Clinical Activity Score for Thyroid Eye Disease Explained

The Clinical Activity Score, or CAS, is a bedside scoring tool that tells doctors whether thyroid eye disease is in an active inflammatory phase or has burned out into a stable one. Developed in the late 1990s, it assigns one point for each sign of orbital inflammation a clinician observes, producing a score from 0 to 7 (or 0 to 10 in its modified form). That single number drives almost every major treatment decision: whether to start immunosuppressive therapy, whether to wait, and when surgery becomes appropriate. The score sounds straightforward, but how it is used, where it falls short, and what may eventually replace or supplement it are all worth understanding if you or someone you know is navigating this disease.

What the Score Actually Measures

The CAS was designed around a simple idea: active inflammation looks different from old scarring, and you can tell the two apart by checking for the classical signs of inflammation at the bedside. The original version tested four of those signs (pain, redness, swelling, and impaired function) and was validated in a double-blind study to see whether it could predict who would benefit from immunosuppressive treatment.¹ Over time, a modified 10-item version became the standard in clinical practice and research. Its items cover two categories of findings: inflammatory signs you can see right now, and measurements suggesting the disease is getting worse.

The 10 items on the modified CAS are:

Spontaneous orbital pain: an aching or pressure sensation behind the eye without any provocation.
Pain on eye movement: discomfort when looking up, down, or to the sides.
Eyelid redness: visible reddening of the skin of the lids.
Conjunctival redness: injection (redness) of the white of the eye.
Caruncle swelling: puffiness of the small pink tissue in the inner corner of the eye.
Eyelid swelling: visible edema of the upper or lower lid.
Chemosis: swelling of the conjunctiva, the clear membrane over the white of the eye.
Increasing proptosis: the eye bulging further forward compared with a prior measurement.
Decreasing eye movement: worsening restriction when trying to move the eye.
Decreasing visual acuity: measurable loss of vision.

Each item scores either 0 (absent) or 1 (present), so the total ranges from 0 to 10.² In most clinical guidelines, a CAS of 3 or higher out of 7 (using only the first seven items at a single visit, since the last three require comparison over time) is the threshold that labels the disease “active.” That label is what opens or closes the door to anti-inflammatory treatments.

How the Score Guides Treatment

The European Group on Graves’ Orbitopathy (EUGOGO), the most influential international body on this disease, explicitly bases its treatment recommendations on the combination of clinical activity and severity.³ In practice, that means the CAS acts as a gatekeeper. If your disease is mild but active, you might get watchful waiting with supportive measures. If it is moderate-to-severe and active (CAS ≥ 3), the standard first-line treatment is intravenous glucocorticoids, typically given as a pulsed course over several months. If the disease is severe but inactive, meaning the inflammation has burned out but left behind bulging eyes, strabismus, or lid retraction, the path is surgical rehabilitation rather than drugs aimed at suppressing inflammation.

Research into predicting who will respond to intravenous glucocorticoids has found that CAS is one of the strongest individual predictors. A machine-learning study that tested various clinical and imaging features found that CAS and lid aperture were the two variables most consistently selected as key predictors of steroid response.⁴ That is intuitive: if the score reflects how much active inflammation is present, people with more inflammation have more inflammation to suppress.

When the Score Misses Active Disease

One of the most clinically important problems with the CAS is that it can read as low even when significant inflammation is happening behind the scenes. Some patients score below 3 on the clinical exam, which would normally classify them as inactive, yet MRI scans reveal swollen, inflamed orbital tissues. A study of 91 such patients, all with a CAS below 3 but MRI-confirmed intraorbital inflammation, found that a course of intravenous glucocorticoids still produced meaningful improvement in many of them.⁵ The researchers were able to build a prediction model combining clinical and radiological features that performed reasonably well in identifying which low-CAS patients would benefit from treatment.

This gap matters because a patient whose CAS is 2 might be told their disease is inactive and that steroids are not indicated, when in reality the inflammation is simply happening deeper in the orbit where a bedside exam cannot see it. The growing recognition of this mismatch is one reason many specialists now consider MRI or other imaging an important complement to the CAS rather than an optional add-on. Quantitative MRI techniques can measure things like water content and blood flow within the extraocular muscles, providing a more granular picture of what is happening inside the orbit.⁶

Reliability Problems Between Examiners

Even setting aside the question of hidden inflammation, the CAS has a reproducibility problem. Because several of its items rely on subjective visual judgment, different doctors looking at the same patient can arrive at different scores. A study that measured inter-observer agreement across the individual CAS components found that overall agreement was only moderate, with the total score earning a reliability statistic of about 0.53 on a scale where 1.0 would be perfect agreement.⁷ Some items performed much worse than others: lid redness had the lowest agreement, while spontaneous pain, which is based on what the patient reports rather than what the examiner sees, had the highest.

