Pediatr Blood Cancer 2015;62:1799–1804
Pediatric Second Primary Malignancies After Retinoblastoma Treatment Petra Temming, MD,1,2* Anja Viehmann, MSc,3 Marina Arendt, MSc,3 Lewin Eisele, MD,3 Claudia Spix, PhD,4 €ckel, PhD,3 and Dietmar R. Lohmann, MD2,7 Norbert Bornfeld, MD,2,5 Wolfgang Sauerwein, MD,6 Karl-Heinz Jo Background. Children with retinoblastoma carry a high risk to develop second primary malignancies in childhood and adolescence. This study characterizes the type of pediatric second primary malignancies after retinoblastoma treatment and investigates the impact of different treatment strategies and prognostic factors at presentation. Procedure. All national patients treated for retinoblastoma at the German referral center with a current age of 6–27 years were invited to participate in a study to characterize late effects. Results. Data on pediatric second primary malignancies were recorded from 488 patients. Ten developed a malignancy before the age of 18 years. For children with heterozygous oncogenic RB1 alteration (heritable retinoblastoma), the cumulative incidence to develop a second malignancy at the age of 10 years was 5.2% (95% CI 1.7; 8.7%). This results in an elevated risk for sarcoma (n ¼ 4) (SIR
147.98; 95% CI 39.81; 378.87) and leukemia (n ¼ 4) (SIR 41.38; 95% CI 11.13; 105.95). Neither the functional type of the RB1 alteration nor its origin showed a significant impact. Treatment modality influenced incidence, latency, and type of malignancy. Previous radiotherapy increased the risk for solid tumors and 3 of 91 children developed acute leukemia after chemotherapy. However, 2 of 10 malignancies were diagnosed in patients with heritable retinoblastoma but without previous chemotherapy or external beam radiotherapy. Conclusions. Screening for second primary malignancy is an important part of pediatric oncological follow-up in patients with heritable retinoblastoma. For patients with sporadic unilateral retinoblastoma, genetic information influences treatment decisions and allows tailoring of follow-up schedules. Pediatr Blood Cancer 2015;62:1799–1804. # 2015 Wiley Periodicals, Inc.
Key words: chemotherapy; mosaic; radiotherapy; RB1; retinoblastoma gene 1; sarcoma
INTRODUCTION Second primary malignancies (SPM) are the major cause of death among survivors of retinoblastoma treatment.[1–3] Most of the survivors with SPM have bilateral retinoblastoma. An increased relative risk is observed mainly for soft tissue sarcoma, bone tumors, melanoma, brain tumors, and epithelial cancers.[4–7] This risk is already increased during childhood and adolescence. The broad spectrum of type and location of SPM complicates the design of adequate follow-up schedules especially for children. Incidence and type of SPM are influenced by treatment modalities. External beam radiotherapy (EBRT) was shown to significantly increase the risk of SPM compared to survivors of bilateral retinoblastoma without eye-preserving radiotherapy.[4,8– 11] For this reason, systemic chemoreduction has almost completely replaced primary EBRT in eye-preserving therapy since the mid 1990s. However, chemotherapy treatment itself can induce secondary leukemia, best described for topoisomerase inhibitors and alkylating agents.[12,13] Furthermore, an increased risk for bone tumors and leiomyosarcoma was reported after treatment for retinoblastoma with alkylating agents in addition to radiotherapy.[14] The current approach to apply chemotherapy locally aims to minimize the systemic exposure to potentially mutagenic agents, but the consequence on the incidence of SPM remains undefined.[15,16] Almost all children with bilateral retinoblastoma are heterozygous for an oncogenic alteration of the RB1 gene (germline mutations). By contrast, oncogenic alterations of the RB1 gene are confined to the tumor (somatic mutations) in most children with isolated unilateral retinoblastoma. As the risk of SPM depends vastly on laterality and family history, it is to be expected that genetic factors influence the type and incidence of SPM in addition to treatment-related factors. In fact, in a recent study it was found that risk of SPM is higher in bilateral retinoblastoma survivors with a presumed inherited germline mutation compared to patients who had a germline mutation of de novo origin.[17] However, as this study had no access to results of genetic testing and in view of the frequency of incomplete penetrance in heritable retinoblastoma, C
2015 Wiley Periodicals, Inc. DOI 10.1002/pbc.25576 Published online 13 May 2015 in Wiley Online Library (wileyonlinelibrary.com).
