A 72-year-old white woman requests urgent evaluation for increasing achiness of 6 weeks’ duration.
Symptoms began shortly after a 3 day family reunion, during which time she “cooked and baked for 20 people.” Initially attributed to overuse, symptoms have steadily worsened, unresponsive to ibuprofen in doses up to 1800 mg per day. Aching prominently involves the upper arms but has also recently affected the back of the neck and the posterior thighs. She awakens at night with pain, feels “like the tin man” on arising in the morning, and doesn’t “loosen up” until 1:00 or 2:00 o’clock in the afternoon. She can no longer hook her bra in the back, and must be assisted with donning a blouse by her husband.
Past medical history includes hypercholesterolemia and hypertension. Her general health has been good. She has been on stable doses of atorvastain and hydrochlorthiazide for several years.
Musuculoskeletal examination is notable for limited abduction and internal rotation at both shoulders.
The erythrocyte sedimentation rate is 49 mm/hour, the C-reactive protein is 24.7 mg/L (normal 0–3).
Polymyalgia rheumatica is suspected, and prednisone 10 mg every morning, is started. She calls 1 week later, “50% better”; in another week, she reports that she is virtually symptom free with no nocturnal pain, minimal morning stiffness, and no functional limitations.
Diagnostic criteria for polymyalgia rheumatica (PMR) have not been validated, although they are now under prospective evaluation by an international work group. Previously formulated criteria for diagnosis have been empirically derived, and include older age; recent onset of symptoms; prominent morning stiffness; proximally distributed symptoms; elevated acute phase reactants; and, in the Healey criteria, the important stipulation of a brisk therapeutic response to low-dose corticosteroids (CS).
PMR is the quintessential systemic rheumatic disease of older adults: the mean age of onset is 73 years ; it occurs rarely in patients younger than the age of 50, if at all.
The onset of symptoms is recent and discrete, sometimes even precipitous—in contrast, for example, to the long duration of symptoms that is typical for the presentation of the patient with fibromyalgia. The requirement that symptoms be of more to 2 to 4 weeks in duration is intended to differentiate PMR from self-limited syndromes such as arthralgias and myalgias or viral etologies. However, because the onset of symptoms in PMR may be abrupt, insistence on a disease duration of 1 month as a prerequisite for diagnosis may inappropriately delay efforts at treatment.
Morning stiffness is invariable in PMR. The gel phenomenon—stiffness with inactivity—is typically of notable intensity in PMR, and manifests as nocturnal pain, marked and prolonged morning stiffness (often resulting in the need for assistance with such activities of daily living as morning dressing), and stiffness with inactivity during the day.
Proximal achiness in PMR is symmetric, affecting the upper arms, posterior neck, and thighs. Achiness of the upper arms, with the complaint of difficulty raising the arms above the shoulders, is especially characteristic. Early in the disease, the lower extremities may be spared; accordingly, in an older adult, the new onset of symmetric aching of the upper arms should encourage consideration for a possible diagnosis of PMR. Physical examina- tion may show limited range of motion about the shoulders and hips, especially the shoulders, owing to a pathophysiology that involves synovitis and bursitis. The term “poly myalgia rheumatica,” although embedded in the medical literature, is hence a misnomer. Range of motion may have returned to normal if the patient is examined later in the day. Muscle tenderness to palpation is a nonspecific sign of no clinical use. The motor examination, although sometimes problematic in the patient who is in significant pain, shows no weakness.
Elevations of the acute phase reactants are usual. The most familiar and most widely referenced of the acute phase reactants is the erythrocyte sedimentation rate (ESR), which, although sometimes elevated to triple figures, has a wide range in PMR. An increased ESR has generally been viewed as essential to the diagnosis; indeed, an ESR greater than 40 mm/h is specified in all proposed diagnostic criteria. But it is clear that PMR can present in the context of a “normal” ESR. The reported frequency of this occurrence varies considerably. Myklebust and Gran found ESRs less than 20 mm/h in only 2.3% of patients with PMR. ESRs less than 40 mm/h have been reported in 7.3% to 22% of other series. The spread among these percentages is difficult to reconcile, but clinical observation suggests that one important consideration is the length of time between when a patient develops symptoms and when he or she is seen and evaluated by a physician.
