Introduction

Giant cell arteritis (GCA) is a granulomatous vasculitis involving large- and middle-sized arteries, with a predilection for cranial arteries [1]. It is the most common primary vasculitis in Europe and North America in patients over 50 years of age, and it affects two to three times more frequently the female gender. The typical features of GCA are new onset of headache accompanied by constitutional symptoms, scalp tenderness, jaw claudication, visual disturbances, and increased inflammatory markers. The most common ophthalmic manifestation is anterior ischemic optic neuropathy (AION), which can lead to irreversible blindness. Prompt diagnosis and treatment are essential to prevent visual loss [2].

Ultrasound (US) of the temporal and axillary arteries, showing a non-compressible “halo sign,” and/or a biopsy of the temporal artery (TAB), presenting transmural inflammation associated with marked disruption of the internal elastic membrane, are the current recommended tests to diagnose patients in whom there is a clinical suspicion of predominantly cranial GCA [3]. However, the diagnostic sensitivity and specificity of ultrasound is very operator-dependent, and although the specificity of TAB is very high (near 100%), it lacks good sensitivity for diagnosis (only around 40%) [4]. Therefore, the presence of raised inflammatory markers—erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP)—is a crucial aid to support the diagnosis of GCA [5]. An elevated ESR is one of the five criteria defined by the American College of Rheumatology to classify GCA [6]. In addition, high CRP has proven to be more sensitive than ESR to diagnose CGA [7]. Nevertheless, there have been some reports of patients with GCA and normal or low inflammatory markers before starting treatment, and a low inflammatory response has been associated with a higher risk of developing irreversible ischemic complications, such as vision loss [8, 9].

We aim to describe a case of a biopsy- and ultrasound-proven GCA, presenting with AION and normal CRP and ESR at diagnosis and during follow-up, and to review the literature on GCA cases with low inflammatory markers, with a particular focus on the presence of eye involvement.

Patients and methods

To address the clinical and prognostic implications of normal acute phase reactants in GCA, particularly the occurrence of eye symptoms in these patients, we performed a literature search between January 1975 and July 2019 in PubMed/MEDLINE, Embase, and Cochrane Library databases with the MeSH terms “giant cell arteritis,” “temporal arteritis,” “inflammatory markers,” “C-reactive protein,” “erythrocyte sedimentation rate,” “normal,” and “low.” Reference lists were additionally hand-searched. The search was performed by two authors and was restricted to the English language, resulting in a total of 194 articles as illustrated in the flowchart (Fig. 1). All retrieved articles were screened by title and abstract, and the eligible ones were kept for full-text review, if accessible. Case reports, small case series, and restrospective/prospective studies that described GCA cases with low or normal inflammatory parameters were considered for inclusion. Review articles with no clinical case reports, case reports with elevated inflammatory markers, editorials, and congress reports were excluded. Moreover, two papers were additionally excluded due to inaccessibility, despite the attempt to contact the corresponding authors. Forty-five articles were obtained, summarized in Tables 1 and 3.

Fig. 1
figure 1

Flowchart of the literature review methodology. *Three manuscripts were included in both case reports/series and retrospective studies

Full size image

Case report

A 76-year-old man, with a background history of left eye cataract, type 2 diabetes mellitus, hypertension, and atrial fibrillation on anticoagulant treatment, presented to the Ophthalmology Emergency Department with a 1-week history of subacute decrease in visual acuity (VA) of the right eye (RE), with a recent fast progression to full blindness in the previous 48 h. Additionally, the patient mentioned bilateral temporal headache. There was no accompanied weight loss, fatigue, jaw claudication, or joint pain/stiffness. On physical examination, temporal arteries were pulsatile, non-tender, and non-indurated. Ophthalmic examination disclosed RE best corrected visual acuity (BCVA) of “counting fingers” (CF) at 30 cm with a right relative afferent pupillary defect and BCVA of 6/10 on the left eye (LE). Biomicroscopy revealed bilateral cataracts. Fundus examination disclosed blurred optic disc margins and discrete disc hemorrhages on the RE and defined contours of the left optic disc with bilateral arteriolar narrowing (Fig. 2a). Optical coherence tomography (OCT) of the peripapillary retinal nerve fiber layer (RNFL) showed diffuse increased thickness with a mean thickness of 224 μm and LE with a mean thickness of 102 μm, confirming right optic disc edema (Fig. 2b).

