Article Text

AB0542 Poor value (lack of specificity) of na-f pet scan to predict which patients referred for possible spondylo-arthropathy, but not fulfilling asas criteria, will respond to an anti-tnf trial.
  1. C. Darrieutort1,
  2. C. Ansquer2,
  3. B. L. Goff1,
  4. E. Soltner1,
  5. Y. Maugars1,
  6. J.-M. Berthelot1
  1. 1Rheumatology
  2. 2Nantes University Hospital, Nantes, France


Background It has been claimed that F-Na PET-scan was more sensitive than technetium bone scan and MRI to detect aseptic osteitis of SAPHO syndrome, and the same might hold true for SpA enthesitis (1).

Objectives We sought to study the ability of F-Na PET-scan to enhance physican’s confidence in the diagnosis of SpA and select patients who could respond to anti-TNF among those not fulfilling ASAS or other SpA criteria.

Methods 32 patients referred as possible SpA were offered the possibility of a F-Na PET-scan if they had a negative MRI scan and did not fulfilled ASAS criteria for SpA (as well as Amor’s, ESSG, and CASPAR criteria). A second visit (V2) was scheduled quickly after the PET-scan, and anti-TNF was prescribed if both physician’s confidence in SpA diagnosis and patient’s expectation of anti-TNF were strong enough. Physician’s confidence in the diagnosis of SpA was recorded before F-Na PET-scan (V0), following F-Na PET-scan (V1), after V2, and, for those who tested it, after 3 months of anti-TNF (V3). FM criteria (WPI and SS scores), BASDAI, BASFI, and AS-DAS scores were collected at V0, V2 and V3.

Results Na-F PET-scans showed abnormal uptake of at least one enthesis (including spinal sites) in 30/32 patients: mean of 6.0 +/- 4.67 [0 to 17] (out of the 42 possible sites). None showed signs of osteitis. Those 32 patients (12 males), were 40.3 +/- 11.7 years old, and only 7/32 (22%) were positive for HLA-B27. Mean +/- SD BASDAI at baseline was 5.99 +/- 1.76; mean BASFI 4.13 +/- 2.75 and mean AS-DAS 3.03 +/- 0.7. CRP (3.74 +/- 5.7 mg/dl) was normal in 78%. Mean WPI and SS scores were 9.1 +/- 4.2 and 5.5 +/- 2.8, respectively, and 44% patients fulfilled criteria for FM at V0. Physician’s confidence in SpA diagnosis was rather low at V0 (3.86 +/- 2.14). However, as F-Na PET-scan was abnormal in 30/32 cases, this confidence raised to 5.20 +/- 2.53 after the PET-scan, and 5.42 +/- 2.56 at the end of V2 (mean increase since V0 = 1.34 +/-2.1: increment in 24/32 cases). An anti-TNF trial was then proposed to 19 patients, of whom 17 were still willing to experiment the drug. A good response (according to ASAS criteria) was observed in only 5/17 patients at V3. Physician’s confidence in the diagnosis of SpA had dropped back to 3.54 +/- 2.4 at V3 (whole group). After 6 months, only 2/17 patients were still good responders to anti-TNF. 2/12 non-responders at V3 were reluctant to stop anti-TNF despite lack of significant changes in BASDAI/BASFI/AS-DAS, and 2 were given another anti-TNF despite failure of the first one, so that at 6 months 6/32 patients were still treated by anti-TNF, although only 2/32 had responded. Finally, 16/32 patients were diagnosed as FM (despite F-Na PET-scan uptakes in 15/16, mean = 6.1 +/- 4.4 sites), and only 2/32 as definite SpA (F-Na PET-scan uptakes in 10.5 +/- 9.2 sites).

Conclusions Although F-Na PET scans showed uptakes of enthesis in 94% of patients, and transiently raised physician’s confidence in the diagnosis of SpA in 24/32 cases, only 2 of the 17 patients thereafter treated by anti-TNF trial were improved by the drug. F-Na PET is very sensitive to detect ossifying processes but much probably lack specificity for SpA, and seems more often misleading than helpful.

References Poisson T et al. Rev Rhum 77 (Suppl 3).A63

Disclosure of Interest None Declared

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.