misclassification cannot be ruled out, and, consequently this finding needs to be confirmed.[17] The spectrum of RB1 gene variants that predispose to retinoblastoma is very wide (http://rb1-lsdb.d-lohmann.de). It is conceivable that allelic heterogeneity has an influence on type and incidence of SPM. In a retrospective cohort, survivors that had heterozygous oncogenic RB1 gene variants with partial functional inactivation or with whole gene deletions showed a lower risk for SPM.[18] Moreover, patients heterozygous for recurrent nonsense CpG-transitions showed an increased risk for SPM.[18] These genotype–phenotype correlations could have an impact on surveillance protocols if confirmed in independent data sets. Additional Supporting Information may be found in the online version of this article at the publisher’s web-site. Abbreviations: ALL, acute lymphatic leukemia; AML, acute myeloid leukemia; CI, confidence interval; EBRT, external beam radiotherapy; IR, incidence ratio; MPNST, malignant peripheral nerve sheath tumor; NK-L, acute natural killer cell leukemia; SPM, second primary malignancy; SD, standard deviation 1
Department of Pediatric Hematology and Oncology, University Hospital Essen, Essen, Germany; 2Eye Oncogenetics Research Group, University Hospital Essen, Essen, Germany; 3Institute for Medical Informatics, Biometry and Epidemiology, University Hospital Essen, Essen, Germany; 4German Childhood Cancer Registry, Institute of Medical Biostatistics, Epidemiology and Informatics, University Medical Centre of the Johannes Gutenberg University Mainz, Mainz, Germany; 5Department of Ophthalmology, University Hospital Essen, Essen, Germany; 6Department of Radiotherapy, University Hospital Essen, Essen, Germany; 7Institute of Human Genetics, University Hospital Essen, Essen, Germany Conflict of interest: Nothing to declare. Karl-Heinz J€ockel and Dietmar R. Lohmann shared senior authorship.
Correspondence to: Petra Temming, Department of Pediatric Hematology and Oncology, University Hospital Essen, Hufelandstrasse 55, 45122 Essen, Germany. E-mail: [email protected] Received 2 February 2015; Accepted 31 March 2015
1800
Temming et al.
Here we present the results of a follow-up study focusing on the incidence of pediatric SPM in a cohort of retinoblastoma survivors aged 6–27 years. Exact knowledge of the genetic cause in a large proportion of these patients allowed us to investigate the impact of the type of RB1 mutation along with the treatment factors for this age group. Because of the change in treatment strategy, this cohort includes children treated with primary chemoreduction, primary EBRT, as well as a combination of both. The study aims to obtain detailed knowledge about the clinical appearance of pediatric SPM in view of the molecular genetic information. This may influence the design of therapeutic regimens and allow tailoring long-term follow-up schedules.
METHOD Patient Cohort The university hospital in Essen is the national referral center for retinoblastoma in Germany treating more than 90% of German children with retinoblastoma. Ethics approval was obtained from the ethics committee of the University of Duisburg-Essen, Germany. All survivors of retinoblastoma treated at the German referral center with a current age between 6 and 27 years were invited to participate in a follow-up study (n ¼ 656). We obtained data on SPM from 488 patients (74.4%). This includes data of two patients who were deceased due to SPM. The other causes of death were metastatic or trilateral retinoblastoma (n ¼ 8), accident (n ¼ 1), and infection (n ¼ 1). In the group without data, 117 were non-contactable, 41 refused to participate, and 10 were deceased (Fig. 1). The cohorts of participants and non-participating patients were comparable regarding gender, laterality, and median age at diagnosis, but the time of follow-up was significantly longer in the group of participants (Supplementary Table SI).
