The diagnostic value of additional research in giant cell arteritis

For the diagnosis of giant cell arteritis the criteria established by the
American College of Reumatology in 1990 are often used. However, these criteria
(Annex 2A) are research criteria and are not always applicable in clinical
practice. In the presence of 3 of 5 criteria a sensitivity of 93.5% and a
specificity of 91.2% is achieved. However, if these criteria are used in
clinical practice, these numbers will be lower, and reaches a sensitivity of
75% and a specificity of 92%. For the population these criteria have a positive
predictive value of only 29% . Clinicians should therefore be cautious in
applying these criteria in clinical practice. A combination of history,
physical examination and additional research is needed for diagnosis.
Additional criteria are hereby desirable. Jaw claudication (positive likelihood
ratio 4.2, 95% confidence interval 2.8 to 6.2) and diplopia (positive
likelihood ratio 3.4, 95% confidence interval 1.3 to 8.6), for example, are not
included in these criteria, but are strong positive predictors for the
diagnosis. Already proposed criteria by Alberts et al, in which a risk
assessment is made, seem more applicable in clinical practice. (Appendix 2B)
Meisner et al have proposed an alternative on the same base. Regarding these
criteria, an algorithm for clinical practice can be made. This algorithm
describes what should be done for further diagnosis and / or treatment.

The additional diagnostic tests includes blood tests, in which an elevated ESR
and CRP can be seen. The ESR has a sensitivity of 76-82%, but only a
specificity of 9%. A normal ESR is seen in 0-20% of the patients. 60% of
patients have a ESR> 100 mm / h, 89% ESR> 50mm/ h, 95% a ESR> 40 mm / hour. The
CRP has a sensitivity up to 97.5%. The CRP may be solely increased , without
increasing ESR. Also, anemia, thrombocytopenia and elevation of liver enzymes
are seen (20-30%).
For diagnosis, the biopsy of the superficial temporal artery is still the gold
standard. The biopsy should be at least 2-3 cm in length because of the
probability of false negative results. The probability of a positive biopsy is
dependent on the length of the biopsy, and is 19% with a biopsy specimen of
less than 5 mm, increasing to 89% at a biopsy of more than 20mm. Partly due to
the presence of skip lesions. The sensitivity of the biopsy, 87%, specificity
increases to 100%. A false negative result is seen in 5-44% of patients,
depending on biopsy length. If a biopsy is negative, a contralateral biopsy can
be contributing, the sensitivity of the procedure would thus be increased 12.7%
compared to a unilateral biopsy. The chance that a biopsy on the contralateral
side is contributing is small, 1-3%, and is therefore not recommended. The
complication rate of biopsy is low, 0.5% .

The use of corticosteroids prior to biopsy rarely influences the outcome of the
biopsy. After starting corticosteroids a positive biopsy can be seen at least
14 days after start of therapy. In 40% of patients the biopsy was positive even
after 4 weeks of treatment with corticosteroids. If giant cell arteritis is
suspected therapy with corticosteroids should therefore be initiated
immediately.

Currently, the research criteria of the American College of Reumatology are
still widely used. Given the varying sensitivity, specificity and positive
predictive value of these criteria in the general population, we examine
whether there are other predictive factors for the diagnosis of giant cell
arteritis. We will focus here on data from the clinical history and
abnormalities on physical examination.

There is increasing awareness about the additional value of ultrasonography of
the temporal arteries in the diagnosis of giant cell arteritis. By
ultrasonography halo's can be seen, increased flow and stenosis and occlusions.
Halos have a sensitivity ranging from 40-86% and a specificity ranging from
78-100%. There is a positive predictive value of 75-92% and a negative
predictive value of 64-96%. It is important that the presence of halos increase
the chance of diagnosing giant cell arteritis strongly, with an odds ratio of>
65. The specificity increases to 100%. The thickness of this halo is also
important, a halo of> 1mm enhances the chance of diagnosing temporal arteritis.
Ultrasonography may be false positive by infection, malignancy, vasculitis, and
absent flow in the adjacent vein. Ultrasonography can be false-negative in the
early (if no vascular edema) and in the late phase of inflammation (decreased
vascular edema).

There are studies that suggest that the use of ultrasonography instead of a
doing a biopsy does not lead to an increase in morbidity. In a patient with a
high suspicion of giant cell arteritis and halos at ultrasonography
examination, a biopsy should no longer be indicated. This is particularly true
for the presence of bilateral halos. If ultrasonography has indeed the same (or
better) diagnostic value as a biopsy, this would be a great advantage for
clinical practice because no more invasive surgery has to happen, or only when
in doubt about the diagnosis. Ultrasonography is not stressful for the patient,
quicker accessible and cost effective. Hellman and Hunder already proposed an
algorithm for the diagnostic approach of giant cell arteritis in 2005 (Annex
3A). Recent meta-analyzes have shown that ultrasonography is not yet superior
to the biopsy.
It is also known that ultrasonography of the larger vessels, including carotid
artery, subclavian artery and femoral artery show abnormalities, particularly
vascular edema / halos, indicating vasculitis. This is described in 29% of the
patients with a giant cell arteritis. If these vessels are also examined in
suspected giant cell arteritis, the sensitivity of the examination can be
increased.

