To the Editor:
Whipple disease is a rare systemic infection caused by Tropheryma whipplei. T. whipplei can also induce focal infections in the joints, brain, and heart, but spinal involvement is extremely rare.1 Here we report 2 cases of spondylodiscitis caused by T. whipplei infection.
Case 1. A 79-year-old White man presented with a recent worsening of chronic low back pain, morning stiffness, and nighttime awakenings. The level of C-reactive protein (CRP) was elevated (67 mg/L), with erythrocyte sedimentation rate (ESR) at 65 mm/h; white blood cell count was 17,440/mm3, with a neutrophil level of 65.5%.
Magnetic resonance imaging (MRI) showed mirror images of the L4 and L5 vertebrae, of low intensity on T1-weighted images and high intensity on T2-weighted images, and an enhancement after gadolinium injection, thereby suggesting spondylodiscitis. The scan-guided disco-vertebral biopsy was germ-free. However, specific quantitative PCR was strongly positive for T. whipplei. T. whipplei–specific PCR was positive in the stool, but negative in blood, saliva, and cerebrospinal fluid. Duodenal biopsies showed no histological evidence of Whipple disease and were negative for T. whipplei as revealed by specific PCR. Isolated T. whipplei spondylodiscitis was diagnosed. Treatment with doxycycline (200 mg/day) and hydroxychloroquine (600 mg/day) was effective, with resolution of the clinical symptoms and normalization of the acute-phase reactants by 3 months (ESR at 11 mm/h and CRP at 2 mg/L). An MRI carried out 11 months later showed resolution of bone and disc signal abnormalities (Figure 1).
Case 1: Lumbar MRI before and after treatment. A. T1-weighted image before treatment. B. Fat-saturated T2-weighted image before treatment. C. T1-weighted image with gadolinium injection after 11 months. D. T2-weighted image after 11 months. MRI: magnetic resonance imaging.
Case 2. A 57-year-old White man experienced intermittent migratory oligoarthritis of the shoulders, hands, and ankles, and low back pain since 2009. Blood acute-phase reactants were elevated. In 2011, the patient presented with asthenia and weight loss, but no digestive symptoms. CRP value was elevated (21 mg/L) but ESR was 20 mm/h; white blood cell count was 7100/mm3, with a neutrophil level of 53.2%. A spinal computed tomography scan showed erosions of L4 and L5 vertebrae (Figure 2). Quantitative detection of T. whipplei DNA by real-time PCR was positive in feces but negative in saliva and blood samples.
Case 2. A. CT scan of the spine in 2011: erosions and osteocondensation of L4 and L5 vertebrae. B, C. PET scan in 2016 showing increased FDG uptake of L3-L4 disc with the eroded adjacent vertebra. There was also increased FDG uptake of T12 with lytic aspect and erosions of T11 and L1. FDG: 18F-fluorodeoxyglucose; PET: positron emission tomography.
Five years later, there was no improvement and there was a persistent rise of CRP values (41 mg/L). Specific real-time PCR of T. whipplei in feces and saliva was positive but led to negative results in duodenal biopsies. Imaging with 18F-fluorodeoxyglucose–positron emission tomography scan demonstrated increased L3-L4 disc uptake at the vertebral mirror erosions, and a heterogeneous lytic aspect of T12 with an eroded/condensed appearance of adjacent T11 and L1 vertebrae (Figure 2). Spondylodiscitis caused by T. whipplei was diagnosed. Treatment with doxycycline (200 mg/day) and hydroxychloroquine (600 mg/day) was administered. The patient gained 5 kg, with resolution of arthralgia, an improvement in back pain, and normalization of CRP values at 1.2 mg/L, with ESR at 7 mm/h.
In contrast to the systemic classic Whipple disease, the focal T. whipplei infection is characterized by normal duodenal biopsies. Joint manifestations are found in 80% of cases and occur at a mean of 6 years before diagnosis, whereas axial involvement is much less common than peripheral arthritis or arthralgia.1
To the best of our knowledge, only 9 cases of spondylodiscitis caused by T. whipplei have been reported in the literature (Table).2-10 Patients were mainly male (10 male:1 female), with a median age of 57 years, and a median symptom duration of 5 years at the time of diagnosis. Eight patients had systemic involvement. Two had isolated spondylodiscitis defined as no other organ involved, and digestive biopsies with negative histology and PCR results. Four patients (including our case 2) had multifocal spondylodiscitis. Ten patients had elevated acute-phase reactants.
Characteristics of ten cases of Tropheryma whipplei spondylodiscitis from the literature.
Eight patients underwent spinal biopsy, with all interpretable biopsies positive for T. whipplei with specific PCR. Five biopsies were analyzed with universal 16S RNA PCR, 4 of which were positive. The sensitivity of T. whipplei–specific PCR seems better than that of the universal 16S RNA PCR, but the latter allows detection and identification of bacteria other than T. whipplei (by sequencing).
Feces and saliva screening PCRs were performed for 5 patients, 4 of whom had at least 1 positive localization at the time of diagnosis. Only 1 study reported PCR detection of T. whipplei from a spinal biopsy, as in the case of patient 1.
All patients, including those with focal disease, were treated with antibiotics for at least 1 year, mainly with a combination of doxycycline and hydroxychloroquine, or cotrimoxazole (Table).
There is no consensus on the optimal treatment of T. whipplei infection. However, the antibiotics used should demonstrate intracellular efficacy. The most commonly used molecules are doxycycline in combination with hydroxychloroquine or cotrimoxazole.
In conclusion, when faced with a culture-negative spondylodiscitis, it seems important to screen T. whipplei by specific PCR, which is more sensitive than 16S RNA PCR. Because PCR results for T. whipplei can be negative in stool and saliva samples in focal infections, it is recommended to perform specific PCR on local samples of the suspected affected organ to diagnose focal presentation and differentiate it from systemic disease.
Footnotes
CONTRIBUTIONS
AT, PP, GC: case. PP: writing. XP: reviewing.
FUNDING
The authors declare no funding or support for this work.
COMPETING INTERESTS
The authors declare no conflicts of interest relevant to this article.
- Copyright © 2026 by the Journal of Rheumatology
