinflammatory therapy among patients with polymyalgia rheumatica. Arthritis Rheum 1997;40:1873–8, http://dx.doi.org/10.1002/1529–0131 (199710)40:10o1873:AID-ART22 43.0.CO;2-V. [18] Tobón GJ, Youinou P, Saraux A. The environment, geo-epidemiology, and autoimmune disease: rheumatoid arthritis. Autoimmun Rev 2010;9:A288–92, http://dx.doi.org/10.1016/j.autrev.2009.11.019. [19] Souza A, Okamoto K, Abrantes F, Schau B, Bacchiega AB, Shinjo SK. Giant cell arteritis: a multicenter observational study in Brazil. Clinics 2013;68:317–22, http://dx.doi.org/10.6061/clinics/2013(03)OA06. [20] National Osteoporosis Guideline Group. Guideline for the diagnosis and management of osteoporosis in postmenopausal women and men from the age of 50 years in the UK. 〈http://www.shef.ac.uk/NOGG/NOGG_Pocket_Gui de_for_Healthcare_Professionals.pdf〉; 2014. [21] Nordborg E, Nordborg C. Giant cell arteritis: epidemiological clues to its pathogenesis and an update on its treatment. Rheumatology (Oxford) 2003;42:413–21. 〈.http://www.ncbi.nlm.nih.gov/pubmed/12626790〉 [accessed 10.02.15]. [22] Loddenkemper T, Sharma P, Katzan I, Plant GT. Risk factors for early visual deterioration in temporal arteritis. J Neurol Neurosurg Psychiatry 2007;78:1255–9, http://dx.doi.org/10.1136/jnnp.2006.113787. [23] Gran JT, Myklebust G, Wilsgaard T, Jacobsen BK. Survival in polymyalgia rheumatica and temporal arteritis: a study of 398 cases and matched population controls. Rheumatology (Oxford) 2001;40:1238–42. 〈.http:// www.ncbi.nlm.nih.gov/pubmed/11709607〉 [accessed 10.02.15]. [24] Doran MF, Crowson CS, O’Fallon WM, Hunder GG, Gabriel SE. Trends in the incidence of polymyalgia rheumatica over a 30 year period in Olmsted County, Minnesota, USA. J Rheumatol 2002;29:1694–7. 〈.http://www.ncbi.nlm.nih. gov/pubmed/12180732〉 [accessed 10.02.15]. [25] González-Gay MA, Blanco R, Abraira V, et al. Giant cell arteritis in Lugo, Spain, is associated with low longterm mortality. J Rheumatol 1997;24:2171–6. 〈http://www.ncbi.nlm.nih.gov/pubmed/9375880〉 [accessed 10.02.15]. [26] Matteson EL, Gold KN, Bloch DA, Hunder GG. Long-term survival of patients with giant cell arteritis in the American College of Rheumatology giant cell arteritis classification criteria cohort. Am J Med 1996;100:193–6. 〈http://www. ncbi.nlm.nih.gov/pubmed/8629654〉 [accessed 10.02.15]. [27] Ninan J, Nguyen AM, Cole A, et al. Mortality in patients with biopsy-proven giant cell arteritis: a south australian population-based study. J Rheumatol 2011;38:2215–7, http://dx.doi.org/10.3899/jrheum.101254.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]] [28] Baslund B, Helleberg M, Faurschou M, Obel N. Mortality in patients with giant cell arteritis. Rheumatology (Oxford) 2015;54:139–43, http://dx.doi.org/ 10.1093/rheumatology/keu303. [29] Mohammad AJ, Nilsson J-Å, Jacobsson LT, Merkel PA, Turesson C. Incidence and mortality rates of biopsy-proven giant cell arteritis in southern Sweden. Ann Rheum Dis 2014. http://dx.doi.org/10.1136/annrheumdis-2013-204652. [30] Uddhammar A, Eriksson A-L, Nyström L, Stenling R, Rantapää-Dahlqvist S. Increased mortality due to cardiovascular disease in patients with giant cell arteritis in northern Sweden. J Rheumatol 2002;29:737–42. 〈http://www.ncbi. nlm.nih.gov/pubmed/11950015〉 [accessed 10.02.15]. [31] Jick SS, Kaye JA, Vasilakis-Scaramozza C, et al. Validity of the general practice research database. Pharmacotherapy 2003;23:686–9. 〈http://www.ncbi.nlm. nih.gov/pubmed/12741446〉 [accessed 31.10.14]. [32] Crow RW, Katz BJ, Warner JEA, et al. Giant cell arteritis and mortality. J Gerontol A Biol Sci Med Sci 2009;64:365–9, http://dx.doi.org/10.1093/gerona/gln030. [33] Kyle V, Hazleman BL. Treatment of polymyalgia rheumatica and giant cell arteritis. II. relation between steroid dose and steroid associated side effects. Ann Rheum Dis 1989;48:662–6, http://dx.doi.org/10.1136/ard.48.8.662. [34] Nesher G, Sonnenblick M, Friedlander Y. Analysis of steroid related complications and mortality in temporal arteritis: a 15-year survey of 43 patients.
