Early Diagnosis

Introduction

Diagnosis of ankylosing spondylitis/axial spondyloarthritis (AS/axSpA) at an early stage is not a simple task and is often missed or markedly delayed, especially in a primary care setting, unless the physician has ample clinical experience and a high index of suspicion. One of the difficulties is the nonspecific nature of some of the early disease symptoms. The presenting symptom is most often chronic back pain and stiffness; however, some patients may underrate these symptoms and may simply adjust to living with them instead of seeking treatment. Also, there are no validated or established criteria for early diagnosis. Radiologic changes, the hallmark of AS, are frequently not evident until later in the course of the disease. Thus the finding of normal sacroiliac joints (SIJs) on plain pelvis films should not mislead the physician to prematurely rule out AS if there is sufficient clinical suspicion.

A physician who is not fully aware of the varied clinical presentation of AS might overlook this diagnosis when a teenager or a young adult presents with chronic back pain and stiffness, even though this is a typical presentation. One reason is that back pain is prevalent in the general population, and AS is not the most common cause of back pain. It may take years from onset of inflammatory back pain (IBP) to development of radiographic sacroiliitis to diagnose AS, and so the absence of this evidence during the early years of disease should certainly not be used to rule out the diagnosis.

For example, in a study of patients with symptoms and signs consistent with early AS/axSpA, but radiographically normal SIJs, only 36% developed radiographic evidence of sacroiliitis in 5 years and 59% did so in 10 years. This was also shown in a study of HLA-B27–positive relatives of patients with AS, where radiographic sacroiliitis was found in 16% of patients younger than 45 years and in 38% of patients older than 45 years. In another family study, radiographic evidence of sacroiliitis was found in 40% of patients with spondyloarthritis (SpA) with duration of symptoms of 10 years, in 70% with symptoms for 10 to 19 years and in 86% with symptoms for 20 years or more. Recent use of magnetic resonance imaging (MRI) has confirmed that active inflammation of the SIJs and/or the spine is present long before the appearance of unequivocal sacroiliitis on plain radiography.

The average delay in diagnosis of AS can vary from 3 to 11 years from onset of symptoms, depending on the type of symptoms and the clinical training of the healthcare provider seeing the patient. A nationwide German study using health insurance data calculated a mean diagnostic delay of 5.7 years (median 2.3) in patients with axSpA. Despite all recent developments, diagnostic delay has not changed between the periods of 1996-2005 (mean 6.3 years; median 2.6 years) and 2006-2015 (mean 7.4 years; median 2.7 years). In UK, the mean diagnostic delay prior to and after 2009 ranged between 8 to 9 years (median 5 years). Data from 288 participating US patients in the International Map of Axial Spondyloarthritis (IMAS) survey revealed a diagnostic delay of 5.2 years for men and 11.2 years for women. The delay in diagnosis is in general significantly longer in women than in men (Figure 11-1), in children and adolescents when compared with older patients, and in HLA-B27–negative patients than in HLA-B27–positive patients. Presence of psoriasis is an additional predictor of diagnostic delay. A diagnostic delay may occur because of misdiagnosis, i.e. when axSpA is confused with another condition (for example, diffuse idiopathic skeletal hyperostosis, osteitis condensans ilii, or fibromyalgia) and because of underdiagnosis, i.e. when axSpA is simply not recognized. Additional factors that contribute to diagnostic delay include a lack of awareness among non-rheumatologists, absence of validated diagnostic criteria or referral recommendations and a comparative lack of diagnostic biomarkers and clear radiologic definitions. The longer the diagnosis is delayed, the worse the functional outcome may be, especially with juvenile-onset AS. Although much rarer, it should be noted that overreliance on inflammatory back pain, a common symptom of axSpA that has low specificity (25-43%), may lead to overdiagnosis.

