AML with erythroid differentiation (M6), is usually caused by leukaemic transformation of MDS or following drug therapy. AML is rarely associated with diabetes insipidus(DI). Numerous explanations for the link between DI and M6 have been proposed, the most common being leukaemic infiltration of the pituitary gland. Two cases of M6 with MRI evidence of pituitary damage are presented, in one an over expression of EVI-1 and in both a low circulating level of TGF-ß1 was observed. Interestingly, following chemotherapy the DI resolved.

#1, a 27 year old man with an 8 month history of DI was diagnosed of M6 and monosomy 7. Brain-MRI showed an empty sella. Serological investigations prior to chemotherapy showed; normal GnRH and TRH-TSH stimulation tests, elevated FSH, elevated erythropoietin, low Vitamin D3, normal calcemia and parathormone and low TGF- ß 1. Partial morphological remission and resolution of DI were observed after AML chemotherapy. An over expression of EVI-1 locus was found on RT-PCR. The patient had further AML therapy, followed by a sibling BMT. He died 3 months later from interstitial pneumonitis.

#2, a 51 years old woman admitted with anaemia and a one-year history of polyuria and polydipsia. She was diagnosed with M6 with complex karyotype, a diagnosis of DI was made post serological investigations. Brain MRI showed smooth alterations of the hypophyseal peduncle. Investigations post admission revealed: normal GnRH and TRH-TSH stimulation tests, elevated erythropoietin, low Vitamin D3, normal calcemia and parathormone and low TGF-ß1. Partial morphological remission and resolution of DI was observed following standard AML chemotherapy. Despite further chemotherapy and an autologous BMT with pulses of interleukin 2, progressive disease was observed. The patient died of sepsis after BMT.

To our knowledge, this is the first documented case of AML-M6 linked with DI and monosomy-7. We have shown for the first time a significant diminished TGF-ß1 serum levels. Furthermore, the higher EVI-1 gene expression level in patient 1 may indicate a selective and additional block of TGF-ß1 signal transduction pathway (STP). Kurokawa et al. have shown that EVI-1gene product can bind and inhibit Smad-3, a mediator of STP. These findings suggest an interesting association in these leukemic patients. Monosomy 7 may be a more general cytogenetic abnormality associated with AML and DI. Also, Monosomy 7 may play an important role in the pathobiology of M6 associated with DI. Haemopoietic stem cells (HSC’s) with such marker could be committed into different sub lineages, as shown in mixed lineage leukemia’s (Rovigatti et al.). Interestingly, human myeloid precursor cells display CD34+ on their surface, as receptor for TGF-ß1 and both described cases were CD-34+. The significant lower TGF-ß1 serum levels could explain why HSC’s in this compartment stop cycling and are allowed to shift to and to be committed towards erythroid blasts. High levels of erythropoietin in these patients are consistent with this shift. EVI-1 gene over expression in patient #1 could also inhibit TGF-ß1 STP, favouring shifting into the erythroid compartment.

In conclusion, we suggest an interesting association between TGF-ß1 STP/hypophyseal damage and AML-M6 + DI with recurrent abnormal karyotypic patterns. Further studies are warranted to dissect the underlying molecular mechanism.

Author notes

Corresponding author

Sign in via your Institution