[35]
[36]
[37]
[38]
[39]
7
J Rheumatol 1994;21:1283–6. 〈http://www.ncbi.nlm.nih.gov/pubmed/7966070〉 [accessed 27.10.14]. Fainaru M, Friedman G, Friedman B. Temporal arteritis in Israel. A review of 47 patients. J Rheumatol 1979;6:330–5. 〈http://www.ncbi.nlm.nih.gov/pubmed/ 490527〉 [accessed 26.02.16]. Graham E, Holland A, Avery A. Prognosis in giant-cell arteritis. Br Med J (Clin Res Ed) 1981;282:269–71. 〈http://www.pubmedcentral.nih.gov/articlerender. fcgi?artid=1504093&tool=pmcentrez&rendertype=abstract〉 [accessed 26.02.16]. Khalifa M, Karmani M, Jaafoura NG, et al. Epidemiological and clinical features of giant cell arteritis in Tunisia. Eur J Intern Med 2009;20:208–12, http://dx. doi.org/10.1016/j.ejim.2008.07.030. Alba MA, García-Martínez A, Prieto-González S, et al. Relapses in patients with giant cell arteritis: prevalence, characteristics, and associated clinical findings in a longitudinally followed cohort of 106 patients. Medicine (Baltimore) 2014;93:194–201, http://dx.doi.org/10.1097/MD.0000000000000033. National Institute for Health and Care Excellence (NICE). NICE Clinical Guideline CG146. Osteoporosis: assessing the risk of fragility fracture. 〈http:// publications.nice.org.uk/osteoporosis-assessing-the-risk-of-fragility-fractur e-cg146/guidance〉; 2013.
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Incidence of outcomes potentially associated with corticosteroid therapy in patients with giant cell arteritis☆ Jessica C. Wilson, PhDa, Khaled Sarsour, PhD, MPHb, Neil Collinson, PhDc, Katie Tuckwell, PhDc, David Musselman, MDb, Micki Klearman, MDb, Pavel Napalkov, MDb, Susan S. Jick, DScd, John H. Stone, MD, MPHe, Christoph R. Meier, PhD, MSa,d,f,n a
Basel Pharmacoepidemiology Unit, University of Basel, Basel, Switzerland Genentech, South San Francisco, CA, USA Roche Products Ltd., Welwyn Garden City, UK d Boston Collaborative Drug Surveillance Program, Boston University School of Public Health, Lexington, MA, USA e Massachusetts General Hospital Rheumatology Unit, Harvard Medical School, Boston, MA, USA f Hospital Pharmacy, University Hospital Basel, Basel, Switzerland b c
a r t i c l e in fo
Keywords: Corticosteroids Giant cell arteritis Health services research Outcomes research Serious adverse events
a b s t r a c t Objective: Giant cell arteritis (GCA) is an inflammatory disorder of blood vessels that preferentially affects large- and medium-sized arteries. High-dose oral corticosteroids (CS) are the mainstay of GCA therapy. Using data from the UK Clinical Practice Research Datalink, we quantified and compared the incidence of selected potentially CS-associated adverse outcomes in patients with and without GCA. Methods: We conducted a retrospective follow-up study of GCA and non-GCA patients to examine the incidence of adverse outcomes attributable to CS use. Eligibility criteria for the GCA group included a first-time diagnosis of GCA at age 50 years or older with receipt of Z1 prescription(s) for prednisolone. GCA patients were matched to a GCA-free comparison group of equal size on age, sex, general practice, and calendar time. We estimated incidence rates and incidence rate ratios (IRRs) for diabetes, osteoporosis, glaucoma, fractures, serious infection requiring hospitalization, and death for GCA and non-GCA patients and compared all-cause hospitalizations between the two groups. Results: The cohort consisted of 5011 GCA and 5011 matched non-GCA patients. Approximately 74% were women, and mean age at GCA diagnosis was 72.9 years. The IR for all outcomes was greater in the GCA group than the non-GCA group. IRRs [95% confidence intervals (CIs)] were as follows: diabetes 1.4 (1.2– 1.7), osteoporosis 2.4 (2.1–2.8), fractures 1.4 (1.2–1.6), glaucoma 2.0 (1.6–2.5), serious infection requiring hospitalization 1.5 (1.3–1.7), and death 1.2 (1.0–1.3). Conclusion: Compared with age- and sex-matched non-GCA patients, patients with GCA were at increased risk for diabetes, osteoporosis, fracture, and glaucoma and at a marginally increased risk for death. & 2016 Elsevier Inc. All rights reserved.
Introduction
☆
Sarsour, Musselman, Klearman, and Napalkov are employees of Genentech, a member of the Roche Group. Collinson and Tuckwell are employees of Roche Products Ltd. Stone are Principal Investigator in a Roche-funded clinical trial of a potential steroid-sparing agent in GCA. Wilson, Jick, and Meier have nothing to disclose. F. Hoffmann-La Roche Ltd. developed the initial concept for the study and provided funding for the research. n Corresponding author at: Basel Pharmacoepidemiology Unit Hospital Pharmacy, University Hospital Basel University of Basel, Spitalstrasse 26, CH-4031 Basel, Switzerland. E-mail address: [email protected] (C.R. Meier). http://dx.doi.org/10.1016/j.semarthrit.2016.10.001 0049-0172/& 2016 Elsevier Inc. All rights reserved.
