Why do temporal arteries go wrong? Principles and pearls from a clinician and a pathologist

Why do temporal arteries go wrong? Principles and pearls from a clinician and a pathologist Abstract Early diagnosis and treatment of GCA is essential to prevent complications of the disease, including permanent vision loss. Temporal artery biopsy has been intrinsically linked with the diagnosis of GCA for several decades. A negative predictive value of > 90% has been reported for temporal artery biopsy; however, a negative result does not reliably indicate the absence of GCA because inflammation of the temporal artery is not always evident because of segmental involvement or other reasons. This is demonstrated by a case study of a patient hospitalized following acute vision loss to the right eye whose glucocorticoid treatment was suspended after temporal artery biopsy revealed no evidence of GCA. The patient subsequently lost sight in the left eye 6 weeks after stopping glucocorticoid therapy. The specificity of temporal artery biopsy for the diagnosis of GCA is variable and influenced by many factors, including length of biopsy specimens, vasculitis in vessels other than the temporal artery (ophthalmic, retinal and posterior ciliary vessels), unilateral versus bilateral biopsy, expertise of the surgeon, interpretation of histology, effects of treatment and confounding factors such as atherosclerosis or other non-GCA diseases that can affect the temporal artery. Considering the limitations of temporal artery biopsy, collaboration and education between the clinician, the pathologist and the patient, taking into account a thorough examination of patient history, recognizing signs and symptoms, and potentially involving newer imaging studies with trained technicians and physicians, are essential in confirming or eliminating diagnosis of GCA. giant cell arteritis, histopathology, primary care rheumatology Rheumatology key messages Temporal artery biopsy has been a cornerstone in diagnosing giant cell arteritis for several decades. A negative temporal artery biopsy result does not reliably preclude the presence of GCA. Collaboration and education between clinicians, pathologists and patients are essential in diagnosing or eliminating GCA. Introduction The first temporal artery biopsy was performed at the Mayo Clinic in 1932 [1], >40 years after GCA was first described in the medical literature [2], indicating how slow progress has been in this disease. Temporal artery biopsy has become intrinsic to the diagnosis of GCA; however, although >85 years have passed since the first procedure was performed, there is much uncertainty regarding the performance, interpretation and test characteristics of temporal artery biopsy. How could a patient have GCA without documentable (and therefore diagnostic) inflammation of the temporal artery? After all, GCA is often referred to as temporal arteritis. Unfortunately, there are many reasons this might happen. Appropriate care of patients with GCA depends critically on gaining insight into these reasons. The stakes are high in making a diagnosis of GCA; the patient’s vision and life are at risk if a diagnosis is missed. Conversely, an incorrect diagnosis of GCA can lead to inappropriate glucocorticoid use and associated morbidity, as well as to undertreatment if the patient has another form of vasculitis (e.g. an ANCA-associated vasculitis). For these reasons, understanding the power and limitations of temporal artery biopsy is crucial to the diagnosis and treatment of patients with suspected GCA. Experience and caution are paramount in the interpretation and reporting of temporal artery biopsy findings and there is great danger in attaching too much certainty to a negative temporal artery biopsy result. This is illustrated by the following case. An illustrative case A 77-year-old man was admitted to hospital twice within a few weeks. The first admission occurred after acute vision loss in the right eye following symptoms of a moderate retro-orbital headache on the left side and amaurosis fugax on the left side, reported as the sensation that his left eyebrow was obscuring his vision. He also noted that his face got tired while chewing food. Two days after the onset of these left-sided symptoms, he awoke blind in his right eye. The patient was evaluated by an ophthalmologist and underwent fluorescein angiography, which revealed poor choroidal circulation and bilateral disc oedema with peripapillary haemorrhages. The patient was diagnosed as having anterior ischaemic optic neuropathy (AION) and was admitted urgently to the hospital for further evaluation. Laboratory examinations revealed normal platelet and white blood cell counts but normochromic, normocytic anaemia [haematocrit 29.4% (normal 36–46%)] and strikingly elevated acute phase reactants, with an ESR of 98 mm/h (normal <20 mm/h) and a CRP level of 63 mg/l (normal <5 mg/l). Empirical treatment with prednisone 60 mg/day was initiated immediately, and his headache abated. The diagnosis of GCA in this patient was potentially confounded by several comorbidities. He had confirmed vasculitis with several risk factors for atherosclerotic disease, a history of hypertensive nephropathy that led to end-stage renal disease, and he was receiving peritoneal dialysis. He was also a former smoker and was hyperlipidaemic. The patient had a history of childhood rheumatic heart disease and had undergone placement of a mechanical aortic valve several years previously that necessitated long-term anticoagulation therapy with warfarin. Computed tomographic angiography confirmed the presence of atherosclerosis and calcific vascular disease in the large vessels of the patient’s head and neck. There was a dense calcific plaque along the aortic arch, accompanied by probable intimal ulceration. There was also significant ossification of the major cervical vessels and severe calcification of the distal right and left common carotid arteries. Although many features of the clinical presentation suggested GCA, non-arteritic AION was also possible. It is important for internists, rheumatologists, neurologists and pathologists to understand that arteritic and non-arteritic AION cannot be distinguished on slit-lamp examination; certain clues may sway ophthalmologists’ thinking toward one diagnosis or the other, but definitive diagnosis by examination of the patient’s retina is not possible. In this case, one test—temporal artery biopsy—offered the possibility of distinguishing arteritic from non-arteritic AION. The medical team decided to proceed with performing a temporal artery biopsy even though the patient was receiving warfarin. The attending physician noted: ‘Despite the patient’s warfarin therapy, I believe that any potential haemorrhagic risk from a temporal artery biopsy is outweighed by the potential thrombotic risk of his mechanical valves were his warfarin to be stopped’. The surgeon performed bilateral temporal artery biopsies—1.4 cm on the right and 0.9 cm on the left—without any bleeding complications. Interpretation of the biopsy specimens was unequivocal. Although >100 cross-sections were examined from each artery, there was no evidence of GCA in either biopsy specimen. Rather, both specimens showed calcific atherosclerosis. The attending physician noted: ‘I am glad we have appropriately excluded giant cell arteritis, which can cause progression of loss of vision despite treatment.’ The patient’s prednisone was stopped, and he was discharged from the hospital. Six weeks later, the patient was re-admitted to hospital. His headache had recurred after he had stopped prednisone treatment. He had lost ∼4.5 kg in body weight because jaw claudication made chewing painful. He also developed new symptoms of right arm claudication and a region of tenderness in his posterior scalp when resting his head on a pillow after being discharged. Physical examination at the time of the second admission to hospital revealed a nodular cord in an occipital artery. The patient’s acute-phase reactant levels remained high. The new medical team was alarmed by what now seemed to be obvious GCA, and the patient was given pulse methylprednisolone immediately. However, it was too late; that evening, while watching television in his hospital room, the patient went blind in his left eye. Ophthalmology evaluated the patient the following day and confirmed bilateral vision loss because of AION. The only possible recommendation was to refer the patient to the State Commission for the Blind. The loss of sight in his remaining eye was devastating for this patient. He became profoundly depressed. He chose to discontinue dialysis, and he died shortly thereafter. What led to the patient’s blindness and, secondarily, to his death? Where did this case go wrong? How could a patient have gone blind in one eye and then in the other in this era of effective therapies for GCA? The crux of the matter is this: the clinicians placed too much emphasis on the finding of negative temporal artery biopsy results. The key diagnostic test was not the temporal artery biopsy specimens; rather, it was the patient’s history. Evidence of GCA in the patient’s history was substantial (if not overwhelming) at the time of his first episode of vision loss. Failure to place sufficient weight on that history and the existing clinical evaluation led to the cessation of glucocorticoid therapy that very likely would have preserved the vision in the patient’s remaining good eye. Let us review the evidence. At the time of the patient’s first episode of vision loss, he had experienced new headaches that were atypical for him, diplopia and jaw claudication. In addition, he had normochromic, normocytic anaemia and marked elevation of acute-phase reactants. Finally, bilateral AION was confirmed by fluorescein angiography. Although correct diagnosis often seems all too easy in retrospect, substantial evidence of GCA was discarded because of negative temporal artery biopsy results, and the diagnosis was simply missed—very much to the patient’s detriment. The take-home lesson is that, in the end, GCA is a clinical diagnosis not a pathological one. Clinicians always feel reassured by a positive temporal artery biopsy, but in reality they may not have the benefit of that confirmation. Therefore, it is important to understand what temporal artery biopsies can tell us and why they often fall short of our expectations. Excessively high expectations The problem of excessively high expectations is summed up by a statement and its corollary, both attributed to Dr Oscar Rambo, professor emeritus at the University of California-San Francisco as follows: ‘A mystic (and dangerous) assumption prevails in the mind of the clinician that the pathologist can produce a statement of absolute truth based on a small piece of tissue’ [3]. The corollary to this statement is that: ‘Equally dangerous to mankind is the pathologist who thinks the same.’ Many clinicians believe that temporal artery biopsy is the ultimate arbiter of whether a patient has GCA and that the diagnosis of GCA cannot be excluded without this procedure. Rambo’s quotation underscores that pathologists must understand not only the value of temporal artery biopsies but also their potential shortcomings. What are the test characteristics of temporal artery biopsy? Although some data (now >30 years old) suggest that the negative predictive value of bilateral temporal artery biopsy is > 90% [4], the specificity of such a finding is probably far lower. The 90% value was derived from a retrospective study at the Mayo Clinic, in which an extremely rigorous approach to the performance of temporal artery biopsy was undertaken. Biopsies at the Mayo Clinic were performed for many decades by a single, highly experienced surgeon who routinely obtained between 3 and 4 cm of temporal artery from the symptomatic side (if symptoms were localized) [5]. In addition, the contralateral side was biopsied if a frozen section failed to demonstrate GCA in the first biopsy. Unfortunately, this level of rigor is rarely achieved at most other medical institutions, and the yield of temporal artery biopsies at other centres is almost certainly far lower. Although it would be most useful to understand both the sensitivity and the negative predictive value of temporal artery biopsy for each medical centre, we simply do not know these test characteristics because there is no true gold standard. In fact, estimates of the sensitivity of temporal artery biopsy vary greatly among studies, and high false-negative rates are commonly reported [6–8]. Many experts consider sensitivity estimates of 60–70% to be accurate, but this is only a guess. The sensitivity of temporal artery biopsy is substantially affected by many variables, which are discussed herein. How we are led astray Inadequate biopsy specimen length Ask any pathologist what he or she most often requests, having agonized inconclusively at the microscope over a tiny (2 mm2) piece of temporal artery tissue, and the answer will be, Give me more tissue; that is, a longer segment of the artery. However, the trend in recent years has been to reduce the size of biopsy specimens as a push toward minimally invasive diagnostic procedures [9]. So significant is this issue that pathologists are now commonly asked to report on the adequacy of the biopsy specimen before giving their diagnostic opinion [10]. Although the value of a (positive) temporal artery biopsy has been accepted for decades, the ideal biopsy length to ensure optimal sensitivity is still a matter of debate. The often-cited rule of thumb that a temporal artery biopsy specimen should not be < 2.0 cm is seldom accomplished in practice. Moreover, a temporal artery biopsy specimen that is 2.0 cm upon removal in the operating room often shrinks substantially after processing in the pathology laboratory. The optimal biopsy length associated with increased diagnostic yield varies between studies from >0.5 cm (post-fixation) to 0.7 or ⩾1.0 cm [11–13]. Whatever the minimum acceptable length of a biopsy specimen, the adequacy of the sample should receive as much attention as the diagnosis itself in the setting of a negative temporal artery biopsy result. Inadequate histological work-up Histopathological work-up of vascular specimens, like other small biopsy samples, usually involves sectioning or serial sectioning of the sample in addition to specific stains. The idea behind the technique of sectioning through the tissue (or at least part of it) is to minimize the possibility of false-negative results, specifically, in patients with GCA, to avoid missing skip lesions. Early studies highlighted the need for examination of multiple histological sections (in addition to adequate biopsy length) to detect skip lesions [14]. Although specific guidelines dictating the number of sections that is appropriate for temporal artery biopsy specimens have not been published, consensus statements on the processing of cardiovascular surgical specimens are available [15]. These recommend three or more sections for tissue samples, which is routinely practised by most histopathology laboratories. However, it should be taken into consideration that sectioning still means that the pathologist will see only a fraction of the sample (a biopsy sample 1.5 cm in length consisting of six sections of 3-µm thickness equates to 0.12% of the entire sample). Temporal arteries not involved The temporal arteries are not part of the vascular distribution in which arterial compromise leads to blindness in GCA. Blindness results from involvement of the ophthalmic, retinal and posterior ciliary vessels. These vessels run a course behind the eye and track along the optic nerve. The temporal artery stems from a different branch of the external carotid artery and traces a path anterior to the tragus of the ear, near the course of branches of the facial nerve, up and over the temple, lying just under the skin and therefore accessible to biopsy. The temporal artery is biopsied because it is comparatively easy to do so, not because it necessarily reflects what is going on elsewhere in the large- and medium-sized vessel circulation or that its involvement predicts patients at higher or lower risk for vision loss. In short, temporal arteries that are free from inflammation do not exclude the presence of active GCA in other vascular distributions. Involved side not biopsied In most cases, if a temporal artery biopsy is worth doing, it is worth performing on both sides. Occasionally, a patient may have a classic history and floridly abnormal temporal arteries (tender, swollen, nodular lesions). Such cases might justify unilateral biopsy, but these are in the minority. Most of the time, the importance of getting it right and the data available on the usefulness of bilateral biopsy argue for taking biopsy specimens from both sides. A retrospective study performed by Boyev et al. [16] evaluated 186 patients who underwent bilateral temporal artery biopsy at Johns Hopkins Hospital. Only six patients (3%) had unilateral arteritis. Although this might be regarded as a small percentage not worth the additional effort and expense of bilateral procedures, those six patients represented at least 20% of the total number of patients whose GCA was diagnosed through biopsy. Similar conclusions were reached in another retrospective study that evaluated bilateral biopsy [17]. Among the 60 patients examined in that study, eight (13%) had GCA identified on only one side. However, those eight patients comprised 40% of the total number of patients diagnosed with GCA by biopsy. These studies suggest that once the decision has been made to pursue temporal artery biopsy, routine performance of bilateral procedures is justified. Inadequate surgical experience and performing biopsy on the wrong portion of the vessel When performing temporal artery biopsy, surgeons attempt to locate the artery by finding a pulse. However, a pulsatile segment of the vessel is not necessarily the portion of the artery that is most likely to yield a diagnosis, and therefore is not the most desirable segment to biopsy. Given that a pulseless artery is sometimes difficult to locate, the skill of the surgeon is an important determinant of the success of temporal artery biopsy. Unfortunately, at many centres, the least-experienced surgeon is usually sent to perform these procedures. The pulse of the temporal artery is often most easily palpable just anterior to the tragus of the ear. Surgeons often perform biopsy here, but this can be problematic for two reasons: first, as mentioned previously, an artery with a robust pulse is not the best site for biopsy; and second, biopsy of the artery in this area risks damage to the facial nerve that courses through the same region. A more fruitful and less risky biopsy targets the frontal branch of the superficial temporal artery just above the temple. Some patients with suspected temporal arteritis have been sent to undergo biopsy only to return with a statement from the surgeon that no pulsatile vessel could be identified. Difficulty finding a pulse should cause the surgeon to redouble efforts to locate a branch of the temporal artery, not abandon the procedure. Some surgeons use intra-operative US to assist in the identification of arteries with faint pulses. Pathological interpretation wrong Pathologists are often hampered by inadequate biopsy specimen length, as noted previously. Less commonly, they miss the diagnosis because of human error. Experience plays a major role in the pathologist’s ability to discern more subtle features that might be compatible with GCA. There are two common causes of just missing it: first, early changes in which incipient inflammation is not clear and, second, inflammation in the temporal artery (temporal arteritis) that is not GCA. A diagnosis of temporal arteritis is not synonymous with GCA. Some of the variations in the histological specimens of cross sections through temporal artery biopsies are depicted in Fig. 1. Advances in the treatment of individual forms of vasculitis now dictate the importance of precision in diagnosis. Fig. 1 View largeDownload slide Histological images of cross sections through temporal artery biopsies (A) and (B) (haematoxylin and eosin staining) show a case of florid GCA with transmural infiltration of lymphocytes, macrophages and the presence of several giant cells (black arrows). (C) (haematoxylin and eosin staining) shows a case of clinically positive GCA without noticeable inflammation on histological work-up. The lumen [as in the case in (A)] is compromised by significant circumferential intimal hyperplasia. An elastica van Gieson stain (D) shows minimal media fibrosis as well as evident disruption of the internal elastic lamina (black arrow), a subtle but not entirely specific sign of possible GCA. Fig. 1 View largeDownload slide Histological images of cross sections through temporal artery biopsies (A) and (B) (haematoxylin and eosin staining) show a case of florid GCA with transmural infiltration of lymphocytes, macrophages and the presence of several giant cells (black arrows). (C) (haematoxylin and eosin staining) shows a