Does tocilizumab eliminate inflammation in GCA? A cohort study on repeated temporal artery biopsies

Abstract

Background

Vascular inflammation persists in temporal artery biopsy (TAB) of giant cell arteritis (GCA) patients even after prolonged glucocorticoid (GC) therapy. We aimed to evaluate the histological impact of adding tocilizumab (TCZ) to GCs.

Methods

We enrolled all consecutive GCA patients with an inflammed TAB at diagnosis who were treated with TCZ and GCs for ≥6 months and followed from December 2017 to December 2023. Within 2 weeks, all patients underwent a second TAB, positron emission 18-fluorodeoxyglucose tomography/CT (PET/CT) and vessel colour Doppler ultrasonography (CDUS). Results were compared with pretreatment findings.

Results

13 patients repeated TAB after a median TCZ treatment of 2.4 years (Q1–Q3: 1.2–3.9 years). The first TAB showed transmural inflammation (TMI) in 11/13 patients (84.6%), inflammation limited to adventitia (ILA) in one patient (7.7%) and small vessel vasculitis (SVV) in another (7.7%). On repeated TABs, five patients (38.5%) still showed some degree of inflammation. Among the 11 patients with initial TMI, 2 had ILA, 1 had TMI, 1had SVV and 1 had vasa vasorum vasculitis at the second TAB. Nine patients had active vasculitis at baseline PET/CT, and three (33.3%) still showed activity at the last PET/CT, with a relevant reduction in mean PET vascular activity score (−6.5; 95% CI 1.54 to 11.45; p=0.017). The repeated quantitative CDUS revealed altered parameters suggestive of vasculitis in temporal arteries in about one-third of the patients.

Conclusion

Our study, using pathological and imaging assessments, revealed that after TCZ and GCs, over one-third of patients still presented with vascular inflammation.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• In giant cell arteritis (GCA), vascular inflammation in the temporal artery (TA) may persist for over a year or longer despite glucocorticoid (GC) treatment. Data on persistence of inflammation in TA for those on combined tocilizumab (TCZ) and GC therapy are lacking.

WHAT THIS STUDY ADDS

• This study found persistent TA inflammation in over one-third of GCA patients on TCZ and GCs for at least 6 months. Positron emission tomography/CT and ultrasound confirmed presumable vascular inflammation in one-third of these patients.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• TCZ and GC treatments do not fully eliminate vascular inflammation in GCA, underscoring the need for new, more effective anti-inflammatory therapeutic strategies.

Giant cell arteritis (GCA) is the most common vasculitis in individuals over 50 years old in Western countries. It typically affects large and medium-sized arteries, particularly the proximal aorta and its main branches, leading to a wide spectrum of disease phenotypes, from cranial-predominant to large vessel disease. Permanent visual loss is a feared complication in the early stages, while thoracic aorta aneurysm is a significant concern in the late stages of the disease.^{1 2}

Temporal artery biopsy (TAB) is the gold standard for diagnosing GCA. Both the European Alliance of Associations for Rheumatology (EULAR) and American College of Rheumatology (ACR) recommendations emphasise the importance of TAB in diagnosing suspected GCA, as a positive TAB provides the highest specificity for classifying GCA patients.³,⁵

The typical histopathological picture of GCA is characterised by a transmural infiltrate where inflammation is often more pronounced between the media and adventitia. This infiltrate primarily consists of T lymphocytes and macrophages. Giant cells are prominently found along the internal elastic lamina and are present in approximately 75% of TAB specimens. Fragmentation of the internal elastic lamina is common, and there is often thickening of the intima due to cellular myofibroblastic proliferation, which can variably occlude the lumen.⁶ In about 20% of inflamed TABs, inflammation is restricted to the adventitia, surrounding vasa vasorum or periadventitial small vessels, without the involvement of the media.⁷

Glucocorticoids (GCs) are the standard treatment for GCA. High-dose GCs are initially recommended to prevent severe ischaemic complications and induce remission. However, relapse rates range from 35% to 75% in GCA patients treated with GCs.⁸,¹³ Each disease relapse typically prompts an increase in GC dosage.^{3 4} The high cumulative dose of GCs and its associated side effects highlight the necessity for effective treatments capable of sparing or substituting GCs.