Data Acquisition Data on diagnosis of SPM were collected during an examination at the university hospital and/or a standardized telephone interview. Data were validated by contacting the physicians in charge after obtaining informed consent. Data on retinoblastoma treatment were extracted from patient’s records. Four treatment groups were defined. All patients without treatment, with enucleation alone or
focal treatment alone (laser coagulation, cryotherapy, or brachytherapy) were included in group 1. Patients in group 2 were treated with chemotherapy alone, in group 3 with EBRT alone and in group 4 with chemotherapy and EBRT. EBRT was performed according to the previously published technique.[19,20] Chemotherapy was defined as adjuvant chemotherapy, chemoreduction, thermochemotherapy, or combination of these. The standard chemotherapy regimen in Germany consists of six cycles of chemotherapy resulting in cumulative doses of 1.2 g/m2 carboplatin, 4.8 g/m2 cyclophosphamide, 9 mg/m2 vincristine, and 1.8 g/m2 etoposide. Genetic testing was performed on blood and tumor tissue on the request of individuals or their legal guardians with the aim to identify the oncogenic alterations of the RB1 gene. Mutation identification analysis was performed on DNA from fresh-frozen tumor samples or DNA from blood as reported previously.[21–24] This included one or more of the following methods: analysis of allele loss in tumors, cytogenetic analysis, denaturing high performance liquid chromatography, exon-by-exon sequencing, multiplex ligation-dependent probe amplification, methylationsensitive PCR, quantitative fluorescent multiplex PCR, quantitative real-time PCR, and single strand conformation polymorphism. Diagnosis of mutational mosaicism was based on the finding of a skewed signal ratio of the variant to the normal allele relative to the ratio obtained from heterozygous DNA. In patients with mosaicism, DNA from a second tissue such as buccal mucosa was analyzed, if available.
Data Analysis The data were subjected to statistical analyses with IBM SPSS Statistics version 22.0 (Chicago, Illinois) and SAS 9.4 (SAS Institute, Cary, North Carolina). Data were truncated at 18 years of age because of the focus on pediatric SPM. The two-tailed Fisher’s exact test was applied for group comparison of categorical data (gender, family history of retinoblastoma, CpG transition). The student’s t-test was used to compare equality of the means for unrelated subgroups of continuous variables (age at diagnosis, number of retinoblastoma at diagnosis). Kaplan–Meier survival analysis was used to estimate survival and cumulative incidence. Standardized Incidence Ratio (SIR) was calculated using SAS 9.4 according to Breslow and Day, the confidence interval (CI) was calculated according to the Byar‘s approximation.[25] SIR was calculated separately for sarcoma and leukemia by comparing the
Patients treated for retinoblastoma 644 survivors 6-27 years of age excluded because no contact (117) no consent (41)
12 deceased
10 w/o information on SPM 486
2
488 patients with information on SPM
Fig. 1. Patient accrual. SPM, second primary malignancy; w/o, without. Pediatr Blood Cancer DOI 10.1002/pbc
Pediatric Second Malignancies After Retinoblastoma observed number of SPM with the expected number from data of the German childhood cancer registry (DKKR). Acute lymphocytic and acute myeloid leukemia were combined. Malignant peripheral nerve sheet tumor, rhabdomyosarcoma, leiomyosarcoma, and osteosarcoma were referred to as sarcomas. Time interval was defined as the time from date of retinoblastoma diagnosis to the date of SPM. As only reference data up to the age of 14 were available, we truncated the data at 15 years for SIR analysis. These analyses were performed on the subset of subjects with a heterozygous RB1 gene mutation. P-values