Fig. 2
figure 2

Temporal artery biopsy. a and b Cross section of muscular right artery with irregular thickening and intimal fibrosis (blue line), lumen reduction (L), and presence of recent thrombus (RT) (H&E stain, ×40). c and d Disappearance (green arrow), reduplication, and fragmentation (blue arrow) of the elastic fibers of the internal elastic membrane (H image amplification), stained black in Verhoeff stain (×40). e Reduction of thickness of the middle layer (black arrow) and slight chronic inflammation in the artery wall (H&E, ×100). f and g Rare multinucleated giant cells, in relation to elastic fibers (red arrow) (H & E, 200X). H&E, hematoxylin and eosin stain

Full size image

Laboratory investigation was unremarkable for full blood count; ESR was 5 mm/h and CRP 0.5 mg/dL. Computed tomography of the brain and orbits excluded relevant abnormalities, and US of the carotid arteries documented atherosclerotic plaques without hemodynamic repercussion. Due to ocular involvement, the patient was started on methylprednisolone pulses of 1 g for 3 days, followed by 60 mg of oral prednisolone with progressive tapering. A biopsy specimen of the right temporal artery was obtained which was compatible with vasculitis, revealing luminal narrowing, mononuclear cell infiltration, giant cells, and destruction of the internal elastic membrane (Fig. 3). While waiting for the TAB results, US of the temporal and axillary arteries was performed and showed a halo sign on the left common superficial temporal artery and frontal and parietal left branches with a positive compression sign characteristic of vasculitis (the right temporal artery had already been biopsied) (Fig. 4). A diagnosis of GCA with RE arteritic AION was established.

Fig. 3
figure 3

Ultrasound of the temporal arteries. a Presence of halo sign (arrows) in the parietal branch of the temporal artery, in longitudinal view. b Presence of halo sign (arrows) in the frontal branch of the temporal artery, in transverse view

Full size image
Fig. 4
figure 4

Retinography and optical coherence tomography (OCT) during follow-up. a Retinography at presentation, note right eye with blurred disc margins. b OCT of the optic disc, at presentation, the retinal fiber layer thickness (RNFL) is increased on the right eye (RE). c Retinography 1 month later, note RE optic disc margins more defined. d OCT of the optic disc, 1 month later, the RNFL is still increased in the RE although less thickened compared to the image above. e OCT of the optic disc, 4 months later, the RNFL thickness was decreased in the RE

Full size image

One month after treatment, there was a mild symptomatic improvement. The patient no longer complained of headache, and the BCVA of the RE improved from CF at 30 to 50 cm. Fundus examination disclosed improvement of the right optic disc edema (Fig. 2c), and after 2 months, the RE presented a temporal optic disc pallor. The OCT showed a progressive decrease of the mean RNFL thickness from 224 to 112 μm after 1 month (Fig. 2d), and after 2 months, optic atrophy appeared, with a mean thickness of 61 μm. The inflammatory parameters remained normal throughout treatment. Reevaluation with US at 4 months revealed full disappearance of the halo sign in all observed branches, and the OCT of the optic disc showed progressive atrophy of the RNFL of the RE (Fig. 2e).

The patient was started on methotrexate 15 mg/week as a steroid-sparing agent and tapered prednisolone to 15 mg/day 6 months after diagnosis.

Review of the literature

The literature search identified 31 case reports and small case series, describing a total of 41 patients with the diagnosis of GCA and low or normal inflammatory markers, 15 retrospective studies which included 1533 patients with GCA, and two prospective studies including a total of 103 patients with GCA [7, 10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53]. In 3/15 of the retrospective studies, detailed case reports were also extracted.