The well-known vagaries of the ESR (e.g., its susceptibility to variation with gender, age, anemia, and abnormal immunoglobulins) have played a role in an increased use of the C-reactive protein (CRP). Prospective head-to-head studies on the diagnostic value of the CRP versus the ESR in PMR, however, are remarkably few. However, growing clinical experience does suggest that the CRP may have superiority over the ESR for both the diagnosis of and follow-up of PMR. Practically, it is common to assess both acute phase reactants, and then to rely on the one that seems better to reflect the clinical activity of a given patient’s disease.
Initial Treatment: The Therapeutic Trial of Low-Dose Corticosteroids
The prompt and virtually complete abolition of symptoms with low-dose CS may be the most striking clinical feature of PMR. Nonsteroidal anti-inflammatory drugs are not indicated because they are of minimal, if any, value for symptomatic management in this disease. Despite the indubitable benefit of CS for the treatment of PMR, controlled studies on the appropriate dose of this drug for initial management have not been done. Studies comparing various CS preparations (e.g., prednisone, prednisolone, and intramuscular methylprednisolone ) are similarly lacking. Doses between 10 and 20 mg of prednisone per day are normative. Higher doses of CS of course are also effective, but they produce nonspecific symptomatic improvement in many musculoskeletal diseases, including those with local pathology, such as rotator cuff disease and osteoarthritis, whereas it is the exquisite symptomatic response to low-dose CS that has diagnostic value in the patient with possible PMR.
The response to low-dose CS may be immediate and seen after only a single dose, but it is common that a slightly longer period of time is required before a complete symptomatic response becomes fully manifest, especially in cases in which the inflammatory reaction is intense. In such cases, the use of divided doses of prednisone may be helpful. Studies of serum interleukin 6 (IL-6) levels in PMR suggest a physiologic basis for this approach. These levels are commonly elevated in PMR, often significantly so; they are promptly suppressed after a dose of CS, but within hours will begin to rebound. This rapid increase in IL-6 levels probably underlies the nocturnal recrudescence of aching and stiffness in the patient whose PMR remains active when under treatment with a single morning dose of CS.
The definition of remission in PMR has not been formalized; practically, however, if there has not been total or near-complete resolution of aching and stiffness after a 14- to 21-day trial of low-dose CS, the diagnosis of PMR should be strongly reconsidered and further diagnostic work-up undertaken. There are several approaches to the conduct of the diagnostic and therapeutic trial of CS, one of which could begin with prednisone 10 mg in a single morning dose. Because it is the immediate symptomatic response to CS that is of such diagnostic import, the patient in instructed to call in 1 week. If the therapeutic response is incomplete, the prednisone is split to 5 mg bid, and at the end of week 2, the patient is again told to call. If symptoms persist, the dose is increased to 10 mg in the morning and 5 mg at night. However, if there is not marked improvement by the end of week 3, further evaluation is advised. Very occasionally, if there is intense inflammation, a prednisone dose of more than 20 mg per day may be needed for initial treatment.
The differential diagnosis of diffuse achiness in an older adult includes other systemic rheumatic diseases, especially rheumatoid arthritis (RA). Distal symptoms and signs, however, may occur in PMR, and differentiation between PMR and RA with a polymyalgic presentation in an older adult can be difficult initially (see below, The Problem of Peripheral Synovitis). Pain may be part of the presentation of polymyositis or other myopathies, but such diseases manifest more often with weakness than with pain, without morning stiffness. Measurement of the serum creatine kinase (CK), which is always normal in PMR, will clarify the diagnosis.
