In December 2005, a 78-year-old woman presented to the emergency department with acute onset of severe lumbar back pain radiating down both posterior thighs, bilateral leg weakness, and paraesthesiae of the feet on getting out of bed that morning.

On examination, lower limb tone was normal, but power was reduced in all movements of the knee and ankle. Reflex testing showed reduced knee jerks, absent ankle jerks and downgoing plantar reflexes. Although anal tone was reduced, sacral sensation was intact. In keeping with cauda equina syndrome, she was unable to pass urine and required a urinary catheter for 2 days.

Two similar episodes had occurred previously.

In 2003, she had been diagnosed with lumbar discitis, after presenting with severe lumbar back pain and subsequently developing urinary retention and lower limb weakness. A magnetic resonance image (MRI) scan showed 50% spinal canal compromise, at multiple levels, due to marked osteophytosis of the facet joints, ligamentum flavum hypertrophy and spondylolisthesis; the spinal cord was normal, but an abnormal signal was detected in the L3/L4 disc. Cerebrospinal fluid (CSF) aspirated under computed tomography (CT) guidance was sterile, but calcium pyrophosphate crystals were present (these were deemed to be not significant). A bone scan showed patchy increased uptake of a radioactive tracer in the lumbar spine, thought to be consistent with discitis. The level of the inflammatory marker C-reactive protein (CRP) was raised (282 mg/L; reference range, < 20 mg/L), the erythrocyte sedimentation rate (ESR) was 79 mm/h (reference range, < 45 mm/h), and the patient was febrile. She was treated with intravenous and then oral flucloxacillin for 6 weeks. Levels of inflammatory markers returned to normal after 2 months.

In 2004, the patient had presented with severe neck pain radiating to both shoulders and fever up to 39°C, with a CRP level of 332 mg/L, ESR of 108 mm/h and white cell count of 18.0×109/L (reference range, 4–11 × 109/L). A neurological examination was normal. An MRI scan showed diffuse enhancement around the cervical spinal canal, suggesting inflammation or infection, and prominence of the ligamentum flavum. CSF microscopy was normal, and culture of CSF for tuberculosis was negative. The patient was diagnosed with cervical spine osteomyelitis, based on the MRI findings, fevers, and raised levels of inflammatory markers, and was commenced on intravenous flucloxacillin. Over 6 weeks of hospitalisation, she gradually improved, and levels of inflammatory markers fell to normal. She continued to take oral antibiotics for 6 months. Her previous history had included longstanding osteoporosis, hypertension, osteoarthritis of the hands, a gastric ulcer (in 1999), and diverticulosis. However, before this second admission she had been well.

The most recent admission followed a similar pattern to the earlier admissions. Initial investigations included an MRI scan of the lumbar spine (Box), which was essentially unchanged from 2003. A bone scan showed increased uptake at several sites, including L1, the patellae, the right foot, and the cervical spine, which was thought to reflect degenerative changes only.

She began to spike fevers up to 39°C, the CRP level rose to 393 mg/L, and she developed a normocytic anaemia (haemoglobin level, 88 g/L; mean cell volume, 86 fL). Blood and urine cultures were repeatedly negative. A CT scan of the abdomen showed no focus of infection, and a chest x-ray was normal. On the basis of the pain, fevers, and raised inflammatory marker levels, she was treated with antibiotics for presumed recurrence of lumbar osteomyelitis/discitis.

After 10 days, she developed painful bilateral knee effusions. X-rays of the knees revealed chondrocalcinosis, and aspirated fluid contained calcium pyrophosphate crystals. We made a diagnosis of systemic calcium pyrophosphate deposition disease (CPPD), involving the spine and causing intermittent cauda equina claudication. The patient was started on colchicine 0.5 mg twice daily and prednisone 40 mg, and the pain settled within 36 hours. A later attempt to withdraw these medications precipitated a flare in pain and a rise in the level of inflammatory markers; she responded promptly to their re-introduction. The woman was discharged home after a month of rehabilitation, and remains well on a small dose of colchicine.

Subsequent review of her extensive previous laboratory data revealed that the disc aspirate from 2003 contained numerous calcium pyrophosphate crystals, and, in retrospect, the two previous diagnoses of lumbar and cervical spine osteomyelitis/discitis were probably episodes of CPPD.

Our case is notable because of its rare combination of factors: systemic CPPD, and CPPD involving the intervertebral disc and ligamentum flavum, which, with spondylolisthesis and facet joint hypertrophy, caused intermittent cauda equina syndrome, in a patient who had not previously had peripheral involvement. This made the diagnosis challenging, despite the fact that calcium pyrophosphate crystals had previously been identified in her lumbar disc. We found no such cases reported in the Medline, EMBASE, CINAHL or Google Scholar databases.

CPPD was first described in the early 1960s by McCarty et al1 and Zitnan and Sit’aj.2 Both reports noted inflammatory joint involvement accompanied by fever, anorexia, weight loss and raised ESR. CPPD is characterised by the clinical features of arthritis (pseudogout), radiographic chondrocalcinosis in both hyaline and fibrocartilage, and identification of calcium pyrophosphate crystals in either synovial fluid or excised tissue.3

There are several distinctive clinical patterns, the most common being acute monoarthritis of the larger joints such as the knee, elbow, shoulder and hip (seen in 35%–60% of patients). The onset of severe pain, swelling, stiffness and erythema typically occurs over a few hours, reaching maximal intensity after 6–24 hours.3

The incidence of calcium pyrophosphate deposits in the intervertebral disc varies from 3.1%4,5 to 14%6 in postmortem and surgical series, respectively. In the surgical series,6 17% of samples had crystals in the ligamentum flavum and disc. In the same study, the incidence of crystal deposition in the ligamentum flavum was 24.5% in patients undergoing laminectomy (contributing to their lumbar spinal stenosis) and 5% in control cadavers.

There have been case reports of calcium pyrophosphate deposits in the ligamentum flavum causing chronic myelopathy.7,8 In a case series from Japan, five patients had intermittent cauda equina claudication and sciatica as a result of anterior spondylolistheses and hypertrophy of the medial facet joint and ligamentum flavum. All patients had calcium pyrophosphate deposition in the ligamentum flavum.9

A case of CPPD mimicking infection in the lumbar spine10 was reported in an elderly woman with progressive lumbar pain, fevers, and elevated levels of inflammatory markers. She underwent surgery before a diagnosis of CPPD was made. The authors comment on the difficulty of differentiating infection from CPPD because of the similar clinical signs.

Like spinal CPPD, systemic CPPD is an uncommon diagnosis. Several case series have described patients who presented with fever and raised inflammatory markers and underwent expensive and invasive investigations and lengthy delays in diagnosis until they developed peripheral arthritis, with CPPD being subsequently diagnosed.11-13 Bong and Bennett11 comment that fever is not a well recognised sign of CPPD, causing diagnostic confusion, over-investigation, and delay in appropriate therapy.

By the time of correct diagnosis, our patient had had 32 plain x-rays, six MRI scans, two bone scans, three CT scans (including two CT-guided biopsies under general anaesthesia), two lumbar punctures and six blood culture sets, and had spent 4 months in hospital. This emphasises the need to consider CPPD in the differential diagnosis of any case of back pain or pyrexia of unknown origin, even if there has been no previous peripheral involvement.

Magnetic resonance image of the L4/L5 disc, showing 50% canal stenosis due to facet joint hypertrophy, spondylolisthesis and ligamentum flavum hypertrophy

Anterolisthesis of L4 on L5 (white arrow). Facet joint hypertrophy (grey arrow). Ligamentum flavum hypertrophy (black arrow).