The most serious long-term complications of anti-tumour therapy are secondary malignancies. After treatment of haematological malignancies such as NHL, Hodgkin’ disease and multiple myeloma as well as solid tumours like breast cancer, lung cancer, soft tissue cancer and ovarian carcinoma therapy-related leukaemia occur in 0.5 to 10% of pts. Parameters which might allow an estimation of the individual risk to develop a therapy-induced neoplasia are urgently needed. Referring to recent findings defects in the DNA-repair-system and in detoxification genes are candidate predictive factors. We have examined 125 cases of t-AML/MDS in a retrospective study. The primary neoplasias were: breast cancer (33%), NHL (22%), M. Hodgkin (7%), multiple myeloma (8%) or other miscellaneous (30%). Cytogenetic data of therapy-induced malignancies were available in 116 of 125 pts. in 66% (n=78) of whom clonal abnormalities were detected. The most frequent single abnormalities were -7/7q- (15%), inv(16)(12%), +8 (8%), t(15;17) and del(20q) (5% each). Aiming to identify possible risk factors we used this patient cohort to study genetic polymorphisms influencing the activity of DNA-repair- and detoxification enzymes. For this purpose analyses of XRCC3, RAD51, GSTT1/M1/A1/P1 and NQO1 were performed. Our patient cohort (n=125) attracted attention that patients with a deficiency for GSTT1 or with combined defects of GSTT1 and M1 had a 1.7-fold (p=0.03625) and a 2.4-fold (p=0.00921) risk resp. of developing t-AML/MDS as compared to healthy controls (n=381). Noticeable is as well that the inactivating polymorphism of GSTA1*B was significantly more frequent in the control group (p=0.01847) in comparison to our patient cohort. Subsequently we generated different categories for statistical analysis according to the primary malignancy: As yet, focusing on GSTT1/M1 revealed an overrepresentation (p<0.0005) (OR=4.0) of double null genotypes for GSTT1/M1 in pts. with t-AML/MDS after adjuvant treatment for breast cancer (n=41). Preliminary results suggest that patients with tAML/MDS following NHL (n=22) who have a defect for XRCC3 have a 4.2-fold risk to develop a tAML/MDS (p=0.02960). The polymorphism of GSTA1*B was significant only in our control group in comparison to patients who showed NHL as a primary malignancy (n=22). We suppose that in this cohort the GSTA1*B polymorphism might protect against t-AML/MDS. This could be an explanation for the overrepresentation of GSTA1*B-defect in our control group as compared to our patient cohort. Other statistical analyses did not show any significance. Taken together polymorphism of GSTA1*B and XRCC3 might be associated with the risk to develop t-AML/MDS after NHL. Defects of GSTT1 and M1 can be taken as a first hint to a risk factor for t-AML/MDS following breast cancer.

Author notes

Disclosure: No relevant conflicts of interest to declare.