Achiness may be among the nonspecific constitutional symptoms that can initially accompany various infections; fever, however, is rare in uncomplicated PMR and, if present, especially in the context of arthralgis and myalgias of recent onset, could indicate a diagnosis such as endocarditis and the need for blood cultures. Fever in true PMR raises concern for concurrent giant cell arteritis (GCA).
Diffuse pain may result from osseous disease, as in multiple myeloma. The possibility of widespread skeletal metastases must be borne in mind in the context of prior or even remote malignancy. Electrophoretic studies of the serum and urine and work-up for malignancy may be indicated. PMR is not a paraneoplastic syndrome, is not associated with an increased risk of cancer, and does not require a malignancy work-up. That said, atypical features in the patient with suspected PMR, such as adenopathy or pulmonary infiltrates, must call the diagnosis into question. There appears to be an association of PMR with myelodysplastic syndromes, but in this unusual situation, PMR responds to low-dose CS in typically brisk fashion.
Statin-related myopathy is commonly considered by both patient and physician in the differential diagnosis. The serum CK is elevated in patients with a bona fide statin myopathy; patients with statin-related myalgias and a normal CK are differentiated from PMR by the absence of a prominent gel phenomenon and the lack of symmetric and proximal localization of symptoms. Significantly symptomatic fibromyalgia is rather uncommon in older adults, and symptoms are generally of long duration, not of recent onset. Multifocal degenerative disease may generate an array of symptoms reminiscent of PMR. For example, rotator cuff disease is common in older adults, and if bilateral and accompanied by degenerative disc disease of the cervical spine or osteoarthritis at the hips, could suggest PMR. In this setting, morning stiffness is usually brief in duration, and acute phase reactants are not elevated. The stiffness that accompanies neurologic disease such as Parkinson’s, can be distinguished from PMR by history and physical examination.
Although commonly included in the differential diagnosis, hypothyroidism rareIy presents with symptoms that are confounded with PMR. Other conditions said to have the potential for diagnostic confusion with PMR have included calcium pyrophosphate dehydrate deposition disease, purported late-onset spondyloarthropathy, and malignancy, but in clinical practice, such occurrences are also rare.
The Problem of Peripheral Synovitis
Distal symptoms and signs, which include carpal tunnel syndromes, peripheral arthritis, and distal swelling, occur in one half of patients with PMR.
Carpal tunnel syndromes occur in as many as one of six patients with PMR and often remit with low-dose steroids.
A modest peripheral arthritis, which is never erosive, may affect the wrists, metacarpophalangeal joints, and knees, and responds with the same singular sensitivity to low-dose steroids as does the proximal symptomatology. Obviously, however, the presence of distal symptoms and signs could suggest a diagnosis of classic rheumatoid disease, which occasionally presents in older adults. The finding of peripheral arthritis in a patient with suspected PMR warrants measurement of rheumatoid factor or anticyclic citrullinated antibodies. The discovery of erosions, of course, indicates rheumatoid disease.
It is not unusual that overt polyarthritis of abrupt onset in an older adult is seronegative, thus confronting the clinician with the dilemma of distinguishing PMR from ostensible seronegative rheumatoid arthritis. Such patients have been described under the rubrics of late-onset rheumatoid arthritis or elderly-onset rheumatoid arthritis, but efforts to identify features that could distinguish at presentation among PMR, classic rheumatoid disease, or other entities have been largely unsuccessful. Practically, the distinction is made by vigilant follow-up, which may require 3 to 6 months before a clinical diagnosis can be firmly settled. The repeated appearance of overt peripheral synovitis during attempts to taper the prednisone dose to below 10 mg/day points to a diagnosis of rheumatoid disease rather than PMR.
Occasionally in older adults, prominent swelling over the dorsa of the hands and wrists presents explosively, which has been dubbed “RS3PE” (remitting symmetrical seronegative synovitis with peripheral edema) or the puffy edematous hand syndrome. This presentation is usually found in the setting of PMR. The syndrome, which may dominate proximal symptoms and also manifest as a relapse of PMR, responds to low-dose CS with characteristic briskness. Though sometimes construed as a unique entity, the puffy edematous hand syndrome ( Fig. 15-1 ) seems most sensibly viewed as part of the spectrum of PMR.