Giant cell arteritis (GCA) is a form of primary systemic vasculitis that predominantly affects large and medium-sized arteries, with a predilection for the aorta and vessels to the head and neck. Vascular involvement in GCA, characterized by granulomatous inflammation of the aorta and its main branches, results in arterial narrowing and blood flow obstruction [1,2]. The disease almost exclusively affects persons older than 50 years of age, with incidence rates (IRs) peaking in the eighth decade of life [3–5]. Women are more likely to develop GCA than men. To date, the etiology of GCA is unknown. Genetic and environmental factors
2
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
are thought to be involved, and it is widely accepted that both the adaptive and the innate immune systems play important roles in the pathophysiology of this disease [3]. High-dose systemic corticosteroids (CS), predominantly prednisolone in the United Kingdom, are the cornerstone of GCA therapy. The duration of CS treatment varies by patient, but, because of disease relapses when patients are on or off treatment, the mean therapy duration is 5–6 years [6]. Long-term CS use is associated with a variety of adverse events (AEs), including weight gain, osteoporosis, fractures, hypertension, diabetes, cataracts, and myopathy [7,8]. Few studies, however, have examined the incidence of serious AEs specifically in GCA patients, a patient population often characterized by advanced age and many attendant comorbidities. A recent UK population-based study that examined the incidence and characteristics of GCA reported a variety of prevalent comorbidities associated with GCA, including osteoporosis, infections, cardiovascular diseases, diabetes, hypokalemia, and intraocular lens replacement. Many of these comorbidities are related to CS use [9]. We investigated the IRs of a series of adverse outcomes considered to be associated with prolonged CS therapy in GCA patients. We focused on diabetes, osteoporosis, fractures, glaucoma, and infections requiring hospitalization and compared the rates of these events with rates of the same events occurring in non-GCA patients. We also assessed mortality rates and the incidence of all-cause hospitalizations among GCA patients and a non-GCA comparison group.
Methods
their GCA diagnosis. We defined study entry as the date of this first prescription. For the comparison group, we randomly identified patients without GCA who were matched 1:1 to the GCA patients on age (year of birth), sex, general practice, study entry year, and years of history in the CPRD before study entry. All patients were required to have at least 3 years of recorded medical history before entry. Follow-up (i.e., study entry) began on the date of first prednisolone prescription for each GCA patient and the same date for the corresponding matched patient in the GCA-free comparison group. We excluded patients in both the GCA and the comparison groups if they had a diagnosis of any type of cancer, alcoholism, drug abuse, or HIV diagnosis before study entry. To further validate the identified GCA group, we conducted a sensitivity analysis using a more stringent, previously validated definition of GCA as described by Smeeth et al. [12]. Here GCA Table 1 Baseline cohort characteristics for GCA and non-GCA comparison patients Characteristics
GCA group, n ¼ 5011
Non-GCA group, n ¼ 5011
Sex Male Female
1298 (25.9) 3713 (74.1)
1298 (25.9) 3713 (74.1)
Age group, years 50–59 60–69 70–79 80–89 90þ
470 1205 2071 1158 108
466 1208 20 1161 107
(9.4) (24.0) (41.3) (23.1) (2.2)
(9.3) (24.1) (41.3) (23.2) (2.1)
χ2
–
–
Data sources
Mean age (SD)
The Clinical Practice Research Datalink (CPRD) provides anonymized health care information on approximately 8 million patients in the United Kingdom, with data extending back to 1987. Information on demographics (age, sex, weight, and height), consultation outcomes and diagnoses, specialist referrals, details on all prescribed medications, and lifestyle factors (e.g., tobacco and alcohol use) is recorded by specially trained general practitioners in the CPRD. READ codes, a disease coding system utilized in the UK National Health Service (NHS), are used to classify medical diagnoses. Recorded information on diagnoses and drug exposure has been validated repeatedly and has proven to be of high quality [10,11]. The study protocol was reviewed and approved by the Independent Scientific Advisory Committee for Medicines and Healthcare Products Regulatory Agency Database Research. Hospital Episode Statistics (HES) data contain details of all patient admissions to NHS hospitals in England only (not Northern Ireland, Scotland, or Wales), constituting approximately 55% of patients in the CPRD. The data collected include patient demographics and clinical and administrative details coded using the International Classification of Diseases, 10th Revision (ICD-10) system. The most up-to-date HES data available were used for this study, covering the years 1997–2012.
BMI 12–18.4 18.5–24.9 25–29.9 30–60 Missing
99 1624 1653 891 744
(2.0) (32.3) (33.0) (17.8) (14.9)
92 1596 1575 876 872
(1.8) (31.9) (31.4) (17.5) (17.4)
0.0131
Smoking status Never Current Former Missing
2415 778 1479 339
(48.2) (15.5) (29.5) (6.8)
2517 665 1362 467
(50.2) (13.3) (27.2) (9.3)
o0.0001
Alcohol use Never Current Former Missing
1244 3102 81 584
(24.8) (61.9) (1.6) (11.7)
1058 3158 93 702
(21.1) (63.0) (1.9) (14.0)
o0.0001
Comorbidity Rheumatologic disease Polymyalgia rheumatica Renal disease Peripheral vascular disease Peptic ulcer disease Myocardial infarction Mild liver disease Moderate liver disease Hemiplegia Diabetes Diabetes with complications Dementia Congestive heart disease Chronic pulmonary disease Cerebrovascular disease
1090 911 446 296 299 305 1 32 25 516 119 32 300 1249 521
(21.8) (18.2) (8.9) (5.9) (6.0) (6.1) (0.02) (0.64) (0.5) (10.3) (2.4) (0.6) (6.0) (24.9) (10.4)
274 127 358 224 245 255 2 12 21 484 105 58 228 952 376
(5.5) (2.5) (7.1) (4.5) (4.9) (5.1) (0.04) (0.24) (0.4) (9.7) (2.1) (1.4) (4.6) (19.0) (7.5)
o0.0001 o0.0001 0.0012 0.0012 0.0173 0.0297 0.5636 0.0025 0.5544 0.2862 0.3441 0.0003 0.0013 o0.0001 o0.0001
Average number comorbidities
1.03
Study population The study population was composed of an initial cohort of all patients in the CPRD 50 years of age and older with a first-time READ code for GCA between January 1, 1995, and August 31, 2013. In order to increase the proportion of validated and confirmed GCA cases in the cohort, patients were required to have at least one recorded prednisolone prescription at or within the 6 months after
72.9 (9.1)
72.9 (9.1)
0.70
–
–
BMI, body mass index; GCA, giant cell arteritis. Values in bold are statistically significant.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
patients were required to have received at least two prednisolone prescriptions, one prescription at or within the 6 months following their GCA diagnosis, with the second prescription being within 6 months of the first. As with the initial GCA group, a comparison group was matched 1:1 to the “twice criteria” GCA group.