An international reliability study of specific soft tissue features found a similar pattern. Eyelid and conjunctival edema (swelling) could be reliably scored by different raters using a simple present-or-absent scale. But eyelid redness and conjunctival redness were reliable only on a binary scale, not on a graded one, and some features like superior bulbar redness and caruncle swelling were not reliably measured by any scale tested.⁸ The implication is that your CAS could shift by a point or two depending on which doctor examines you, and that shift can be enough to cross the threshold between “active” and “inactive,” potentially changing whether treatment is offered.

What Is Happening Biologically When the Score Is High

A high CAS is not just a clinical label. It reflects a real biological process: inflammatory immune cells infiltrating the tissues around the eye, releasing signaling molecules that cause swelling, redness, and tissue remodeling. Researchers studying orbital tissue samples have found that several specific inflammatory markers track closely with the CAS. A study correlating tissue-level cytokines with clinical scores found that four molecules in particular, along with the density of certain immune cells and fat cells in the orbital tissue, were strongly linked to a higher CAS.⁹ Separately, another study showed that a pro-inflammatory molecule called IL-36α is elevated in the blood and orbital tissues of people with thyroid eye disease, and its levels correlate with the CAS. When the signaling pathway driving IL-36α production was blocked in laboratory experiments, the downstream inflammatory response was effectively shut down.¹⁰

This kind of research matters beyond academic curiosity. If specific inflammatory pathways correlate with the CAS, they become potential drug targets. And understanding that the CAS reflects a measurable biological state, not just a checklist of symptoms, helps explain why it works as a predictor of treatment response and why it sometimes fails when the inflammation is present but not yet visible at the surface.

Biologic Therapies and the Question of Activity

The arrival of teprotumumab, the first FDA-approved drug specifically for thyroid eye disease, has complicated the traditional CAS-based treatment framework. Teprotumumab works by blocking a receptor involved in the signaling pathway that drives orbital tissue expansion. In its pivotal trials, it was tested primarily in patients with active disease. But a post-hoc analysis specifically looked at patients with longstanding disease and low inflammatory activity, the kind of patients who would traditionally be considered past the window for anti-inflammatory treatment and headed toward surgery. Teprotumumab still produced a meaningful reduction in eye bulging in this group: about 2.4 mm of proptosis reduction compared with roughly 0.9 mm in the placebo group at 24 weeks, with about 62% of treated patients achieving a clinically significant proptosis response versus 25% on placebo.¹¹

This finding challenges the binary active-versus-inactive framework that the CAS enforces. If a drug can improve proptosis in patients whose CAS says they are inactive, then either the CAS is missing ongoing low-grade disease activity, or the drug works through mechanisms that are relevant even after inflammation has resolved. Either way, the rigid CAS threshold of 3 does not capture the full picture of who might benefit from treatment.

The CAS and Surgical Timing

For patients who need surgical rehabilitation, typically orbital decompression to reduce eye bulging, strabismus surgery to correct double vision, or eyelid surgery, the CAS plays a role in deciding when to operate. The general principle is that surgery should wait until the disease has been stable and inactive for at least six months, because operating on an actively inflamed orbit risks a worse outcome and the need for repeat procedures. In practice, “inactive” is defined as a CAS below 3 and stable measurements over time.¹² Surgeons reviewing charts often have to reconstruct whether the disease was truly inactive at the time of the decision, relying on documented exam findings and proptosis measurements when the CAS was not explicitly recorded.

The exception is sight-threatening disease. If the optic nerve is being compressed (dysthyroid optic neuropathy), urgent decompression surgery may be necessary regardless of the activity score, because waiting for the inflammation to settle on its own risks permanent vision loss. In those cases the CAS becomes secondary to the clinical emergency.

Impact on Quality of Life

Thyroid eye disease affects far more than the eyes themselves. Patients frequently report significant impairment in how they see, how they feel about their appearance, and how they function socially and professionally. Studies using the Graves’ Ophthalmopathy Quality of Life questionnaire (GO-QoL), the standard patient-reported outcome tool for this disease, have consistently found that higher CAS scores correlate with worse quality-of-life scores, particularly in the domain of visual functioning.¹³ Inflammatory activity was also associated with impairment in the appearance domain, and that connection was especially pronounced in women.¹⁴

This correlation is worth knowing about because the CAS is a clinician-facing tool. It was not designed to capture how a patient feels about their disease. A patient with a CAS of 2 may still be profoundly bothered by their appearance or struggling with double vision, and their quality-of-life scores may reflect that even though their disease is classified as inactive. The GO-QoL fills a gap the CAS was never meant to address, and ideally both tools are used together to get a complete picture of how someone is doing.