Among GC-sparing agents, tocilizumab (TCZ) is recommended either from the disease onset or at relapse, according to various guidelines.^{3 4} TCZ is a recombinant monoclonal antibody directed against the interleukin-6 receptor (IL-6R), blocking the signalling of this proinflammatory cytokine, which is crucial in the pathogenesis of the disease. Preclinical and clinical studies have shown increased serum levels of IL-6 and high expression of IL-6 in the TAB specimens of GCA patients.^{14 15} Two randomised controlled trials have demonstrated that TCZ is effective in achieving sustained remission and has a significant GC-sparing effect in both relapsing and new-onset GCA.^{16 17}

Evaluating disease activity in GCA patients treated with TCZ is particularly challenging since laboratory tests, specifically inflammatory markers like C reactive protein (CRP), are affected by the biological activity of TCZ and are usually normal. Additionally, more than 50% of patients in remission experienced flares when this biological agent was stopped after 1 year of therapy, suggesting that TCZ temporarily controls but does not resolve vascular inflammation in these patients.¹⁸,²¹ Similarly, despite clinical and laboratory remission, 60% of patients with biopsy-proven GCA showed persistent vascular inflammation in second TAB specimens during GC therapy.²² No study has yet evaluated the inflammatory state of temporal arteries at different times using a second TAB in patients during TCZ and GC therapy.

The aim of this study was to assess the residual degree and composition of vessel inflammation in GCA patients treated with TCZ, in addition to GCs, by repeating histological evaluations of the temporal artery and conducting large vessel imaging, along with clinical evaluations. These patients received TCZ for at least 6 months prior to the repeated TAB.

Patients and methods

Study design and patients

This retrospective single-centre cohort study included consecutive patients with a diagnosis of biopsy-confirmed GCA treated with GCs and TCZ, who were followed prospectively at the Rheumatology Unit of Reggio Emilia Hospital (Italy). Although this study is retrospective, all patients were uniformly evaluated and treated prospectively. The observational prospective cohort of GCA patients at the Reggio Emilia Hospital was established in 1986 to study the course and prognosis of patients with GCA. For this study, all TABs performed for a suspected GCA between December 2017 and December 2023 at our institution were reviewed and patients satisfying all the following criteria were selected:

• Evidence of inflammation at TAB.

• Baseline 18-fluorodeoxyglucose positron emission tomography/CT (18FDG PET/CT).

• Ongoing treatment with TCZ for more than 6 months.

All patients satisfying the inclusion criteria were contacted and were proposed to participate in this study by repeating contralateral TAB, PET/CT scan and colour Doppler ultrasonography (CDUS) of the temporal arteries, supra-aortic trunks and axillary arteries. All patients agreed to participate and signed informed consent.

All proposed examinations were performed within 2 weeks. The results of the repeated biopsy, PET/CT and CDUS imaging were compared with those obtained at diagnosis.

Clinical and laboratory methods

Medical records of the study participants were retrospectively reviewed from the date of diagnosis to the end of the study period (31 December 2023). Information about demographics, clinical manifestations, laboratory exams and treatment data was collected at diagnosis and during each follow-up visit.

Since 2017, all patients referred to Reggio Emilia Hospital for suspected GCA are typically assessed within 24 hours, and CDUS of the temporal and axillary arteries is carried out at the first visit. TAB and PET/CT are routinely performed within 14 days from the initial referral in all patients where the clinical suspicion of GCA is confirmed. Follow-up includes rheumatological visits and acute phase reactant determinations every 3 months, along with yearly imaging evaluations using CDUS, CT angiography, MR angiography or PET/CT to assess large-vessel involvement. Additional imaging is performed if flares or vascular damage occur. Our standard protocol includes acute phase reactant determination and a physical examination by a rheumatologist 24–48 hours before imaging.