3
We calculated crude IRs with 95% CIs for all outcomes of interest among the GCA group and the non-GCA comparison group, stratified by age and sex. We also estimated crude incidence rate ratios (IRRs) with 95% CIs.
Results Follow-up analysis of outcomes of interest Study population and characteristics We followed up all patients in the study population and accumulated person-time from study entry until the first event in the following list of events: diagnosis of an outcome of interest, death, end of the patient's medical record, or end of the study (August 31, 2013). Outcomes of interest were incident diabetes or a new prescription for a drug to treat diabetes, incident osteoporosis, incident glaucoma, or a newly prescribed therapy to lower intraocular pressure, bone fracture, serious infection requiring hospitalization, all-cause hospitalizations, and death (Supplementary Table S1). We examined each outcome separately; thus, a patient could have been included in more than one analysis. For the subset of patients with HES linkage, we combined CPRD and HES data in order to assess “all-cause hospitalizations” and “infections requiring hospitalization.” We excluded patients from this analysis if they had an entry date outside the dates of available HES data (and the corresponding date in the matched non-GCA group). To be included in the final HES-linked cohort, both GCA and matched non-GCA comparison patients were required to have available HES data linkage, thus ensuring equal access to outcomes information for both groups. For the outcomes diabetes, osteoporosis, and glaucoma, we established separate sub-cohorts from which we excluded GCA and non-GCA patients if they had a previous diagnosis of the outcome of interest before study entry. Definitions of all outcomes are provided in Supplementary Table S1.
We identified 5011 patients 50 years of age and older who had an incident diagnosis of GCA during the study period. All GCA patients received at least one prednisolone prescription within the 6 months after the GCA diagnosis. We also identified at random a matched comparison group of 5011 patients with no GCA diagnosis. Table 1 displays the basic characteristics of patients included in the study population. Most (74%) GCA patients were women, and the mean age at disease onset was 72.9 years. The median number of prednisolone prescriptions prescribed to GCA patients during follow-up was 17 (Q1 ¼ 6, Q3 ¼ 38). Less than 18% of nonGCA patients had a prednisolone prescription, the median number during follow-up was 3 (Q1 ¼ 1, Q3 ¼ 12). Figure 1 displays the final numbers for each of the separate cohorts following the exclusion of patients with a prevalent outcome. Incidence of the clinical outcomes of interest During follow-up, 594 patients (6.6%) in the GCA and the nonGCA groups combined developed diabetes, 778 (8.4%) received an osteoporosis diagnosis, 387 (4.2%) had an incident glaucoma diagnosis, 733 (7.3 %) sustained a fracture, 822 (8.2%) were hospitalized for a severe infection, and 1234 (12.3%) died. In the GCA group, the median time (in years) for the occurrence of diabetes, glaucoma, osteoporosis, infection, fractures, and death
Initial Cohort 5,011 GCA group 5,011 Non-GCA group
Diabetes Cohort 4,479 GCA group 4,517 Non-GCA group 532 GCA patients and 494 non-GCA patients with a history of diabetes were excluded Osteoporosis Cohort 4,567 GCA group 4,681 Non-GCA group 444 GCA patients and 330 non-GCA patients with a history of osteoporosis were excluded
Glaucoma Cohort 4,600 GCA group 4,662 Non-GCA group 411 GCA patients and 349 non-GCA patients with a history of glaucoma were excluded Fractures and Death Cohorts 5,011 GCA group 5,011 Non-GCA group No previous exclusions
HES-Linked Cohort 2,491 GCA group 2,491 Non-GCA group No previous exclusions
Fig. 1. Exclusions before follow-up.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
4
Table 2 IRs (per 1000 person-years) and IRR with 95% CIs for all outcomes of interest in GCA and non-GCA patients GCA group Outcomes, n
