Quality of life of patients treated for giant cell arteritis: a case-control study

Giant cell arteritis (GCA) is a large vessel systemic vasculitis that affects preferably Caucasian women over 70 years old [1]. Complications of GCA may be visual, neurological, or cardiovascular in 15% of cases [2‐4]. GCA treatment relies on a prolonged oral corticosteroid therapy (CS), enabling rapid resolution of clinical symptoms and favorable long-term outcome if initiated early [5‐7]. However, high dose CS may induce metabolic, cardiovascular, skeletal, and infectious side effects, especially in the seniors [5‐8]. Such side effects, along with GCA complications, may cause significant disability, impairing patients’ quality of life (QOL). So far, few studies have assessed QOL in patients with GCA. For Kupersmith et al., there was no correlation between “Activities of Daily Vision Scale” (ADVS) score, SF36 score, and visual impairment 1 year after GCA diagnosis [9]. Hellmann et al. focused on the importance of several altered domains of QOL in GCA patients. In their study, the three most important domains of impaired QOL were vision, motor skills, and personal hygiene [10]. To date, no study has evaluated QOL of GCA patients after withdrawal of CS. In our clinical practice, we estimated that a significant number of patients treated for GCA had an impaired QOL after CS withdrawal. We speculated that impaired QOL may be related to nonspecific symptoms such as muscle weakness and pain. To demonstrate whether this clinical impression is a reality, we conducted a case-control study in which the aim was to evaluate the QOL of GCA patients following high dose of CS. The second aim of the study was to determine whether the complications of GCA or CS may influence QOL in this group of patients.

Between March and June 2015, we included 30 patients with GCA and 60 matched controls. Table 1 describes the clinical characteristics of both groups.

Twenty-five patients (83.3%) had stopped CS for at least 3 months and 5 patients (16.7%) were taking long-term low doses of CS (≤ 5 mg/day). The median duration of treatment, for patients free of CS, was 18 months (14–21). Regarding the 5 patients taking long-term low-doses of CS, the median time between telephone interview and date of decision to keep long-term treatment was 12 months (10–15). For all patients, median delay from GCA diagnosis to telephone interview was 31 months (24.25–44.75). Four patients had received CS boluses at diagnosis and the median maximum dose of CS was 57.5 mg per day out bolus. Fifteen patients (50%) had at least one complication related to GCA. Twelve patients (40%) had ocular complications, 5 patients (16.7%) had cardiovascular complications and 1 patient (3.3%) had neurological complications. Twenty-three patients (76.7%) had at least one complication related to CS therapy. The most frequent complications were skin fragility (40%), weight gain (36.7%), and bone fractures (30%), followed by repeated infections (13.3%), adrenal insufficiency (13.3%), severe infections (10%), psychiatric disorders (6.7%), and steroid-induced diabetes (3.3%) (Table 2).

Regarding GCA specific questionnaire, the data is resumed in Table 3. Seventeen patients (56.7%) estimated that their condition was better since withdrawal of CS. Functional complaints of patients were dominated by walking difficulties (53.3%), appearance changes (36.7%), and falls (26.7%). Visual disturbances corresponded to a median impairment score of 3 (or “moderately”). Eating difficulties, appearance changes, walking difficulties, and falls corresponded to a median impairment score of 2 (or “a little”). Only hair loss did not provide any impairment with a median score of 1 (or “not at all”).

Regarding SF36 questionnaire data (Table 4), GCA patients’ SF 36 PCS and MCS were not statistically different from controls. SF36 scores for different QOL domains were normal in patients with GCA and generally lower in controls with a significantly lower difference for the “Vitality” (VT) domain (48.7 vs 43.5, p = 0.018).

Concerning the subgroup analysis, PCS and MCS and scores of the different SF 36 domains were not different between patients who had stopped treatment and those who took long term low doses of CS (≤ to 5 mg/day) and between patients with or without PMR. Regarding patients who were no longer taking CS, negative correlations were found between total duration of CS therapy and the QOL domain “Physical Functioning” (PF) score (p = 0.0156) and between the average duration of CS and the QOL domain “Mental Health” (MH) score (p = 0.035). No correlation with the duration of CS therapy was demonstrated for the other SF36 QOL domain’s scores, including PCS and MCS. PCS and MCS among patients with all cause complications related to the GCA (visual, cardiovascular and/or neurological) or with all cause complications related to CS therapy (skin fragility, weight gain, bone fracture, severe infections, repeated infections, adrenal insufficiency, psychiatric disorders, and steroid-induced diabetes) were not different from those who had no complications. Regarding the different SF 36 QOL domains, the domain “Bodily Pain” (BP) score was significantly lower in patients who had repeated infections (41.26 vs 53.07, p = 0.002) and the domain VT score was significantly lower in patients who had gained weight (37.30 vs 49.17, p = 0.013).

