INTRODUCTION
Rosai–Dorfman disease (RDD), also known as sinus histiocytosis with massive lymphadenopathy, is an uncommon histiocytic proliferative disorder that commonly involves the lymph nodes but may present as extranodal disease. Central nervous system (CNS) involvement is extremely rare, occurring in approximately 5% of cases, and among those, spinal involvement accounts for only 20–25% of the cases.[1] The spine-restricted RDD is an extremely rare variant, posing significant diagnostic challenges and often leading to misdiagnosis due to its non-specific clinical presentation and radiologic resemblance to infectious or neoplastic spinal pathologies.
We present the case of a 32-year-old gentleman with a 1-year history of back pain, initially managed as ankylosing spondylitis (AS), who later developed spastic paraparesis. Only then was he diagnosed with this rare variant of RDD. To the best of our knowledge, only around 60 cases of spine-restricted RDD have been reported in the literature,[14] and none have described the coexistence of AS.
This case may suggest a potentially shared immunopathology between the two entities. We report what we believe is the first case of spine-limited RDD associated with AS, highlighting the diagnostic pitfalls and emphasizing the value of interdisciplinary evaluation.
CASE REPORT
A 32-year-old, right-hand dominant, gentleman, known case of hypertension, presented with a history of back pain followed by progressive bilateral lower limb weakness for a year. He had previously tested positive for human leukocyte antigen B27 (HLA-B27) and had been evaluated by multiple rheumatologists, with a presumed diagnosis of AS, leading to a delay in diagnosis. However, given the progression of his lower limb weakness and lack of symptom improvement, he presented to our hospital for further evaluation.
There was no history of upper limb weakness, bowel or bladder dysfunction, trauma, any constitutional symptoms, or systemic illness. His family history was also negative for any genetic disorders or neoplasms.
On general examination, he was a moderately built, young male. Examination of the lower limbs revealed mildly increased tone with grade 4/5 power in all muscle groups. Deep tendon reflexes were brisk (+3), and plantar reflexes were upgoing bilaterally.
Upper limb examination was unremarkable, higher mental functions were intact, and there were no cranial nerve deficits or cerebellar signs noted.
Routine blood investigations, including complete blood count and renal function tests, were within normal limits. Magnetic resonance imaging (MRI) of the thoracolumbar spine with contrast revealed a contrast-enhancing, extradural lesion extending from thoracic 4 to lumbar 2 levels, causing anterior displacement of the spinal cord [Figure 1]. Based on the radiological appearance, the differential diagnosis included lymphoma versus metastatic disease.
A contrast-enhanced CT scan of the chest, abdomen, and pelvis (CT CAP) was done, which showed no evidence of lymphadenopathy or distant metastasis; however, an incidental finding of a featureless descending colon raised suspicion for inflammatory bowel disease (IBD).
Given the motor deficits attributable to spinal cord compression, the decision of surgical intervention was made. The patient underwent D4-7 laminectomy with maximum safe resection of the lesion. Intraoperatively, a highly vascular, firm to hard extradural lesion was identified, adherent to dura with difficult dissection planes.
Intraoperative frozen section revealed multiple fragments of soft tissue with mixed inflammation comprising neutrophils, plasma cells, and histiocytes. No evidence of granuloma formation, malignancy, lymphoma, or any metastatic disease was identified.
Postoperatively, the patient recovered well and was discharged on the 2nd day with general postoperative care and physiotherapy. The initial histopathologic read noted fibrocollagenous tissue with a mixed lymphoplasmacytic inflammatory infiltrate and scattered histiocytes suggestive of inflammation. No well-formed granuloma or any evidence of malignancy were observed.
Special stains such as PASD and Zeihl–Nelson stains were also negative for fungal and mycobacterial organisms, respectively.
Given the inconclusive histopathology, the patient was then referred to heme-oncologist for further evaluation. An autoimmune workup was performed, which was negative except for persistent HLA-B27 positivity. Based on the CT CAP findings, a tumor marker CEA level was done, which was within normal limits, and a colonoscopy was done which did not show any evidence of inflammatory or neoplastic disease.
Histopathology slides were re-reviewed by a neuropathologist (A.G.), and additional workup was conducted. This time, S100-positive histiocytes with emperipolesis (engulfment of intact lymphocytes and plasma cells by macrophages) were noted, leading to the diagnosis of RDD. Immunohistochemical staining was positive for S100, cyclin D2, and CD68 in histiocytes, while MUM1, immunoglobulin G (IgG), and IgG4 were negative [Figure 2].
These findings confirmed the diagnosis of RDD.
Follow-up MRI revealed progression of the residual disease [Figure 3].
After confirmation of histopathology, the case was discussed in a multidisciplinary neuro-oncology meeting and re-resection was advised. He was then readmitted and underwent D7-L2 laminectomy with further resection of the lesion. Intraoperative findings were consistent with the initial surgery.
A positron emission tomography scan at this stage revealed a hypermetabolic focus in the caudate lobe of the liver, suspicious for a lymphogenous deposit.
Postoperatively, the case was discussed again in the neurooncology multidisciplinary tumor board meeting. He was referred again to the hemato-oncologist and was started on systemic therapy with mercaptopurine and methotrexate. His neurological status has remained stable since, and he continues with regular physiotherapy along with physical rehabilitation.
