Dose-response relationship between chemotherapy for childhood cancer and the risk for a carcinoma or malignant melanoma as subsequent neoplasm - a scoping review of the literature and a case-control study
Date issued
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License
Abstract
In Germany, 8.3% of CCS were diagnosed with a SPN within 35 years after diagnosis. One third of them were CSPNs. The therapy of the FPN is a risk factor for a SPN. There is much evidence for the effect of radiotherapy, but chemotherapy was also found to increase SPN risk. However, the evidence for a CSPN risk increase is scarce, especially for a dose-dependent risk association. This thesis investigated the dose-response-relationship between chemotherapeutic treatment for cancer in children and adolescents and the risk for a carcinoma or malignant melanoma as subsequent primary neoplasm.
Two approaches were used. First, a systematic literature search in the manner of a scoping review was performed using the search results from three different literature databases. Inclusion of articles stopped in July 2019. The results of 14 original research articles were used to describe the dose-response-relationship between different chemotherapeutic groups or substances and the risk for several carcinomas (thyroid cancer, breast cancer, salivary gland carcinomas, colorectal cancer, basal cell carcinoma, or gastrointestinal cancer).
Second, a case-control study was conducted with 272 patients with a CSPN and one to four matched controls with a first neoplasm during childhood. This is a subgroup analysis of the case-control study STATT (Second Tumour After Tumour Therapy) carried out by the German Childhood Cancer Registry on all patients with a subsequent primary neoplasm and matched controls registered between 1980 and 2014 (n = 1244). Only patients with intention-to-treat therapy data were included in the analyses of this thesis. Logistic regression analysis with the principle of fractional polynomials with a spike at zero was used to exploratively model the dose-response relationship between different chemotherapeutic groups or substances and the CSPN risk. Substances were pooled into pharmacological groups, their doses were converted using equivalence ratios into the dose of a reference drug of the group, and all doses within a group were summed.
There is evidence for both a risk-increasing and a protective effect of certain chemotherapeutic groups or substances on CSPN occurrence.
The low number of studies with this topic published over a period of 29 years shows the importance of adding to the knowledge base. There was some evidence from the scoping review that anthracyclines and specifically doxorubicin might increase CSPN risk. Procarbazine, an alkylating agent, might be a risk factor, but the results for alkylating agents in general were inconsistent. Cyclophosphamide might decrease the risk, at least up to a relatively high dose. In some studies, the risk by chemotherapeutic substances became only or especially visible in patient groups with little or no radiotherapy treatment.
The case-control study found a risk decrease with increasing cyclophosphamide dose. Methotrexate and other antimetabolites might have a protective effect as well. CSPN risk increased with increasing epipodophyllotoxin and antibiotic dose whereas low doses seemed protective. There were no associations with other common substance groups such as anthracyclines or platinum compounds.
Taken together, cyclophosphamide might be protective, but the association with alkylating agents in general is unclear. Methotrexate and other antimetabolites might have a protective effect with regard to CSPN risk. The association between epipodophyllotoxin dose and CSPN risk is unclear even though there were first indications for a risk-increasing effect. Anthracyclines seem to increase the risk in a dose-dependent manner. Antibiotics showed a significant association with CSPN risk only in the explorative case-control study, which is why the level of evidence is very low. All in all, there were significant associations for few groups or substances.
Nevertheless, it is worthwhile investigating the dose-response relationship between chemotherapy treatment of an FPN and the CSPN risk in CCS, at least for some substances. However, recommendations for actions are hard to draw from the current evidence. This thesis pointed out directions for future research on this topic; patients might prospectively benefit from this research both during FPN treatment and for CSPN screening.