After analysis of SF 36 scores based on the specific GCA questionnaire answers, patients who thought their condition was better after stopping CS had PCS and MCS and scores for the different SF 36 QOL domains similar to patients who thought their condition was identical or worse before and after CS therapy. Only patients who estimated that their condition had not changed before or after taking CS had a PF score significantly higher than other GCA patients (55.15 vs 47.46, p = 0.0431). PCS, MCS and scores of the different SF36 QOL domains were not different between patients who reported visual impairment, eating difficulties, appearance changes, hair loss and those who did not report these complaints. Patients who considered having walking difficulties had a PCS significantly lower (45.33 vs 51.57, p = 0.0340), with a significant impairment of “General Health” (GH) and PF (47.60 vs 55.26, p = 0.005 and 45.94 vs 51.39, p = 0.0298, respectively) domains. These results were found in patients who thought their walking difficulties were related to balance disorders or pain and not to muscle weakness (Fig. 1). PCS and MCS were not statistically different between patients who had fallen since withdrawal of CS and those who had not fallen. Only patients who had fallen 3 times or more had a significantly lower MCS score (40.25 vs 50.76 p = 0.0460). Regarding the different SF36 QOL domains, patients who had fallen, regardless of the number of falls, and patients who had fallen 3 times or more had a score for the VT domain significantly lower (39.97 vs 51.92, p = 0.0058). Patients who had fallen only once had a significantly lower score for the domain BP (39.42 vs 49.22, p = 0.0493) (data not shown).

Regarding analysis of SF36 scores in all study subjects who had associated chronic disease (data not shown), PCS was significantly lower in subjects with chronic cardiovascular disease (43.16 vs 48.09, p 0.0114) with a similar MCS. Subjects with diabetes, musculoskeletal, neurological, digestive, respiratory, psychiatric chronic diseases, and subjects who had a previous history of neoplasia in remission showed no significant alteration of the SF36 summarized scores. For the different SF36 QOL domains, only subjects with diabetes had a GH score and VT score significantly lower (42.34 vs 50.87 p 0.0006 and 40.96 vs 46.83 p 0.0161, respectively) and those who had chronic cardiovascular disease (atrial fibrillation, ischemic heart disease) had a PF score (40.53 vs 48.22 p = 0.0046) and a “Social Functioning” (SF) score (48.43 vs 52.09 p = 0.0395) significantly lower.

The results of our study suggest that in patients with GCA, QOL may be preserved following high dose of CS. To the best of our knowledge, this work is the first to evaluate QOL in such patients. Previous studies have evaluated QOL but only in GCA patients under treatment. According to Hellmann et al., vision, motor skills, and personal hygiene are the most important domains of QOL for GCA patients [10]. However, Kupersmith et al. showed no correlation between the SF 36 scores and the presence or not of visual loss at diagnosis. Only the MH domain score was significantly correlated to visual performance. Their results were obtained while patients were still under CS treatment and concerned almost only vision without any results for the other domains of QOL qualified as important by GCA patients.

Regarding the influence of disease complications or treatment side effects on QOL, our results suggest that GCA patients have no significant post-treatment impairment of QOL as compared to controls. Despite significant complications related to GCA and common side effects associated with CS, QOL was generally better among patients with GCA than controls, with a significant difference for the VT domain score. In addition, 57% of our GCA patients felt that their general condition had improved after the withdrawal of CS. It seems that in our clinical practice, GCA patients at time of diagnosis have an impaired QOL related to GCA complications or to the presence of systemic inflammatory and cephalic symptoms. Hellmann et al. joined this impression as, in their study, the items “presence of unexplained prolonged febrile episodes”, “pain” and “difficulty chewing, eating” were among the 10 most important items of impaired QOL for GCA patients [10]. Therefore, GCA seems to affect patients’ QOL at diagnosis and for which CS therapy allows them to recover a similar or even better QOL compared to controls. Experience of disease could result, once recovery is obtained, in a positive change of life, priorities and way of being. These factors may explain the significant increase of the VT domain’s score in our GCA patients. This phenomenon has already been identified in patients over 60 years old in remission from breast cancer [18].

As GCA is a large vessels vasculitis, it seems relevant to compare our results to those obtained in other vasculitis. Abularrage et al.’s study evaluated the SF 36 QOL of 158 patients with Takayasu arteritis, 40% of them being in remission. Results showed a significant association between disease remission and improvement of SF 36 PCS and MCS. However, scores were below the normal range, as compared to the general population [19]. Similar data have been found in patients with microscopic polyangiitis and granulomatosis with polyangiitis. Indeed, patients in remission of such diseases improved their QOL compared to diagnosis period, although they keep values of SF36 scores below normal [20‐22]. These findings differ from our data, but could be explained by the fact that ANCA vasculitis cause more severe complications than GCA, and involve more side effects related to immunosuppressive therapies. In our study, patients requiring immunosuppressive therapy were not included. Thus, we could not determine whether the use of immunosuppressive therapies could alter QOL in GCA.

The second objective of our study was to perform a subgroup analysis of SF36 scores depending on the different characteristics of GCA patients. Patients with complications related to GCA had no significant impairment of their SF36 QOL. In addition, patients with visual complications and those who reported visual problems in our specific questionnaire did not have lower scores in the specific SF36 QOL domains. These data are similar to Kupersmith et al.’s study, as their results did not show any significant QOL impairment in patients with visual loss at one year of GCA diagnosis [9]. However, patients who reported visual disturbances in our GCA specific questionnaire had a daily impairment score of 3 which corresponds to a “moderate” impairment and was the highest impairment score regarding all items. This difference could be explained by the fact that there is no specific item about vision in the SF36 questionnaire.