In the typical chronic low back pain patient population presenting to a primary care physician, back pain that meets the criteria for IBP increases the probability of AS or axial SpA by a factor of 3, rising from <5% at baseline to only >14%. So a single clinical feature is not sufficient to make the diagnosis: the more features present that are typical of the disease, the higher the diagnostic probability. In a nested case-control study using anonymized primary care electronic records from Scotland, distinct back pain episodes with intervals >6 months has emerged as the most predictive variable for reaching an early diagnosis of AS. If the patient has three or more additional clinical features (“red flags”) listed in Table 11-1, the probability of presence of AS/axSpA can reach or exceed a comfort zone of diagnosis (>90% probability). An analysis of probands with axSpA and their first-degree relatives (FDR) reported that acute anterior uveitis (AAU) occurred more commonly in FDR with axSpA (38.2%) than those without axSpA (11.6%), suggesting that the presence of AAU may be a useful indicator to screen for axSpA. Another analysis of data from this cohort found that an index combining chronic inflammatory back pain, thoracic spine pain or discomfort and anterior chest wall pain or discomfort has good sensitivity (83.1%) and specificity (87.2%) for axSpA and may be a useful diagnostic tool pending validation.

Thus, there are clinical situations where one can strongly suspect spondylitic disease without radiographic evidence of sacroiliitis that requires a careful and thorough clinical history (including a family history and inquiry about smoking) and physical examination that support the diagnosis, determine disease severity and exclude “look-alikes.” The physical examination should include looking for tenderness over the SIJs or eliciting pain by maneuvers to stress these joints, as discussed in the section on physical examination, in addition to checking for impaired spinal mobility and chest expansion, and signs of enthesitis. Physical findings due to enthesitis that are present in many patients, especially those with juvenile onset of SpA, but are often overlooked, include tenderness over vertebral spinal processes of the spine, iliac crest, anterior chest wall, calcaneus (plantar fasciitis and/or Achilles tendinitis), ischial tuberosities, greater trochanters, tibial tubercle, and sometimes the superior and inferior margins of patella. Women tend to have a milder or slower disease course, and their main initial manifestations, such as neck and thoracic pain and/or tenderness and peripheral joint involvement, may resemble fibromyalgia or early rheumatoid arthritis (RA). This delay in diagnosis of AS in women, as in men, may shorten with increasingly greater awareness and knowledge of this disease.

The final diagnosis incorporates clinical, laboratory, and imaging features that have been discussed in the preceding chapters. Magnetic resonance imaging (MRI) can detect active inflammation of the SIJs and/or the spine long before the appearance of unequivocal sacroiliitis on plain radiography. Interpretation of imaging findings has become even more important as the new Assessment of Spondyloarthritis International Society (ASAS) classification criteria include both active sacroiliitis by MRI and radiographic sacroiliitis in their definition of axSpA. A consensus-based definition of positive MRI in AS/axSpA has been published, that have helped achieve earlier disease recognition. But it needs to be emphasized that the ASAS classification criteria set is meant to decide which of the patients diagnosed by a rheumatologist as suffering from axSpA will be suitable for inclusion in clinical trials, and it gives a clear yes or no answer. In clinical practice, although the same clinical features are relied on to diagnose axSpA, there is a level of flexibility in diagnostic certainty based, in part, on the number of clinical features (red flags) present. Moreover, the clinician has to take negative findings into account and also exclude other causes (“look-alike”) for the patient’s clinical presentation because the classification criteria do not contain these components (i.e., exclusion criteria). The clinician cannot simply checklist boxes of the components of these classification criteria that are not designed for application to an individual patient in a clinical setting.

In some patients, with an initial normal MRI of their SIJs, the diagnostic uncertainty may remain despite extensive investigations. Such patients can be managed with symptomatic treatment and a wait-and-see attitude requiring regular follow-up. In such patients the MRI findings may become abnormal after a year or so, especially among male HLA-B27–positive patients.