Non-GCA group PT, years
IR (95% CI)
Outcomes, n
PT, years
IR
IRR (95% CI)
Diabetes Male Female
340 85 255
24,017.8 5646.3 18,371.5
14.2 (12.7–15.7) 15.1 (12.2–18.6) 13.9 (12.3–15.7)
254 75 179
25,395.7 6055.9 19,339.8
10.0 (8.9–1.3) 12.4 (9.9–15.5) 9.3 (8.0–10.7)
1.4 (1.2–1.7) 1.2 (0.9–1.7) 1.5 (1.2–1.8)
Osteoporosis Male Female
532 63 469
23,203.1 6409.5 16,793.6
22.9 (21.1–24.9) 9.8 (7.7–12.6) 27.9 (25.5–30.5)
246 28 218
26,124.4 6932.0 19,192.4
9.4 (8.3–10.7) 4.0 (2.8–5.8) 11.4 (10.0–13.0)
2.4 (2.1–2.8) 2.4 (1.6–3.8) 2.5 (2.1–2.9)
Fractures Male Female
433 57 376
26,227.3 6605.4 19,621.9
16.5 (15.0–18.1) 8.6 (6.7–11.2) 19.2 (17.3–21.2)
340 29 311
27,707.9 7064.5 20,643.4
12.3 (11.0–13.6) 4.1 (2.9–5.9) 15.1 (13.5–16.8)
1.4 (1.2–1.6) 2.1 (1.3–3.3) 1.3 (1.1–1.5)
Glaucoma Male Female
253 60 193
24,837.9 6,000.9 18,837.0
10.2 (9.0–11.5) 10.0 (7.8–12.9) 10.3 (8.9–11.8)
134 37 97
26,451.0 6491.4 19,959.6
5.1 (4.3–6.0) 5.7 (4.1–7.9) 4.9 (4.0–5.9)
2.0 (1.6–2.5) 1.8 (1.2–2.6) 2.1 (1.7–2.7)
Serious infection Male Female
476 120 356
12,559.8 3099.4 9460.5
37.9 (34.7–41.4) 38.7 (32.5–46.1) 37.6 (34.0–41.7)
346 103 243
13,454.1 3331.0 10,123.1
25.7 (23.2–28.5) 30.9 (25.6–37.4) 24.0 (21.2–27.2)
1.5 (1.3–1.7) 1.3 (1.0–1.6) 1.6 (1.3–1.9)
Mortality Male Female
653 197 456
27,932.0 6826.6 21105.3
23.4 (21.7–25.2) 28.9 (25.1–33.1) 21.6 (19.7–23.7)
581 196 385
28,724.6 7117.8 21,606.8
20.2 (18.7–21.9) 27.5 (24.0–31.6) 17.8 (16.1–19.7)
1.2 (1.0–1.3) 1.1 (0.9–1.3) 1.2 (1.1–1.4)
2101 548 1553
4587.0 1114.4 3472.6
1876 492 1384
6758.5 1663.1 5095.4
Hospitalization Male Female
458.0 (443.6–472.5) 491.7 (462.5–521.01 447.2 (430.7–463.8)
277.6 (267.0–288.4) 295.8 (274.4–318.2) 271.6 (259.6–284.0)
1.7 (1.6–1.8) 1.7 (1.5–1.9) 1.7 (1.5–1.8)
CI, confidence interval; GCA, giant cell arteritis; IR, incidence rate; IRR, incidence rate ratio; PT, person-time. Values in bold are statistically significant.
was 1, 1.5, 2, 2, 3, and 4 respectively. The incidence of all outcomes was higher in the GCA group than in the non-GCA comparison group (Table 2). All IRRs for diabetes, osteoporosis, glaucoma, fracture, infection requiring hospitalization, and death were between 1.2 and 2.0 and were significantly higher in the GCA group than the non-GCA cohort (Table 2). The sensitivity analysis revealed marginal difference in the reported IRs and IRRs for the outcomes in the more stringent “twice criteria” GCA group (Supplementary Table S2). After stratification of the analyses by sex, there was little evidence of effect modification for any outcome except fracture (Table 2). The IRR for fracture in patients with GCA compared with those without GCA was substantially higher for men (IRR ¼ 2.1, 95% CI: 1.3–3.3) than for women (IRR ¼ 1.3, 95% CI: 1.1–1.5). In the GCA and non-GCA group, a greater proportion of females (77% GCA:25% non-GCA) were prescribed osteoporosis medications compared to males (63% GCA:11% non-GCA). IRs for most outcomes across most age groups were higher for GCA patients than for the relevant non-GCA patients (Figs. 2 and 3). Incidence of all-cause hospitalization During follow-up, 2101 GCA patients and 1876 non-GCA patients each had at least one hospitalization. GCA patients were at greater risk for hospitalization than non-GCA patients (IRR ¼ 1.7, 95% CI: 1.6–1.8). The risk was similar in men and women (Table 2). Based on ICD-10 classifications, more patients in the GCA group than the non-GCA group (overall P o 0.001) were hospitalized for diseases of the circulatory system (ICD-10 codes I00I99) (26.0% vs 18.9%), digestive system (ICD-10 codes K00-K93) (24.9% vs 21.1%), eye and adnexa (ICD-10 codes H00-H59) (20.5% vs 17.0%), respiratory system (ICD-10 codes J00-J99) (15.3% vs 11.4%), endocrine, nutritional, and metabolic systems (ICD-10 codes E00E90) (3.7% vs 2.2%), and blood and blood-forming organs,
including disorders involving the immune mechanism (ICD-10 codes D50-D89) (4.8% vs 3.1%).