Clinical active disease was defined by at least one of the following criteria: the presence of signs or symptoms of GCA or polymyalgia rheumatica (PMR); elevation of inflammatory indices, specifically CRP ≥10 mg/L or erythrocyte sedimentation rate (ESR) ≥40 mm/1st hour, if these elevations were considered by the investigator to be due to GCA; or the necessity for the introduction or increase in GCs dose. Patients whose disease was judged as clinically active at follow-up visits or whose imaging was suggestive of active vasculitis were considered relapsing. Remission was defined as the absence of any clinical symptoms directly attributable to vasculitis, including normalisation of CRP and ESR.

Histopathological specimen examination

TABs were performed by the same vascular surgeon (GM), under local anaesthesia on the contralateral temporal artery to the previously biopsied one. All patients provided informed consent for the surgical procedure. The biopsies, ranging in length from 0.5 cm to 0.8 cm, were then transversally sectioned and routinely fixed in formalin. All histological samples, including the first biopsy, were reviewed by a single pathologist (AC) who is an expert in vasculitides. Inflamed TABs were classified into four patterns based on the localisation of inflammation: transmural inflammation (TMI), inflammation limited to adventitia (ILA), vasa vasorum vasculitis (VVV) and small vessel vasculitis (SVV).⁷ The presence or absence of giant cells, laminar necrosis and calcifications was evaluated for each biopsy specimen by the same pathologist. The severity of intimal hyperplasia was graded semiquantitatively as mild (<25% reduction in lumen diameter), moderate (25%–75% reduction) or severe (>75% reduction). The entity of inflammation was graded on a qualitative scale into mild, moderate and severe.

18-FDG PET/CT and CDUS

All patients underwent whole-body 18-FDG PET/CT scans at baseline and at the time of the second biopsy. All PET-CT examinations were acquired by using a hybrid PET/CT machine (Discovery, GE Healthcare, Chicago, Illinois, USA) with a 2 min emission scan per bed position and CT attenuation correction. Patients were required to fast for at least 4 hours before intravenous injection of 2.5 MBq of 18F-FDG per kg of body weight. Blood glucose levels before tracer injection were <200 mg/mL in all cases. A low-dose, non-contrast-enhanced CT scan was performed for PET coregistration.

All available PET/CT scans, including those performed at the time of the first biopsy, were reviewed by a single nuclear medicine physician expert in large vessel vasculitis (LVV) (AV). This physician determined whether the findings were consistent with active or inactive vasculitis using the visual 0–3 vascular to liver FDG uptake grading scale: 0=no uptake; 1=low-grade uptake (<liver); 2=intermediate-grade uptake (=liver) and 3=higher-grade uptake (>liver).²³ Scans showing a grade ≥2 FDG uptake in at least one vascular district, other than femoral arteries, were considered consistent with active LVV.²⁴

Additionally, a visual assessment of FDG uptake relative to liver uptake was performed for nine arterial territories (ascending aorta, aortic arch, descending thoracic aorta, abdominal aorta, subclavian arteries, carotid arteries and brachiocephalic trunk). The degree of FDG uptake in these arteries was visually assessed relative to liver FDG uptake, and a PET vascular activity score (PETVAS) of 0–27 points was calculated by adding the qualitative scores of the nine arterial territories.²⁵ In a recent study involving two centres, two nuclear medicine specialists (including the one involved in the present study) reviewed the PET/CT scans of patients with LVV by using the 0–3 visual grading scale and the PETVAS, inter-reader reliability was >95%.²⁶

