Histopathological correlation supports the use of x-rays in the diagnosis of hand osteoarthritis
- Ilse-Gerlinde Sunk1,
- Love Amoyo-Minar1,
- Birgit Niederreiter1,
- Afschin Soleiman2,
- Franz Kainberger3,
- Josef S Smolen1,
- Klaus Bobacz1
- 1Department of Internal Medicine III, Division of Rheumatology, Medical University of Vienna, Vienna, Austria
- 2Department of Pathology, Medical University of Vienna, Vienna, Austria
- 3Department of Radiology and Osteology, Medical University of Vienna, Vienna, Austria
- Correspondence to Klaus Bobacz and Ilse-Gerlinde Sunk, Department of Internal Medicine III, Division of Rheumatology, Medical University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria; ;
Contributors I-GS: study design, data collection, data analysis, data interpretation, writing of the manuscript. LA-M, BN, AS: data collection. FK: data collection and data interpretation. JSS: study design, data analysis, data interpretation and writing of the manuscript. KB: study design, data analysis, data interpretation, figures and writing of the manuscript.
- Received 18 October 2011
- Accepted 14 April 2012
- Published Online First 12 May 2012
Objective To correlate histopathological and radiographic features of distal and proximal interphalangeal (DIP and PIP) joints in order to test whether the use of an x-ray examination would be beneficial to the classification/diagnosis process of hand osteoarthritis (OA).
Methods DIP and PIP joints were obtained from post mortem specimens (n=40). Plain x-rays of the DIP and PIP joints were taken and radiographic OA was determined by the Kellgren and Lawrence classification. Individual radiographic features were scored according to the method described by Altman. Joint samples were prepared for histological analysis; cartilage damage was graded according to the Mankin scoring system. Spearman's correlation was applied to examine the relationship between histological and radiographical changes. Differences between groups (bony swelling vs no bony swelling) were determined by Student t test.
Results A highly significant correlation was found between histological (Mankin score) and radiographic (Kellgren/Lawrence score) changes in the investigated DIP (rs=0.87, p<0.0001) and PIP (rs=0.79, p<0.0001) joints. A subgroup of patients (37.5% for DIP and 18.8% for PIP joints) showed advanced radiographic changes (Kellgren/Lawrence score ≥2) in joints without clinical bony swelling. Histologically, the mean Mankin scores accounted for 11±1.66 for DIP and 9.67±2.4 for PIP joints.
Conclusion On the basis of histopathological changes of DIP and PIP joints, this investigation demonstrates the validity of x-ray examinations and supports the use of plain radiography in the diagnosis of hand OA and in the classification of hand OA in clinical trials.
Osteoarthritis (OA) of the hand is a very common disease and the prevalence of radiographic changes is quite high (29–76% of the population).1 ,2 Symptomatic hand OA, however, is less prevalent, affecting up to 16% of the population,1 ,3 but may lead to significant impairment of physical function and quality of life.4,–,8 While a number of drugs are useful in symptomatic hand OA, there is currently no unequivocal evidence that structural changes can be modified.9 Nevertheless, the correct diagnosis of hand OA is important given the various possible differential diagnoses.10 The prospect that disease pathologies may be unravelled and structure-modifying agents may be developed in the near future likewise requires a reliable diagnosis of hand OA.
Diagnosing hand OA appears to be quite simple. According to the current American College of Rheumatology (ACR) criteria for the classification and reporting of hand OA, a clinical examination is sufficient to identify patients with this disease,11 requiring the presence of hand pain, aching or stiffness and hard tissue enlargement of two or more of 10 selected joints among several items. This tool is claimed as very accurate in identifying patients with hand OA, with a sensitivity of 94% and a specificity of 87%. Nevertheless, several items might confound the use of these criteria in epidemiological studies and clinical practice, such as the need for joint pain/aching/stiffness together with bony swelling and also the omission of x-rays.12 In contrast, the European League Against Rheumatism recommends plain x-rays as the gold standard for morphological assessment of hand OA.10 Given this discrepancy in the approach to hand OA, the question arises as to whether or not plain radiography is useful. This pertains to diagnosis since in clinical practice classification criteria are also frequently employed for diagnostic purposes, and also to classification (which is usually done for trial purposes) since diseases should be classified optimally when using such criteria, especially when therapies are to be tested in future trials.
