We present the case of a 18-year-old female with 10 years history of SLE referred to pain clinic for progressive, localized, dull and non-irradiated pain in both knees for over 12 months, with difficulty weight bearing particularly on the right side leading to limping and impaired walk. The pain was constant during the day and often at night, aggravated during walk and stair climbing, and partially relieved with rest. No neurological symptoms including numbness or paresthesias in lower extremities or bowel and bladder dysfunction were reported. One year before BMAC treatment, the patient was neither a drinker nor smoker, and did not report any noticeable weight change.

The patient was diagnosed in 2009 with SLE and when first seen was on a protocol of 10 mg of prednisone daily. SLE flare ups endured for periods of 4 weeks, then reduced gradually over 4 weeks and then halted. Patient had an average of 1 flare up every 6 months for a total average annual cumulative dose of 784 mg/years. Past medical history included glucocorticoid induced osteoporosis, cataracts and macrophage activation syndrome. There was no known history of significant traumas to the knees. Patient's past surgical history was negative. Patient's medications were hydroxychloroquine 200 mg po daily, levetiracetam 750 mg po twice a week for SLE related seizures, prednisone 10 mg po once daily, calcium carbonate 1,250 mg po twice daily, cholecalciferol 2000 IU po daily, ferrous fumarate 300 mg po twice daily, ramipril 7.5 mg po as needed when elevated blood pressure. Before the treatment patient had Erythrocyte Sedimentation Rate: 34 mm.h, C-Reactive Protein: 3.5 mg/L, Complement component 3: 1.2 g/L, Complement component 4: 0.23 g/L, ANA titers within normal limits, dsDNA: 13.4 IU/ml, Rheumatoid factor autoantibody: 10 μ/ml, and anti-Cyclic Citrullinated Peptide: 8 μ/ml.

Physical examination of the lower extremities showed swelling in the right knee with bilateral pronounced effusion in both knees. Palpation revealed extreme generalized tenderness in the right knee more pronounced than the left. Mobility of the right knee showed a reduced range of motion (ROM) with a lack of 70 degrees on flexion and 25 degrees on extension, inducing pain with further movement. The left knee showed neither active nor passive limitations. Muscle weakness was noted bilaterally, however, power was 5/5 bilaterally on flexion and extension. No sensory or reflex alterations were found. No joint stability (ligament) issues were observed in either knee. Upper and lower joints were in normal limits.

Based on the underlying SLE, the sustained corticosteroid therapy and the clinical presentation, the patient was suspected to have developed secondary ON in both knees. The diagnosis was confirmed by MRI, which showed extensive subchondral fluid on both lateral femoral condyles and anterior weight-bearing surfaces of the medial femoral condyles, more pronounced in the right knee. These findings were associated with impending collapse and fragmentation of the articular surface. Therefore, extensive ON involving bilateral medial and lateral femoral condyles was confirmed and staged as Ficat IV. Furthermore, areas of bone infarction in the distal femoral diaphysis and proximal tibial diaphysis were also observed, with no evidence of ON concerning the subarticular proximal tibia. Based on an initial clinical assessment, the right knee was confirmed as more advanced ON than the left (Supplemental Video 1).

Bone marrow (BM) was aspirated five times (initial three times for the right knee injection and two final for the left knee injections), with intervals of 1–2 weeks and alternating the harvesting site. Under sterile conditions, the BM aspiration site (Posterior Superior Iliac Spine—PSIS) was marked upon visualization with ultrasound guidance, followed by injection of 2% lidocaine in the soft tissue and periosteum (Figure 1). The BM tissue was collected using a 14G trocar needle, with an entry point created with the introducer needle and the bone drilled through the periosteum and cortex into the spongy bone into a depth of 4 cm. Subsequently, BM was aspirated slowly at a rate of 1–2 cc per level and withdrawn gradually using heparinized syringes. The aspirated BM (i.e., BMA) was then passed through a customized size-based cell filtration system (Chondrostem device, manufactured by CCRT, Toronto, Canada. Patent pending) similar to other concentrators (Ito et al., 2009; Otsuru et al., 2013) to enrich the mononuclear cells (MNCs) fraction, especially the mesenchymal stem/stromal cells (MSC) component present within that fraction and described as the culture-expanded progeny of in vivo connective tissue progenitors (CTP) (Sivasubramaniyan et al., 2018).

The injections were performed under local anesthesia, sterile conditions and ultrasound guidance, using a 18-gauge 3.5-inch needle inserted 2 cm deep into the condyle bone (Figure 1). The processed BMAC was injected in five sessions for right and left knees independent treatments (Table 1), including two intra-articular injections only to the right knee given its early signs of articular cartilage involvement.

in vitro data showed that Chondrostem filtration result in CD45⁻CD44⁺CD90⁺CD105⁺ MSC/EPC subpopulation enrichment (43% of total) (Figure 2A). MSC-related phenotypic profile and tri-potentiality (bone-fat-cartilage) was established in BMAC cells cultured for two passages in vitro (Figure 2B).

The patient was followed-up 1 week after each procedure at the same time with the following injection, then at 1, 3, 6, 9, and 12 months while undergoing post-surgical rehabilitation. At each visit joint effusions, swelling and tenderness to palpation at the joint line level, and range of motion (ROM) were assessed. Patient-reported outcomes measurement (PROM) were assessed using clinically validated scoring systems (VAS, KOOS, and WOMAC).

Four weeks after the final procedure pain was almost absent on both sides (1–2/10), with weight-bearing the left leg (not the right one) and able to walk with crutches. On examination, there was only mild effusion on the left knee with a full ROM, whereas the right side had limited 10 degrees on extension and 35 degrees on flexion. At 6 months, the patient was pain-free (0/10 on the left and 1/10 on the right side) with a full ROM on both sides. She was able to bear weight, walk and even dance without the support of crutches. Her improvement continued until 9 and 12 months, when she was pain-free with full ROM and with the ability to run (Supplemental Video 2). Clinically, the PROM showed significant improvement over time, starting right after the end of the sequential therapy one (1) month after its start and for a 24-months follow-up (Figure 3). MRI at 12 months showed bone maturation (with granulation tissue) in active ON areas, less marked BM edema bilaterally and significant reduction in subchondral fluid, which proved radiological improvement supporting the positive clinical progression. With exception of the lateral condyle of the right knee, which showed persistent irregularities and fragmentation/collapse of the central and posterior weight bearing surfaces, the rest of structures did not show signs of collapse or fragmentation (Figure 4). At 24 months follow-up the patient was stable in terms of pain and function, able to cycle and dance, no effusion in the knees with full active and passive range of motion on physical examination.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.