Discussion In this retrospective population-based follow-up study, the IRs of diabetes, glaucoma, osteoporosis, fractures, and serious infections were all higher in GCA patients than in a non-GCA comparison group. Increased risks for adverse outcomes such as diabetes, serious infections, and fractures have been reported to be between 1.3 and 3 times higher in GCA patients than in non-GCA patients [6,9], which is consistent with IRRs found in the current study. It is well recognized that CS therapy increases the risk for these outcomes and that these outcomes are often attributable to AEs resulting from long-term CS therapy. Increased risk for diabetes, fractures, osteoporosis, infection, and gastrointestinal effects has been reported in patients receiving prolonged CS treatment [12–16]. A small-scale, retrospective, US study in 120 GCA patients found that a substantial proportion of patients treated with CS (86%) developed AEs attributable to CS therapy [6]. Few studies have examined the occurrence of serious AEs of CS therapy by sex in GCA patients. The IRRs for diabetes, glaucoma, serious infections, and death in GCA patients compared with nonGCA patients did not differ substantially between women and men. In contrast to other studies that have suggested greater risk for osteoporosis and CS-related fragility fractures among women [17,18]; however, our study found the IRR for fracture was considerably higher among men. Given that men are generally considered to have fewer risk factors for insufficiency fracture, our finding is consistent with the concept that the presence of one major risk factor in GCA patients compared with non-GCA patients —namely, CS use—disproportionately increases the fracture risk for men. The United Kingdom has no recognized population screening
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
B
20 15 10 5 0
80 70 60 50 40 30 20 10 0
16
70
14
60
12
50
10 8 6 4 2 0
50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
40 30 20 10
Number of Glaucoma Cases
25
90
Incidence Rate (1,000 person-years)
30
Number of Diabetes Cases
Incidence Rate (1,000 person-years)
A
5
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
30
140
20
100
15
80
10
60 40
5 0
20 0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
35
140
30
120
25
100
20
80
15
60
10
40
5
20
0
Number of Fracture Cases
120
Number of Osteoporosis Cases
25
160
Incidence Rate (1,000 person-years)
D Incidence Rate (1,000 person-years)
C
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+ Age Group, years Cases GCA Cases Non-GCA Rate GCA Rate Non-GCA
Fig. 2. IRs (per 1000 person-years) and number of cases in GCA and non-GCA patients stratified by age at GCA diagnosis for (A) diabetes, (B) glaucoma, (C) osteoporosis, and (D) fractures. GCA, giant cell arteritis; IR, incidence rate.
system in place to identify patients who have osteoporosis or who are at high risk for fracture. Instead, osteoporosis is identified opportunistically through a case-finding strategy based on history of fragility fracture or on the presence of clinical risk factors that include but are not limited to age, sex, and current glucocorticoid treatment for more than 3 months [19]. The high IRR observed in male GCA patients may highlight the need for additional clinical awareness of fracture risk in this patient group. An increased incidence of all-cause hospitalizations was observed in GCA patients compared with non-GCA patients. The most common causes of hospitalization in the GCA group were diseases of the cardiovascular system, closely followed by diseases of the digestive system and of the eyes. It appears likely that this observation reflects both steroid-related AEs and the pathophysiology of GCA, in which increased risks for cardiovascular disease [9,20] and vision impairment [4,21] have been observed independently of CS use in patients with GCA. We observed a modest increase in the risk for death in GCA patients compared with the matched non-GCA group. These results are consistent with most, but not all, studies in the literature that have generally demonstrated mortality rates comparable to or slightly higher than those of the general population [22–26] and that have reported increased mortality rates among GCA patients in the first years after GCA diagnosis [26–29]. Compared with the comparison group, GCA patients had a higher frequency of comorbidities before study entry, which might have contributed to the observed increased risk for mortality. Our study has a number of important strengths. The CPRD is a well-respected primary care database of high quality and completeness [10,11,30]. We were rigorous in our assessment of
incident cases of GCA to prevent misclassification, and strict exclusion criteria were applied to ensure that everyone included had a valid GCA diagnosis. The finding of little difference in results of the additional sensitivity analysis comprising patients with a stricter GCA definition further validates our GCA group and supports the presented findings. Moreover, this is one of few studies to examine the incidence of outcomes likely associated with prolonged CS therapy by sex. Finally, previous studies of GCA with regard to the development of serious AEs have been small, each involving fewer than 150 GCA patients [6,31–38]. The current study, which includes data on more than 5000 GCA patients, is one of the largest to date to examine the incidence of multiple, potentially CS-related effects in patients with this disease. The findings add further weight to the perceived potential clinical burden of prolonged CS treatment in GCA patients. Our study also has some potential limitations. There might have been a bias for osteoporosis diagnosis in the GCA group because of an increased likelihood of referring patients with GCA for fracture risk assessment. This potential bias might have been minimized, however, because the incidence of GCA peaks between 70 and 80 years of age, coinciding with the recommended age for fracture risk assessment: 65 years of age and older for women and 75 years of age and older for men [39]. Similar considerations apply to a potential bias in the detection of glaucoma. We did not exclude non-GCA patients who had a prednisolone prescription during follow-up; however, the percentage of users was low ( o 18%). In addition, the median number of prescriptions during follow-up was considerably lower than in the GCA exposed patients and was likely to have minimal influence on the reported IRs in the nonGCA group.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]
6
Incidence Rate (1,000 person-years)
120
90 80
100
70
80
60 50
60
40
40
30 20
20
10
0
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+
Cases GCA
Age Group, years Cases Non-GCA Rate GCA
Number of Serious Hospital Infection Cases
A
Rate Non-GCA
90
180
80
160
70
140
60
120
50
100
40
80
30
60
20
40
10
20
Number of Death Cases
Incidence Rate (1,000 person-years)
B
0
0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90+
Cases GCA
Age Group, years Cases Non-GCA Rate GCA
Rate Non-GCA
Fig. 3. IRs (per 1000 person-years) and number of cases in GCA and non-GCA patients stratified by age at GCA diagnosis for (A) serious infections and (B) death outcomes of interest. GCA, giant cell arteritis; IR, incidence rate.
In conclusion, the findings of this large population-based study indicate that patients with GCA are at increased risk for diabetes, osteoporosis, and glaucoma, are more likely to sustain a fracture, and are more likely to require all-cause hospitalization than are patients without GCA. The risk for death was found to be only marginally higher in patients with GCA than in an age- and sexmatched comparison group.