case of clinically positive GCA without noticeable inflammation on histological work-up. The lumen [as in the case in (A)] is compromised by significant circumferential intimal hyperplasia. An elastica van Gieson stain (D) shows minimal media fibrosis as well as evident disruption of the internal elastic lamina (black arrow), a subtle but not entirely specific sign of possible GCA. Confounders in diagnosing GCA: previous therapy It is widely accepted that the diagnosis of GCA can be made by temporal artery biopsy even 2 weeks after initiation of glucocorticoids and sometimes much longer [5]; however, there is little doubt that glucocorticoid treatment alters the underlying histopathology in the blood vessel wall. Both the pathological interpretation of the temporal artery biopsy and the clinical decision-making that follows must take this into account. The full impact of glucocorticoid administration on histological results is a matter of debate. Although some studies clearly show a time-dependent reduction in positive biopsy results after glucocorticoid treatment [18–20], others have not been able to demonstrate a clear relationship between glucocorticoid use and reduction of inflammation [5, 21, 22]. Giant cells are important but not required for this diagnosis The presence of giant cells in a temporal artery biopsy specimen is strongly suggestive of GCA and giant cells are probably the most carefully sought feature when GCA is suspected. However, the frequency with which giant cells are identified is substantially lower than the name of the disease suggests [23–25]. In some cases, multiple cuts of an artery are necessary to identify giant cells, and giant cells are not detected in ∼12% of clinically diagnosed cases of GCA [26]. Giant cells not only have a high positive predictive value for GCA, they may also predict the occurrence of cranial ischaemic events [27]. The presence of giant cells in biopsy specimens has also been strongly associated with the presence of jaw claudication and elevated serological markers of inflammation, suggesting that the presence of giant cells reflects a more acute and aggressive form of the disease [28]. In the absence of giant cells, more subtle histopathological changes must suffice as clues to possible presence of GCA [29]. Experience on the part of the pathologist is crucial to the proper interpretation of intimal thickening, focal infiltration by lymphocytes and macrophages, and localized scarring in the arterial wall, all of which are compatible with diagnosis but not necessarily diagnostic themselves in the absence of careful clinical correlation. Clinicians and pathologists should sit down at the microscope together whenever possible. Atherosclerosis causing intimal changes poses challenges in differentiation Another confounding factor is atherosclerosis, which can cause intimal hyperplasia similar to that of GCA. Arterial hypertension, dyslipidaemia and diabetes all affect the integrity of the vessel wall and may complicate histological interpretation in a patient with suspected GCA. Although atherosclerotic changes were observed in a cohort of 40 steroid-treated patients with GCA, the changes were not significantly different from those in age-matched controls [30]. Premature atherosclerosis has been reported in several pro-inflammatory rheumatological conditions and may be accelerated in vasculitis [31]. Much attention is focused on the internal elastic lamina and its fragmentation in the setting of GCA. However, fragmentation of the elastic laminae by elastolysis is a feature common to both atherosclerosis and GCA [32]. Variations in the appearance of laminar fragmentation and its relationship to fibrosis, as well as nuances of intimal hyperplasia, may provide important clues to help differentiate changes associated with atherosclerosis from those caused by vasculitis. In many cases, such histological subtleties make the pathologist’s job as much an art as a science. What is healed arteritis and what is its relationship to GCA? The term healed arteritis was used many years ago by Allsop and Gallagher [33] to distinguish a small subgroup of patients who did not fall into either the large category of patients with florid inflammation or the category of purely arterio-/atherosclerotic patients. Although studies are limited, results of some investigations of patients with so-called healed arteritis imply that this histopathology predicts a lower risk for disease recurrence, suggesting that the finding of healed arteritis is a biologically relevant state or subtype of vasculitis (or a different disease state completely) that does not necessitate the use of the high-dose glucocorticoids typically used to treat GCA [34]. In a recent study by Maleszewski et al. [35], patients with an initial positive result from temporal artery biopsy who were receiving treatment underwent a second biopsy after 3 months to 1 year. The multi-nucleated giant cells and granulomatous inflammation observed in the majority of positive biopsy results at presentation disappeared over time (and with treatment). Although signs of vascular remodelling were common to both the first and the second biopsies, medial fibrosis seemed to indicate a long-term healing process. Unfortunately, healed arteritis is an elusive histopathological finding. None of the histological findings often attributed to healed arteritis—reduplication or rupture of the internal and external elastic laminae, calcification, fibrosis and neovascularization—are specific for GCA that is no longer active. In short, reliable differentiation between healed (i.e. quiescent) GCA and arteriosclerosis may simply not be possible by routine histopathology [36]. When confronted with the finding of healed arteritis on a temporal artery biopsy report, the clinician must rely more on the clinical features presented than on the histopathological findings when deciding whether to treat the patient for GCA. Utility of staining for CD68-positive macrophages Correctly identifying macrophages by routine histology is something that the most junior of pathologists learn, and immunohistochemical staining is rarely performed in routine cases of inflammation to confirm the presence of macrophages outside of research protocols. However, when the presence of (isolated) macrophages can tip the scale towards one diagnosis or another, the use of an antibody against CD68, a glycoprotein on monocytes and macrophages [37], may aid in visualization of the number and distribution of tissue macrophages. In theory, the identification of macrophages around the internal elastic lamina, a finding facilitated by CD68 stains, provides additional support for the diagnosis of GCA, particularly in cases in which the histopathology is otherwise unconvincing. However, this theory has not been studied sufficiently. A retrospective study of 60 patients aimed to predict the biopsy characteristics suggestive of response to glucocorticoid treatment and characteristics distinguishing GCA, PMR or arteriosclerosis using immunophenotyping. In that study, CD68 positivity was a predictor of the success of glucocorticoid treatment in patients with GCA and distinguished GCA from PMR in the absence of GCA [38]. All patients examined had histologically proven GCA; therefore, the value of immunohistochemistry in ambivalent cases or cases with subtle histological abnormalities was not clarified by this study. In a larger retrospective study that used a three-tier system to distinguish among positive, indeterminate or negative biopsies, the value of CD68 lay in aiding a decision in indeterminate cases [25]. The authors concluded correctly that the ultimate diagnosis of GCA in such cases depends heavily on clinical judgement and clinicopathological correlation rather than simply on pathology. Although CD68 staining facilitates the identification of macrophages within the arterial wall [39], it does not resolve all problems related to the diagnosis of GCA. Histologically equivocal cases Dealing with histologically equivocal cases is part of the daily life of a pathologist. Although vasculitis is defined simply (or over-simply) as ‘inflammation of blood vessel walls at least at some time during the course of the disease’ [40], correctly identifying this histologically is often tricky. At what time point are aggregated lymphocytes considered inflammatory infiltrates? Does the presence of inflammatory infiltrates equate to disease? A recent study by Jia et al. [41] revealed that patients with lymphocytic inflammation involving small blood vessels or the adventitia of the temporal artery in the absence of transmural arteritic damage were not at increased risk for temporal arteritis-like adverse events. Therefore, it is frequently difficult for pathologists to know where to draw the line between disease and possible disease. Distinguishing perivasculitis from inflammation that truly targets and damages the blood vessel wall also frequently raises challenges. Mimickers of GCA These challenges aside, there is no doubt that other bona fide forms of vasculitis can affect the temporal artery. That is, diseases other than GCA can cause temporal arteritis and be misdiagnosed by pathologists who are not aware of this possibility. Non-GCA diseases most likely to affect the temporal arteries include the ANCA-associated vasculitides (granulomatosis with polyangiitis, microscopic polyangiitis and eosinophilic granulomatosis with polyangiitis), polyarteritis nodosa and mixed cryoglobulinaemia [42]. What are the solutions? The growing role of imaging Data from the Tocilizumab in Giant Cell Arteritis (GiACTA) trial [43] emphasize the growing role that imaging plays in the diagnosis of GCA. In that trial, 38% of patients either did not undergo temporal artery biopsy or the biopsy results were negative, highlighting the importance of imaging modalities as tools for the diagnosis of GCA. Conversely, many patients who have negative results on imaging studies will have positive temporal artery biopsy results, so the approaches to diagnosis are complementary. In some cases, neither temporal artery biopsy nor imaging yields a compelling case for the diagnosis of GCA. Therefore, advances in diagnostic imaging have supplanted pathology in a minority of GCA cases—particularly those characterized by large-vessel involvement—but temporal artery biopsy remains the mechanism of diagnostic confirmation in most patients. Careful history taking For a clinician, the key diagnostic test is neither the temporal artery biopsy nor large-vessel imaging, but the patient’s history; the patient’s history is the clinician’s scalpel and microscope. Skill in taking a patient’s history is crucial to suspecting GCA in the first place and to making the decision to order temporal artery biopsy or any other test that might confirm or refute the diagnostic suspicion. Communication The development of a single diagnostic test that is perfect, or close to perfect, be it clinical, serological, radiological or pathological, is not imminent. Communication between clinicians, surgeons, pathologists, radiologists, imaging technicians and patients is fundamental to determining the best solution in difficult cases. The clinician must devote the time necessary to listen to the patient, and to interpret the patient’s words wisely. In some cases, the clinician and radiologist can guide the surgeon with regard to the proper site for temporal artery biopsy. Following a biopsy, the likelihood of confirming or refuting the diagnosis of GCA in an accurate manner is then heightened by a process of collegial clinicopathological correlation, best conducted in person between the clinician and the pathologist sitting at two heads of the same microscope. Acknowledgements The manuscript was prepared by the authors, with professional writing and editorial assistance provided by Sara Duggan, PhD, on behalf of F. Hoffmann-La Roche Ltd. Supplement: This supplement is funded by F. Hoffmann-La Roche Ltd. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this manuscript. Disclosure statement: J.H.S. reports grants and consulting fees from Roche, Genetech and Xencor. The other author has declared no conflicts of interest. References 1 Horton BT, Magath TB, Brown GE. An undescribed form of arteritis of the temporal vessels. Mayo Clin Proc  1932; 7: 700– 1. 2 Hutchinson J. Disease of the arteries. On a peculiar form of thrombotic arteritis of the aged which is sometimes productive of gangrene. Arch Surg  1890; 1: 323– 9. 3 Rambo ON. The limitations of histologic diagnosis. Progr Radiat Ther  1962; 2: 215– 24. Google Scholar PubMed  4 Hall S, Persellin S, Lie JT et al.   The therapeutic impact of temporal artery biopsy. Lancet  1983; 2: 1217– 20. Google Scholar CrossRef Search ADS PubMed  5 Achkar AA, Lie JT, Hunder GG, O'Fallon WM, Gabriel SE. How does previous corticosteroid treatment affect the biopsy findings in giant cell (temporal) arteritis? Ann Intern Med  1994; 120: 987– 92. Google Scholar CrossRef Search ADS PubMed  6 Ashton-Key MR, Gallagher PJ. False-negative temporal artery biopsy. Am J Surg Pathol  1992; 16: 634– 5. Google Scholar CrossRef Search ADS PubMed  7 Borchers AT, Gershwin ME. Giant cell arteritis: a review of classification, pathophysiology, geoepidemiology and treatment. Autoimmun Rev  2012; 11: A544– 54. Google Scholar CrossRef Search ADS PubMed  8 Luqmani R, Lee E, Singh S et al.   The role of ultrasound compared to biopsy of temporal arteries in the diagnosis and treatment of giant cell arteritis (TABUL): a diagnostic accuracy and cost-effectiveness study. Health Technol Assess  2016; 20: 1– 238. Google Scholar CrossRef Search ADS PubMed  9 Fernandez EM, Helm T, Ioffreda M, Helm KF. The vanishing biopsy: the trend toward smaller specimens. Cutis  2005; 76: 335– 9. Google Scholar PubMed  10 Welsch MJ, Ioffreda MD, Clarke LE, Helm KF. Assessing and communicating adequacy of biopsy specimens and clinical information. J Am Acad Dermatol  2012; 66: 509– 10. Google Scholar CrossRef Search ADS PubMed  11 Breuer GS, Nesher R, Nesher G. Effect of biopsy length on the rate of positive temporal artery biopsies. Clin Exp Rheumatol  2009; 27(1 Suppl 52): S10– 3. 12 Ypsilantis E, Courtney ED, Chopra N et al.   Importance of specimen length during temporal artery biopsy. Br J Surg  2011; 98: 1556– 60. Google Scholar CrossRef Search ADS PubMed  13 Taylor-Gjevre R, Vo M, Shukla D, Resch L. Temporal artery biopsy for giant cell arteritis. J Rheumatol  2005; 32: 1279– 82. Google Scholar PubMed  14 Klein RG, Campbell RJ, Hunder GG, Carney JA. Skip lesions in temporal arteritis. Mayo Clin Proc  1976; 51: 504– 10. Google Scholar PubMed  15 Stone JR, Basso C, Baandrup UT et al.   Recommendations for processing cardiovascular surgical pathology specimens: a consensus statement from the Standards and Definitions Committee of the Society for Cardiovascular Pathology and the Association for European Cardiovascular Pathology. Cardiovasc Pathol  2012; 21: 2– 16. Google Scholar CrossRef Search ADS PubMed  16 Boyev LR, Miller NR, Green WR. Efficacy of unilateral versus bilateral temporal artery biopsies for the diagnosis of giant cell arteritis. Am J Ophthalmol  1999; 128: 211– 5. Google Scholar CrossRef Search ADS PubMed  17 Pless M, Rizzo JF3rd, Lamkin JC, Lessell S. Concordance of bilateral temporal artery biopsy in giant cell arteritis. J Neuroophthalmol  2000; 20: 216– 8. Google Scholar CrossRef Search ADS PubMed  18 Allison MC, Gallagher PJ. Temporal artery biopsy and corticosteroid treatment. Ann Rheum Dis  1984; 43: 416– 7. Google Scholar CrossRef Search ADS PubMed  19 Font RL, Prabhakaran VC. Histological parameters helpful in recognising steroid-treated temporal arteritis: an analysis of 35 cases. Br J Ophthalmol  2007; 91: 204– 9. Google Scholar CrossRef Search ADS PubMed  20 Narvaez J, Bernad B, Roig-Vilaseca D et al.   Influence of previous corticosteroid therapy on temporal artery biopsy yield in giant cell arteritis. Semin Arthritis Rheum  2007; 37: 13– 9. Google Scholar CrossRef Search ADS PubMed  21 Ray-Chaudhuri N, Kine DA, Tijani SO et al.   Effect of prior steroid treatment on temporal artery biopsy findings in giant cell arteritis. Br J Ophthalmol  2002; 86: 530– 2. Google Scholar CrossRef Search ADS PubMed  22 Bury D, Joseph J, Dawson TP. Does preoperative steroid treatment affect the histology in giant cell (cranial) arteritis? J Clin Pathol  2012; 65: 1138– 40. Google Scholar CrossRef Search ADS PubMed  23 Parker F, Healey LA, Wilske KR, Odland GF. Light and electron microscopic studies on human temporal arteries with special reference to alterations related to senescence, atherosclerosis and giant cell arteritis. Am J Pathol  1975; 79: 57– 80. Google Scholar PubMed  24 Lie JT. Temporal artery biopsy diagnosis of giant cell arteritis: lessons from 1109 biopsies. Anat Pathol  1996; 1: 69– 97. Google Scholar PubMed  25 Zhou L, Luneau K, Weyand CM et al.   Clinicopathologic correlations in giant cell arteritis: a retrospective study of 107 cases. Ophthalmology  2009; 116: 1574– 80. Google Scholar CrossRef Search ADS PubMed  26 Regan MJ, Wood BJ, Hsieh YH et al.   Temporal arteritis and Chlamydia pneumoniae: failure to detect the organism by polymerase chain reaction in ninety cases and ninety controls. Arthritis Rheum  2002; 46: 1056– 60. Google Scholar CrossRef Search ADS PubMed  27 Muratore F, Boiardi L, Cavazza A et al.   Correlations between histopathological findings and clinical manifestations in biopsy-proven giant cell arteritis. J Autoimmun  2016; 69: 94– 101. Google Scholar CrossRef Search ADS PubMed  28 Ting KH, Lester S, Dunstan E, Hill CL. Association between histological features and clinical features of patients with biopsy positive giant cell arteritis. Clin Exp Rheumatol  2016; 34(3 Suppl 97): S40– 3. 29 Stacy RC, Rizzo JF, Cestari DM. Subtleties in the histopathology of giant cell arteritis. Semin Ophthalmol  2011; 26: 342– 8. Google Scholar CrossRef Search ADS PubMed  30 Gonzalez-Juanatey C, Lopez-Diaz MJ, Martin J, Llorca J, Gonzalez-Gay MA. Atherosclerosis in patients with biopsy-proven giant cell arteritis. Arthritis Rheum  2007; 57: 1481– 6. Google Scholar CrossRef Search ADS PubMed  31 Tervaert JW. Translational mini-review series on immunology of vascular disease: accelerated atherosclerosis in vasculitis. Clin Exp Immunol  2009; 156: 377– 85. Google Scholar CrossRef Search ADS PubMed  32 O'Brien JP. Destruction of elastic tissue (elastolysis) as a link between atherosclerosis and the temporal arteritis/polymyalgia rheumatica syndrome. Observations on an actinic factor in vascular disease. Pathol Biol  1984; 32: 123– 38. Google Scholar PubMed  33 Allsop CJ, Gallagher PJ. Temporal artery biopsy in giant-cell arteritis. A reappraisal. Am J Surg Pathol  1981; 5: 317– 23. Google Scholar CrossRef Search ADS PubMed  34 ter Borg EJ, Haanen HC, Seldenrijk CA. Relationship between histological subtypes and clinical characteristics at presentation and outcome in biopsy-proven temporal arteritis. Identification of a relatively benign subgroup. Clin Rheumatol  2007; 26: 529– 32. Google Scholar CrossRef Search ADS PubMed  35 Maleszewski JJ, Younge BR, Fritzlen JT et al.   Clinical and pathological evolution of giant cell arteritis: a prospective study of follow-up temporal artery biopsies in 40 treated patients. Mod Pathol  2017; 30: 788– 96. Google Scholar CrossRef Search ADS PubMed  36 Cox M, Gilks B. Healed or quiescent temporal arteritis versus senescent changes in temporal artery biopsy specimens. Pathology  2001; 33: 163– 6. Google Scholar CrossRef Search ADS PubMed  37 Holness CL, Simmons DL. Molecular cloning of CD68, a human macrophage marker related to lysosomal glycoproteins. Blood  1993; 81: 1607– 13. Google Scholar PubMed  38 Braun N, Fritz P, Rieth A et al.   Predictors for treatment success and expression of glucocorticoid receptor in giant cell arteritis and polymyalgia rheumatica. J Rheumatol  2009; 36: 2269– 76. Google Scholar CrossRef Search ADS PubMed  39 Wang AL, Raven ML, Surapaneni K, Albert DM. Studies on the histopathology of temporal arteritis. Ocul Oncol Pathol  2017; 3: 60– 5. Google Scholar CrossRef Search ADS PubMed  40 Jennette JC, Falk RJ, Bacon PA et al.   2012 revised International Chapel Hill Consensus Conference Nomenclature of Vasculitides. Arthritis Rheum  2013; 65: 1– 11. Google Scholar CrossRef Search ADS PubMed  41 Jia L, Couce M, Barnholtz-Sloan JS, Cohen ML. Is all inflammation within temporal artery biopsies temporal arteritis? Hum Pathol  2016; 57: 17– 21. Google Scholar CrossRef Search ADS PubMed  42 Cavazza A, Muratore F, Boiardi L et al.   Inflamed temporal artery: histologic findings in 354 biopsies, with clinical correlations. Am J Surg Pathol  2014; 38: 1360– 70. Google Scholar CrossRef Search ADS PubMed  43 Tuckwell K, Collinson N, Dimonaco S et al.   Newly diagnosed vs. relapsing giant cell arteritis: Baseline data from the GiACTA trial. Semin Arthritis Rheum  2017; 46: 657– 64. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Rheumatology Oxford University Press

Why do temporal arteries go wrong? Principles and pearls from a clinician and a pathologist

Rheumatology , Volume 57 (suppl_2) – Feb 1, 2018

Loading next page...
