Cartilage calcification of the ankle joint is associated with osteoarthritis in the general population

Osteoarthritis (OA) is a major health problem that affects about 15% of the global population [1]. While OA in weight-bearing hip and knee joints is relatively common, in the ankle joint (AJ) it affects only 1% of the population [2]. It is often hypothesized that the development of AJ OA is mostly related to previous trauma [3, 4]. Valderrabano et al. reported a high prevalence of post-traumatic AJ OA (in 78% of cases) [4], while, other studies have shown a considerably lower prevalence (only 14%) [5]. The real impact of trauma on the development of AJ OA is yet to be fully understood, and the ability to accurately predict which patients will develop AJ OA in the future requires further investigations.

It is likely that the variations in individuals’ articular cartilage composition will play a role in the development of AJ OA. Eckstein et al. reported surprisingly high variability in the quantitative distribution of cartilage in the ankles of patients [6], whereas Quinn et al. found intraindividual variations in the cartilage cells and matrix morphologies of knees and ankle joints [7].

Another possible explanation for the development of AJ OA could be the occurrence of calcification within the hyaline articular cartilage, also known as chondrocalcinosis [8]. A high prevalence of cartilage calcification (CC) as well as a significant correlation between CC and OA has been reported in both weight-bearing hip and knee joints as well as in the first metatarsophalangeal joint (MTP-I joint) [9‐13]. Furthermore, in vitro studies have shown that calcium phosphate crystals can alter cartilage tissue via biomechanical [14, 15] and pro-inflammatory biochemical processes [16‐19], all of which can lead to degeneration of the affected joint.

The prevalence of ankle joint cartilage calcification (AJ CC) in the general population is reported at around 4.7%, and is based on only one cross-sectional study in which the occurrence of calcification on the talar surface was analysed macroscopically [20] and an association between CC and OA of the talus was reported. However, early signs of CC are only measurable in the nano- to micrometre ranges, thus raising the possibility of underestimation with conventional imaging techniques. In order to detect the onset of CC, high-resolution imaging techniques like digital contact radiography (DCR) are required [21]. Taking this into account, the precision of the previously reported cross-sectional study might be called into question [20].

Therefore, the primary goal of this study was to evaluate and quantify the prevalence of AJ CC using high-resolution DCR. Secondly, we examined the correlations between the observed CC with age, BMI and the histological grade of osteoarthritis.

Both ankle joints (n = 160) of 80 donors were obtained from an unselected cohort who underwent autopsy at the Department for Legal Medicine, University Medical Center Hamburg-Eppendorf [22]. Only donors with bilaterally intact ankle joints with no signs of any other diseases (except for OA) were included in this study. Donors with a history of previous ankle surgery, tumours, infections and/or rheumatic diseases were excluded. The study was approved by the local Ethics Committee (PV 4570) and is in compliance with the Helsinki Declaration.

The mean age of the study population was 62.4 years (SD ±17.7, range 23–95 years). Thirty-four of the donors were female, whereas 46 were male. Biometric characteristics of the study population are presented in Table 1.

Independent of the donor’s gender and side, an unexpectedly high prevalence of AJ CC (51.3%) was found in this study. Intraindividual correlations existed between the amount of CC of the left with that of the right AJ, as well as between the distal tibial and the talar cartilage. These findings underline the systemic appearance of ankle-related CC. Furthermore, since calcification has already been found in intact AJ cartilage and the amount of CC correlated with the histological OA grade, we hypothesize that CC may play a causative role in the pathogenesis of AJ OA.

