The prevalence and clinical characteristics of nonradiographic axial spondyloarthritis among patients with inflammatory back pain in rheumatology practices: a multinational, multicenter study

Background Patients with ankylosing spondylitis (AS), who by definition have radiographic sacroiliitis, typically experience symptoms for a decade or more before being diagnosed. Yet, even patients without radiographic sacroiliitis (i.e., nonradiographic axial spondyloarthritis [nr-axSpA]) report a significant disease burden. The primary objective of this study was to estimate the prevalence and clinical characteristics of nr-axSpA among patients with inflammatory back pain (IBP) in rheumatology clinics in a number of countries across the world. A secondary objective was to estimate the prevalence of IBP among patients with chronic low back pain (CLBP). Methods Data were collected from 51 rheumatology outpatient clinics in 19 countries in Latin America, Africa, Europe, and Asia. As consecutive patients with CLBP (N = 2517) were seen by physicians at the sites, their clinical histories were evaluated to determine whether they met the new Assessment of SpondyloArthritis international Society criteria for IBP. For those who did, their available clinical history (e.g., family history, C-reactive protein [CRP] levels) was documented in a case report form to establish whether they met criteria for nr-axSpA, AS, or other IBP. Patients diagnosed with nr-axSpA or AS completed patient-reported outcome measures to assess disease activity and functional limitations. Results A total of 2517 patients with CLBP were identified across all sites. Of these, 974 (38.70 %) fulfilled the criteria for IBP. Among IBP patients, 29.10 % met criteria for nr-axSpA, and 53.72 % met criteria for AS. The prevalence of nr-axSpA varied significantly by region (p < 0.05), with the highest prevalence reported in Asia (36.46 %) and the lowest reported in Africa (16.02 %). Patients with nr-axSpA reported mean ± SD Ankylosing Spondylitis Disease Activity Scores based on erythrocyte sedimentation rate and CRP of 2.62 ± 1.17 and 2.52 ± 1.21, respectively, indicating high levels of disease activity (patients with AS reported corresponding scores of 2.97 ± 1.13 and 2.93 ± 1.18). Similarly, the overall Bath Ankylosing Spondylitis Disease Activity Index score of 4.03 ± 2.23 for patients with nr-axSpA (4.56 ± 2.17 for patients with AS) suggested suboptimal disease control. Conclusions These results suggest that, in the centers that participated in the study, 29 % of patients with IBP met the criteria for nr-axSpA and 39 % of patients with CLBP had IBP. The disease burden in nr-axSpA is substantial and similar to that of AS, with both groups of patients experiencing inadequate disease control. These findings suggest the need for early detection of nr-axSpA and initiation of available treatment options to slow disease progression and improve patient well-being.

proportion of axial SpA patients with nr-axSpA is unknown, as estimates vary considerably, ranging from 23 % to 80 % in a recent review of patients with axial SpA [4]. In part, this is due to different methods of assessment and the fact that these studies were not specifically designed to follow patients with undifferentiated SpA and nr-axSpA [4].
The lack of data on nr-axSpA and AS is even more pronounced in emerging countries in Latin America, Europe, Africa, and Asia. The primary objective of this study was to provide prevalence estimates on the presence of nr-axSpA among patients with inflammatory back pain (IBP) in rheumatology clinics across a number of emerging countries. In this study, we also sought to describe the clinical characteristics associated with both nr-axSpA and AS. A secondary objective was to estimate the prevalence of IBP among patients with chronic low back pain (CLBP), given that IBP represents an increasingly important part of identifying patients with axial SpA [2,3].

Study design
A noninterventional, cross-sectional study was conducted to estimate the prevalence of nr-axSpA among patients with IBP in 51 rheumatology outpatient clinics from 19 countries in Latin America (Colombia, Costa Rica, Mexico, and Peru), Europe inclusive of western Asia (Hungary, Israel, Kazakhstan, Poland, Romania, Russia, and Turkey), Africa (Algeria, Morocco, and South Africa), and Asia focused on southern and eastern Asia (Bangladesh, China, India, Malaysia, and Taiwan). Patient recruitment took place from January through December 2014. The protocol and study materials received institutional review board approval at each participating site; the specific sites are listed in the Acknowledgements section.
Consecutive patients with CLBP were seen by participating physicians at the study sites and were evaluated clinically to determine whether they met the criteria for IBP. Patients who met 2009 ASAS criteria for IBP had their clinical histories further evaluated to determine their eligibility for the medical record abstraction portion of this study. The inclusion criteria were age ≥18 years, CLBP ≥3 months, and four of five of the following parameters: age of onset <40 years, insidious onset, improvement with exercise, no improvement with rest, and pain at night. The exclusion criteria were noninflammatory back pain, a condition that could mimic IBP (e.g., fibromyalgia), the presence of a neuropathic component, unexplained weight loss of >10 kg within the past 6 months, persistent fever, urinary incontinence or retention, saddle anesthesia, decreased anal sphincter tone or fecal incontinence, bilateral lower extremity weakness or numbness, or progressive neurologic deficit.
For patients who met the appropriate inclusion criteria and provided written informed consent, a medical record abstraction using a case report form (CRF) was performed to determine whether the patients met the criteria for AS, nr-axSpA, or other forms of IBP ("other IBP" hereafter) based on ASAS criteria for axial SpA and the Modified New York (Modified NY) criteria for AS ( Fig. 1). More specifically, patients who did not meet ASAS criteria were classified as having other IBP. Patients who met ASAS criteria for axial SpA but did not meet Modified NY criteria were classified as having nr-axSpA. Patients who met both ASAS criteria for axial SpA and Modified NY criteria were classified as having AS.
Patients who were diagnosed with AS or nr-axSpA were provided with a brief survey to evaluate their patient-reported outcomes (PROs). It is important to note that diagnosed refers to the physician's classification of that patient as indicated in the medical record. It is not known what information was used to make this assessment. The classification of patients based on ASAS criteria for axial SpA and Modified NY criteria for AS, as described above, was an analytical exercise performed by the authors after data collection; it was not conducted in real time as patients were enrolled into the study. As a result, the study depended upon the physician's classification (referred hereafter as diagnosis to distinguish the methods) in order to identify patients eligible to receive the survey, even if this differed from our methods of classification using the ASAS criteria for axial SpA and the Modified NY criteria for AS. This survey was completed entirely by the patient in the waiting room to avoid any influence from the investigator. At the end of subject recruitment, all completed materials were  Fig. 1 Summary of inflammatory back pain (IBP) group classifications. AS ankylosing spondylitis, ASAS Assessment of SpondyloArthritis international Society, axSpA axial spondyloarthritis, CRP C-reactive protein, HLA-B27 human leukocyte antigen B27, Modified NY criteria Modified New York criteria for ankylosing spondylitis, nr-axSpA nonradiographic axial spondyloarthritis, NSAID nonsteroidal anti-inflammatory drug, SpA spondyloarthritis collected on-site and checked for completion, with the exception of the patient questionnaire, which remained confidential.

Measures
The CRF assessed information on each patient's demographics (

Statistical analysis
The analysis of the primary objectives (prevalence and clinical characteristics of nr-axSpA) was focused on patients with IBP who met the inclusion or exclusion criteria described above. The analysis of the secondary objective (prevalence of IBP among patients with CLBP) was focused on all patients with CLBP. Frequencies, percentages, and 95 % CIs were reported for binary and/or categorical variables. Counts, means, and SDs were reported for continuous variables. Statistical differences across geographical regions were analyzed using chisquare tests and one-way analysis of variance for categorical and continuous variables, respectively.

Results
Prevalence of nr-axSpA overall, by region, and by sex A total of 2517 patients with CLBP were identified across all sites (Fig. 2). Of these, 974 (38.70 %) fulfilled the criteria for IBP and were advanced to the CRF portion of the study for assessment of IBP group status (i.e., nr-axSpA, AS, or other IBP). Overall, 29.10 % (95 % CI 26.15-32.05 %) of patients with IBP met the criteria for nr-axSpA (

Demographic characteristics of patients with nr-axSpA
Patients with nr-axSpA had a mean age of 34.75 years (SD 10.03); 36.47 % were female, and 55.64 % were white ( Table 2). These figures contrasted with patients with AS, who had a mean age of 39.03 years (SD 11.38); 28.72 % were female, and 68.23 % were white. Minimal regional differences were observed with respect to the demographic characteristics of patients with nr-axSpA (Table 3). Patients with nr-axSpA were oldest in Africa (36.67 years) and youngest in Asia (33.24 years) (p < 0.05); however, no differences in sex, BMI, years since CLBP presentation, or age of IBP onset were observed across regions.

Clinical characteristics of patients with nr-axSpA
Patients with nr-axSpA experienced CLBP for 6.32 ± 7.61 years (compared with 10.91 years for patients with  Several clinical characteristics varied across regions among those with nr-axSpA (Table 5). Patients in Europe were the most likely to have a positive HLA-B27 test result (84.85 %), and patients in Asia were the least likely (62.24 %) (p < 0.05). Patients in Asia were the most likely to have an elevated ESR value (48.89 %), and patients in Europe were the least likely (20.48 %) (p < 0.05). No differences in family history or NSAID response were observed.

Delay from IBP and CLBP to nr-axSpA or AS diagnosis
For patients who received a diagnosis of nr-axSpA, there was a mean delay of 5.21 ± 7.69 years between the presentation of IBP and diagnosis (Table 4). There was a mean delay of 6.48 ± 8.53 years between the presentation of IBP and diagnosis for patients with AS.