Author contributions All authors reviewed and revised the manuscript drafts and approved the final version for publication. CRM had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Study conception and design: All authors. Drafting the manuscript: Wilson, Jick, and Meier.Acquisition of data: Wilson, Jick, and Meier. Analysis and interpretation of data: All authors.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.semarthrit.2016. 10.001.
References [1] Seetharaman M, Paget S, Leibowitz E. Giant Cell Arteritis. 〈http://emedicine. medscape.com/article/332483-overview〉; 2012. [2] Gonzalez-Gay MA, Martinez-Dubois C, Agudo M, Pompei O, Blanco R, Llorca J. Giant cell arteritis: epidemiology, diagnosis, and management. Curr Rheumatol Rep 2010;12:436–42, http://dx.doi.org/10.1007/s11926-010-0135-9. [3] Chew SSL, Kerr NM, Danesh-Meyer HV. Giant cell arteritis. J Clin Neurosci 2009;16:1263–8, http://dx.doi.org/10.1016/j.jocn.2009.05.002. [4] Gonzalez-Gay MA, Vazquez-Rodriguez TR, Lopez-Diaz MJ, et al. Epidemiology of giant cell arteritis and polymyalgia rheumatica. Arthritis Rheum 2009;61:1454–61, http://dx.doi.org/10.1002/art.24459. [5] Villa-Forte A. Giant cell arteritis: suspect it, treat it promptly. Cleve Clin J Med 2011;78:265–70, http://dx.doi.org/10.3949/ccjm.78a.10131. [6] Proven A, Gabriel SE, Orces C, O'Fallon WM, Hunder GG. Glucocorticoid therapy in giant cell arteritis: duration and adverse outcomes. Arthritis Rheum 2003;49:703–8, http://dx.doi.org/10.1002/art.11388. [7] Kaushik P, Cote D. Giant cell arteritis. Indian J Rheumatol 2009;4:20–4, http: //dx.doi.org/10.1016/S0973-3698(10)60157-4. [8] Royal College of Physicians. Diagnosis and management of giant cell arteritis: concise guidance to good practice series; 2010. [9] Petri H, Nevitt A, Sarsour K, Napalkov P, Collinson N. Incidence of giant cell arteritis and characteristics of patients: data-driven analysis of comorbidities. Arthritis Care Res (Hoboken) 2014. http://dx.doi.org/10.1002/acr.22429. [10] Jick H, Jick SS, Derby LE. Validation of information recorded on general practitioner based computerised data resource in the United Kingdom. Br Med J 1991;302:766–8, http://dx.doi.org/10.1136/bmj.302.6779.766. [11] Khan NF, Harrison SE, Rose PW. Validity of diagnostic coding within the General Practice Research Database: a systematic review. Br J Gen Pract 2010;60:e128–36, http://dx.doi.org/10.3399/bjgp10X483562. [12] Smeeth L, Cook C, Hall AJ. Incidence of diagnosed polymyalgia rheumatica and temporal arteritis in the United Kingdom, 1990–2001. Ann Rheum Dis 2006;65:1093–8, http://dx.doi.org/10.1136/ard.2005.046912. [13] Saag KG, Koehnke R, Caldwell JR, et al. Low dose long-term corticosteroid therapy in rheumatoid arthritis: an analysis of serious adverse events. Am J Med 1994;96:115–23, http://dx.doi.org/10.1016/0002-9343(94)90131-7. [14] Blackburn D, Hux J, Mamdani M. Quantification of the Risk of Corticosteroidinduced Diabetes Mellitus Among the Elderly. J Gen Intern Med. 2002;17:717–20. [15] Van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C. Oral corticosteroids and fracture risk: relationship to daily and cumulative doses. Rheumatology (Oxford) 2000;39:1383–9. 〈http://www.ncbi.nlm.nih.gov/pubmed/11136882〉. [16] Van Staa TP, Cooper C, Leufkens HG, Bishop N. Children and the risk of fractures caused by oral corticosteroids. J Bone Miner Res 2003;18:913–8, http://dx.doi.org/10.1359/jbmr.2003.18.5.913. [17] Gabriel SE, Sunku J, Salvarani C, O'Fallon WM, Hunder GG. Adverse outcomes of antiinflammatory therapy among patients with polymyalgia rheumatica. Arthritis Rheum 1997;40:1873–8, http://dx.doi.org/10.1002/1529–0131 (199710)40:10o1873:AID-ART22 43.0.CO;2-V. [18] Tobón GJ, Youinou P, Saraux A. The environment, geo-epidemiology, and autoimmune disease: rheumatoid arthritis. Autoimmun Rev 2010;9:A288–92, http://dx.doi.org/10.1016/j.autrev.2009.11.019. [19] Souza A, Okamoto K, Abrantes F, Schau B, Bacchiega AB, Shinjo SK. Giant cell arteritis: a multicenter observational study in Brazil. Clinics 2013;68:317–22, http://dx.doi.org/10.6061/clinics/2013(03)OA06. [20] National Osteoporosis Guideline Group. Guideline for the diagnosis and management of osteoporosis in postmenopausal women and men from the age of 50 years in the UK. 〈http://www.shef.ac.uk/NOGG/NOGG_Pocket_Gui de_for_Healthcare_Professionals.pdf〉; 2014. [21] Nordborg E, Nordborg C. Giant cell arteritis: epidemiological clues to its pathogenesis and an update on its treatment. Rheumatology (Oxford) 2003;42:413–21. 