All patients underwent CDUS of the temporal arteries (common, parietal and frontal branches), supra-aortic trunks (common carotid, subclavian arteries) and axillary arteries at the time of the second biopsy. All examinations were performed by the same ultrasonographer (PMacchioni) with extensive experience in the field. CDUS exams performed at the time of the first biopsy were also reviewed. The exams were performed using an Esaote MyLab ClassC device (Genua, Italy) with high-frequency probes (12–22 MHz for the temporal arteries and 6–13 MHz for the supra-aortic trunks). The maximum intima–media thickness (IMT) was recorded in all vascular segments.²⁷ IMT was measured longitudinally in the wall distal to the probe and in the area with the greatest thickness, avoiding areas with atherosclerotic appearance.²⁸ An abnormal vessel wall thickness indicating vasculitis was defined as >0.4 mm in temporal artery segments and >1.0 mm in the axillary artery.^{27 29} Each halo thickness was scored according to the halo score grading, and the total halo score was calculated as the sum of the halo scores of the temporal branches plus the axillary arteries, multiplied by three. The halo score correlated positively with CRP levels, and a total halo score ≥10 firmly established the diagnosis of GCA, also showing potential to discriminate between remission and relapse in GCA.^{30 31}

The Outcome Measures in Rheumatology GCA ultrasonography score (OGUS) was calculated for all available US examinations by summing the IMT in each segment and dividing it by the rounded cut-off values of IMT for each segment.³² The rounded cut-off values are 0.4 mm for the common trunk of the superficial temporal arteries, 0.3 mm for the parietal and frontal branches and 1.0 mm for the axillary arteries. The total sum was then divided by the number of segments available. An OGUS >1 indicates consistency with vasculitis from an US standpoint.³¹,³³

The nuclear medicine physician and the sonographer were blinded to the clinical data at the time of the second biopsy.

Statistical analysis

A descriptive analysis of the data was conducted. Continuous variables were expressed as mean and SD for normally distributed data and as median and first and third quartiles (Q1–Q3) for non-normally distributed variables. Categorical variables were presented as percentages and frequencies. To compare PETVAS at the time of the second biopsy with baseline values, the Wilcoxon signed-rank test for paired samples was used. Due to the absence of a predefined sample size and power calculation, no specific significance threshold for p values was set to reject or accept the null hypothesis. Therefore, p values should be interpreted as continuous variables representing the probability that an observed difference would occur under the null hypothesis.

Results

Characteristics of the patients

13 GCA biopsy-positive patients were included. All patients satisfied the ACR/EULAR classification criteria for GCA.⁵ The characteristics of the patients with GCA diagnosis are reported in table 1. Pre-existing comorbidities were present in 11 patients (84.6%), with dyslipidaemia being the most prevalent (46.1%), followed by arterial hypertension (38.4%). At disease onset, cranial manifestations were present in 10 patients (76.9%), including visual symptoms in 3 patients (23.1%) and permanent visual loss in 1 patient (7.7%). None of the patients experienced cerebrovascular events. Systemic symptoms were recorded in 10 patients (76.9%). Articular symptoms, including PMR and peripheral synovitis, were observed in six patients (46.1%).

Table 1. Characteristics and histological findings of the 13 GCA patients at diagnosis.

Variables	Study cohort
Age at disease onset, mean (SD), years	69.3 (8.1)
Male/female	6 (46.1)/7 (53.8)
Comorbidities	11 (84.6)
Dyslipidaemia	6 (46.1)
Arterial hypertension	5 (38.4)
Diabetes	1 (7.6)
Ever smoker	4 (30.7)
Articular symptoms*	6 (46.1)
Cranial manifestations	10 (76.9)
Visual manifestations†	3 (23.1)
Systemic manifestations‡	10 (76.9)
ESR, mean (SD), mm/hour	48.2 (38.6)
CRP, mean (SD), mg/dL	10.25 (7.1)
Active PET/CT§	9 (69.2)
PETVAS mean (SD)	14.5 (8.6)
OGUS mean (SD)¶	1.13 (0.28)
Halo score, mean (SD)¶	27.1 (9.1)
TMI	11 (84.6)
ILA	1 (7.7)
SVV	1 (7.7)

Values are the number (%), except where otherwise indicated.