Pediatric Second Primary Malignancies After Retinoblastoma Treatment Petra Temming, MD,1,2* Anja Viehmann, MSc,3 Marina Arendt, MSc,3 Lewin Eisele, MD,3 Claudia Spix, PhD,4 €ckel, PhD,3 and Dietmar R. Lohmann, MD2,7 Norbert Bornfeld, MD,2,5 Wolfgang Sauerwein, MD,6 Karl-Heinz Jo Background. Children with retinoblastoma carry a high risk to develop second primary malignancies in childhood and adolescence. This study characterizes the type of pediatric second primary malignancies after retinoblastoma treatment and investigates the impact of different treatment strategies and prognostic factors at presentation. Procedure. All national patients treated for retinoblastoma at the German referral center with a current age of 6–27 years were invited to participate in a study to characterize late effects. Results. Data on pediatric second primary malignancies were recorded from 488 patients. Ten developed a malignancy before the age of 18 years. For children with heterozygous oncogenic RB1 alteration (heritable retinoblastoma), the cumulative incidence to develop a second malignancy at the age of 10 years was 5.2% (95% CI 1.7; 8.7%). This results in an elevated risk for sarcoma (n ¼ 4) (SIR
147.98; 95% CI 39.81; 378.87) and leukemia (n ¼ 4) (SIR 41.38; 95% CI 11.13; 105.95). Neither the functional type of the RB1 alteration nor its origin showed a significant impact. Treatment modality influenced incidence, latency, and type of malignancy. Previous radiotherapy increased the risk for solid tumors and 3 of 91 children developed acute leukemia after chemotherapy. However, 2 of 10 malignancies were diagnosed in patients with heritable retinoblastoma but without previous chemotherapy or external beam radiotherapy. Conclusions. Screening for second primary malignancy is an important part of pediatric oncological follow-up in patients with heritable retinoblastoma. For patients with sporadic unilateral retinoblastoma, genetic information influences treatment decisions and allows tailoring of follow-up schedules. Pediatr Blood Cancer 2015;62:1799–1804. # 2015 Wiley Periodicals, Inc.
Key words: chemotherapy; mosaic; radiotherapy; RB1; retinoblastoma gene 1; sarcoma
INTRODUCTION Second primary malignancies (SPM) are the major cause of death among survivors of retinoblastoma treatment.[1–3] Most of the survivors with SPM have bilateral retinoblastoma. An increased relative risk is observed mainly for soft tissue sarcoma, bone tumors, melanoma, brain tumors, and epithelial cancers.[4–7] This risk is already increased during childhood and adolescence. The broad spectrum of type and location of SPM complicates the design of adequate follow-up schedules especially for children. Incidence and type of SPM are influenced by treatment modalities. External beam radiotherapy (EBRT) was shown to significantly increase the risk of SPM compared to survivors of bilateral retinoblastoma without eye-preserving radiotherapy.[4,8– 11] For this reason, systemic chemoreduction has almost completely replaced primary EBRT in eye-preserving therapy since the mid 1990s. However, chemotherapy treatment itself can induce secondary leukemia, best described for topoisomerase inhibitors and alkylating agents.[12,13] Furthermore, an increased risk for bone tumors and leiomyosarcoma was reported after treatment for retinoblastoma with alkylating agents in addition to radiotherapy.[14] The current approach to apply chemotherapy locally aims to minimize the systemic exposure to potentially mutagenic agents, but the consequence on the incidence of SPM remains undefined.[15,16] Almost all children with bilateral retinoblastoma are heterozygous for an oncogenic alteration of the RB1 gene (germline mutations). By contrast, oncogenic alterations of the RB1 gene are confined to the tumor (somatic mutations) in most children with isolated unilateral retinoblastoma. As the risk of SPM depends vastly on laterality and family history, it is to be expected that genetic factors influence the type and incidence of SPM in addition to treatment-related factors. In fact, in a recent study it was found that risk of SPM is higher in bilateral retinoblastoma survivors with a presumed inherited germline mutation compared to patients who had a germline mutation of de novo origin.[17] However, as this study had no access to results of genetic testing and in view of the frequency of incomplete penetrance in heritable retinoblastoma, C
2015 Wiley Periodicals, Inc. DOI 10.1002/pbc.25576 Published online 13 May 2015 in Wiley Online Library (wileyonlinelibrary.com).