Patient-reported outcomes in nr-axSpA and AS
The mean disease activity levels for patients with nr-axSpA were 2.62 ± 1.17 and 2.52 ± 1.21 for the ESR and CRP versions of the ASDAS, respectively, suggesting a high level of disease activity (i.e., ≥2.1) (

Discussion
In this study, 39 % of patients referred to rheumatology clinics with CLBP met the ASAS criteria for IBP. Further, 29 % of patients with IBP met the criteria for nr-axSpA. The proportion of axial SpA patients with nr-axSpA was within the range (23-80 %) reported in the literature [4,8], though on the lower end of prior estimates.
Our data suggest a higher percentage of males among patients with nr-axSpA (64 %) relative to the percentages in other published noninterventional studies (34-50 %) [9,10] and most clinical trials (48-64 %) [11][12][13][14][15][16]. There were a number of methodological differences across these studies (e.g., inclusion and exclusion criteria, country), but it is unclear which of these factors would help to explain the differences in results. Further research is necessary.
We found that patients with nr-axSpA were the youngest, and they experienced CLBP for the shortest duration at slightly over 6 years compared with nearly 11 years for patients with AS. Although no age differences were found in past literature reviews between patients with nr-axSpA and those with AS [4,8], several prior studies have found a longer symptom duration for patients with AS [4-8, 10, 17, 18]. This is to be  The p values represent the omnibus statistical comparison of percentages (or means) across IBP groups based on the chi-square (or F-test) values "Other IBP" refers to patients who did not meet the Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis expected, given that nr-axSpA and AS likely represent a progression in the spectrum of the same disease. Among patients who had been diagnosed with nr-axSpA, there was a delay of approximately 5 years between presentation of IBP and diagnosis. This finding was consistent with the study by Poddubnyy and colleagues, who which also found a gap of slightly more than 5 years between back pain and the assessment of nr-axSpA [17]. However, it should be noted that many patients who met the criteria for nr-axSpA were not diagnosed even after this period of several years, reinforcing the importance of increasing the awareness of, and adherence to, ASAS classification criteria for timely diagnosis and initiation of treatment.
Among those with HLA-B27 test results, 83 % of the patients with nr-axSpA had positive results. This finding is consistent with what was summarized by Boonen et al. [4] and reported directly in two recent noninterventional studies (73 % in a study by Poddubnyy and colleagues [11] and 86 % in a study by Kiltz and colleagues [18]. Over two-thirds (68 %) of patients with nr-axSpA had elevated CRP values (≥3.5 mg/L), which was nearly as high as the percentage of patients with AS who had elevated CRP values (77 %).
Data derived from the patient surveys demonstrated a high level of disease activity and a suboptimal level of disease control, as assessed using the ASDAS and BAS-DAI, respectively, for patients with nr-axSpA. It is important to note that even patients with nr-axSpA [19], who are in an earlier phase in the course of axial SpA, exhibited a significant burden that was comparable to that of patients with AS. BASDAI scores in our study were lower than those reported in a review of clinical trial results of biologic treatments [8], suggesting a less severe patient-reported burden in this real-world patient population. The levels of functional impairment (BASFI) and limitations (BASMI) were comparable to those reported for clinical trial populations [4]. Given the early age of onset for nr-axSpA, these impairment data suggest that patients can experience a substantial level of burden for many years. This further illustrates the importance of identifying patients early in order to slow disease progression.

Limitations
Limitations of this epidemiological study include the use of a single assessment with a questionnaire and CRF that may not adequately capture a comprehensive medical history for a particular patient. It is also important to mention that this was an observational study, so not all patients had complete information available. This could have affected the classification of patients and therefore Abbreviations: AS ankylosing spondylitis, CRP C-reactive protein, ESR erythrocyte sedimentation rate, HLA-B27 human leukocyte antigen B27, IBP inflammatory back pain, nr-axSpA nonradiographic axial spondyloarthritis, NSAID nonsteroidal anti-inflammatory drug, SpA spondyloarthritis The p values represent the omnibus statistical comparison of percentages (or means) across IBP groups based on the chi-square values the prevalence estimates. For example, because a positive HLA-B27 test is one way to classify a patient as having nr-axSpA instead of other IBP, missing HLA-B27 data would underestimate the number of nr-axSpA patients relative to other IBP patients. Another limitation was the lack of available information on the other IBP group. Although this group also had poor outcomes based on the PRO data, the explanation for this finding is unclear without knowing more about the composition of the other IBP group. Patient surveys were administered only to patients who were diagnosed with AS and nr-axSpA, so patients who were not diagnosed with either condition, even if they met the appropriate classification criteria, did not provide PRO data. The external validity of the study is dependent on the extent to which patients at the selected rheumatology practices are representative of all IBP patients in these countries. Because these sites were selected for being major centers for the treatment of SpA, it is possible the patients who are managed by these sites are fundamentally different (e.g., more severe disease).