〈.http://www.ncbi.nlm.nih.gov/pubmed/12626790〉 [accessed 10.02.15]. [22] Loddenkemper T, Sharma P, Katzan I, Plant GT. Risk factors for early visual deterioration in temporal arteritis. J Neurol Neurosurg Psychiatry 2007;78:1255–9, http://dx.doi.org/10.1136/jnnp.2006.113787. [23] Gran JT, Myklebust G, Wilsgaard T, Jacobsen BK. Survival in polymyalgia rheumatica and temporal arteritis: a study of 398 cases and matched population controls. Rheumatology (Oxford) 2001;40:1238–42. 〈.http:// www.ncbi.nlm.nih.gov/pubmed/11709607〉 [accessed 10.02.15]. [24] Doran MF, Crowson CS, O’Fallon WM, Hunder GG, Gabriel SE. Trends in the incidence of polymyalgia rheumatica over a 30 year period in Olmsted County, Minnesota, USA. J Rheumatol 2002;29:1694–7. 〈.http://www.ncbi.nlm.nih. gov/pubmed/12180732〉 [accessed 10.02.15]. [25] González-Gay MA, Blanco R, Abraira V, et al. Giant cell arteritis in Lugo, Spain, is associated with low longterm mortality. J Rheumatol 1997;24:2171–6. 〈http://www.ncbi.nlm.nih.gov/pubmed/9375880〉 [accessed 10.02.15]. [26] Matteson EL, Gold KN, Bloch DA, Hunder GG. Long-term survival of patients with giant cell arteritis in the American College of Rheumatology giant cell arteritis classification criteria cohort. Am J Med 1996;100:193–6. 〈http://www. ncbi.nlm.nih.gov/pubmed/8629654〉 [accessed 10.02.15]. [27] Ninan J, Nguyen AM, Cole A, et al. Mortality in patients with biopsy-proven giant cell arteritis: a south australian population-based study. J Rheumatol 2011;38:2215–7, http://dx.doi.org/10.3899/jrheum.101254.
J.C. Wilson et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]] [28] Baslund B, Helleberg M, Faurschou M, Obel N. Mortality in patients with giant cell arteritis. Rheumatology (Oxford) 2015;54:139–43, http://dx.doi.org/ 10.1093/rheumatology/keu303. [29] Mohammad AJ, Nilsson J-Å, Jacobsson LT, Merkel PA, Turesson C. Incidence and mortality rates of biopsy-proven giant cell arteritis in southern Sweden. Ann Rheum Dis 2014. http://dx.doi.org/10.1136/annrheumdis-2013-204652. [30] Uddhammar A, Eriksson A-L, Nyström L, Stenling R, Rantapää-Dahlqvist S. Increased mortality due to cardiovascular disease in patients with giant cell arteritis in northern Sweden. J Rheumatol 2002;29:737–42. 〈http://www.ncbi. nlm.nih.gov/pubmed/11950015〉 [accessed 10.02.15]. [31] Jick SS, Kaye JA, Vasilakis-Scaramozza C, et al. Validity of the general practice research database. Pharmacotherapy 2003;23:686–9. 〈http://www.ncbi.nlm. nih.gov/pubmed/12741446〉 [accessed 31.10.14]. [32] Crow RW, Katz BJ, Warner JEA, et al. Giant cell arteritis and mortality. J Gerontol A Biol Sci Med Sci 2009;64:365–9, http://dx.doi.org/10.1093/gerona/gln030. [33] Kyle V, Hazleman BL. Treatment of polymyalgia rheumatica and giant cell arteritis. II. relation between steroid dose and steroid associated side effects. Ann Rheum Dis 1989;48:662–6, http://dx.doi.org/10.1136/ard.48.8.662. [34] Nesher G, Sonnenblick M, Friedlander Y. Analysis of steroid related complications and mortality in temporal arteritis: a 15-year survey of 43 patients.
[35]
[36]
[37]
[38]
[39]
7
J Rheumatol 1994;21:1283–6. 〈http://www.ncbi.nlm.nih.gov/pubmed/7966070〉 [accessed 27.10.14]. Fainaru M, Friedman G, Friedman B. Temporal arteritis in Israel. A review of 47 patients. J Rheumatol 1979;6:330–5. 〈http://www.ncbi.nlm.nih.gov/pubmed/ 490527〉 [accessed 26.02.16]. Graham E, Holland A, Avery A. Prognosis in giant-cell arteritis. Br Med J (Clin Res Ed) 1981;282:269–71. 〈http://www.pubmedcentral.nih.gov/articlerender. fcgi?artid=1504093&tool=pmcentrez&rendertype=abstract〉 [accessed 26.02.16]. Khalifa M, Karmani M, Jaafoura NG, et al. Epidemiological and clinical features of giant cell arteritis in Tunisia. Eur J Intern Med 2009;20:208–12, http://dx. doi.org/10.1016/j.ejim.2008.07.030. Alba MA, García-Martínez A, Prieto-González S, et al. Relapses in patients with giant cell arteritis: prevalence, characteristics, and associated clinical findings in a longitudinally followed cohort of 106 patients. Medicine (Baltimore) 2014;93:194–201, http://dx.doi.org/10.1097/MD.0000000000000033. National Institute for Health and Care Excellence (NICE). NICE Clinical Guideline CG146. Osteoporosis: assessing the risk of fragility fracture. 〈http:// publications.nice.org.uk/osteoporosis-assessing-the-risk-of-fragility-fractur e-cg146/guidance〉; 2013.