The Question of Giant Cell Arteritis
Clinically explicit, biopsy-proved GCA is customarily said to affect 15% of patients with PMR. A recent prospective study reported GCA in 6% of 248 patients with PMR, a figure that seems closer to clinical experience.
Evidence of arteritis has also been found on temporal artery (TA) biopsy in occasional PMR patients with no symptoms referable to GCA. Evidence for the presence of subclinical GCA has also been found by recent imaging techniques. Using duplex ultrasonography, the halo sign (thought to represent edema of the vascular wall) or stenosis was seen in the temporal arteries of eight of 102 patients with PMR who were clinically asymptomatic for GCA, of whom four subsequently had TA biopsies positive for arteritis. Using 18-fluorodeoxyglucose positron emission tomography (FDG-PET), an increased uptake in the subclavian arteries was noted in one third of 35 patients with PMR. The prognostic and therapeutic significance of such subclinical arteritis is completely unknown.
TA biopsy is not routinely recommended in patients with PMR who have no clinical symptoms of GCA. Biopsy should be strongly considered if there are arteritic symptoms, especially new onset headache, jaw claudication, or amaurosis, or if fever is present. Biopsy should also be considered if physical examination demonstrates asymmetric blood pressures in the arms, vascular bruits by auscultation, or abnormalities of the temporal arteries to palpation. Similar consideration should be given to the occurrence of any of these symptoms and signs in the PMR patient already on treatment, because low-dose CS therapy is not a bulwark against the emergence of GCA. Finally, it should be remembered that the clinical courses of PMR and GCA may not be synchronous, and that GCA may present in patients whose PMR is no longer active, and vice-versa.
Subsequent Treatment and Clinical Course
As is the case with the selection of the initial dose of CS for PMR, there are no evidence-based guidelines for the ensuing dosing of this drug, either for maintenance or as a taper. Whether CS treatment actually affects the natural history of PMR (i.e., whether it is remission inducing) is unknown; conceivably, it provides symptomatic management until the disease itself undergoes spontaneous remission. Accordingly, subsequent management of PMR is an exercise in empiricism, the goal of which is the identification of the lowest dose of CS that keeps the given patient comfortable and functional. Generally, after a patient has been asymptomatic for 1 month, dose reductions are initiated, every 2 to 4 weeks. Above 10 mg of prednisone per day, 2.5-mg decrements may be possible, but because symptoms in PMR are so steroid sensitive, the use of 1-mg decrement once the dose is below 10 mg/day is an effective strategy.
Some patients with PMR can be tapered smoothly and rather quickly off CS over a matter of mere months, but subsequent treatment is commonly punctuated by symptomatic relapses in up to 50% of cases, necessitating upward adjustments in the dose of CS. Consequently, CS therapy can be prolonged. One recent study reported that a third of patients required CS treatment for more than 6 years, whereas at another center, the median duration of some CS therapy for PMR was 37 months. The data imply that there are two large subsets of PMR, one with milder disease, for which 1 to 2 years of treatment with CS suffices, and another with more protracted, relapsing disease, for which several years of such treatment is necessary. The potential length of treatment also implies that, when CS treatment is begun for PMR, the clinician must be prepared to commit to the care of an older adult for what can be an extended period of time with a medication that has an array of potential toxicities. Finally, in addition to relapses, the clinical course of PMR can also be marked by overt recurrences, that is, renewed activity of disease years after CS treatment has been stopped.
Normalization of the ESR and CRP usually follows remission of symptoms in PMR. Increases in these tests may accompany relapses of symptoms, but not always, and a sizeable portion of symptomatic flares of PMR are not attended by such elevations, especially in the case of the ESR. Rises in the acute phase reactants in the absence of worsened symptoms, on the other hand, are not grounds for a reflexive escalation in the CS dose.