 
/lp/ou_press/why-do-temporal-arteries-go-wrong-principles-and-pearls-from-a-IrI0K8O8U9
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1462-0324
eISSN
1462-0332
D.O.I.
10.1093/rheumatology/kex524
Publisher site
See Article on Publisher Site

Abstract

Abstract Early diagnosis and treatment of GCA is essential to prevent complications of the disease, including permanent vision loss. Temporal artery biopsy has been intrinsically linked with the diagnosis of GCA for several decades. A negative predictive value of > 90% has been reported for temporal artery biopsy; however, a negative result does not reliably indicate the absence of GCA because inflammation of the temporal artery is not always evident because of segmental involvement or other reasons. This is demonstrated by a case study of a patient hospitalized following acute vision loss to the right eye whose glucocorticoid treatment was suspended after temporal artery biopsy revealed no evidence of GCA. The patient subsequently lost sight in the left eye 6 weeks after stopping glucocorticoid therapy. The specificity of temporal artery biopsy for the diagnosis of GCA is variable and influenced by many factors, including length of biopsy specimens, vasculitis in vessels other than the temporal artery (ophthalmic, retinal and posterior ciliary vessels), unilateral versus bilateral biopsy, expertise of the surgeon, interpretation of histology, effects of treatment and confounding factors such as atherosclerosis or other non-GCA diseases that can affect the temporal artery. Considering the limitations of temporal artery biopsy, collaboration and education between the clinician, the pathologist and the patient, taking into account a thorough examination of patient history, recognizing signs and symptoms, and potentially involving newer imaging studies with trained technicians and physicians, are essential in confirming or eliminating diagnosis of GCA. giant cell arteritis, histopathology, primary care rheumatology Rheumatology key messages Temporal artery biopsy has been a cornerstone in diagnosing giant cell arteritis for several decades. A negative temporal artery biopsy result does not reliably preclude the presence of GCA. Collaboration and education between clinicians, pathologists and patients are essential in diagnosing or eliminating GCA. Introduction The first temporal artery biopsy was performed at the Mayo Clinic in 1932 [1], >40 years after GCA was first described in the medical literature [2], indicating how slow progress has been in this disease. Temporal artery biopsy has become intrinsic to the diagnosis of GCA; however, although >85 years have passed since the first procedure was performed, there is much uncertainty regarding the performance, interpretation and test characteristics of temporal artery biopsy. How could a patient have GCA without documentable (and therefore diagnostic) inflammation of the temporal artery? After all, GCA is often referred to as temporal arteritis. Unfortunately, there are many reasons this might happen. Appropriate care of patients with GCA depends critically on gaining insight into these reasons. The stakes are high in making a diagnosis of GCA; the patient’s vision and life are at risk if a diagnosis is missed. Conversely, an incorrect diagnosis of GCA can lead to inappropriate glucocorticoid use and associated morbidity, as well as to undertreatment if the patient has another form of vasculitis (e.g. an ANCA-associated vasculitis). For these reasons, understanding the power and limitations of temporal artery biopsy is crucial to the diagnosis and treatment of patients with suspected GCA. Experience and caution are paramount in the interpretation and reporting of temporal artery biopsy findings and there is great danger in attaching too much certainty to a negative temporal artery biopsy result. This is illustrated by the following case. An illustrative case A 77-year-old man was admitted to hospital twice within a few weeks. The first admission occurred after acute vision loss in the right eye following symptoms of a moderate retro-orbital headache on the left side and amaurosis fugax on the left side, reported as the sensation that his left eyebrow was obscuring his vision. He also noted that his face got tired while chewing food. Two days after the onset of these left-sided symptoms, he awoke blind in his right eye. The patient was evaluated by an ophthalmologist and underwent fluorescein angiography, which revealed poor choroidal circulation and bilateral disc oedema with peripapillary haemorrhages. The patient was diagnosed as having anterior ischaemic optic neuropathy (AION) and was admitted urgently to the hospital for further evaluation. Laboratory examinations revealed normal platelet and white blood cell counts but normochromic, normocytic anaemia [haematocrit 29.4% (normal 36–46%)] and strikingly elevated acute phase reactants, with an ESR of 98 mm/h (normal <20 mm/h) and a CRP level of 63 mg/l (normal <5 mg/l). Empirical treatment with prednisone 60 mg/day was initiated immediately, and his headache abated. The diagnosis of GCA in this patient was potentially confounded by several comorbidities. He had confirmed vasculitis with several risk factors for atherosclerotic disease, a history of hypertensive nephropathy that led to end-stage renal disease, and he was receiving peritoneal dialysis. He was also a former smoker and was hyperlipidaemic. The patient had a history of childhood rheumatic heart disease and had undergone placement of a mechanical aortic valve several years previously that necessitated long-term anticoagulation therapy with warfarin. Computed tomographic angiography confirmed the presence of atherosclerosis and calcific vascular disease in the large vessels of the patient’s head and neck. There was a dense calcific plaque along the aortic arch, accompanied by probable intimal ulceration. There was also significant ossification of the major cervical vessels and severe calcification of the distal right and left common carotid arteries. Although many features of the clinical presentation suggested GCA, non-arteritic AION was also possible. It is important for internists, rheumatologists, neurologists and pathologists to understand that arteritic and non-arteritic AION cannot be distinguished on slit-lamp examination; certain clues may sway ophthalmologists’ thinking toward one diagnosis or the other, but definitive diagnosis by examination of the patient’s retina is not possible. In this case, one test—temporal artery biopsy—offered the possibility of distinguishing arteritic from non-arteritic AION. The medical team decided to proceed with performing a temporal artery biopsy even though the patient was receiving warfarin. The attending physician noted: ‘Despite the patient’s warfarin therapy, I believe that any potential haemorrhagic risk from a temporal artery biopsy is outweighed by the potential thrombotic risk of his mechanical valves were his warfarin to be stopped’. The surgeon performed bilateral temporal artery biopsies—1.4 cm on the right and 0.9 cm on the left—without any bleeding complications. Interpretation of the biopsy specimens was unequivocal. Although >100 cross-sections were examined from each artery, there was no evidence of GCA in either biopsy specimen. Rather, both specimens showed calcific atherosclerosis. The attending physician noted: ‘I am glad we have appropriately excluded giant cell arteritis, which can cause progression of loss of vision despite treatment.’ The patient’s prednisone was stopped, and he was discharged from the hospital. Six weeks later, the patient was re-admitted to hospital. His headache had recurred after he had stopped prednisone treatment. He had lost ∼4.5 kg in body weight because jaw claudication made chewing painful. He also developed new symptoms of right arm claudication and a region of tenderness in his posterior scalp when resting his head on a pillow after being discharged. Physical examination at the time of the second admission to hospital revealed a nodular cord in an occipital artery. The patient’s acute-phase reactant levels remained high. The new medical team was alarmed by what now seemed to be obvious GCA, and the patient was given pulse methylprednisolone immediately. However, it was too late; that evening, while watching television in his hospital room, the patient went blind in his left eye. Ophthalmology evaluated the patient the following day and confirmed bilateral vision loss because of AION. The only possible recommendation was to refer the patient to the State Commission for the Blind. The loss of sight in his remaining eye was devastating for this patient. He became profoundly depressed. He chose to discontinue dialysis, and he died shortly thereafter. What led to the patient’s blindness and, secondarily, to his death? Where did this case go wrong? How could a patient have gone blind in one eye and then in the other in this era of effective therapies for GCA? The crux of the matter is this: the clinicians placed too much emphasis on the finding of negative temporal artery biopsy results. The key diagnostic test was not the temporal artery biopsy specimens; rather, it was the patient’s history. Evidence of GCA in the patient’s history was substantial (if not overwhelming) at the time of his first episode of vision loss. Failure to place sufficient weight on that history and the existing clinical evaluation led to the cessation of glucocorticoid therapy that very likely would have preserved the vision in the patient’s remaining good eye. Let us review the evidence. At the time of the patient’s first episode of vision loss, he had experienced new headaches that were atypical for him, diplopia and jaw claudication. In addition, he had normochromic, normocytic anaemia and marked elevation of acute-phase reactants. Finally, bilateral AION was confirmed by fluorescein angiography. Although correct diagnosis often seems all too easy in retrospect, substantial evidence of GCA was discarded because of negative temporal artery biopsy results, and the diagnosis was simply missed—very much to the patient’s detriment. The take-home lesson is that, in the end, GCA is a clinical diagnosis not a pathological one. Clinicians always feel reassured by a positive temporal artery biopsy, but in reality they may not have the benefit of that confirmation. Therefore, it is important to understand what temporal artery biopsies can tell us and why they often fall short of our expectations. Excessively high expectations The problem of excessively high expectations is summed up by a statement and its corollary, both attributed to Dr Oscar Rambo, professor emeritus at the University of California-San Francisco as follows: ‘A mystic (and dangerous) assumption prevails in the mind of the clinician that the pathologist can produce a statement of absolute truth based on a small piece of tissue’ [3]. The corollary to this statement is that: ‘Equally dangerous to mankind is the pathologist who thinks the same.’ Many clinicians believe that temporal artery biopsy is the ultimate arbiter of whether a patient has GCA and that the diagnosis of GCA cannot be excluded without this procedure. Rambo’s quotation underscores that pathologists must understand not only the value of temporal artery biopsies but also their potential shortcomings. What are the test characteristics of temporal artery biopsy? Although some data (now >30 years old) suggest that the negative predictive value of bilateral temporal artery biopsy is > 90% [4], the specificity of such a finding is probably far lower. The 90% value was derived from a retrospective study at the Mayo Clinic, in which an extremely rigorous approach to the performance of temporal artery biopsy was undertaken. Biopsies at the Mayo Clinic were performed for many decades by a single, highly experienced surgeon who routinely obtained between 3 and 4 cm of temporal artery from the symptomatic side (if symptoms were localized) [5]. In addition, the contralateral side was biopsied if a frozen section failed to demonstrate GCA in the first biopsy. Unfortunately, this level of rigor is rarely achieved at most other medical institutions, and the yield of temporal artery biopsies at other centres is almost certainly far lower. Although it would be most useful to understand both the sensitivity and the negative predictive value of temporal artery biopsy for each medical centre, we simply do not know these test characteristics because there is no true gold standard. In fact, estimates of the sensitivity of temporal artery biopsy vary greatly among studies, and high false-negative rates are commonly reported [6–8]. Many experts consider sensitivity estimates of 60–70% to be accurate, but this is only a guess. The sensitivity of temporal artery biopsy is substantially affected by many variables, which are discussed herein. How we are led astray Inadequate biopsy specimen length Ask any pathologist what he or she most often requests, having agonized inconclusively at the microscope over a tiny (2 mm2) piece of temporal artery tissue, and the answer will be, Give me more tissue; that is, a longer segment of the artery. However, the trend in recent years has been to reduce the size of biopsy specimens as a push toward minimally invasive diagnostic procedures [9]. So significant is this issue that pathologists are now commonly asked to report on the adequacy of the biopsy specimen before giving their diagnostic opinion [10]. Although the value of a (positive) temporal artery biopsy has been accepted for decades, the ideal biopsy length to ensure optimal sensitivity is still a matter of debate. The often-cited rule of thumb that a temporal artery biopsy specimen should not be < 2.0 cm is seldom accomplished in practice. Moreover, a temporal artery biopsy specimen that is 2.0 cm upon removal in the operating room often shrinks substantially after processing in the pathology laboratory. The optimal biopsy length associated with increased diagnostic yield varies between studies from >0.5 cm (post-fixation) to 0.7 or ⩾1.0 cm [11–13]. Whatever the minimum acceptable length of a biopsy specimen, the adequacy of the sample should receive as much attention as the diagnosis itself in the setting of a negative temporal artery biopsy result. Inadequate histological work-up Histopathological work-up of vascular specimens, like other small biopsy samples, usually involves sectioning or serial sectioning of the sample in addition to specific stains. The idea behind the technique of sectioning through the tissue (or at least part of it) is to minimize the possibility of false-negative results, specifically, in patients with GCA, to avoid missing skip lesions. Early studies highlighted the need for examination of multiple histological sections (in addition to adequate biopsy length) to detect skip lesions [14]. Although specific guidelines dictating the number of sections that is appropriate for temporal artery biopsy specimens have not been published, consensus statements on the processing of cardiovascular surgical specimens are available [15]. These recommend three or more sections for tissue samples, which is routinely practised by most histopathology laboratories. However, it should be taken into consideration that sectioning still means that the pathologist will see only a fraction of the sample (a biopsy sample 1.5 cm in length consisting of six sections of 3-µm thickness equates to 0.12% of the entire sample). Temporal arteries not involved The temporal arteries are not part of the vascular distribution in which arterial compromise leads to blindness in GCA. Blindness results from involvement of the ophthalmic, retinal and posterior ciliary vessels. These vessels run a course behind the eye and track along the optic nerve. The temporal artery stems from a different branch of the external carotid artery and traces a path anterior to the tragus of the ear, near the course of branches of the facial nerve, up and over the temple, lying just under the skin and therefore accessible to biopsy. The temporal artery is biopsied because it is comparatively easy to do so, not because it necessarily reflects what is going on elsewhere in the large- and medium-sized vessel circulation or that its involvement predicts patients at higher or lower risk for vision loss. In short, temporal arteries that are free from inflammation do not exclude the presence of active GCA in other vascular distributions. Involved side not biopsied In most cases, if a temporal artery biopsy is worth doing, it is worth performing on both sides. Occasionally, a patient may have a classic history and floridly abnormal temporal arteries (tender, swollen, nodular lesions). Such cases might justify unilateral biopsy, but these are in the minority. Most of the time, the importance of getting it right and the data available on the usefulness of bilateral biopsy argue for taking biopsy specimens from both sides. A retrospective study performed by Boyev et al. [16] evaluated 186 patients who underwent bilateral temporal artery biopsy at Johns Hopkins Hospital. Only six patients (3%) had unilateral arteritis. Although this might be regarded as a small percentage not worth the additional effort and expense of bilateral procedures, those six patients represented at least 20% of the total number of patients whose GCA was diagnosed through biopsy. Similar conclusions were reached in another retrospective study that evaluated bilateral biopsy [17]. Among the 60 patients examined in that study, eight (13%) had GCA identified on only one side. However, those eight patients comprised 40% of the total number of patients diagnosed with GCA by biopsy. These studies suggest that once the decision has been made to pursue temporal artery biopsy, routine performance of bilateral procedures is justified. Inadequate surgical experience and performing biopsy on the wrong portion of the vessel When performing temporal artery biopsy, surgeons attempt to locate the artery by finding a pulse. However, a pulsatile segment of the vessel is not necessarily the portion of the artery that is most likely to yield a diagnosis, and therefore is not the most desirable segment to biopsy. Given that a pulseless artery is sometimes difficult to locate, the skill of the surgeon is an important determinant of the success of temporal artery biopsy. Unfortunately, at many centres, the least-experienced surgeon is usually sent to perform these procedures. The pulse of the temporal artery is often most easily palpable just anterior to the tragus of the ear. Surgeons often perform biopsy here, but this can be problematic for two reasons: first, as mentioned previously, an artery with a robust pulse is not the best site for biopsy; and second, biopsy of the artery in this area risks damage to the facial nerve that courses through the same region. A more fruitful and less risky biopsy targets the frontal branch of the superficial temporal artery just above the temple. Some patients with suspected temporal arteritis have been sent to undergo biopsy only to return with a statement from the surgeon that no pulsatile vessel could be identified. Difficulty finding a pulse should cause the surgeon to redouble efforts to locate a branch of the temporal artery, not abandon the procedure. Some surgeons use intra-operative US to assist in the identification of arteries with faint pulses. Pathological interpretation wrong Pathologists are often hampered by inadequate biopsy specimen length, as noted previously. Less commonly, they miss the diagnosis because of human error. Experience plays a major role in the pathologist’s ability to discern more subtle features that might be compatible with GCA. There are two common causes of just missing it: first, early changes in which incipient inflammation is not clear and, second, inflammation in the temporal artery (temporal arteritis) that is not GCA. A diagnosis of temporal arteritis is not synonymous with GCA. Some of the variations in the histological specimens of cross sections through temporal artery biopsies are depicted in Fig. 1. Advances in the treatment of individual forms of vasculitis now dictate the importance of precision in diagnosis. Fig. 1 View largeDownload slide Histological images of cross sections through temporal artery biopsies (A) and (B) (haematoxylin and eosin staining) show a case of florid GCA with transmural infiltration of lymphocytes, macrophages and the presence of several giant cells (black arrows). (C) (haematoxylin and eosin staining) shows a case of clinically positive GCA without noticeable inflammation on histological work-up. The lumen [as in the case in (A)] is compromised by significant circumferential intimal hyperplasia. An elastica van Gieson stain (D) shows minimal media fibrosis as well as evident disruption of the internal elastic lamina (black arrow), a subtle but not entirely specific sign of possible GCA. Fig. 1 View largeDownload slide Histological images of cross sections through temporal artery biopsies (A) and (B) (haematoxylin and eosin staining) show a case of florid GCA with transmural infiltration of lymphocytes, macrophages