We would like to investigate what the additional value of ultrasonography, of
the temporal arteries, is compared with the biopsy. If ultrasonography shows
abnormalities appropriate to diagnose giant cell arteritis, a biopsy must be
taken, guided by ultrasonographic abnormalities. This should increase the
sensitivity of the biopsy. Research has already shown that the probability of a
positive biopsy is increased with an ultrasound guided biopsy. With a strong
suspicion of giant cell arteritis and a negative ultrasound examination, a
biopsy will still be performed, given the chance of false-negative ultrasound
examination. In 1 in 10, biopsy proven giant cell arteritis, ultrasonography
can be negative. This means that all patients will undergo a biopsy and
ultrasonography of the temporal artery.
The literature has shown that ultrasonographic abnormalities are seen for 16
(7-56) days after starting treatment with corticosteroids. We would like to
investigate how long abnormalities can be seen after starting treatment in our
population. We would also like to investigate how the ultrasonographic
abnormalities change during treatment or by recurrent symptoms increase in
inflammatory parameters .

By using new criteria and other complementary imaging techniques a risk
assessment can be made. By implementing this risk assessment in the already
existing algorithm of Hellmann and Hunder it is possible to improve clinical
practice with a new algorithm for clinical practice. Meisner et al in 2011
have already made a proposal for practice (Annex 3B).

This study aims to investigate the sensitivity and specificity of the history
and physical examination for the diagnosis of giant cell arteritis, compared
with the superficial temporal artery biopsy, and get more insight into the
positive and negative predictors of the disease. The additional value of
ultrasonography of the temporal arteries in suspected giant cell arteritis will
also be examined. We will examine the sensitivity and specificity of
ultrasonography compared with the superficial temporal artery biopsy. We will
also examine the duration that the ultrasonographic abnormalities persist after
starting therapy. The behaviour of ultrasonographic abnormalities under therapy
(decrease and disappearence) and recurrent symptoms or increase in blood
disorders (increased inflammatory parameters) will also be followed.

In this prospective, cohort study, we will examine all patients aged over 50
years with a newly developed unilateral headache (temporal or occipital) and /
or visual complaints (ie loss of vision, amaurosis fugax, diplopia, or
blindness), where there is suspicion of giant cell arteritis. Patients seen in
the St. Elisabeth Hospital Tilburg and TweeSteden hospital by a neurologist,
internist, rheumatologist or ophthalmologist, patients seen in the Atrium
Medical Center Heerlen by a neurologist and patients seen in the Gelderse
Vallei Hospital Ede by a neurologist, are included. Initially, 125 patients
with suspected giant cell arteritis are prospectively studied.

Flowchart:
1. All patients aged over 50 years with a newly developed unilateral headache
and / or visual difficulties which the suspicion of giant cell arteritis seen
by the neurologist, internist, rheumatologist or ophthalmologist.
*
2. ultrasonography of the temporal artery on both sides (every patient)
*
3. After ultrasonography, the patient is referred to the neurology clinic
where he urgently (same or next business day) will be seen by the 'principal
investigator' (Drs TWH Alleman, St. Elisabeth Hospital, Tilburg, drs. JMP
Rovers, Gelderse Vallei Hospital, Ede or drs. P. Janssen, Atrium Medical
Center, Heerlen). The patient is informed about the study.
*
4. An additional history and physical examination is performed.
*
5. Blood tests according to protocol
*
6. Temporal artery biopsy on the symptomatic side, if possible ultrasound
guided (signed), within 5 working days. All biopsies will be assessed by the
pathologist, blinded to any clinical information and duplex results.
*
7. When giant cell arteritis is diagnosed, than further treatment according
to the local guidelines
*
8. If ultrasonography of the temporal arteries shows presence of halos, a
control ultrasonography is performed after 4 weeks and 8 weeks after initiation
of therapy with corticosteroids. This will be repeated every 8 weeks until the
halos are disappeared. If there is an exacerbation of giant cell arteritis
(increased symptoms or increased inflammatory parameters) a control
ultrasonography will be performed (focussed on the place where in previous
research the halos were seen).
*
9. Outpatient control after the last ultrasonography, according to the local
guidelines
*
10. Analysis

This is a prospective multicenter cohort study, where patients with suspected
giant cell arteritis are examined. Given the regular therapy occurs (blood
examination, biopsy and ultrasonography), the patient is not exposed to
additional risks.
However, all patients undergo additional ultrasonography investigation after
initiation of therapy, by 4 weeks and 8 weeks. And then every 8 weeks until the
halos have disappeared. This is not an invasive procedure, but can be perceived
as a burden. Further follow up will be according to the standard guidelines.