Regarding the case reports and case series (Table 1), there was a predominance of female individuals (63%), mean age 69.4 ± 12.5 years, with headache and eye involvement being the most frequent clinical manifestations reported (81% and 59%, respectively). Diplopia and visual loss were the most common eye symptoms described, and 14/24 (58%) patients with visual involvement were reported to have visual loss. Less than one quarter of the patients complained of jaw claudication, polymyalgia rheumatica (PMR), or constitutional symptoms (Table 2). The average ESR reported was 20.5 ± 11.9 mm/h; however, its definition of normal vs. low varied from case to case. CRP was only widely available for most centers around the 1990s; therefore, just 11/41 cases reported both values of ESR and CRP: 6/11 with CRP bellow 0.8 mg/dL, in which 5/6 also presented ESR < 50 mm/h. To support the diagnosis of GCA, TAB was performed in 39/41 cases, with a result compatible with vasculitis in 38/39 cases. In the most recent publications, US was also included as a diagnostic modality, having been performed in three patients (all positive).

Table 2 Clinical features of patients with GCA described in the case reports/series (n = 41 patients)
Full size table
Table 1 Clinical features of patients with GCA and low or normal inflammatory markers: case reports and small case series
Full size table

Considering the retrospective studies (Table 3), the presence of decreased or normal inflammatory markers varied from 1.4 to 22.5%; in the study by Wong and Korn, all 36 patients included had an ESR ≤ 40 mm/h [18]. The total number of untreated patients with GCA and ESR < 50 mm/h was reported in 12/15 studies and corresponded to 96/1175 patients (8.2%). In only 5/15 studies, CRP was also assessed together with ESR; in 2/5 studies, the total number of untreated patients with low CRP was reported, corresponding to 4/138 (2.9%) cases [49, 51]. In five studies [7, 18, 46, 47, 50], differences in visual symptoms between patients with normal/low inflammatory markers and elevated inflammatory markers were evaluated showing no significant results. In addition, in the study by Ellis and Ralston, there was no significant correlation between presenting ESR and development of visual loss [44]; however, Hayreh and Zimmerman reported that patients without visual loss had significantly higher ESR than patients with visual loss, although the same could not be found for CRP [48]. Moreover, in three studies [21, 43, 49], all patients with normal/low inflammatory markers were reported to have visual disturbances, although no formal comparison was made between patients with low and high inflammatory markers.

Regarding the two prospective studies, 2/103 (1.9%) patients presented normal or low values of ESR; CRP was only assessed in one study with normal values reported in 1/54 (1.9%) patient [52]. No reference to association of eye involvement with ESR/CRP was mentioned.

Discussion

One of the most feared complications in GCA is visual loss, which has already been associated with less constitutional symptoms and lower inflammatory markers [8, 9, 54]. In clinical practice, physicians tend to heavily rely on inflammatory markers to diagnose patients with GCA and therefore decide on treatment initiation, given TAB results are not readily available and not all centers have quick access to high-quality US. However, up to 22.5% of patients with GCA may present with an ESR ≤ 30mm/h [44].