Ultimately, dose reductions of CS in PMR must be individualized. If, as often occurs, the CS taper cannot be uneventfully consummated, a minimum dose adequate for symptomatic relief is located. The frequency with which further dose reductions are attempted will vary, depending on the comfort level of the physician and patient with the given CS dose and the extent to which CS-related toxicities ensue. Because the CS dose is adjusted against the patient’s symptoms, frequent assessment of the acute phase reactants seems of little practical use, unless a marked change in clinical status supervenes.
Toxicities of Corticosteroids
The toxicities of CS are well known and can occur even in the low doses used to treat PMR.
Axial bone loss can begin early in the course of CS treatment, for which aggressive prophylaxis with appropriate supplements of calcium and vitamin D is warranted in most patients, both women and men. Baseline bone mineral densitometry should be obtained. The value of bisphosphonates for reducing fracture risk due to CS use is well documented, and in clinical practice, the proactive deployment of such therapy has significantly decreased the incidence of fractures occurring during CS treatment for PMR.
An underacknowledged side effect of even low-dose CS in older adults is that of increased capillary fragility, which may result in widespread ecchymoses and in slowly healing superficial lacerations, a susceptibility exacerbated by concurrent treatment with aspirin, warfarin, or other anticoagulants. There is no treatment for this toxicity, aside from reduction in the CS dose.
The Difficult CS Taper
The taper of CS in PMR may be difficult for an assortment of reasons. PMR itself is prone to frequent relapses, as discussed earlier; the persistence of distal synovitis may prompt a revision of the operational diagnosis to rheumatoid disease; and CS-related toxicities may ensue. In addition, CS may produce capricious symptomatic improvement in a mixture of noninflammatory musculoskeletal conditions, such as osteoarthritis and rotator cuff disease. As CS are tapered, it is important to differentiate recrudescence of symptoms due to these local conditions from those arising from a true relapse of PMR, in order to avoid overtreatment.
Adjunctive Treatments for Polymyalgia Rheumatica
Although symptomatic treatment of PMR is invariably effective with the use of low-dose CS, such management in older adults may engender unacceptable toxicity, as has been discussed. Methotrexate (MTX) and tumor necrosis factor (TNF) alpha blockers have been explored as possible adjunctive treatments.
An earlier randomized control trial showed no benefit from the use of adjunctive MTX at a dose of 7.5 mg/week. The more recent randomized control trial (RCT) allocated 72 patients to treatment with prednisone alone vs prednisone and MTX at a dose of 10 mg/week. Prednisone was administered in an initial dose of 25 mg/week for 4 weeks, and then tapered by 2.5-mg decrements every 4 weeks, unless a flare (defined as increased symptoms of PMR and a rise in the ESR to greater than 30 or in the CRP to greater than 5 mg/L) prompted an increased dose. MTX was given for 48 weeks; patients were followed for 76 weeks. More patients treated in the MTX-treated group achieved the primary endpoint, discontinuation of CS ( P = 0.003); they also had fewer flares and required a lower total cumulative CS dose.
The extent to which these conclusions can be generalized to practice is uncertain. The mean reduction in the prednisone dose in the MTX-treated group was little more than 1 mg/day; there were no differences in potential CS-related toxicities between the two groups; and the MTX dosing was low, by current standards.
The only RCT of TNF alpha blockade for the treatment of PMR involved infliximab (IFX). In that study, 52 patients were randomized to treatment with prednisone alone versus prednisone and IFX, given at a dose of 3 mg/kg at weeks 0, 2, 6, 14, and 22. Prednisone was administered at an initial dose of 15 mg/day, and then tapered off by 16 weeks, unless a flare (defined as in the MTX study described above) required an increase in dose. Patients were followed for 52 weeks. The study was negative: there was no significant effect of IFX on any of the outcome variables, including numbers of relapse-free patients, patients no longer taking prednisone, the number of flares, the duration of CS treatment, and cumulative CS dose.