and the presence of several giant cells (black arrows). (C) (haematoxylin and eosin staining) shows a case of clinically positive GCA without noticeable inflammation on histological work-up. The lumen [as in the case in (A)] is compromised by significant circumferential intimal hyperplasia. An elastica van Gieson stain (D) shows minimal media fibrosis as well as evident disruption of the internal elastic lamina (black arrow), a subtle but not entirely specific sign of possible GCA. Confounders in diagnosing GCA: previous therapy It is widely accepted that the diagnosis of GCA can be made by temporal artery biopsy even 2 weeks after initiation of glucocorticoids and sometimes much longer [5]; however, there is little doubt that glucocorticoid treatment alters the underlying histopathology in the blood vessel wall. Both the pathological interpretation of the temporal artery biopsy and the clinical decision-making that follows must take this into account. The full impact of glucocorticoid administration on histological results is a matter of debate. Although some studies clearly show a time-dependent reduction in positive biopsy results after glucocorticoid treatment [18–20], others have not been able to demonstrate a clear relationship between glucocorticoid use and reduction of inflammation [5, 21, 22]. Giant cells are important but not required for this diagnosis The presence of giant cells in a temporal artery biopsy specimen is strongly suggestive of GCA and giant cells are probably the most carefully sought feature when GCA is suspected. However, the frequency with which giant cells are identified is substantially lower than the name of the disease suggests [23–25]. In some cases, multiple cuts of an artery are necessary to identify giant cells, and giant cells are not detected in ∼12% of clinically diagnosed cases of GCA [26]. Giant cells not only have a high positive predictive value for GCA, they may also predict the occurrence of cranial ischaemic events [27]. The presence of giant cells in biopsy specimens has also been strongly associated with the presence of jaw claudication and elevated serological markers of inflammation, suggesting that the presence of giant cells reflects a more acute and aggressive form of the disease [28]. In the absence of giant cells, more subtle histopathological changes must suffice as clues to possible presence of GCA [29]. Experience on the part of the pathologist is crucial to the proper interpretation of intimal thickening, focal infiltration by lymphocytes and macrophages, and localized scarring in the arterial wall, all of which are compatible with diagnosis but not necessarily diagnostic themselves in the absence of careful clinical correlation. Clinicians and pathologists should sit down at the microscope together whenever possible. Atherosclerosis causing intimal changes poses challenges in differentiation Another confounding factor is atherosclerosis, which can cause intimal hyperplasia similar to that of GCA. Arterial hypertension, dyslipidaemia and diabetes all affect the integrity of the vessel wall and may complicate histological interpretation in a patient with suspected GCA. Although atherosclerotic changes were observed in a cohort of 40 steroid-treated patients with GCA, the changes were not significantly different from those in age-matched controls [30]. Premature atherosclerosis has been reported in several pro-inflammatory rheumatological conditions and may be accelerated in vasculitis [31]. Much attention is focused on the internal elastic lamina and its fragmentation in the setting of GCA. However, fragmentation of the elastic laminae by elastolysis is a feature common to both atherosclerosis and GCA [32]. Variations in the appearance of laminar fragmentation and its relationship to fibrosis, as well as nuances of intimal hyperplasia, may provide important clues to help differentiate changes associated with atherosclerosis from those caused by vasculitis. In many cases, such histological subtleties make the pathologist’s job as much an art as a science. What is healed arteritis and what is its relationship to GCA? The term healed arteritis was used many years ago by Allsop and Gallagher [33] to distinguish a small subgroup of patients who did not fall into either the large category of patients with florid inflammation or the category of purely arterio-/atherosclerotic patients. Although studies are limited, results of some investigations of patients with so-called healed arteritis imply that this histopathology predicts a lower risk for disease recurrence, suggesting that the finding of healed arteritis is a biologically relevant state or subtype of vasculitis (or a different disease state completely) that does not necessitate the use of the high-dose glucocorticoids typically used to treat GCA [34]. In a recent study by Maleszewski et al. [35], patients with an initial positive result from temporal artery biopsy who were receiving treatment underwent a second biopsy after 3 months to 1 year. The multi-nucleated giant cells and granulomatous inflammation observed in the majority of positive biopsy results at presentation disappeared over time (and with treatment). Although signs of vascular remodelling were common to both the first and the second biopsies, medial fibrosis seemed to indicate a long-term healing process. Unfortunately, healed arteritis is an elusive histopathological finding. None of the histological findings often attributed to healed arteritis—reduplication or rupture of the internal and external elastic laminae, calcification, fibrosis and neovascularization—are specific for GCA that is no longer active. In short, reliable differentiation between healed (i.e. quiescent) GCA and arteriosclerosis may simply not be possible by routine histopathology [36]. When confronted with the finding of healed arteritis on a temporal artery biopsy report, the clinician must rely more on the clinical features presented than on the histopathological findings when deciding whether to treat the patient for GCA. Utility of staining for CD68-positive macrophages Correctly identifying macrophages by routine histology is something that the most junior of pathologists learn, and immunohistochemical staining is rarely performed in routine cases of inflammation to confirm the presence of macrophages outside of research protocols. However, when the presence of (isolated) macrophages can tip the scale towards one diagnosis or another, the use of an antibody against CD68, a glycoprotein on monocytes and macrophages [37], may aid in visualization of the number and distribution of tissue macrophages. In theory, the identification of macrophages around the internal elastic lamina, a finding facilitated by CD68 stains, provides additional support for the diagnosis of GCA, particularly in cases in which the histopathology is otherwise unconvincing. However, this theory has not been studied sufficiently. A retrospective study of 60 patients aimed to predict the biopsy characteristics suggestive of response to glucocorticoid treatment and characteristics distinguishing GCA, PMR or arteriosclerosis using immunophenotyping. In that study, CD68 positivity was a predictor of the success of glucocorticoid treatment in patients with GCA and distinguished GCA from PMR in the absence of GCA [38]. All patients examined had histologically proven GCA; therefore, the value of immunohistochemistry in ambivalent cases or cases with subtle histological abnormalities was not clarified by this study. In a larger retrospective study that used a three-tier system to distinguish among positive, indeterminate or negative biopsies, the value of CD68 lay in aiding a decision in indeterminate cases [25]. The authors concluded correctly that the ultimate diagnosis of GCA in such cases depends heavily on clinical judgement and clinicopathological correlation rather than simply on pathology. Although CD68 staining facilitates the identification of macrophages within the arterial wall [39], it does not resolve all problems related to the diagnosis of GCA. Histologically equivocal cases Dealing with histologically equivocal cases is part of the daily life of a pathologist. Although vasculitis is defined simply (or over-simply) as ‘inflammation of blood vessel walls at least at some time during the course of the disease’ [40], correctly identifying this histologically is often tricky. At what time point are aggregated lymphocytes considered inflammatory infiltrates? Does the presence of inflammatory infiltrates equate to disease? A recent study by Jia et al. [41] revealed that patients with lymphocytic inflammation involving small blood vessels or the adventitia of the temporal artery in the absence of transmural arteritic damage were not at increased risk for temporal arteritis-like adverse events. Therefore, it is frequently difficult for pathologists to know where to draw the line between disease and possible disease. Distinguishing perivasculitis from inflammation that truly targets and damages the blood vessel wall also frequently raises challenges. Mimickers of GCA These challenges aside, there is no doubt that other bona fide forms of vasculitis can affect the temporal artery. That is, diseases other than GCA can cause temporal arteritis and be misdiagnosed by pathologists who are not aware of this possibility. Non-GCA diseases most likely to affect the temporal arteries include the ANCA-associated vasculitides (granulomatosis with polyangiitis, microscopic polyangiitis and eosinophilic granulomatosis with polyangiitis), polyarteritis nodosa and mixed cryoglobulinaemia [42]. What are the solutions? The growing role of imaging Data from the Tocilizumab in Giant Cell Arteritis (GiACTA) trial [43] emphasize the growing role that imaging plays in the diagnosis of GCA. In that trial, 38% of patients either did not undergo temporal artery biopsy or the biopsy results were negative, highlighting the importance of imaging modalities as tools for the diagnosis of GCA. Conversely, many patients who have negative results on imaging studies will have positive temporal artery biopsy results, so the approaches to diagnosis are complementary. In some cases, neither temporal artery biopsy nor imaging yields a compelling case for the diagnosis of GCA. Therefore, advances in diagnostic imaging have supplanted pathology in a minority of GCA cases—particularly those characterized by large-vessel involvement—but temporal artery biopsy remains the mechanism of diagnostic confirmation in most patients. Careful history taking For a clinician, the key diagnostic test is neither the temporal artery biopsy nor large-vessel imaging, but the patient’s history; the patient’s history is the clinician’s scalpel and microscope. Skill in taking a patient’s history is crucial to suspecting GCA in the first place and to making the decision to order temporal artery biopsy or any other test that might confirm or refute the diagnostic suspicion. Communication The development of a single diagnostic test that is perfect, or close to perfect, be it clinical, serological, radiological or pathological, is not imminent. Communication between clinicians, surgeons, pathologists, radiologists, imaging technicians and patients is fundamental to determining the best solution in difficult cases. The clinician must devote the time necessary to listen to the patient, and to interpret the patient’s words wisely. In some cases, the clinician and radiologist can guide the surgeon with regard to the proper site for temporal artery biopsy. Following a biopsy, the likelihood of confirming or refuting the diagnosis of GCA in an accurate manner is then heightened by a process of collegial clinicopathological correlation, best conducted in person between the clinician and the pathologist sitting at two heads of the same microscope. Acknowledgements The manuscript was prepared by the authors, with professional writing and editorial assistance provided by Sara Duggan, PhD, on behalf of F. Hoffmann-La Roche Ltd. Supplement: This supplement is funded by F. Hoffmann-La Roche Ltd. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this manuscript. Disclosure statement: J.H.S. reports grants and consulting fees from Roche, Genetech and Xencor. The other author has declared no conflicts of interest. References 1 Horton BT, Magath TB, Brown GE. An undescribed form of arteritis of the temporal vessels. Mayo Clin Proc  1932; 7: 700– 1. 2 Hutchinson J. Disease of the arteries. On a peculiar form of thrombotic arteritis of the aged which is sometimes productive of gangrene. Arch Surg  1890; 1: 323– 9. 3 Rambo ON. The limitations of histologic diagnosis. Progr Radiat Ther  1962; 2: 215– 24. Google Scholar PubMed  4 Hall S, Persellin S, Lie JT et al.   The therapeutic impact of temporal artery biopsy. Lancet  1983; 2: 1217– 20. Google Scholar CrossRef Search ADS PubMed  5 Achkar AA, Lie JT, Hunder GG, O'Fallon WM, Gabriel SE. How does previous corticosteroid treatment affect the biopsy findings in giant cell (temporal) arteritis? Ann Intern Med  1994; 120: 987– 92. Google Scholar CrossRef Search ADS PubMed  6 Ashton-Key MR, Gallagher PJ. False-negative temporal artery biopsy. Am J Surg Pathol  1992; 16: 634– 5. Google Scholar CrossRef Search ADS PubMed  7 Borchers AT, Gershwin ME. Giant cell arteritis: a review of classification, pathophysiology, geoepidemiology and treatment. Autoimmun Rev  2012; 11: A544– 54. Google Scholar CrossRef Search ADS PubMed  8 Luqmani R, Lee E, Singh S et al.   The role of ultrasound compared to biopsy of temporal arteries in the diagnosis and treatment of giant cell arteritis (TABUL): a diagnostic accuracy and cost-effectiveness study. Health Technol Assess  2016; 20: 1– 238. Google Scholar CrossRef Search ADS PubMed  9 Fernandez EM, Helm T, Ioffreda M, Helm KF. The vanishing biopsy: the trend toward smaller specimens. Cutis  2005; 76: 335– 9. Google Scholar PubMed  10 Welsch MJ, Ioffreda MD, Clarke LE, Helm KF. Assessing and communicating adequacy of biopsy specimens and clinical information. J Am Acad Dermatol  2012; 66: 509– 10. Google Scholar CrossRef Search ADS PubMed  11 Breuer GS, Nesher R, Nesher G. Effect of biopsy length on the rate of positive temporal artery biopsies. Clin Exp Rheumatol  2009; 27(1 Suppl 52): S10– 3. 12 Ypsilantis E, Courtney ED, Chopra N et al.   Importance of specimen length during temporal artery biopsy. Br J Surg  2011; 98: 1556– 60. Google Scholar CrossRef Search ADS PubMed  13 Taylor-Gjevre R, Vo M, Shukla D, Resch L. Temporal artery biopsy for giant cell arteritis. J Rheumatol  2005; 32: 1279– 82. Google Scholar PubMed  14 Klein RG, Campbell RJ, Hunder GG, Carney JA. Skip lesions in temporal arteritis. Mayo Clin Proc  1976; 51: 504– 10. Google Scholar PubMed  15 Stone JR, Basso C, Baandrup UT et al.   Recommendations for processing cardiovascular surgical pathology specimens: a consensus statement from the Standards and Definitions Committee of the Society for Cardiovascular Pathology and the Association for European Cardiovascular Pathology. Cardiovasc Pathol  2012; 21: 2– 16. Google Scholar CrossRef Search ADS PubMed  16 Boyev LR, Miller NR, Green WR. Efficacy of unilateral versus bilateral temporal artery biopsies for the diagnosis of giant cell arteritis. Am J Ophthalmol  1999; 128: 211– 5. Google Scholar CrossRef Search ADS PubMed  17 Pless M, Rizzo JF3rd, Lamkin JC, Lessell S. Concordance of bilateral temporal artery biopsy in giant cell arteritis. J Neuroophthalmol  2000; 20: 216– 8. Google Scholar CrossRef Search ADS PubMed  18 Allison MC, Gallagher PJ. Temporal artery biopsy and corticosteroid treatment. Ann Rheum Dis  1984; 43: 416– 7. Google Scholar CrossRef Search ADS PubMed  19 Font RL, Prabhakaran VC. Histological parameters helpful in recognising steroid-treated temporal arteritis: an analysis of 35 cases. Br J Ophthalmol  2007; 91: 204– 9. Google Scholar CrossRef Search ADS PubMed  20 Narvaez J, Bernad B, Roig-Vilaseca D et al.   Influence of previous corticosteroid therapy on temporal artery biopsy yield in giant cell arteritis. Semin Arthritis Rheum  2007; 37: 13– 9. Google Scholar CrossRef Search ADS PubMed  21 Ray-Chaudhuri N, Kine DA, Tijani SO et al.   Effect of prior steroid treatment on temporal artery biopsy findings in giant cell arteritis. Br J Ophthalmol  2002; 86: 530– 2. Google Scholar CrossRef Search ADS PubMed  22 Bury D, Joseph J, Dawson TP. Does preoperative steroid treatment affect the histology in giant cell (cranial) arteritis? J Clin Pathol  2012; 65: 1138– 40. Google Scholar CrossRef Search ADS PubMed  23 Parker F, Healey LA, Wilske KR, Odland GF. Light and electron microscopic studies on human temporal arteries with special reference to alterations related to senescence, atherosclerosis and giant cell arteritis. Am J Pathol  1975; 79: 57– 80. Google Scholar PubMed  24 Lie JT. Temporal artery biopsy diagnosis of giant cell arteritis: lessons from 1109 biopsies. Anat Pathol  1996; 1: 69– 97. Google Scholar PubMed  25 Zhou L, Luneau K, Weyand CM et al.   Clinicopathologic correlations in giant cell arteritis: a retrospective study of 107 cases. Ophthalmology  2009; 116: 1574– 80. Google Scholar CrossRef Search ADS PubMed  26 Regan MJ, Wood BJ, Hsieh YH et al.   Temporal arteritis and Chlamydia pneumoniae: failure to detect the organism by polymerase chain reaction in ninety cases and ninety controls. Arthritis Rheum  2002; 46: 1056– 60. Google Scholar CrossRef Search ADS PubMed  27 Muratore F, Boiardi L, Cavazza A et al.   Correlations between histopathological findings and clinical manifestations in biopsy-proven giant cell arteritis. J Autoimmun  2016; 69: 94– 101. Google Scholar CrossRef Search ADS PubMed  28 Ting KH, Lester S, Dunstan E, Hill CL. Association between histological features and clinical features of patients with biopsy positive giant cell arteritis. Clin Exp Rheumatol  2016; 34(3 Suppl 97): S40– 3. 29 Stacy RC, Rizzo JF, Cestari DM. Subtleties in the histopathology of giant cell arteritis. Semin Ophthalmol  2011; 26: 342– 8. Google Scholar CrossRef Search ADS PubMed  30 Gonzalez-Juanatey C, Lopez-Diaz MJ, Martin J, Llorca J, Gonzalez-Gay MA. Atherosclerosis in patients with biopsy-proven giant cell arteritis. Arthritis Rheum  2007; 57: 1481– 6. Google Scholar CrossRef Search ADS PubMed  31 Tervaert JW. Translational mini-review series on immunology of vascular disease: accelerated atherosclerosis in vasculitis. Clin Exp Immunol  2009; 156: 377– 85. Google Scholar CrossRef Search ADS PubMed  32 O'Brien JP. Destruction of elastic tissue (elastolysis) as a link between atherosclerosis and the temporal arteritis/polymyalgia rheumatica syndrome. Observations on an actinic factor in vascular disease. Pathol Biol  1984; 32: 123– 38. Google Scholar PubMed  33 Allsop CJ, Gallagher PJ. Temporal artery biopsy in giant-cell arteritis. A reappraisal. Am J Surg Pathol  1981; 5: 317– 23. Google Scholar CrossRef Search ADS PubMed  34 ter Borg EJ, Haanen HC, Seldenrijk CA. Relationship between histological subtypes and clinical characteristics at presentation and outcome in biopsy-proven temporal arteritis. Identification of a relatively benign subgroup. Clin Rheumatol  2007; 26: 529– 32. Google Scholar CrossRef Search ADS PubMed  35 Maleszewski JJ, Younge BR, Fritzlen JT et al.   Clinical and pathological evolution of giant cell arteritis: a prospective study of follow-up temporal artery biopsies in 40 treated patients. Mod Pathol  2017; 30: 788– 96. Google Scholar CrossRef Search ADS PubMed  36 Cox M, Gilks B. Healed or quiescent temporal arteritis versus senescent changes in temporal artery biopsy specimens. Pathology  2001; 33: 163– 6. Google Scholar CrossRef Search ADS PubMed  37 Holness CL, Simmons DL. Molecular cloning of CD68, a human macrophage marker related to lysosomal glycoproteins. Blood  1993; 81: 1607– 13. Google Scholar PubMed  38 Braun N, Fritz P, Rieth A et al.   Predictors for treatment success and expression of glucocorticoid receptor in giant cell arteritis and polymyalgia rheumatica. J Rheumatol  2009; 36: 2269– 76. Google Scholar CrossRef Search ADS PubMed  39 Wang AL, Raven ML, Surapaneni K, Albert DM. Studies on the histopathology of temporal arteritis. Ocul Oncol Pathol  2017; 3: 60– 5. Google Scholar CrossRef Search ADS PubMed  40 Jennette JC, Falk RJ, Bacon PA et al.   2012 revised International Chapel Hill Consensus Conference Nomenclature of Vasculitides. Arthritis Rheum  2013; 65: 1– 11. Google Scholar CrossRef Search ADS PubMed  41 Jia L, Couce M, Barnholtz-Sloan JS, Cohen ML. Is all inflammation within temporal artery biopsies temporal arteritis? Hum Pathol  2016; 57: 17– 21. Google Scholar CrossRef Search ADS PubMed  42 Cavazza A, Muratore F, Boiardi L et al.   Inflamed temporal artery: histologic findings in 354 biopsies, with clinical correlations. Am J Surg Pathol  2014; 38: 1360– 70. Google Scholar CrossRef Search ADS PubMed  43 Tuckwell K, Collinson N, Dimonaco S et al.   Newly diagnosed vs. relapsing giant cell arteritis: Baseline data from the GiACTA trial. Semin Arthritis Rheum  2017; 46: 657– 64. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

RheumatologyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial