Case Report: Systemic Lupus Erythematosus Complicated by Chronic Nonbacterial Osteitis: A Rare Case Report and Mechanistic Insights

Dear Editor,

A 40‐year‐old woman was admitted in July 2025 with a 3‐month history of progressive neck and left hip pain. The pain was persistent, nocturnally aggravated, and unrelated to physical activity or positional changes. She denied fever, weight loss, recent trauma, livestock exposure, unpasteurized dairy intake, palmoplantar pustulosis, acne vulgaris, or psoriasis.

The patient had been diagnosed with systemic lupus erythematosus (SLE) 17 years earlier, with previous manifestations including cutaneous lupus, lupus nephritis, and Guillain–Barré syndrome. Both renal and neurological involvement responded well to initial glucocorticoids and hydroxychloroquine and remained in sustained remission; renal biopsy was never performed. Prior to maintenance therapy with low‐dose prednisone (5 mg/day) and hydroxychloroquine (0.2 g/day), the patient experienced SLE flares limited to cutaneous involvement and received azathioprine and thalidomide accordingly. Importantly, the patient had no prior symptoms involving bones, tendons, or joints. On admission, her SLE Disease Activity Index 2000 (SLEDAI‐2 K) score was 0, indicating quiescent disease.

Laboratory investigations revealed markedly elevated inflammatory markers, including an erythrocyte sedimentation rate (ESR) of 120 mm/h and a C‐reactive protein (CRP) concentration of 34.2 mg/L. Peripheral blood counts were largely normal. Autoimmune serology revealed a speckled antinuclear antibody pattern (1:100) and mildly elevated anti‐β₂‐glycoprotein I antibodies (29.97 RU/mL; normal range: < 20 RU/mL), while anticardiolipin antibodies and lupus anticoagulant were negative; anti‐double‐stranded DNA antibodies and complement levels remained within normal ranges. To rigorously exclude antiphospholipid syndrome (APS), a repeat serological assay was performed 12 weeks later (October 2025), at which time the anti‐β₂‐glycoprotein I antibodies titer had returned to the normal range (15.1 RU/mL), thus failing to fulfill the 2023 ACR/EULAR APS classification criteria.

Magnetic resonance imaging demonstrated multifocal abnormal marrow signals involving the bilateral proximal femora, sacrum, cervical spine, and thoracic vertebrae (Figure 1A). ¹⁸F‐FDG PET/CT demonstrated extensive hypermetabolic bone lesions in the sternum, clavicle, sacrum, spine, and proximal femora (Figure 1B), raising concern for hematologic malignancy or metastatic disease.

FIGURE 1.

(A) July 2025 T2‐weighted fat‐saturated (STIR) MRI shows bone marrow edema and osteolysis involving the bilateral proximal femora, sacrum, cervical spine, and thoracic vertebrae (arrows). (B) July 2025 PET‐CT shows multifocal radiotracer uptake within the cervical spine, left proximal clavicle, manubrium sterni, thoracic vertebrae, sacrum, and bilateral proximal femora (arrows). (C) August 2025 sternal biopsy histopathology: Necrotic bone with fibrous hyperplasia and lymphocytic infiltration (H&E staining, ×200, scale bar = 50 μm). (D) January 2026 bone‐window CT shows osteosclerosis accompanied by bone destruction in the left proximal clavicle, manubrium sterni, and T11 pedicle (arrows).

A comprehensive infectious work‐up, including T‐SPOT.TB, Brucella serology, fungal biomarkers (galactomannan, 1,3‐β‐D‐glucan, cryptococcal antigen), and nucleic acid testing for Epstein–Barr virus and cytomegalovirus, was negative. Bone marrow aspiration, lymphocyte immunophenotyping, and repeated histopathological examinations excluded lymphoma, leukemia, and metastatic malignancy. Sternal biopsy demonstrated sterile chronic inflammatory infiltrate, fibrous hyperplasia, and necrotic bone, without evidence of malignancy or infection (Figure 1C).

No anterior chest wall hyperostosis, a classic SAPHO‐spectrum imaging feature, was observed on serial multimodal imaging. Serial multimodal imaging captured the characteristic evolution of lesions from bone marrow edema and osteolysis to osteosclerosis and structural destruction (Figure 1D), ultimately supporting the diagnosis of chronic nonbacterial osteitis (CNO). The diagnosis of CNO was formally established in accordance with the 2025 expert consensus recommendations [1], based on: (i) histopathologically confirmed sterile bone inflammation; (ii) multifocal skeletal involvement (anterior chest wall, spine, sacrum, proximal femurs); (iii) imaging evolution from marrow edema/osteolysis to osteosclerosis/destruction; and (iv) thorough exclusion of infection and malignancy.

The patient received sequential lines of treatment, accompanied by dynamic changes in inflammatory biomarkers and subjective pain intensity. The trends of ESR, CRP, and visual analogue scale (VAS) pain score across different therapeutic phases are summarized in Figure 2. She initially responded poorly to antibiotics (Moxifloxacin, Clindamycin), nonsteroidal anti‐inflammatory drugs (NSAIDs), glucocorticoids, and iguratimod. Subsequently, tofacitinib (5 mg twice daily) was added to low‐dose prednisone (10 mg/day) and NSAIDs for 3 months, but failed to alleviate bone pain, and inflammatory markers rose further. Whole‐exome sequencing identified three heterozygous variants in the proband: MEFV E148Q, CFHR5, and CFTR. Sanger segregation analysis revealed that the mother carried heterozygous variants in all three genes, whereas the father carried a homozygous MEFV E148Q variant and tested negative for CFHR5 and CFTR variants. Because of progressive vertebral destruction and persistent inflammatory activity, therapy was escalated to high‐dose prednisone (1.5 mg/kg/day) combined with adalimumab (40 mg once weekly) and methotrexate (10 mg once weekly), resulting in partial improvement. However, upon tapering prednisone to 15 mg/day, bone pain recurred and inflammatory markers increased; adalimumab 40 mg weekly was unable to maintain pain relief beyond 1 week. Consequently, zoledronic acid was added per the 2025 adult CNO consensus for its systemic dual anti‐resorptive and anti‐inflammatory effects on multifocal skeletal lesions; prednisone was increased to 0.5 mg/kg/day, and adalimumab was switched to infliximab. Following therapy optimization, the patient achieved 3 months of sustained remission with complete pain control, with final inflammatory markers (ESR 13 mm/h, CRP 3.4 mg/L) returning to normal on maintenance prednisone 5 mg/day. No follow‐up imaging was performed during this period, as clinical and laboratory remission was achieved.

FIGURE 2.

Serial inflammatory biomarkers (ESR, CRP) and VAS pain scores across five sequential therapeutic stages: (1) Glucocorticoids (GC) + NSAIDs + antibiotics (ABX); (2) NSAIDs + iguratimod (IGU); (3) NSAIDs + tofacitinib (TOFA); (4) GC + adalimumab (ADA) + methotrexate (MTX); (5) GC + infliximab (IFX) + MTX + zoledronic acid (ZOL). ESR, erythrocyte sedimentation rate; CRP, C‐reactive protein; VAS, visual analogue scale.

CNO, a rare autoinflammatory bone disorder marked by sterile osteomyelitis, recurrent bone pain and radiologically evident inflammatory bone destruction [1, 2], arises mainly from innate immune activation via inflammasome signaling and proinflammatory cytokines including interleukin (IL)‐1, IL‐6, and tumor necrosis factor (TNF)‐ α [3]. Although SLE is classically regarded as an adaptive immune‐mediated disorder, emerging studies reveal extensive crosstalk between adaptive and innate immune cascades in lupus pathogenesis. Type I interferon (IFN) signaling and inflammasome activity are particularly involved in this interplay [4, 5]. This case illustrates that concomitant SLE and CNO supports the concept of an overlapping autoimmune‐autoinflammatory disease spectrum.

Of particular interest was the heterozygous MEFV E148Q variant. E148Q is a low‐penetrance, gain‐of‐function polymorphism rather than a monogenic mutation. Its carrier rate reaches up to 45.61% in the Chinese Han population, whereas classical high‐penetrance MEFV mutations are extremely rare [6]. Functional studies show that E148Q potentiates pyrin inflammasome formation and in vitro IL‐1β and IL‐18 secretion [7]. This may partially explain why our patient did not present typical familial Mediterranean fever (FMF) manifestations while still possessing a lowered inflammasome activation threshold. Notably, MEFV mutations have been associated with earlier onset, more extensive bone lesions, and higher inflammatory burden in CNO cohorts [8]. In lupus models, E148Q may downregulate anti‐double‐stranded DNA antibody production, reduce memory B cells, and attenuate renal pathology [9]. We speculate that this immunological skewing toward innate immunity may simultaneously attenuate adaptive autoimmune features of SLE while increasing susceptibility to localized, innate‐mediated bone tissue destruction, such as CNO.

To explain this phenotype, we hypothesize a two‐hit pathogenic model. In our patient, chronic low‐grade innate immune priming inherent to SLE served as the indispensable “Hit one” (even with a SLEDAI‐2 K score of 0). Prolonged Type I IFN exposure primes monocytes for enhanced inflammasome activation by upregulating the baseline expression of pyrin and its molecular adaptors [10], creating a permissive inflammatory background even during clinical remission. Superimposed on this was “Hit two”, the E148Q gain‐of‐function variant that further lowers the pyrin inflammasome threshold, yielding a hyper‐responsive state. In this context, subclinical stimuli, such as mechanical stress at skeletal entheses, may trigger overproduction of IL‐1β and IL‐18, which could ultimately induce sterile osteitis in this patient.

Differentiating CNO from osteonecrosis is critical in SLE. Long‐term glucocorticoid therapy and antiphospholipid antibody raise the risk of osteonecrosis [11, 12], which typically appears as well‐demarcated epiphyseal/diaphyseal lesions with serpiginous T1‐hypointense rims and a T2 double‐line sign [13, 14]. Extensive sternal, clavicular, and axial involvement is highly atypical for osteonecrosis [15, 16]. In our case, multifocal ill‐defined bone lesions were distributed across the axial skeleton and bilateral proximal femora, with none of the typical osteonecrosis imaging signs identified on MRI (Figure 1A). The proximal femora showed mixed iso‐ to slightly hyperintense T1WI signals and heterogeneous T2WI signals, whereas vertebral appendages and sacrum presented diffuse patchy hyperintensity on fat‐saturated T2WI. The patient only had transient low‐titer anti‐β₂‐glycoprotein I antibodies without persistent seropositivity to predispose osteonecrosis, and serial imaging demonstrated the characteristic inflammatory‐to‐sclerotic transition of CNO. Sternal histology confirmed sterile inflammation. Exhaustive work‐up excluded infection and malignancy, confirming the diagnosis of CNO.

SAPHO syndrome lies within the CNO disease spectrum. The patient had no cutaneous SAPHO manifestations (palmoplantar pustulosis, acne, psoriasis), and serial imaging excluded anterior chest wall hyperostosis, consistent with isolated CNO.

The therapeutic course is further consistent with a paradigm of innate‐immune dominance. The JAK inhibitor tofacitinib, a Type I IFN signaling blocker, showed limited efficacy over 3 months of treatment, suggesting that established bone lesions may become less reliant on IFN‐mediated inflammation in this patient. Established sterile osteitis appears autonomously driven by a local IL‐6 and TNF‐α feed‐forward loop [3]. Accordingly, TNF‐α blockade with infliximab induced substantial and sustained improvement, consistent with current evidence for anti‐TNF therapies in refractory CNO [1, 2]. By interrupting the terminal TNF‐α/RANKL osteoclastogenic axis, infliximab provided a mechanistic resolution of osteolytic destruction [17, 18].

Single‐case design, lack of functional validation for the MEFV E148Q variant, and limited 3‐month follow‐up without serial imaging restrict causal inference between SLE and CNO and long‐term skeletal outcome prediction; the two‐hit mechanistic hypothesis requires further validation.

This case offers a key clinical lesson: in patients with longstanding, quiescent SLE and persistent bone pain and elevated inflammatory markers, bone symptoms should not be readily attributed to lupus flares or glucocorticoid‐induced osteonecrosis. After excluding infection and malignancy, CNO merits high differential priority. Early whole‐body MRI or PET/CT enables timely diagnosis and avoids therapeutic delays. Multimodal imaging reliably distinguishes CNO from SLE‐related bone infarction. Clinicians must recognize autoinflammatory‐autoimmune overlap in SLE to deliver prompt, targeted therapy.

Author Contributions

Longchuan Xu drafted the initial version of this case report. Chenglong Fang and Zhen Chen participated in the design of this study and revised the manuscript. Dongming Xu guided the article and provided ideas. All authors contributed to the preparation of the manuscript. All authors read and approved the final manuscript.

Ethics Statement

This work complies with the ethical standards of the institutional research committee.

Consent

Written informed consent was obtained from the patient for publication of this case report and accompanying data.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.