The need for an early diagnosis becomes more and more important because of the availability of effective therapies that are probably even more effective if given early. New strategies are being developed that will assist primary care physicians in screening for these patients so that they can be referred to a rheumatologist when the disease is clinically in its early stages. A recent study has confirmed that among patients with chronic back pain for ≥3 months beginning at ages younger than 45 years, the presence of ≥1 of 3 SpA features (HLA-B27 positivity, current inflammatory back pain and MRI or radiographic evidence of sacroiliitis) is an effective way to identify those with possible axSpA (Figure 11-2). An ASAS-endorsed diagnostic algorithm has been published that helps guide rheumatologists in diagnosing axSpA in a cost-effective manner (Figure 11-3).

Danve and Deodhar pointed out to the disparity in the prevalence and diagnosis rates of AS, and discussed the challenges in early recognition of axSpA in the USA. They pinpointed several issues as the main hurdles to early diagnosis: the commonality of mechanical causes of back pain in the United States, lack of validated diagnostic criteria, limitations of physical examination, lack of reliable biomarkers, patient preferences in seeking medical care for back pain and unfamiliarity of non-rheumatologists in recognizing symptoms of back pain specific to axSpA. Many of these issues were raised also by US primary care physicians involved in a qualitative research study. These problems are not unique to US, as has been nicely reviewed in a recent article from UK (Figure 11-4). A study from UK has demonstrated that the diagnostic delay in axSpA can be successfully reduced to 3 years when an Early IBP Service was embedded within a rheumatology department alongside a local public awareness and educational campaign that directs patients to the “right specialist at the right time.” Multidisciplinary care should include non-physicians, such as physiotherapists, social workers and psychologists, working together to provide a truly patient-centered care.

A multicenter Turkish study reported that only 4% of 393 AS patients sought initial medical care from rheumatologists and they had a significantly shorter diagnostic delay as compared to those consulting other specialties. Approximately one third of the AS patients had received a prior diagnosis of lumbar disc herniation in this study, and this was one of the reasons for their delayed diagnosis. Another study found that among the patients who underwent surgery for lumbar disc herniation, 18% fulfilled the European Spondyloarthropathy Study Group (ESSG) criteria and 9% fulfilled the Amor criteria for SpA. Among the patients who agreed to undergo further evaluation including imaging, 7.3% fulfilled the imaging arm of the ASAS axSpA classification criteria.

A study involving 13 European countries found a diagnostic delay of more than 7 years, on average. Overall, the number of health professionals consulted prior to the accurate diagnosis of AS was most strongly associated with diagnostic delay, suggesting that incorrect diagnoses may delay the time to diagnosis. Thus, it is crucial to increase familiarity of non-rheumatologist physicians for the typical symptoms of axSpA, as well as public awareness for early referral to rheumatologists. With the growing need for care of patients with rheumatic and musculoskeletal problems, models of care are emerging in some countries where some allied health professionals, such as nurse practitioners and physician assistants in the US and extended role practitioners in Canada, are trained as skilled professionals to provide musculoskeletal or rheumatic care and act as the first-contact provider.

A predictive model developed by utilizing machine learning techniques for predicting a diagnosis of AS has performed better than a clinically based model, suggesting feasibility of using such technological tools for early diagnosis of AS. The opportunities and potential benefits of machine learning techniques in early detection of AS/axSpA has been reviewed by Walsh and colleagues.

As discussed in detail in Laboratory Findings and Imaging, there is no specific “diagnostic” or pathognomonic laboratory marker for AS/axSpA, and diagnosis is based on clinical history, physical examination, and appropriate laboratory and imaging findings. Cats and colleagues in the Netherlands, in cooperation with Khan in the US, in 1987 proposed diagnostic criteria for spondylitic disease, including AS (Table 11-2) but these criteria have not been validated as yet.

In conclusion, we need an early diagnosis of axSpA to remove or minimize diagnostic uncertainty, allay patient’s anxiety, prevent wrong treatment and initiate appropriate treatment early with biologics that work better in early disease. It needs to be emphasized that the burden of this disease, that usually starts at the prime of one’s life, is substantial even when diagnosed in early stage.

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Enlarge