^*pPresence of polymyalgia rheumatica and/or peripheral synovitis;.

^†pPresence of at least one of the following: amaurosis fugax, permanent visual loss, diplopia;.presence of at least one of the following: fatigue, anorexia, weight loss of at least Kg, or fever;

^‡ and axillary Color Doppler Ultrasonography at diagnosis was performed in ;Presence of at least one of the following: fatigue, anorexia, weight loss of at least 4 kg or fever.

^§Presence of grade ≥2 FDG uptake in at least one vascular district.

^¶Temporal artery and axillary colour Doppler ultrasonography at diagnosis was performed in 7 patients.

CRPC reactive proteinESRerythrocyte sedimentation rateFDGfluorodeoxyglucoseGCAgiant cell arteritisILAinflammation limited to adventitiaOGUSOutcome Measures in Rheumatology (OMERACT) ultrasonography scorePETpositron emission tomographyPETVASPET Vascular Activity ScoreSVVsmall vessel vasculitisTMItransmural inflammation

All patients received a prednisone (PDN) dose of 50 mg/day at the time of diagnosis, coinciding with when the biopsy results were obtained. TCZ was introduced at diagnosis in three patients (23.1%), at relapse in nine patients (69.2%) and due to GCs intolerance/side effects in one patient (7.7%). Three patients were judged clinically active at the last visit before TAB repetition. At the time of the second biopsy, the median disease duration was 3.7 (Q1–Q3: 2.2–4.3) years. All patients were treated with TCZ at the time of the second biopsy. The median treatment duration with TCZ was 2.4 (Q1–Q3: 1.2–3.9) years, and 10 patients (76.9%) were still on low doses of GCs (mean PDN dose (SD): 2.5 (2.7) mg/day).

Histological evaluation

The histopathological changes of TABs at baseline and after TCZ treatment, and their relationship to clinical disease activity, are shown in figure 1.

Figure 1. Histological modifications of temporal artery biopsies during tocilizumab and glucocorticoid treatment: relationship with disease activity. Data are reported as % (n) . GCA, giant cell arteritis; ILA, inflammation limited to adventitia; NEG, negative; SVV, small vessel vasculitis; TMI, transmural inflammation; VVV, vasa vasorum vasculitis.

At the first TAB, 11 out of 13 patients (84.6%) presented with TMI, 1 patient (7.7%) with ILA and 1 patient (7.7%) with SVV. Nine patients (69.2%) were under GC treatment, with a mean (SD) PDN dose of 17.5 (20.2) mg/day and a median PDN duration of 12 (Q1–Q3: 0–15) days. The degree of inflammation in TMI cases was judged as mild in six patients (54.6%), moderate in three patients (27.3%) and severe in two patients (18.2%). Giant cells, laminar necrosis and calcifications were observed in seven TMI cases (63.7%), three TMI cases (27.3%) and two TMI cases (18.2%), respectively. Intimal thickening was graded as mild in two TMI cases (18.2%), moderate in five TMI cases (45.6%) and severe in four TMI cases (36.4%). No intimal thickening or other pathological features were observed in the ILA and SVV specimens.

A total of 5 out of 13 after-treatment TABs (38.5%) still presented signs of inflammation at the second biopsy. Among the 11 patients with TMI at the first TAB, 5 (45.4%) still had inflammation at the second biopsy: 2 (18.2%) had ILA, 1 (9.1%) had VVV, 1 (9.1%) had SVV and 1 (9.1%) had persistent TMI (figure 2). The two biopsies with inflammation at baseline restricted to adventitial/periadventitial tissue showed no evidence of inflammatory infiltration at the second TAB. Giant cells, laminar necrosis and calcifications were no longer observed in any patients except for the one with persistent TMI, who still presented a mild inflammatory lymphocytic infiltrate with giant cells.

Figure 2. Histological patterns at the first and at the second temporal artery biopsies of the four patients with giant cell arteritis and persistent inflammation of temporal arteries. (A, C, E, G) Evidence of transmural inflammation at the first biopsy in all four patients; (B, D, F, H) Respective histological pattern at the second biopsy: transmural inflammation (B), inflammation limited to adventitia (D), small vessel vasculitis (F) and vasa vasorum vasculitis (H) (H&E, ×200–400).

Among the three patients judged as clinically active at the time of the second biopsy: one TAB showed TMI, while the other two did not display inflammatory signs at histological evaluation.

18-FDG PET/CT and CDUS evaluations

Nine out of 13 patients (69.2%) had evidence of LVV activity at baseline PET/CT. They were all on GCs at the first PET (mean PDN dose (SD): 11.7 (15.4) mg/day; median PDN duration: 7 (Q1–Q3: 0–19) days). PET/CT scans repeated at the time of the second biopsy showed active LVV in 3 out of 13 patients (23.1%) patients. At that time, 9 out of 13 patients (69.2%) patients were still on low-dose GCs (mean PDN dose (SD): 2.1 (2.7) mg/day; median PDN duration: 78 (Q1–Q3: 0–115) days). The difference observed in PETVAS between baseline and the time of the second biopsy was small (mean ∂PETVAS=−3.36; 95% CI −1.67 to 8.39; p=0.167) in the whole cohort and compatible with random fluctuation. Among the nine patients with active vasculitis on PET/CT at baseline, three (33.3%) showed persistent active vasculitis on PET/CT scans performed at the time of the second biopsy. Only one of these three patients had clinically active disease.

When we evaluated the reduction in the mean PETVAS just among patients with an active PET/CT at baseline, we found a reduction that is unlikely to be due to chance (mean ∂PETVAS=−6.5; 95% CI 1.54 to 11.45; p=0.017).

Among the seven patients with baseline CDUS available, six (85.7%) presented temporal artery IMT >0.4 mm, 5 (71.4%) had IMT >1 mm at the axillary arteries, and all presented a total halo score ≥10. Four (57.1%) patients had an OGUS score >1. A significant reduction in the mean (SD) halo score and mean (SD) OGUS was observed at the time of the second biopsy compared with the baseline examination (27.1 (9.11) vs 16.0 (6.8), p=0.002; and 1.13 (0.28) vs 0.82 (0.16), p=0.007, respectively). Considering all 13 patients, at the time of the second biopsy, 12 (92.3%) had a halo score ≥10 and 2 (15.4%) had an OGUS >1. IMT values were indicative of vasculitis in the temporal arteries in four (30.8%) patients and in the axillary arteries in four (30.8%) patients. Six (46.2%) patients did not have IMT indicative of vasculitis in any of the examined segments.

Features of the five patients with a second inflamed TAB

Table 2 outlines the characteristics of the five patients who retained a second TAB with evidence of inflammation. Three patients had PET/CT scans showing an FDG uptake grade of ≥2 in at least one vascular district. Notably, the patient with TMI had an FDG uptake grade of 3, which also correlated with clinical evidence of active vasculitis. All patients had an OGUS <1, along with normal ESR and CRP values.

Table 2. Features of the five patients with an inflamed TAB at the time of the second biopsy.

	Second TAB histology	PET/CT (Meller scale)	CDUS (OGUS)	Clinical disease activity	ESR (mm/hour)	CRP (mg/dL)
Patient 1	TMI	3	0.85	Active	2	0.05
Patient 2	VVV	0	0.85	Inactive	2	0.05
Patient 3	ILA	0	0.95	Inactive	2	0.05
Patient 4	ILA	2	0.86	Inactive	2	0.05
Patient 5	SVV	2	0.82	Inactive	9	0.05

CDUScolour Doppler ultrasonographyCRPC reactive proteinESRerythrocyte sedimentation rateILAinflammation limited to adventitiaOGUSOutcome Measures in Rheumatology (OMERACT) ultrasonography scorePETpositron emission tomographySVVsmall vessel vasculitisTABtemporal artery biopsyTMItransmural inflammationVVVvasa vasorum vasculitis

Discussion

TMI and other histopathological features of GCA in temporal arteries may persist in patients treated with GCs from 1 to 12 months, according to various studies.^{22 34} This is the first study assessing the histopathological changes in temporal arteries in GCA patients treated with both TCZ and GCs. Assessing the impact of TCZ and GC therapy on histopathological changes in GCA patients is critical, as it represents the best way to determine if these drugs can resolve vascular inflammation. There is a lack of studies directly measuring this effect. Our results partially align with those of Maleszewski et al, who found vasculitis in the second TAB in 44% of patients 12 months after diagnosis, when GCA was clinically non-active and patients were treated with a median dose of 5 mg of PDN daily.²²

Despite a longer treatment duration and the addition of TCZ, we still found histological inflammatory signs in temporal arteries in almost 40% of patients. However, some important differences were present between the two studies. We had only one patient with TMI at the second TAB, while the other four patients had inflammation restricted to the adventitial/periadventitial tissue. Therefore, considering that 11 patients had TMI at diagnosis, only 9.1% of our patients had media inflammation at the second TAB. In contrast, in the Mayo Clinic study, all four patients with a 12-month second TAB representative of active arteritis had media inflammation. Isolated inflammation restricted to the adventitial/periadventitial tissue was not considered active arteritis in the Mayo Clinic study, as stated in the 2023 consensus statement on the processing, interpretation, and reporting of TAB for arteritis.³⁵

The significance of this more limited inflammation is controversial and remains to be established.³⁶,³⁸ Some authors consider it as part of the pathological spectrum of GCA because it has been associated with cranial GCA manifestations, including severe ischaemic symptoms, while others view it as a finding associated with ageing, not indicative of an inflammatory process. Some authors also hypothesise that a longer duration of GC treatment and higher GC doses could be associated with these atypical TAB findings.³⁹ To further complicate matters, the prognostic significance of the more limited inflammation found in repeat TABs of patients with established GCA treated with GCs and TCZ is unknown and has never been evaluated. In the present study, we described TABs showing these restricted lesions as inflamed. Supporting this view, three of our patients exhibited adventitial inflammation on their second biopsy, which more closely resembles TMI than periadventitial inflammation in terms of diagnostic value and gene expression profile.^{40 41} Additionally, all of these patients initially presented with TMI on their first biopsy, reinforcing the idea that adventitial inflammation may be part of the GCA pathological spectrum.

Regarding the PET/CT findings, our results are in line with those of Quinn et al, who reported the effectiveness of TCZ in lowering FDG vascular activity, using PETVAS, after 2 years of TCZ treatment in GCA patients with active vasculitis on PET/CT scan at baseline visit.⁴² However, in both studies, one-third of the PET/CT scans showed persistent active vasculitis at the follow-up PET/CT during TCZ therapy. Two other studies suggested that TCZ did not resolve arterial FDG uptake in a significant portion of patients.^{26 43} In a Spanish study, although most LVV-GCA patients achieved clinical remission after 1 year of TCZ therapy, less than one-third showed normalisation of 18F-FDG vascular uptake.⁴³ Similarly, an observational German-Italian retrospective study, demonstrated a comparable reduction in vascular inflammation using PETVAS in patients with new-onset, active LV-GCA under different treatments (PDN monotherapy or in combination with methotrexate or TCZ). However, after 1 year of treatment, only 42% of follow-up PET/CT scans were graded as inactive by nuclear medicine physicians.²⁶