A small, recent case series described the use of etanercept, 25 mg twice weekly, in refractory PMR, defined as patients whose CS dose could not be reduced below 7.5 mg/day after 12 months of treatment, and who had relapsing symptoms of PMR and CS-related adverse events. In five of six patients the CS dose was decreased but not apparently stopped; whether CS-related toxicities were affected was not reported.
The routine use of methotrexate or TNF alpha blockade in the initial management of PMR is not recommended. Whether either of these medications is of value for the subsequent management of the PMR patient who develops CS-related toxicities is unclear. In practice, for patients with intolerable toxicities from even low-dose CS, MTX in contemporary doses of 15 to 25 mg/week seems to have a modestly steroid-sparing benefit, with the caveat that such doses have not been formally studied. The role of TNF alpha blockade at any point in a disease whose treatment involves only low-dose CS is uncertain, at best.
Low doses of CS are the foundation of treatment for PMR. Management should aim for the lowest dose of CS needed to maintain quality of life and functional status for the individual patient. Because of the extraordinary sensitivity of PMR to CS, use of 1-mg decrements for the taper of this drug is advantageous. Both patients and physicians should be aware of the frequency of relapses and of the potential need for long-term treatment.
GIANT CELL ARTERITIS
A 78-year-old woman is referred for emergency evaluation of sudden loss of vision in the left eye.
Two evenings previously, sudden and complete loss of vision in the left eye occurred. Thirty-six hours later, an ophthalmologist finds evidence for an acute ischemic optic neuropathy, prescribes 60 mg of prednisone, and requests emergency rheumatology consultation.
Two months previously, the left side of the head became tender to the touch, to an extent that the patient cancelled an appointment with her hairdresser for a permanent. She also experienced unilateral maxillary pain. Her primary care physician was consulted, who raised the possibility of sinusitis, for which a 2-week course of antibiotics was prescribed. Scalp pain receded, but was replaced with a low-grade, daily, occipital headache, somewhat improved with acetaminophen. Nine days ago, an “odd shape” appeared in the “left corner” of the eye for 10 minutes. She denies morning stiffness, jaw pain, fever, or weight loss.
There is a history of an anxiety disorder, paroxysmal atrial fibrillation, and osteoporosis, for which bisphosphonate therapy had been refused.
Medications include a beta blocker, aspirin 325 mg/day, and diazepam.
Physical examination shows an anxious patient, with no light perception in the left eye. Blood pressures are equal in the arms, 136/76 both arms. There are no vascular bruits or cardiac murmurs. TA pulsations are symmetric and nontender.
The ESR, obtained by the ophthalmologist earlier in the day, is 78.
Intravenous methylprednisone, 500 mg, is administered. The next day, unilateral TA biopsy is performed. Intravenous methylprednisolone is repeated that day and again the following morning. Histopathologic inspection of the TA shows a panarteritis, with disruption of the internal elastic lamina. Headache and facial pain resolve promptly, but visual loss in the left eye remains complete.
Diagnosis: Clinical Features
Like PMR, GCA is a disease of older adults: it rarely occurs in patients younger than 50 years of age, and 80% of the patients are older than 70 years of age. Although any ethnic or racial group can be affected, the disease is unusual in blacks.
The most frequent phenotype of GCA is that of so-called cranial arteritis, which presents as symptoms and signs referable to extracranial arteries, especially branches of the external carotid arteries. Headache in any location, whether it is temporal, frontal, parietal, or occipital, is common, affecting three quarters of patients. Jaw claudication occurs in slightly less than one-half of patients. Permanent visual loss, the most sinister complication of GCA, continues to be reported in 10% to 20% of patients ( Table 15-1 ). More than 90% of instances of vision loss are due to an anterior ischemic optic neuropathy, caused by occlusion of the posterior ciliary artery, which supplies the optic nerve; most other cases of visual loss result from occlusion of the central retinal artery. Involvement of branches of the external carotid arteries can result in scalp tenderness, facial swelling, and pains in the tongue and throat.