Pediatric Considerations

Thyroid eye disease in children is rare, and the CAS was developed and validated in adults. Pediatric cases tend to present differently: a systematic review of over 800 pediatric patients found that the most common feature was exophthalmos, present in nearly all cases, followed by eyelid retraction and dry eye.¹⁵ How well the CAS applies to children is an open question. The inflammatory signs it measures can overlap with other pediatric conditions, and the subjective items (pain, for instance) may be harder to assess reliably in younger children. Most pediatric specialists adapt the same scoring framework but interpret it with caution, relying more heavily on imaging and clinical judgment.

AI-Assisted Scoring and Remote Monitoring

One of the most promising developments around the CAS is the effort to automate it using artificial intelligence applied to facial photographs. Several research groups have built machine-learning systems that analyze photos of patients’ faces and estimate the CAS or detect active disease. One such system, trained on both studio-quality clinical photographs and smartphone selfies, achieved accuracy in detecting active thyroid eye disease that was comparable to oculoplastic specialists and outperformed general ophthalmologists and residents.¹⁶

The practical appeal is obvious. If you could reliably monitor your disease activity at home by snapping a selfie and running it through an app, you could catch a flare early without waiting for a clinic appointment. A validated AI tool embedded in a smartphone app could also help general practitioners and endocrinologists identify patients who need urgent referral to an eye specialist, rather than relying on those non-specialists to perform a CAS assessment they may not be trained in.¹⁷ These tools are still in the validation stage and not yet ready for clinical deployment, but they address two real problems at once: the inter-observer variability of the bedside CAS, and the access barrier for patients who live far from a specialist center.

The same technology has potential for telemedicine. During a video consultation, an AI system running in the background could estimate the patient’s CAS from their video feed, giving the remote physician an objective data point to work with rather than trying to judge redness and swelling through a webcam. Researchers working on these systems have emphasized that the technology needs further confirmatory studies before clinical adoption, but the direction of travel is clear: the subjective bedside assessment is likely to be supplemented, and perhaps partly replaced, by image-based analysis within the next several years.

1
PubMed. Clinical activity score as a guide in the management of patients with Graves’ ophthalmopathy
2
PLoS One. Photographic evaluation of clinical activity score in thyroid eye disease
3
PubMed. The 2021 European Group on Graves’ orbitopathy (EUGOGO) clinical practice guidelines for the medical management of Graves’ orbitopathy
4
PubMed. Integrating ocular and clinical features to enhance intravenous glucocorticoid response prediction in thyroid eye disease: a machine learning approach
5
PubMed. Effectiveness of intravenous glucocorticoids in active thyroid eye disease patients with low clinical activity scores
6
PubMed Central / Wiley Online Library. Application of Quantitative MRI in Thyroid Eye Disease: Imaging Techniques and Clinical Practices
7
PubMed. Inter-observer Variability of Clinical Activity Score: Assessments in Patients With Thyroid Eye Disease
8
Ophthalmic Plastic & Reconstructive Surgery. Soft Tissue Metrics in Thyroid Eye Disease: An International Thyroid Eye Disease Society Reliability Study
9
Investigative Ophthalmology & Visual Science. Orbital Clinicopathological Differences in Thyroid Eye Disease: An Analysis of Cytokines With Histopathological and Clinical Correlation
10
PubMed. IGF-1 upregulates IL-36α expression via NF-κB and p38 MAPK pathways to promote inflammation in thyroid eye disease
11
The Journal of Clinical Endocrinology & Metabolism. Efficacy and Safety of Teprotumumab in Patients With Thyroid Eye Disease of Long Duration and Low Disease Activity
12
PubMed Central. Thyroid-Related Orbital Decompression Surgery: A Multivariate Analysis of Risk Factors and Outcomes
13
PubMed Central. Evaluation of the Graves’ Orbitopathy-Specific Quality of Life Questionnaire in the Mainland Chinese Population
14
Endocrine Abstracts. Thyroid-associated orbitopathy and quality of life: correlation of GO-QoL with clinical activity score and severity in 101 patients
15
PubMed Central. Management and presentation of pediatric thyroid eye disease: A systematic review and meta-analysis
16
PubMed. A Preliminary Evaluation of the Diagnostic Performance of a Smartphone-Based Machine Learning-Assisted System for Evaluation of Clinical Activity Score in Digital Images of Thyroid-Associated Orbitopathy
17
PubMed Central. AI-based assessment of Clinical Activity Score and detection of active thyroid eye disease using facial images: validation of Glandy CAS