At quantitative US examination, almost one-third of our patients had an IMT indicative of vasculitis in the temporal arteries at the time of the second biopsy. Seitz et al demonstrated that TCZ monotherapy resulted in a progressive decrease in IMT of the temporal arteries and had a smaller effect on the axillary and subclavian arteries.⁴⁴ In this study, the reduction of IMT in temporal arteries was accompanied by decreasing symptoms and clinical remission, highlighting the important role of US in monitoring disease activity. As observed in other studies as a result of treatment,^{31 33 45 46} a reduction unlikely to be due to chance in the mean halo score and mean OGUS was observed in our patients at the time of the second biopsy compared with the baseline examination, clearly indicating a reduction in the extent of arterial inflammation due to the combined treatment with TCZ and GCs as quantified by ultrasonography. However, considering all 13 patients at the time of the second biopsy, the presence of an IMT indicative of vasculitis at quantitative US, particularly in temporal arteries in one-third of our patients and the presence of an OGUS >1 in 15.4% of patients confirms that some inflammation is still ongoing in the arterial walls of our patients despite long-term treatment with TCZ and GCs.^{3133 44},⁴⁶ However, these findings should be interpreted with caution, as the presence or persistence of wall thickening in US may indicate active vasculitis, but could also result from unresolved hyperplasia, representing a non-inflammatory finding.

In only one patient with TMI at the second TAB, there was a correspondence between persistent inflammation on biopsy, persistent activity on PET/CT scans and clinical evidence of disease activity (table 2). Two patients, respectively, with ILA and SVV at the repeated biopsy had evidence of persistent activity on PET/CT; however, the GCA was considered clinically inactive. Overall, there was no consistent correlation between TAB, PET/CT, CDUS and clinical evaluation in defining disease activity at the time of the second biopsy.

Although persistent inflammation was observed in one-third of our GCA patients after at least 1 year of therapy with TCZ and PDN, the long-term impact of ongoing vascular inflammation on further vascular damage has not yet been established. While the significance of the persistent vascular inflammation noted on TAB and CDUS as a predictor of vascular damage remains unclear, previous studies have shown that angiographic progression is often preceded by FDG-PET activity in LVV,^{47 48} suggesting that our observation may have potential clinical relevance.

Our study presents several strengths. First, it is the first to assess the efficacy of TCZ on inflammation at the pathological level in TAB of GCA patients. Second, we performed clinical, histological and imaging examinations simultaneously, providing a comprehensive evaluation of disease activity. Furthermore, we used standardised imaging and pathological protocols, with assessments conducted by the same physician expert in vasculitis, to limit interoperator variability and minimise information bias. In addition, the nuclear medicine physician and the sonographer were blinded to the clinical data at the time of the second evaluation. Lastly, all patients had a long-term follow-up exceeding 2 years and a duration of TCZ therapy of at least 1 year. Our study has also some limitations. First, it was a single-centre and retrospective study, subject to selection and referral bias. Most of our patients were started on TCZ at disease relapse, which may have resulted in the selection of a subgroup of patients with more severe disease. On the other hand, we cannot exclude the possibility that an earlier introduction of TZC (ie, at diagnosis) could have a more profound effect on vessel inflammation. Furthermore, the small sample size and the absence of a comparator group of patients treated only with GCs were also limitations, not allowing us to estimate the course of the histological characteristics in the absence of TCZ treatment. Additionally, the percentage of persistent inflammation in the TA at the second TAB may also vary based on differing interpretations of restricted inflammation as active inflammation.

In conclusion, our pathological and imaging study demonstrated that after 1 year of TCZ therapy, at least one-third of patients still exhibit some degree of vascular inflammation. As Weyand and Goronzy have suggested, while TCZ and GCs effectively block the TH17 cell pathway, they may have only limited impact on the chronic, IL-6- and GC-independent, TH1-cell-mediated vasculitis.⁴⁹ Nevertheless, our findings suggest that persistent histological inflammatory alterations are predominantly limited to the adventitial vessels, while TMI signs have regressed in most patients. This observation opens new scenarios for understanding the interaction between TCZ treatment and the persistence of inflammation at the tissue level. Further analysis of the cytokines produced by the persistent inflammatory infiltrate in repeat temporal artery specimens would help better characterise the phenotype of these immune cells and identify new appropriate treatment.

Data availability statement

Data are available on reasonable request.