In GCA, patient’s QOL may be impaired by treatment side effects. Long-term CS therapy is the gold standard of GCA but causes undeniable side effects, with an increased risk when important cumulative doses are given and in older people [5, 6, 23]. Proven et al.’s showed that 86% of their patients experienced complications related to a CS therapy, mostly bone fractures, infections and clinical signs of CS impregnation [6]. In our study, complications of CS affected 77% of the patients and were dominated by skin fragility, weight gain, and bone fractures. So far, no study has evaluated the QOL of patients with complications related to long term CS therapy. In our study, such patients had normal PCS and MCS scores. Patients who gained weight had a significantly lower score for the QOL domain VT. In the literature, overweighted or obese patients have an impaired QOL, especially in PF and “Role Physical” (RP) domains. Weight loss is associated with increased scores in physical QOL domains as well as in VT, GH and MH domains [24, 25]. These results may encourage professionals to screen for weight gain in their GCA patients, and to consider a specific approach in case of excessive weight gain. In our study, about 1/3 of the patients complained about appearance changes without impairment of their QOL according to the SF36 and our specific questionnaire. This result can be explained by the fact that GCA patients have an average age of 82 years old. It is well known that older people feel less stigmatized about their appearance than young subjects [26]. CS therapy causes undeniable side effects but does not seem to have any major impact on the QOL of GCA patients. These data, as well as its effectiveness and rapid action, reinforce the fact that CS remains the gold standard therapy for GCA.

In our study, the most common complaint of GCA patients was related to walking. Fifty-three percent of them considered having walking difficulties related to balance disorders or diffuse pain rather than to muscle weakness. A relationship between gait disorders and muscular atrophy secondary to long term CS therapy could have been expected, but it was not the case. It is also interesting to note that patients who believed having walking difficulties had a lower PCS than controls, with a significant impairment of the domains GH and PF and similar scores for the mental domains. This result suggests that patients with walking disorders have a reduced walking perimeter and difficulties in achieving even moderate physical efforts, explaining the low score for the domain PF with pejorative feelings and decreased score for the domain GH. In our study, 27% of patients had fallen since withdrawal of CS with impaired mental QOL domains. Fall occurrence and walking difficulties are not specific to GCA but mostly explained by the age of our patients. In the literature, 82% of people over 80 years old have walking difficulties and about 1/3 of patients over 65 years old will fall within the year [27, 28]. Cumming et al. showed that scores of the QOL domains PF, MH, BP, and PCS were significantly lower in patients with a Falls Efficacy Scales score (FES) ≤75 meaning with an increased risk of falling [29]. Moreover, Coimbra et al. showed a significant association between the occurrence of falls in people over 60 years old and a decreased SF36 MCS [30]. However, it is uneasy to determine whether the alteration of QOL is the cause or the consequence of falls. Thus, a geriatric evaluation should be proposed to older GCA patients. Moreover, we estimate that vitamin D prescription, occupational therapy evaluation, physiotherapy, and simplification of medications should be routinely performed in the management of older GCA patients. The duration of CS therapy should also be as short as possible. Indeed, in our study, the total duration of CS therapy was negatively correlated to the PF’s score’s domain and the mean duration of CS therapy to the MH’s score’s domain.

Our study has some limitations. SF36 scale is a generic scale and not a specific tool for GCA without any items about vision, eating, or appearance that may be important to assess QOL in such patients. SF36 has been used in many studies with patients over 65 years old, but some questions remain inappropriate and sometimes difficult to understand for an elderly population. A specific version for people over 65 years old would facilitate its use, as proposed by Walters et al. [17]. Another potential limitation of our study concerns the control group. Selected controls were representative of a discharged (>3 months) hospitalized population and had significantly more cardiovascular diseases than GCA patients. However, in our study, subjects with cardiovascular disease had lower scores in all SF36 domains with a significantly lower score in the domains PF, RP, and in the physical summary score. These data are similar to those found in literature. Indeed, patients with ischemic or hypertrophic cardiomyopathy, atrial fibrillation or obliterative arteriopathy have an impaired QOL compared to the general population and patients undergoing valve replacement have a better QOL after surgery [31‐34].

GCA patients after high dose of CS have no significant impairment of their QOL compared to controls. GCA complications including visual impairment do not seem to have any major impact on QOL. Side effects of CS affected 77% of our GCA patients without significant impairment of QOL except for an alteration of the domain VT in patients who gained weight. Walking difficulties and falls are frequently reported probably due to the age of the GCA population and are associated with a significant impairment of physical and mental aspects of QOL. These results as well as the effectiveness and rapid action of CS reinforce the fact that they remain the standard treatment for GCA. Close weight monitoring, development of lifestyle, and dietary rules during consultations as well as specific geriatric assessments and falls prevention measures could be useful in the management of these patients.