misclassification cannot be ruled out, and, consequently this finding needs to be confirmed.[17] The spectrum of RB1 gene variants that predispose to retinoblastoma is very wide (http://rb1-lsdb.d-lohmann.de). It is conceivable that allelic heterogeneity has an influence on type and incidence of SPM. In a retrospective cohort, survivors that had heterozygous oncogenic RB1 gene variants with partial functional inactivation or with whole gene deletions showed a lower risk for SPM.[18] Moreover, patients heterozygous for recurrent nonsense CpG-transitions showed an increased risk for SPM.[18] These genotype–phenotype correlations could have an impact on surveillance protocols if confirmed in independent data sets. Additional Supporting Information may be found in the online version of this article at the publisher’s web-site. Abbreviations: ALL, acute lymphatic leukemia; AML, acute myeloid leukemia; CI, confidence interval; EBRT, external beam radiotherapy; IR, incidence ratio; MPNST, malignant peripheral nerve sheath tumor; NK-L, acute natural killer cell leukemia; SPM, second primary malignancy; SD, standard deviation 1
Department of Pediatric Hematology and Oncology, University Hospital Essen, Essen, Germany; 2Eye Oncogenetics Research Group, University Hospital Essen, Essen, Germany; 3Institute for Medical Informatics, Biometry and Epidemiology, University Hospital Essen, Essen, Germany; 4German Childhood Cancer Registry, Institute of Medical Biostatistics, Epidemiology and Informatics, University Medical Centre of the Johannes Gutenberg University Mainz, Mainz, Germany; 5Department of Ophthalmology, University Hospital Essen, Essen, Germany; 6Department of Radiotherapy, University Hospital Essen, Essen, Germany; 7Institute of Human Genetics, University Hospital Essen, Essen, Germany Conflict of interest: Nothing to declare. Karl-Heinz J€ockel and Dietmar R. Lohmann shared senior authorship.
Correspondence to: Petra Temming, Department of Pediatric Hematology and Oncology, University Hospital Essen, Hufelandstrasse 55, 45122 Essen, Germany. E-mail: [email protected] Received 2 February 2015; Accepted 31 March 2015
1800
Temming et al.
Here we present the results of a follow-up study focusing on the incidence of pediatric SPM in a cohort of retinoblastoma survivors aged 6–27 years. Exact knowledge of the genetic cause in a large proportion of these patients allowed us to investigate the impact of the type of RB1 mutation along with the treatment factors for this age group. Because of the change in treatment strategy, this cohort includes children treated with primary chemoreduction, primary EBRT, as well as a combination of both. The study aims to obtain detailed knowledge about the clinical appearance of pediatric SPM in view of the molecular genetic information. This may influence the design of therapeutic regimens and allow tailoring long-term follow-up schedules.
METHOD Patient Cohort The university hospital in Essen is the national referral center for retinoblastoma in Germany treating more than 90% of German children with retinoblastoma. Ethics approval was obtained from the ethics committee of the University of Duisburg-Essen, Germany. All survivors of retinoblastoma treated at the German referral center with a current age between 6 and 27 years were invited to participate in a follow-up study (n ¼ 656). We obtained data on SPM from 488 patients (74.4%). This includes data of two patients who were deceased due to SPM. The other causes of death were metastatic or trilateral retinoblastoma (n ¼ 8), accident (n ¼ 1), and infection (n ¼ 1). In the group without data, 117 were non-contactable, 41 refused to participate, and 10 were deceased (Fig. 1). The cohorts of participants and non-participating patients were comparable regarding gender, laterality, and median age at diagnosis, but the time of follow-up was significantly longer in the group of participants (Supplementary Table SI).
Data Acquisition Data on diagnosis of SPM were collected during an examination at the university hospital and/or a standardized telephone interview. Data were validated by contacting the physicians in charge after obtaining informed consent. Data on retinoblastoma treatment were extracted from patient’s records. Four treatment groups were defined. All patients without treatment, with enucleation alone or
focal treatment alone (laser coagulation, cryotherapy, or brachytherapy) were included in group 1. Patients in group 2 were treated with chemotherapy alone, in group 3 with EBRT alone and in group 4 with chemotherapy and EBRT. EBRT was performed according to the previously published technique.[19,20] Chemotherapy was defined as adjuvant chemotherapy, chemoreduction, thermochemotherapy, or combination of these. The standard chemotherapy regimen in Germany consists of six cycles of chemotherapy resulting in cumulative doses of 1.2 g/m2 carboplatin, 4.8 g/m2 cyclophosphamide, 9 mg/m2 vincristine, and 1.8 g/m2 etoposide. Genetic testing was performed on blood and tumor tissue on the request of individuals or their legal guardians with the aim to identify the oncogenic alterations of the RB1 gene. Mutation identification analysis was performed on DNA from fresh-frozen tumor samples or DNA from blood as reported previously.[21–24] This included one or more of the following methods: analysis of allele loss in tumors, cytogenetic analysis, denaturing high performance liquid chromatography, exon-by-exon sequencing, multiplex ligation-dependent probe amplification, methylationsensitive PCR, quantitative fluorescent multiplex PCR, quantitative real-time PCR, and single strand conformation polymorphism. Diagnosis of mutational mosaicism was based on the finding of a skewed signal ratio of the variant to the normal allele relative to the ratio obtained from heterozygous DNA. In patients with mosaicism, DNA from a second tissue such as buccal mucosa was analyzed, if available.
Data Analysis The data were subjected to statistical analyses with IBM SPSS Statistics version 22.0 (Chicago, Illinois) and SAS 9.4 (SAS Institute, Cary, North Carolina). Data were truncated at 18 years of age because of the focus on pediatric SPM. The two-tailed Fisher’s exact test was applied for group comparison of categorical data (gender, family history of retinoblastoma, CpG transition). The student’s t-test was used to compare equality of the means for unrelated subgroups of continuous variables (age at diagnosis, number of retinoblastoma at diagnosis). Kaplan–Meier survival analysis was used to estimate survival and cumulative incidence. Standardized Incidence Ratio (SIR) was calculated using SAS 9.4 according to Breslow and Day, the confidence interval (CI) was calculated according to the Byar‘s approximation.[25] SIR was calculated separately for sarcoma and leukemia by comparing the
Patients treated for retinoblastoma 644 survivors 6-27 years of age excluded because no contact (117) no consent (41)
12 deceased
10 w/o information on SPM 486
2
488 patients with information on SPM
Fig. 1. Patient accrual. SPM, second primary malignancy; w/o, without. Pediatr Blood Cancer DOI 10.1002/pbc
Pediatric Second Malignancies After Retinoblastoma observed number of SPM with the expected number from data of the German childhood cancer registry (DKKR). Acute lymphocytic and acute myeloid leukemia were combined. Malignant peripheral nerve sheet tumor, rhabdomyosarcoma, leiomyosarcoma, and osteosarcoma were referred to as sarcomas. Time interval was defined as the time from date of retinoblastoma diagnosis to the date of SPM. As only reference data up to the age of 14 were available, we truncated the data at 15 years for SIR analysis. These analyses were performed on the subset of subjects with a heterozygous RB1 gene mutation. P-values
