Although the association of AIHA and ovarian teratoma has been made since 1938, by West-Watson and Young,
8 it is still a rare paraneoplasic phenomenon in ovarian tumour patients with only 26 cases reported until 2003.
9 The ovarian teratoma is relatively common, but the incidence of haemolytic anaemia in patients with this tumour is, obviously, low and at present we do not know why a very occasional tumour produces this reaction when the vast majority do not.
Payne
et al10 in 1981 analysed 19 case reports previously published of patients with haemolytic anaemia associated with ovarian tumours. Sixteen patients had a successful response with tumour removal alone, one patient had a total response after ooforectomy and splenectomy were performed simultaneously and two patients died before surgery by non-related causes. Only three patients had a partial response to corticotherapy, which was considered negligible, and then went on to have the tumour removed which consequently cured the disorder. Disappearance of autoantibody following tumor removal was documented in many instances with time intervals of 2 weeks to 7 months following surgery.
The haemolysis mechanism has not yet been clarified and further research would be of interest as available data has mainly reported on the past.
4–7,9–11 The mechanisms proposed are: (1) liberation of the substance by the tumour, which alters the red cell surface rendering it antigenic to the host; (2) the tumour or its contents may contain material antigenic to the host, stimulating the production of antibodies that cross-reacts with host erythrocytes; (3) the ovarian tumour itself produces red cell antibody providing it is immunologically competent. Most of the cases tend to support this last hypothesis, because antibody production seems to cease immediately after tumour removal in almost all cases reported, and antibody serum disappearance seems to matches well with the half-life of serum IgG.
In our case report, the extra-vascular haemolysis was refractory to corticotherapy and, due to the clinical instability and the increasing biochemical haemolysis, treatment with splenectomy was performed associated with laparoscopic ooforectomy allowing a further histopathology characterisation of the ovarian lesion. Since both interventions were performed at the same time and complete haemolysis regression was obtained, it seems reasonable to think, from similar case reports and review of the literature, that tumour removal was the major factor in resolving the disorder. Furthermore, the persistence of auto-antibodies up to 3 months after the surgical procedure is also reported in the literature and can be detected up to 10 months after surgery.