Enejd Veizi, Yasin Erdoğan, Başak Sinem Sezgin, Yavuz Karaman, Kasım Kılıçarslan, Ahmet Fırat

Department of Orthopedics and Traumatology, Ankara City Hospital, Ankara, Türkiye

Keywords: Corticosteroids, COVID, hip, knee, osteonecrosis.


Objectives: This study aims to evaluate the incidence of osteonecrosis (ONC), with a special focus on ONC of the femoral head (ONFH), in novel coronavirus disease 2019 (COVID-19) patients two years after the pandemic.

Patients and methods: This prospective study included COVID-19 patients who were admitted to our center between March 2020 and June 2020. A total of 472 patients (289 males, 183 females; mean age: 42.3±12.0 years; range, 18 to 60 years) were arranged in a list according to their date and time of admission and, then, divided into two groups: those not receiving corticosteroid (CS) treatment (Group 1, n=236) and those receiving CS treatment (Group 2, n=236). The patients were evaluated for joint pain based on X-rays and magnetic resonance imaging scans, and the patients were routinely followed. For each patient in Group 2, additional data regarding CS use were recorded. The possible relationship between ONC and risk factors was analyzed.

Results: Both groups were similar in terms of age and sex. Group 2 had a significantly longer hospitalization period. A significant increase in the number of painful joints was observed in Group 2. At two years, 5.1% of the patients in Group 1 complained of at least one painful joint compared to 11.9% of patients in Group 2. Eight patients from Group 2 developed ONC.

Conclusion: The incidence of ONC after CS therapy in COVID-19 patients is on the rise. At two years, 5% of patients receiving various doses of CSs may develop ONC. Residual joint pain is common even after recovering from the virus. No relationship is evident between the duration of treatment, cumulative dosage of medication, maximum one-day dosage received, and the presence of ONC.


The novel coronavirus disease 2019 (COVID-19) is often characterized by a severe acute respiratory syndrome in affected patients.[1] The majority of COVID-19 patients are asymptomatic and very few require medical treatment. On the other hand, severely ill patients are hospitalized, often requiring mechanical ventilation support.[2] A variety of agents have been attempted for the treatment of COVID-19,[3] but only corticosteroids (CSs) have been defined as life-saving in the management of COVID-19.[4]

Widespread usage of CSs has been called out since the beginning of the pandemic, arguing that an uncensored use could lead to a secondary crisis of osteonecrosis (ONC), mainly on the hip and knee joints.[5] While no incidence of large patient cohorts has yet been published, some initial reports have started to emerge regarding ONC lesions of the jaw,[6] knee joint[7] and hip.[8] A study regarding the long-term results of the severe acute respiratory syndrome (SARS) outbreak in 2003 reported ONC rates ranging between 21.1 and 57.7% in patients treated with CSs, the slight majority of which were female.[9] The symptoms appeared as early as one month after treatment with a high dosage and the incidence increased drastically within the first six months.[10] Osteonecrosis is most commonly encountered in the hip joint[9,11] and, considering the debilitating effects, on the quality of life in affected patients, an early diagnosis is extremely important.

Anti-apoptotic effects of thymoquinone may be a promising treatment option for early stage of ONC.[11] Cumulative and maximum dose play a role, but also underlying diseases such as diabetes and autoimmune diseases, influence the final outcome.[12] Early diagnosed patients have the possibility to undergo medical treatment[13] or surgery.[14] Due to the large number of patients who were administered CSs during the pandemic, data regarding previous SARS outbreaks and the relative urgency of early diagnosis, in the present study, we aimed to evaluate the incidence of ONC, with a special focus on ONC of the femoral head (ONFH), in COVID-19 patients two years after the pandemic.

Patients and Methods

This single-center, prospective, longitudinal study was conducted at Ankara City Hospital, Department of Orthopedics and Traumatology between March 2020 and June 2020. Patients were arranged in a list according to their date and time of admission and, then, divided into two groups: those not receiving CS treatment (Group 1) and those receiving CS treatment (Group 2). Inclusion criteria were as follows: age between 18 and 60 years, hospitalization at our center with International Classification of Diseases, 10th Revision (ICD-10) codes related to COVID-19 (U06, U07.0, U07.1 and U07.2) for at least five days, a positive polymerase chain reaction (PCR) result confirming the COVID-19 diagnosis and no prior history of CS usage. Exclusion criteria were as follows: age >60 years, hospitalization in more than one medical center, prior usage of CS medication for any reason (e.g., rheumatoid arthritis, systemic lupus erythematosus, organ transplantation, other autoimmune diseases), a history of hypercoagulability condition (e.g., Factor V Leiden, sickle cell anemia, malignancy), and a prior diagnosis of hip-related deformities or surgeries (e.g., Perthes disease, developmental hip dysplasia, hip fracture). A power analysis with an alpha (α) error of 0.05 and a power of 90% (β error) calculated that 133 patients were required in each group.

The hospitals registry was screened according to the inclusion criteria during the month of July 2020 for eligible patients. Calculating for dropouts and loss to follow-up, we decided to include the first 250 patients in each group. Demographic data and detailed data related to the medications administered during hospitalization were meticulously collected. For each patient in Group 2, the following additional information regarding CS use was recorded: total dose of CS usage (presented as cumulative prednisolone equivalent, after conversion, in mg), maximal dose per day, total duration of treatment (days).

According to their admission date (oldest first), patients were called at intervals of six months and were asked a set of questions relating to recently occurring hip, knee and shoulder pain specifically, and other joint pain in general (Figure 1). Patients describing pain in one or more joints were invited for an outpatient visit. X-rays and magnetic resonance imaging (MRI) were obtained for the evaluation of the painful joints. Patients with painful joints were followed with routine visits every six months and routine X-rays and MRIs were taken, if pain persisted. All patients were contacted by phone up to the second year after admission.

The study was initially started with 250 patients in each group. Thirteen patients in Group 1 and 12 patients in Group 2 refused to fulfill the questionnaire or attend to routine follow-up visits at various intervals, reporting lack of actual joint pain. One patient in Group 2 died due to a jobrelated accident, while one patient from each group failed to show up for the scheduled MRIs. Both patients refused to further participate and reported that the pain they felt was tolerable. Finally, a total of 472 patients (289 males, 183 females; mean age: 42.3±12.0 years; range, 18 to 60 years) including 236 patients in each group completed the study (Figure 2).

Radiological evaluation

Magnetic resonance imaging and X-rays were evaluated by two experienced orthopedic surgeons. The primary focus of the study was on ONFH, while other joints were also examined for ONC lesions. Diagnosis of ONFH was confirmed by a single density "bandlike" lesion with a low rim of signal intensity surrounding the necrotic spot in T1 sequences, and/or by a “double-density sign” with an external low signal intensity rim and a high signal intensity internal rim on the T2 images (Figure 3).

Statistical analysis

Statistical analysis was performed using the IBM SPSS version 20.0 software (IBM Corp., Armonk, NY, USA). Descriptive data were expressed in mean ± standard deviation (SD) or median (min-max) for continuous variables or in number and frequency for categorical variables. Fisher chi-square test, continuity correction chi-square test, and Pearson chi-square test were used to compare categorical variables between groups. The Mann-Whitney U test was used to compare data sets that were not normally distributed, while independent samples t-test was used to compare normally distributed datasets. Univariate and multivariate logistic regression analyses were performed to identify the association between ONC and the risk factors. A p value of <0.05 was considered statistically significant.


Age and sex of the patients were similar between the groups. Patients receiving CSs (Group 2), on the other hand, had a significantly longer hospitalization period (8.9±3.7 days vs. 11.1±7.1 days). Demographic data are shown in Table I.

The patients were administered different dosages of CSs for a mean of 6.8±4.1 (range, 1 to 27) days and the mean cumulative dosage of medication (total cumulative dosage of CS given to patients during hospitalization) used was 1,491.9±2,506.6 (range, 50 to 20,675) mg of prednisolone. Data regarding CS usage is presented on Table II.

Data regarding number of painful joints and signs of ONC on MRI are presented in detail in Table III. The results showed a significant increase in the number of painful joints on the first- and second-year visit, after an apparently similar sixmonth period.

The most frequently affected joint was clearly the hip, with two patients in Group 1 and eight in Group 2 developing ONC at the end of the study period. Data are presented in Table IV. Only one patient in Group 2 was treated with a hip replacement (52 months after hospitalization), since his affected hip joint degenerated in a quick manner (Figure 4). Knee ONC was also managed with a decompression of the necrotic site and injected with a bone marrow aspirate concentrate (data not shown here).

A regression analysis was performed to explore the relationship between CS usage and overall presence of ONC. Univariate and multivariate analyses failed to find a significant relationship between duration of treatment, cumulative dosage of the medication, and the presence of ONC (Table V).


The current study found that more than 5% of patients receiving prednisone for an average of 6.8 days developed ONC in the hip or knee joints. At two years, 11.9% of patients who received CSs had at least a residual painful joint. Duration of treatment, cumulative dosage of the medication, and maximum one-day dosage received were not correlated with the presence of ONC.

Corticosteroids have been the most effective and widely used drug type in the treatment of SARS.[15] The greatest risk of using CSs in the first place, particularly from an orthopedic perspective, is the development of joint ONC and reports from the SARS outbreak of a prevalence of up to 57.7%, should come as no surprise.[16] The results of our study which included 236 consecutive patients hospitalized and treated for COVID-19, showed a 5% incidence of ONC. This is consistent with the results reported by Assouline-Dayan et al.,[17] but is much lower compared to the study of Zhao et al.[10]

Motomura et al.[18] showed a correlation between the maximum daily dose of CSs and ONFH, while Massardo et al.[19] reported that a dose of prednisone greater than 40 mg/day was positively correlated with ONC. Total cumulative dosage also plays a role. Zhao et al.[10] concluded that physicians should avoid using dosages >10 g of methylprednisolone (or equivalents) and possibly allow for “treatment holidays”, since the risk of ONFH increased by a ratio of 1.29 for every 10 days of treatment. In contrast to the literature, this study found no direct relationship between the duration of treatment, overall dosage of the medication, maximum one-day dosage received and the presence of ONC.

Although early diagnosis is of utmost importance, the onset of symptoms may take time to develop. Several studies have described various times to onset, from three weeks after the last CS dosage intake[20] and up to 24 months.[21] General consensus is that between 6 and 12 months after moderate-to-high CS usage, physicians should be vigilant for the possibility of ONFH development and always consider the possibility within the first three years.[17] At 48 months, 28 patients who received different dosages of CS treatment complained of at least one painful joint, but ONC was spotted in only 12 of them. Due to the fact that symptomatic onset may come later that radiological onset, many authors recommend routine screening with X-rays and MRIs.[16,22] Zhao et al.[23] suggested that MRIs be taken at intervals of 3, 6, and 12 months after steroid administration, but not within the first months, since it has been shown to lead to false negativity.[22] Whether MRI screening should become a routine imaging modality after CS treatment is still a matter of debate.

Hofmann et al.[24] proposed that the SARS virus was in, itself, an independent factor for the occurrence of ONFH and the new COVID-19 virus might also have the same effect. Furthermore, the ensuing strong systemic inflammatory response leads to a large number of inflammatory mediators being released, some of which cause vasoconstriction, one of the accepted mechanisms of bone ONC.[25] Li et al.[26] reported that differences in susceptibility levels of patients exist and that patients suffering from the same disease and being treated with the same CS regimen do not show the same rate of complication. Consistent with the literature, a minority of patients not receiving CS therapy developed ONC in our control cohort.

Nonetheless, this study has certain limitations, and the results should be interpreted accordingly. First, we used telephone calls to screen for symptoms, while it is known that clinical complaints appear later than radiological signs. We attempted to overcome this issue by keeping in touch with the patients every six months, trying not to miss any ongoing or newly appearing painful ONC process. The study could have been more effective, if we had screened all included patients with sequential MRIs, which would have surely detected earlier ONC processed; however, we lacked the appropriate funding for such a study design. Second, the study used a non-validated questionnaire to scan for overall joint pain in eligible patients. The questionnaire was used as a tool to standardize the questions asked to patients, although no scores or related values were used in the analysis. While we excluded patients with possible confounding factors such as diabetes, hypercoagulability, or organ transplantation, we did not screen for blood results. Vitamin D and cholesterol values, for instance, have been shown to correlate with ONC in previous studies.[25] We attempted to overcome this by excluding patients with previously known comorbidities based on previously used ICD codes, registry data, and questioning.

Despite these shortcomings, this is the first study to report on the incidence of ONC after a two-year period of CS treatment due to COVID-19. The main strength of this study is the creation of a control group and the relatively large sample size. The results of this study should not, though, refrain healthcare providers from using CSs in critically ill patients. They have been proven to have a lifesaving effect, and that is more important than a possible future joint replacement.

In conclusion, the incidence of ONC after CS therapy in COVID-19 patients is on the rise. At two years, 5% of patients receiving various doses of CSs may develop ONC. Residual joint pain is common even after recovering from the virus. No relationship is evident between the duration of treatment, cumulative dosage of medication, maximum oneday dosage received, and the presence of ONC. Further studies are warranted to draw more reliable conclusions on this subject.

Citation: Veizi E, Erdoğan Y, Sezgin BS, Karaman Y, Kılıçarslan K, Fırat A. The painful joint after COVID-19 treatment: A study on joint osteonecrosis following COVID-19-related corticosteroid use. Jt Dis Relat Surg 2023;34(1):75-83. doi: 10.52312/jdrs.2023.895

Ethics Committee Approval

The study protocol was approved by the Ankara City Hospital Clinical Research Ethics Committee (date: 20.01.2021, no: E1/1429/2020). The study was conducted in accordance with the principles of the Declaration of Helsinki.

Author Contributions

Idea/concept, analysis and/or interpretation, critical review, materials: A.F., E.V.; Design, literature review: Y.E., K.K.; Control/supervision: A.F., K.K.; Data collection and/or processing, writing the article, references and fundings: B.S.S., Y.K.

Conflict of Interest

The authors declared no conflicts of interest with respect to the authorship and/or publication of this article.

Financial Disclosure

The authors received no financial support for the research and/or authorship of this article.


  1. WHO Coronavirus disease Dashboard 2022 Available at: https://covid19.who.int/. [Accessed: 05.07.2022].
  2. Wu Z, McGoogan JM. Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: Summary of a report of 72 314 cases from the Chinese Center for Disease Control and Prevention. JAMA 2020;323:1239-42. doi: 10.1001/jama.2020.2648.
  3. Ferrari F, Martins VM, Fuchs FD, Stein R. ReninAngiotensin-Aldosterone system inhibitors in COVID-19: A review. Clinics (Sao Paulo) 2021;76:e2342. doi: 10.6061/ clinics/2021/e2342.
  4. Powell C, Chang C, Naguwa SM, Cheema G, Gershwin ME. Steroid induced osteonecrosis: An analysis of steroid dosing risk. Autoimmun Rev 2010;9:721-43. doi: 10.1016/j. autrev.2010.06.007.
  5. Zhang B, Zhang S. Corticosteroid-induced osteonecrosis in COVID-19: A call for caution. J Bone Miner Res 2020;35:1828- 9. doi: 10.1002/jbmr.4136.
  6. Al-Mahalawy H, El-Mahallawy Y, Dessoky NY, Ibrahim S, Amer H, Ayad HM, et al. Post-COVID-19 related osteonecrosis of the jaw (PC-RONJ): An alarming morbidity in COVID-19 surviving patients. BMC Infect Dis 2022;22:544. doi: 10.1186/s12879-022-07518-9.
  7. Agarwala SR, Vijayvargiya M, Sawant T. Secondary osteonecrosis of the knee as a part of long COVID-19 syndrome: A case series. BMJ Case Rep 2022;15:e248583. doi: 10.1136/bcr-2021-248583.
  8. Agarwala SR, Vijayvargiya M, Pandey P. Avascular necrosis as a part of 'long COVID-19'. BMJ Case Rep 2021;14:e242101. doi: 10.1136/bcr-2021-242101.
  9. Guo KJ, Zhao FC, Guo Y, Li FL, Zhu L, Zheng W. The influence of age, gender and treatment with steroids on the incidence of osteonecrosis of the femoral head during the management of severe acute respiratory syndrome: A retrospective study. Bone Joint J 2014;96-B:259-62. doi: 10.1302/0301-620X.96B2.31935.
  10. Zhao R, Wang H, Wang X, Feng F. Steroid therapy and the risk of osteonecrosis in SARS patients: A dose-response meta-analysis. Osteoporos Int 2017;28:1027-34. doi: 10.1007/ s00198-016-3824-z.
  11. Dasci MF, Yaprak Sarac E, Gok Yurttas A, Atci T, Uslu M, Acar A, et al. The effects of thymoquinone on steroidinduced femoral head osteonecrosis: An experimental study in rats. Jt Dis Relat Surg 2022;33:553-66. doi: 10.52312/ jdrs.2022.752.
  12. Lai SW, Lin CL, Liao KF. Evaluating the association between avascular necrosis of femoral head and oral corticosteroids use in Taiwan. Medicine (Baltimore) 2020;99:e18585. doi: 10.1097/MD.0000000000018585.
  13. Agarwala S, Banavali SD, Vijayvargiya M. Bisphosphonate combination therapy in the management of postchemotherapy avascular necrosis of the femoral head in adolescents and young adults: A retrospective study from India. J Glob Oncol 2018;4:1-11. doi: 10.1200/JGO.17.00083.
  14. Ranjan R, Rampal S, Jaiman A, Tokgöz MA, Koong JK, Ramayah K, et al. Common musculoskeletal disorders in chronic liver disease patients. Jt Dis Relat Surg 2021;32:818- 23. doi: 10.52312/jdrs.2021.25.
  15. Atik OŞ. Which articles do the editors prefer to publish? Jt Dis Relat Surg 2022;33:1-2. doi: 10.52312/jdrs.2022.57903.
  16. Shetty GM. Double trouble-COVID-19 and the widespread use of corticosteroids: Are we staring at an osteonecrosis epidemic? Indian J Orthop 2021;56:226-36. doi: 10.1007/ s43465-021-00546-8.
  17. Assouline-Dayan Y, Chang C, Greenspan A, Shoenfeld Y, Gershwin ME. Pathogenesis and natural history of osteonecrosis. Semin Arthritis Rheum 2002;32:94-124.
  18. Motomura G, Yamamoto T, Irisa T, Miyanishi K, Nishida K, Iwamoto Y. Dose effects of corticosteroids on the development of osteonecrosis in rabbits. J Rheumatol 2008;35:2395-9. doi: 10.3899/jrheum.080324.
  19. Massardo L, Jacobelli S, Leissner M, González M, Villarroel L, Rivero S. High-dose intravenous methylprednisolone therapy associated with osteonecrosis in patients with systemic lupus erythematosus. Lupus 1992;1:401-5. doi: 10.1177/096120339200100610.
  20. Kubo Y, Yamamoto T, Motomura G, Tsukamoto N, Karasuyama K, Sonoda K, et al. MRI-detected bone marrow changes within 3 weeks after initiation of high-dose corticosteroid therapy: A possible change preceding the subsequent appearance of low-intensity band in femoral head osteonecrosis. Rheumatol Int 2015;35:1909-12. doi: 10.1007/s00296-015-3346-6.
  21. Anderton JM, Helm R. Multiple joint osteonecrosis following short-term steroid therapy. Case report. J Bone Joint Surg [Am] 1982;64:139-41.
  22. Zhao FC, Hu HX, Zheng X, Cang DW, Liu X, Zhang JZ, et al. Clinical analysis of 23 cases of steroidassociated osteonecrosis of the femoral head with normal initial magnetic resonance imaging presentation. Medicine (Baltimore) 2017;96:e8834. doi: 10.1097/ MD.0000000000008834.
  23. Zhao FC, Li ZR, Guo KJ. Clinical analysis of osteonecrosis of the femoral head induced by steroids. Orthop Surg 2012;4:28-34. doi: 10.1111/j.1757-7861.2011.00163.x.
  24. Hofmann H, Geier M, Marzi A, Krumbiegel M, Peipp M, Fey GH, et al. Susceptibility to SARS coronavirus S protein-driven infection correlates with expression of angiotensin converting enzyme 2 and infection can be blocked by soluble receptor. Biochem Biophys Res Commun 2004;319:1216-21. doi: 10.1016/j.bbrc.2004.05.114.
  25. Zhang S, Wang C, Shi L, Xue Q. Beware of steroid-induced avascular necrosis of the femoral head in the treatment of COVID-19-experience and lessons from the SARS epidemic. Drug Des Devel Ther 2021;15:983-95. doi: 10.2147/DDDT. S298691.
  26. Li ZR, Sun W, Qu H, Zhou YX, Dou BX, Shi ZC, et al. Clinical research of correlation between osteonecrosis and steroid. Zhonghua Wai Ke Za Zhi 2005;43:1048-53.