Employing clinical criteria alone at present would identify only patients with symptoms who also have Heberden and/or Bouchard nodes, while patients with more variable symptoms, with or without nodes, would not be classified. Interestingly, clinically apparent (bony swelling) but intermittently symptomatic/asymptomatic hand OA is more prevalent than frequently symptomatic hand OA.13 ,14 Furthermore, Dahaghin and coworkers, who assessed the determinants for hand pain in a cohort of almost 8000 participants, found hand OA to be a very poor explanation for hand pain.6 Given the fact that pain in hand OA is frequently transient,15 ,16 clinical symptoms at the time of examination may not constitute a major criterion in the definition of hand OA.17 This notion is supported by the observation that only 70% of patients with hand OA according to the ACR criteria fulfilled the same criteria 6 months later.18
Examination of symptomatic finger joints aims to distinguish hand OA from other forms of arthritis and also accurately to exclude the diagnosis of hand OA itself. For this purpose, it appears that radiology may provide a better overall assessment for hand OA and may be more useful than clinical examination of the interphalangeal joints alone.19,–,21
Thus, the statement that diagnosing hand OA appears to be quite simple may have to be rephrased by raising the question: ‘How easy is it to classify/diagnose hand OA?’ and expanding it to ask how valid x-rays are in the diagnosis of hand OA. To address these questions we will not primarily focus on clinical examinations, since their advantages and limitations in comparison with radiography have been widely addressed in the literature. Rather, since no study has so far addressed this issue, we will compare radiographic changes with histopathological changes to evaluate whether plain x-rays of the hands—and particularly of the distal and proximal interphalangeal joints (DIP and PIP)—accurately reflect structural joint alterations and improve the classification/diagnosis process.
Post mortem specimens, clinical and radiographic assessment
Joint specimens were obtained post mortem at the Department of Pathology, Medical University of Vienna and comprised 80 interphalangeal joints (40 DIP and 40 PIP) from 40 consecutive individuals (18 women and 22 men). Patients with a documented history of inflammatory joint disease such as rheumatoid arthritis or psoriatic arthritis were excluded. Joints were obtained within 24 h of the subjects' death and were derived from the second or third finger because of the high prevalence of OA at these locations.14 ,16 ,22 To obtain these specimens the skin and subcutaneous tissues were carefully dissected until the DIP and PIP joints as well as the phalangeal bones became visible. The first and third phalangeal bones were then cut above the DIP joint and below the PIP joint so that the PIP and DIP joints could be obtained in full. Of these 80 joints, 76 (37 DIP and 39 PIP joints) could be processed for histological and radiographic investigations.
Before dissection both hands were clinically examined for Heberden and Bouchard nodes (palpation for firm/hard posterolateral rounded swelling and/or joined dorsal bars). Heberden and Bouchard nodes were classified as present or absent. If bony swelling was present, the finger that clinically displayed the worst changes either on the left or the right hand was selected. If no nodes were present, the right hand was always used in accordance with the higher prevalence of OA on this side,23 and a computer program was used to randomise which finger to dissect. Joint assessment was performed by a physician with long-standing rheumatological experience (KB). This study was approved by the ethics committee of the Medical University of Vienna (No 409/2005).
Plain radiography of the interphalangeal joints (posteroanterior and lateral images) of the deceased individuals was performed using a Philips Optimus 80 x-ray generator. A blind assessment was carried out by a musculoskeletal radiologist (FK). To determine radiographic OA, each joint was graded according to the Kellgren and Lawrence scoring method on a scale of 0–4 scale where grades ≥2 indicate the presence of definite OA.24 Additionally, radiographic changes (osteophytes, joint space narrowing (JSN), erosions, subchondral sclerosis and subchondral cysts) were graded according to the scoring system and radiographic atlas of Altman.25 Scoring was performed by displaying the images in DICOM format on a workstation equipped with Agfa Impax software suitable for diagnostic use.
Finger joints were fixed in 4.0% paraformaldehyde for 5 days and then bisected by cutting from the radial to the ulnar side through the central area. Sectioning was performed with caution in order to preserve the intact articulation of the joint head and socket. The samples were then decalcified in 14% EDTA (Sigma-Aldrich; St Louis, Missouri, USA) at 4°C (pH adjusted to 7.2 by addition of ammonium hydroxide (Sigma-Aldrich)) until the bones were pliable and embedded in paraffin. Serial sections (5 µm thick) were cut for histological analysis. The sections were stained with Safranin-O/Fast green and toluidine blue according to standard protocols.
The grade of structural damage of each sample was histomorphologically evaluated (by I-GS) according to the modified Mankin score.26 This scoring system is composed of four categories: cartilage structure (0–6 points), cartilage cells (0–3 points), staining (0–4 points) and tidemark integrity (0–1 point). Scores of each category are summed to give a total score with a possible maximum of 14 points. The samples had total Mankin scores ranging from 2 to 14 for DIP joints and 0 to 14 for PIP joints. A cut-off for classifying a sample as having no or a low degree of damage was defined as a Mankin score of ≤5 (lower third of the Mankin scale); specimens with Mankin scores >5 were considered to be osteoarthritic.27
To examine the relationship between histological and radiographical changes, Spearman rank order correlation was used and expressed as r values (rs). Mankin scores were correlated with the Kellgren/Lawrence score and with the most frequent single x-ray features. The Student t test was used to determine statistically significant differences in histological or radiographical changes between groups (bony swelling vs no bony swelling). A p value <0.05 was considered significant. The sensitivity and specificity of plain radiography was calculated as follows: true positive samples were defined as Mankin score >5 and Kellgren/Lawrence score ≥2; false positive: Mankin score ≤5 and Kellgren/Lawrence score ≥2; true negative: Mankin score ≤5 and Kellgren/Lawrence score <2; false negative: Mankin score >5 and Kellgren/Lawrence score <2.
Characteristics of the post mortem specimens
The gender distributed was 56% men and 44% women and did not significantly influence the extent of joint damage as determined by histology (data not shown). DIP and PIP joints were derived from index (48.7%) and middle fingers (51.3%). There was no difference between the second and third fingers in structural histological changes of either the DIP or PIP joints (data not shown).
Patients' ages ranged from 33 to 96 years (median 66 years). Structural damage increased between the age of 40 and 60, reaching a peak at the age of 70–80 years (figure 1). As expected, increasing age was significantly correlated with the extent of joint damage for both DIP (rs=0.76, p<0.0001) and PIP joints (rs=0.53, p<0.0005).
With regard to individual x-ray features, we assessed the prevalence of JSN, osteophytes, subchondral sclerosis, subchondral cysts and erosions (central and marginal erosions were pooled for this analysis) in DIP and PIP joints displaying a Kellgren/Lawrence score of ≥2. JSN and osteophytes were the most prominent features (figure 2) and were used for separate analysis throughout.
It is of note that, of the specimens defined as radiographical OA, two DIP specimens and two PIP specimens had central erosions (9% and 15%, respectively). The mean±SD Mankin scores were 11.5±3.5 for DIP joints and 13.5±0.7 for PIP joints.
Correlation between structural, histological damage and x-ray changes of the DIP and PIP joints
For the DIP joints examined (n=37) we found a highly significant correlation between histological damage and radiography as determined by the Kellgren/Lawrence score (rs=0.87, p<0.0001; figure 3A). Moreover, when we evaluated individual radiographical features we also found a significant correlation for JSN (rs=0.77, p<0.0001) and osteophytes (rs=0.89, p<0.0001; figure 3B,C).
Similar to the DIP joints, the correlation of Mankin scores with the Kellgren/Lawrence scores in PIP joints (n=39) was significant (rs=0.79, p<0.0001; figure 4A) and, likewise, distinct radiographic features correlated with the amount of histological damage for JSN (rs=0.76, p<0.0001) and osteophytes (rs=0.69, p<0.0001; figure 4B,C).
We next analysed x-ray performance to determine the sensitivity and specificity of x-ray examinations in the evaluation of hand OA. All DIP joint specimens with Kellgren/Lawrence scores ≥2 showed moderate to severe histological changes (Mankin score >5: mean±SD 11.23±1.9); however, 4 out of 15 subjects with Kellgren/Lawrence scores <2 had a Mankin score above the threshold of 5 (6.25±0.5), resulting in an x-ray sensitivity of 84.6% (22 true positive, 4 false negative) and a specificity of 100% (11 true negative, no false positive) for DIP joints. While all PIP joints with Kellgren/Lawrence scores ≥2 (n=13) had Mankin scores >5 (10.15±2.3), 11 out of 26 PIP joints without radiographic changes also had Mankin scores >5 (6.7±1). Thus, the x-ray sensitivity for PIP joints was 54.2% (13 true positive, 11 false negative) and the specificity was 100% (15 true negative, no false positive).
Bony swelling versus no bony swelling in post mortem specimens
All joints were examined clinically for bony swelling of the DIP and PIP joints prior to dissection. Of the DIP joints investigated, 35% had bony swelling and 65% had none while the clinical investigation showed that 18% of the PIP joints had bony swelling and 82% had none. All DIP and PIP specimens with bony swelling had a Kellgren/Lawrence score ≥2 together with substantial histological damage (mean±SD Mankin score 11.38±2.1 and 10.57±2.37, respectively). In the subgroup without bony swelling, however, a Kellgren/Lawrence score ≥2 was found in 37.5% of DIP joints and 18.8% of PIP joints. The mean Mankin scores accounted for 11±1.66 for DIP joints and 9.67±2.4 for PIP joints (figure 5A). There was no significant difference in the extent of structural/histological damage between the ‘bony swelling’ and the ‘no bony swelling’ subgroups either for DIP (p=0.65) or PIP specimens (p=0.51). With regard to x-ray alterations, there was also no marked difference in the degree of radiographic changes as assessed by the severity of the most prominent x-ray features, JSN and osteophytes (figure 5B,C).
