Should patients with FMF undergo BMT?

Isabelle Touitou

Recently, Milledge et al1 reported a patient suffering from both CDA (congenital dyserythropoietic anemia) and presumptive FMF (familial Mediterranean fever). They observed that the patient's symptoms of “FMF” rapidly abated after BMT (bone marrow transplantation) that was required to treat her CDA. This prompted the authors to conclude that through BMT the missing factor in FMF was provided.

I feel that this assertion should be tempered. The fact that the symptoms have stopped after BMT does not necessary mean that the allogenic healthy bone marrow complemented the defect responsible for the clinical expression of the disease. Spontaneous remissions are not rare in FMF, and it is well known that some environmental factors (cold, stress, fatigue, infections, etc) precipitate FMF attacks. The gain in quality of life promoted by the curation of the concomitant CDA could simply account for the improvement of FMF symptoms. The authors, by suggesting that BMT should be considered as a last resort in patients extremely unresponsive to other therapies, could lead clinicians to expeditiously deduce from this single case that BMT is an alternative cure for FMF.

Moreover, from the genetic and clinical data described in this report, this little girl might suffer from an hereditary inflammatory disorder other than FMF, for example hyper-IgD syndrome (HIDS) or chronic infantile neurologic cutaneous and articular syndrome (CINCA), and be simply a carrier for Met680Ile, the most frequent FMF mutation in the Egyptian population. Finally, I would like to mention that the MEFV-encoded protein contains 781 amino acids (not 791, as written in the introduction) and that the “Asp692Ile” mutation does not exist and was probably meant as the rare “Ile692del” deletion (mutation analysis).


Bone marrow transplantation for FMF

Dr Touitou raises questions about the diagnosis of FMF in our patient and the temporal relationship between disappearance of symptoms and the transplantation.

The diagnosis of familial Mediterranean fever (FMF) in our patient was initially a clinical one, although perhaps her disease might better be described as familial paroxysmal polyserositis rather than FMF, as it was the characteristic recurrent serositis rather than the fever that led to the clinical diagnosis at 14 months of age. Her clinical presentation was given in brief in the case report, but we are happy to expand on this here. She presented with recurrent joint swellings of the elbows and knees, which were asymmetrical. She showed exquisite sensitivity to colchicine, with rapid response in her symptoms after 1-2 doses of colchicine and clinically her grunting respiration and tachypnea (taken as evidence of pleuritic involvement) also responded rapidly to this therapy. Although she responded to colchicine, she suffered frequent attacks, and this was not the prime reason for her undergoing bone marrow transplantation (BMT). She did have some symptoms and signs that were not typical of FMF. Her diarrhea was felt to be related to exacerbation of her lactose intolerance with the use of colchicine, and her hepatomegaly was related more to her CDA than FMF. Dr Touitou raises other possible diagnoses. The patient being Egyptian rather than northern European without an elevated IgA makes a diagnosis of hyperimmunoglobulinaemia D with periodic fever syndrome unlikely. Unfortunately we do not have a pre-BMT specimen on which to check her IgD level, but her urinary mevalonic acid was not elevated.

With regards to the possible diagnosis of chronic infantile neurologic cutaneous and articular syndrome (CINCA), our patient was lacking 2 of the cardinal features of the triad, namely a cutaneous rash and chronic meningitis,1-1 and so we believe this diagnosis to be extremely unlikely.

There was a close temporal relationship between the start of the BMT countdown and the disappearance of her elbow swelling, despite her stopping colchicine. None of her symptoms have recurred and the patient failed to show any signs of recurrence when she was stressed by bacterial sepsis at 14 months after BMT, when she was off all immunosuppressive drugs. We feel confident in our diagnosis of FMF in this patient and find it implausible that she had a spontaneous remission of 3 years that coincided with the onset of conditioning for her BMT.

We agree that it is possible that this patient was merely a carrier for the Met680Ile mutation. There are many examples in human genetics of clinically or biochemically well-characterized disorders where a disease-causing mutation is not found in the coding sequence of the gene associated with that disorder, for instance, in ornithine transcarbamylase deficiency1-2 and Rett syndrome.1-3

The author does point out 2 typographical errors, which should have been corrected in the galley proofs. The symbol “Δ ” was used for “del.” This was lost in a font change and was to converted to “D,” which was then expanded to the 3 amino acid code of “Asp”. The MEFV-encoded protein does contain 781 amino acids and not 791.


  1. 1-1.
  2. 1-2.
  3. 1-3.