So far, the prevalence of ankle-related CC has been reported in only one cross-sectional study [20]. However, since calcium phosphate deposition is known to begin in the nano- to micrometre range, it is almost impossible to detect the early stages of calcification through macroscopic analysis or even using standard radiographic techniques. High-resolution imaging techniques such as Digital Contact Radiography (DCR) are therefore necessary for detection of the onset of CC [21]. Using DCR, we were able to establish that CC was prevalent in 51.3% of all cases; this was in stark contrast to previously published studies in which the prevalence of CC was a mere 4.7% [20]. Despite this, DCR has shown that AJ CC is relatively rare when compared to the prevalence of CC in other joints, such as the shoulder (98.9%) [26], the hip (96.6%) [11] or the knee (94.3–100%) [11, 23]. Nonetheless, the reason for the difference remains elusive. Interestingly, calcification is comparably prevalent in both the AJ and the MTP-I joint (48.1%) [12]. It has been theorized that weight-loading promotes pro-mineralization of the joints [27, 28], however, the results of our study contradict this point. Since the ankle generally bears enormous loads (many times greater than body weight), it stands to reason that the degree of calcification should be higher. Moreover, other studies have reported that calcification is more prevalent in non-weight bearing joints such as the shoulder [26] and has even been observed in non-weight-bearing parts of the knee cartilage [29]. Another factor to investigate is the impact of the donor’s BMI on CC (additional mechanical stress induced by increasing BMI). We could not find any association between the donor’s BMI and the mean amount of AJ CC. Given this, it can be assumed that mechanical load is not the predominant factor.

Our study highlighted significant correlations between the mean amount of CC in the left and right AJ, between the left and right distal tibia and talus, as well as correlations between the mean amount of calcification in the distal tibial and talar cartilage. These results underline the theory that the development of calcification is systemic [11, 26, 30].

Another interesting observation was the correlation between CC and the donor’s histological OA grade. CC was detected in donors with histologically intact or almost intact hyaline cartilage (i.e. an OARSI grade <  3) in 12% and in 22% of the distal tibial and talar cartilage respectively. Given this, it can be indicated that CC is already present in the joint before the histological OA is even measurable, and might occur before the OA process initiates. Similar observations of spontaneous OA development have been found in two animal models [31, 32], wherein calcification was detectable before cartilage degeneration occurred.

In comparison with other joints [9‐13, 26, 30, 33], there is also a clear association between CC and OA in the ankle. In our study, CC in the distal tibia was detectable in 82% of the donors with severe OA (i.e. an OARSI grade ≥ 3), whereas talar CC was detected in 92%. Moreover, our quantitative analysis demonstrated that the mean amount of calcification in the distal tibial, as well as in the talar cartilage correlated with the histological degeneration grade. Muehlemann et al. also described an association between the prevalence of CC and macroscopic talar degeneration, even though no quantitative analysis was conducted for their study [20]. Taken together, there seems to be evidence that CC plays a crucial role in the development of AJ OA.

No correlation was found between the mean amount of AJ CC and the donors’ age. This is in line with previously published results for other joints, including the shoulder [26], the hip/knee [11] and the MTP-I joint [12]. In contrast, Mitsuyama et al. [23] observed significant correlations between the mean amount of CC in the knee and age of the general population. However, since no adjustment for the donor’s OA grade was conducted, it is conceivable that this association between CC and donor age might have been a spurious correlation, which would disappear once an adjustment for OA grade would be performed.

Certainly, there are some limitations to this study. There were no information about the donor lifestyle, activity and medical history, in particular ankle complaints. Even though the standardized cartilage-bone specimens of the distal tibia and talus used in this study were representative, they reflected only a small proportion of the ankle articulating surface. Lastly, the calcium-phosphate composition of DCR-detected CC was not thoroughly characterized in our study since such analyses require the use of specific diagnostic methods, e.g. FTIR spectroscopy [34] or X-Ray diffractometry [35], and were not specifically in the scope of our study. Nevertheless, none of these limitations is likely to influence the study’s findings and conclusions.

DCR analysis revealed that the prevalence of ankle-related cartilage calcification is much higher than previously considered in general population. Even though it is independent of the donor’s age and/or BMI, calcification seems to occur in histologically intact ankle cartilage and is linked with the joint’s histological OA grade. These insights indicate that hyaline CC is an early, age-independent element and a possible causative factor in the development of ankle-related osteoarthritis. However, the exact pathophysiological role of CC in osteoarthritis and its subsequent importance in the disease’s molecular mechanisms are yet to be identified and investigated.