|Aiello (Mayo Clinic, 1993)||14.0|
|Font (Barcelona, Spain, 1997)||15.75|
|Gonzalez-Gay (Northwestern Spain, 2000)||14.9|
|Liozon (France, 2001)||13.2|
|Nesher (Israel, 2004)||18.3|
|Salvarani (Italy, 2005)||19.1|
Almost as common as headache in GCA are constitutional symptoms and signs, including anorexia, weight loss, malaise, sweats, and fevers. Fevers may be high-grade, exceeding 39° C, and may account for one of six fevers of unknown origin in older adults. In up to 10% of patients with GCA, systemic symptoms and signs dominate the clinical presentation, leading to evaluations for malignancy or infection. Also common in GCA are musculoskeletal symptoms and signs—not only PMR, which occurs in slightly under one-half of patients with GCA, but also distal findings, including peripheral arthritis, carpal tunnel syndromes, and distal extremity swelling with edema.
Physical examination of the patient with suspected GCA should include careful palpation of the frontal and parietal branches of the superficial temporal arteries for tenderness, nodularity, enlargement, or decreased pulsation. Although these abnormalities are neither fully sensitive nor specific, they may be found in up to one half of patients with GCA. The self-evident converse should also be remembered: the temporal arteries are normal on palpation in at least one-half of patients.
Among laboratory data are anemia and thrombocytosis. Liver function tests may be elevated, most often the alkaline phosphatase, rarely more than twice normal. The ESR is usually elevated, often significantly so, but neither high nor low values can be relied on to confirm or rule out a diagnosis of GCA. The medical literature and clinical experience are replete with biopsy-proved cases of GCA in patients whose ESRs were in a normal range or but modestly increased. In the Mayo Clinic series of 167 patients, 10.8% had ESRs less than 50, and 5.4% less than 40; the risk of visual loss in these patients was the same as patients with higher ESRs. Ranges for the CRP in the initial diagnosis of GCA have not been studied.
Diagnosis: The Temporal Artery Biopsy
The definitive diagnosis of GCA is premised on histopathologic proof. Such proof is usually sought by biopsy of the most accessible of the cranial arteries, the TA. The procedure is simple, brief, and of little risk.
Some clinical features may argue for or against the diagnosis, but none is sufficiently specific or sensitive to be of conclusive diagnostic use to the clinician. The limitations of attempting a clinical diagnosis of GCA are suggested by the instructive meta-analysis assembled by Smetana and Shmerling of 2680 positive and negative TA biopsies from 21 reports. The only symptoms that increased the likelihood of a positive TA biopsy were jaw claudication (likelihood ratio [LR] 4.2; 95% confidence interval [CI], 2.8–6.2) and, interestingly, diplopia (LR 3.4; 95% CI, 1.3–8.6). The sensitivity, however, of these features was low: in biopsy-proved GCA, jaw claudication was present in only 34% (95% CI, 0.29–0.41) of cases, and diplopia in 9% (95% CI, 0.7–0.13). Conversely, the more prevalent clinical features of GCA (e.g., any headache, constitutional symptoms, and PMR) were not associated with increased LRs of a positive biopsy.
Palpation of the TA yielded more robust predictive information than most features of the history: “beaded,” enlarged, or tender temporal arteries were all associated with increased LRs of 4.6 (95% CI, 1.1–18.4), 4.3 (95% CI, 2.1–8.9), and 2.6 (95% CI, 1.9–3.7). An ESR > 100 mm/h had a lesser predictive value for a positive biopsy (LR 1.9, 95% CI, 1.1–3.3) than findings on palpation of the temporal arteries, and the mean value of the ESR was not statistically different between biopsies that were positive and negative ( Tables 15-2 and 15-3 ).
|Symptoms, Signs, and Laboratory Data||Positive LR (CI, 95%)||Negative LR (CI, 95%)|