We presented a case report of a patient with biopsy- and US-proven GCA, with AION, associated with persistently low CRP and ESR. We found similar case reports in the literature (Table 1), but most of them were published many years ago and just mentioned the values of ESR (30/41). In the 11 cases in which CRP measurement was also available, only 5/11 reported low CRP together with low ESR at disease presentation [29, 32,33,34,35]. In addition, in 2/5 cases [33, 35], inflammatory markers increased during follow-up. Thus, it is possible that if the older cases had CRP measurement available, they might have not been considered as having “normal inflammatory parameters.” ESR values are known to vary with age, sex, and race [55, 56] making them hard to standardize. Moreover, there is variability of the ESR cutoffs used to assume normal values: in 11/41 cases of patients considered to have normal ESR, the reported values were ≥ 30mm/h [19, 21, 22, 24, 31, 33, 37], which might have not been actually normal values for those specific individuals, and its decrease after treatment initiation for most patients supports this hypothesis. Furthermore, in many of the reported cases, there was no comment on prior steroid treatment, even if taken for other diseases. For example, in the report by Newman [11], the patient had been taking prednisolone 10 mg/day due to asthma during the 3 years previous to the GCA diagnosis, which might have contributed to the low ESR of 6 mm/h. In most cases, there was no reference to the repetition of the dosing of inflammatory markers without treatment, or if that repetition had even been performed, which could have showed increased values over time [18, 19, 22, 31, 33, 35]. Although it is known that GCA usually affects patients older than 50 years, in 4/41 cases [14, 17, 26, 27], patients were younger, with ages between 30 and 48 years, which is exceptional and raises the possibility of alterative diagnosis. In addition, in contrast to our case, not all patients had biopsy- or US-proven GCA [11, 19, 23], which can be regarded as mandatory to diagnose GCA in the setting of low inflammatory markers.

Nevertheless, like in our case, 24/41 (59%) of patients had ocular involvement, which is a higher percentage than the 20–50% of visual manifestations usually reported for GCA [54, 57,58,59]; vision loss occurred in half of the cases. Systemic manifestations, such as PMR or constitutional symptoms, and jaw claudication were reported less frequently and in contrast with the 50% usually mentioned in the literature [60]. This high percentage of visual involvement with few systemic features in patients and low or normal inflammatory markers could be potentially explained by (1) the localized nature of this type of disease, mostly confined to the eyes, and therefore not leading to an increase in inflammatory markers and (2) delay in recognizing the disease due to low ESR and/or CRP and therefore allowing it to progress to visual loss or other ocular manifestations. However, when comparing patients with normal/low inflammatory markers and patients with elevated inflammatory markers in terms of visual symptoms, no differences were found in five of the retrospective studies retrieved [7, 18, 46, 47, 50]. In addition, in a retrospective study by Liozon et al., including 339 patients with GCA, there was no laboratory variable independently associated with permanent visual loss [58]. Lopez-Diaz et al. showed that only ESR values between 70 and 100 mm/h predicted visual ischemic complications [50]. Only a single retrospective study (Table 3) showed that patients without visual loss had higher ESR than patients with visual loss, although the same could not be verified for CRP [48]. Therefore, although many of the case reports with low/normal laboratory values showed visual symptoms and there have been some previous associations between eye involvement and low inflammatory markers [8, 9, 54], the link between low ESR/CRP and visual manifestations is still debatable and needs further testing, preferably in prospective studies, with TAB- or US-confirmed GCA, and using CRP together with ESR given its higher sensitivity for diagnosis [7, 49, 57].

Table 3 GCA with low or normal inflammatory markers: restrospective and prospective studies
Full size table

When there is a suspicion of GCA, physicians should not exclude this diagnosis based on normal/decreased values of ESR or CRP. All efforts should be made for patients to undergo a TAB and/or US and be observed by ophthalmology, given the high percentage of visual involvement in this setting. In our case, US allowed a fast track diagnosis of GCA, given TAB results were not readily available. If the patient had been referred sooner, for example, immediately after the first visual symptoms started, progression to full blindness might have been avoided [61, 62]. In addition, given that the negative inflammatory markers precluded correct monitoring of treatment response during follow-up, US and regular ophthalmologic observation with retinography and OCT were crucial.

Conclusions

In summary, both CRP and ESR are important in the diagnosis of GCA; occurrence of this disease with normal acute phase reactants is unusual but does not rule out the diagnosis. Headache and visual manifestations were the most common symptoms reported in these cases; therefore, patients with the clinical suspicion of GCA require immediate US and/or TAB (with US having the advantage of allowing prompt results) for correct diagnosis and treatment initiation in order to avoid permanent visual loss.