Radiographic examination would therefore potentially identify subgroups of patients characterised by the absence of bony swelling but severe structural damage which, however, would not be recognised if only clinical criteria for hand OA were applied.11 In our study cohort this subgroup accounted for 25% and 15%, respectively, of all investigated DIP and PIP joints.
Based on histological changes, x-ray sensitivity and specificity was 100% for DIP and PIP joint specimens with bony swelling while, for samples without bony swelling, the sensitivity was 69% for DIP joints and 35% for PIP joints and the specificity was 100% for both DIP and PIP joints.
Hand OA is a very common disease and its correct diagnosis is pivotal because several chronic inflammatory diseases may mimic it. Likewise, classifying hand OA correctly is important in clinical trials since agents that modify the structural changes of OA which often relate to functional impairment are eagerly awaited in order to relieve—and ideally reverse—the effects of the disease. The current ACR classification criteria for hand OA11 are the subject of debate because they do not require x-ray examination12 ,28 while x-rays are regarded as a gold standard by the EULAR recommendations for defining hand OA10 based on a general association between radiographic and clinical changes.11 ,20 ,29,–,31 However, the ultimate correlation with structural changes in DIP and PIP joints as determined by histology has not been studied to date.
We report a highly significant correlation between histological and x-ray changes in DIP and PIP joints, indicating that radiography is indeed a valid tool for detecting OA changes in interphalangeal joints. A considerable benefit of radiography for classifying/diagnosing hand OA was seen in our subgroup analysis assessing interphalangeal joints with or without clinical signs of bony swelling. Generally, patients with bony joint swelling can easily be identified clinically by their frequently impressive Heberden/Bouchard nodes and do not necessarily need an imaging method for diagnosis. Nevertheless, some forms of hand OA lack the bony swelling of the DIP and PIP joints10 ,32 and patients with hand OA without bony enlargement would therefore be excluded by the current ACR classification criteria.28 We therefore tested whether x-ray examination would potentially identify patients without clinical signs of bony swelling but apparent structural damage who would, despite any pain/aching/stiffness, remain undetected using clinical examination alone. Our findings clearly demonstrate profound radiographic changes in joints displaying advanced histological damage in the absence of nodal deformation, and this subgroup accounted for an appreciable proportion of our patient cohort.
Another issue to be addressed relates to the notion that classical radiographic features such as JSN and osteophytes are sensitive (sensitivity 75–100%) but not very specific for hand OA (specificity 18–71%).10 However, these analyses were performed on the basis of clinical findings rather than on structural histological changes. Our data, which are based on histological abnormalities, revealed a 100% sensitivity and specificity for both DIP and PIP joints displaying bony swelling. However, the sensitivity of x-rays was lower for joints without any bony enlargement (69% for DIP and 35% for PIP joints). This might be due to the fact that most of the false-negative subjects showed structural changes only slightly above the threshold Mankin score of 5 (mean±SD score of 6.25±0.5 for DIP joints and 6.7±1 for PIP joints). Within this grey zone of supposedly early OA changes or ‘pre-radiological OA’, the use of plain radiography might be limited. The striking difference in the sensitivity of x-rays between DIP and PIP joints, however, may be due to distinct anatomical and mechanical properties of the PIP joints that might influence tissue remodelling and mask structural damage. Nevertheless, plain radiography was 100% specific in both DIP and PIP joints without bony swelling using histology as a standard for comparison.
At present the treatment of hand OA is confined to the use of symptomatic measures and disease modification is still an aspirational goal. However, research is ongoing and it is to be expected that agents that modify structural changes will be more intensively studied and eventually become available. The candidates for such drugs, in contrast to those focusing on symptoms, may not be confined to patients with symptoms of hand OA and may also be beneficial to patients who do not show significant levels of pain at a given point in time but in whom structural changes may lead to future symptoms. However, even in patients with symptoms the presence of structural changes may allow for a more reliable classification of the disease.
Our histological study was limited in that the sample number was relatively small and data on hand pain were unavailable from the deceased subjects. However, we were able to show a significant correlation between histological and radiographic changes in DIP and PIP joints as well as the possibility of identifying subgroups of patients with hand OA without clinical signs of bony swelling by x-ray assessment. This study therefore indicates that x-rays should be used to diagnose hand OA and to classify hand OA in trials, cohorts and surveys.
The opportunity to carry out this investigation was provided under the auspices of the Joint and Bone Center (Schwerpunkt muskuloskelettale Erkrankungen) of the Medical University of Vienna.
Funding This trial was an independent academic study founded by and performed at the Medical University of Vienna.
Competing interests None.
Ethics approval Ethics approval was obtained from the Ethics Committee of the Medical University of Vienna.
Provenance and peer review Not commissioned; externally peer reviewed.