DISCUSSION
The histological features of RDD were first described by Pierre-Paul Destombes in 1965 in a series of cases titled “Adenitis with Lipid-laden Histiocytes.” Later, in 1965, Juan Rosai and Ronald F. Dorfman published a series of four cases that highlighted both the clinical presentation and the histopathological patterns of the disease, giving it the eponym RDD.[3,11] RDD is a rare, non-Langerhans cell histiocytic disorder, often considered part of the myeloproliferative spectrum. It classically presents as bilateral, massive, painless cervical lymphadenopathy, typically in children and young adults, with a slight male predominance.[6,14]
The exact pathophysiology of this disease remains unclear. Several studies have postulated potential roles for viral infections, such as Epstein-Barr virus, human herpesviruses, human immunodeficiency virus (HIV), and hepatitis B virus, while others have proposed autoimmune dysregulation or familial cases associated with germline mutations and several activating kinase mutations.[9] In our patient, HIV and hepatitis serologies were negative, and an autoimmune workup was unremarkable except for HLA-B27 positivity.
To the best of our knowledge, this is the first reported case of spine-restricted RDD in a patient with AS. The co-existence of AS, a chronic HLA-B27-associated inflammatory spondyloarthropathy, may suggest a shared immune dysregulation pathway or antigen-driven histiocytosis. HLA-B27 positivity has been previously described in nodal and other extranodal RDD forms, supporting this hypothesis.
Literature review revealed only a few prior reported cases, none of which described isolated spine involvement in a patient with concurrent AS.[5,10,15]
Interestingly, our patient also had incidental findings suggestive of colonic involvement, specifically a featureless descending colon, raising the suspicion of IBD. While gastrointestinal involvement in RDD is rare, it has been documented in isolated case reports involving the colon, rectum, and appendix, often mimicking IBD or colorectal malignancies.[4] Moreover, a retrospective review of cutaneous RDD found a higher than expected prevalence of Crohn’s disease, suggesting a possible immunologic association between RDD and IBD.[13] In the context of our patient’s HLA-B27 positivity and concurrent AS, both conditions are known to be associated with gut inflammation, which raises the possibility of a broader systemic immune dysregulation contributing to the pathogenesis of RDD. While speculative, this observation warrants further exploration and may inform future surveillance strategies in such patients.
While RDD is most commonly nodal, extranodal involvement can occur involving the skin, deep soft tissues, orbit, other viscera, bone, and less commonly, the CNS. CNS involvement is rare, accounting for only 5% of the cases, with spine-restricted form representing only 9–25% of these.[1] The clinical presentation depends on the location of the disease and the most common location in the spine is thoracic segment followed by cervical and lumbar segments.[14]
Lesions are typically extradural or intradural extramedullary; intramedullary lesions and osseous involvement are rarer.[14] Rao et al. highlighted a rare but documented potential of this disease for simultaneous involvement of the extradural and intradural compartments at the same spinal level, a nuance important for surgical planning and differential diagnosis.[12]
Radiologically, spine-restricted RDD lesions are typically isointense on T1, iso- to hypointense on T2, with homogenous contrast enhancement on post-contrast images. CT scan may show hyperdense lesions and, in some cases, concurrent bone destruction.[1,6] The main differential diagnoses of spine-restricted RDD based on imaging include lymphoma, meningioma, tuberculosis, pachymeningitis, IgG4 disease, and other histiocytoses, as well as other inflammatory or neoplastic lesions.[2] Infectious etiologies must also be considered in the presence of bony involvement.[8] In our case, lymphoma was initially suspected. Notably, preoperative misdiagnosis is common, with a reported misdiagnosis rate as high as 93%.[6]
Grossly, these lesions are often highly vascular, firm to hard in consistency, and adherent to surrounding dura, as observed intraoperatively in our patient. Histologically, emperipolesis, the histiocytic engulfment of lymphocytes, is a hallmark feature. However, this can also be seen in some lymphomas, emphasizing the need for immunohistochemical confirmation.[1] RDD histiocytes typically stain positive for S-100 and CD68, and negative for CD1a, distinguishing this from Langerhans cell histiocytosis.[14] Our patient’s diagnosis was confirmed by the presence of emperipolesis and S-100 and CD68 positivity on immunohistochemistry.
There are no standardized treatment guidelines for the management of spine-restricted RDD. Surgical resection remains the mainstay, especially for symptomatic or compressive lesions, with the goal of maximum safe resection. Inaccessible lesions or those not amenable to surgery may be treated with steroids, radiotherapy, or immunosuppressive therapy.
James et al. reported a case of a longitudinally extensive extradural RDD lesion extending from C3 to L1 successfully treated with a course of steroids, showing both clinical and radiological improvement.[7] Similarly, another case involving a paravertebral RDD lesion with intraspinal epidural extension at the C7-T4 level responded well to high-dose steroids.[8] However, our patient was on a prolonged course of steroids while waiting for the final histopathology, but his symptoms and radiology did not improve with it.
The role of chemotherapy in spine-restricted RDD is yet to be discovered, with current data limited to anecdotal case reports.[9] Our patient, who experienced disease progression post-surgery, was started on mercaptopurine and methotrexate with clinical stabilization. A recent systematic review reported a 70% improvement rate and 15% recurrence rate, with a mean time to recurrence of approximately 19.5 months. Intradural extramedullary and intramedullary lesions carry a higher risk of recurrence.[14]
CONCLUSION
Spine-restricted RDD is an exceptionally rare entity having no pathognomonic clinical or radiological features, making early and accurate diagnosis often challenging. Due to limited awareness, diagnosis is frequently delayed, even following initial histopathological evaluation. Surgical resection remains the cornerstone of management, particularly in patients with neurological deficits, with reduced risk of recurrence and overall better neurological prognosis.
However, the role of adjuvant modalities remains unclear and warrants further investigation. Recognizing this rare entity and its potential associations with immune-mediated conditions, such as with HLA-B27-related disorders and possibly IBD, is essential to avoid misdiagnosis and guide individualized, multidisciplinary treatment strategies.