Second Malignancies in Children Following Treatment for Neuroblastoma

Introduction

Treatment for childhood cancer has become progressively more effective in the last few decades leading to a remarkable increase in both 5- and 10-year overall survival for these patients, which now approaches 80% overall ( ). However, the mortality due to recurrence or progression of the original tumor after these time periods remains a relevant problem ( ). In addition, long-term survivors of cancer diagnosed in childhood do have a significant risk of developing multiorgan complications and psychosocial problems ( ). Finally, the chance that a second malignant neoplasia (SMN) might develop over time in these individuals has led to great concern ( ). Improvement in survival has also involved children affected by neuroblastoma, although it has been less evident compared to other childhood malignancies, particularly leukemias, lymphomas, and nephroblastoma ( ). In the experience of the Italian Neuroblastoma Group, the cure rate of neuroblastoma patients has increased from 35% in the early 1980s to a remarkable 64% recorded between 2000 and 2005 ( ). Even for neuroblastoma survivors, the occurrence of SMNs is no longer considered an exceptional event, although it remains less frequent than for leukemias and lymphomas ( ). This may be due in part to the peculiar clinical behavior of neuroblastoma itself, including the ability to undergo spontaneous regression and the chance of being cured by surgery alone. Furthermore, it has to be taken into consideration that treatment for neuroblastoma has considerably changed over time following a number of remarkable discoveries regarding the tumor’s clinical and biological features ( ). However, the correlation between the occurrence of an SMN in children diagnosed with neuroblastoma and the previously administered treatment remains unclear.

We report our personal experience and the results of a search of the literature to describe the incidence and type of SMNs in neuroblastoma as related to the type of treatment, possible genetic implications, clinical implications and survival after the diagnosis of an SMN.

Risk of SMN in Children Treated for Neuroblastoma

The British Childhood Cancer Survivor Study analyzed a large population-based cohort of 5-year survivors of childhood cancer who had been diagnosed before 15 years of age between 1940 and 1981 in Great Britain. Among 17 981 such patients who had been followed up for a median of 24.3 years, a total of 1354 developed an SMN. The most frequently observed SMNs were brain tumors, non-melanoma skin cancer, and gastrointestinal, genitourinary and breast carcinomas, as well as bone tumors. The standardized incidence ratio (SIR) for these patients was four times more than expected, while it was 2.8 times higher for neuroblastoma patients ( ). A similar investigation was carried out in the USA by the Childhood Cancer Survivor Study (CCSS). Of 14 359 5-year survivors of childhood cancer diagnosed between 1970 and 1986 and followed up for a median of 30 years, a total of 1402 developed at least one SMN for a total of 2703 neoplasms with a cumulative incidence of 20.5% at 30 years from the first malignancy and an SIR of 6.0. The highest SIR was found for Hodgkin’s lymphoma (8.7; 95% CI, 7.7–8.8) and Ewing’s sarcoma (8.5; 95% CI, 6.2–11.7), while it was 6.9 for neuroblastoma (95% CI, 4.9–9.7), ( ). The CCSS also reported data on a selected group of 954 neuroblastoma patients who had survived more than 5 years from diagnosis. The cumulative incidence of any relevant chronic health condition for these neuroblastoma patients was 41% with an 8.3-fold increased risk as compared to their siblings. Nearly 6% of these patients died, most often of disease recurrence (n=43, 5% of all deaths) and SMN (n=13). Cumulative mortality was associated with older age at diagnosis. The cumulative incidence of SMN after 5 years from diagnosis was 3.5% at 25 years from diagnosis. The most frequent SMN was thyroid cancer (n=8) followed by renal cell carcinoma (n=5), ( ). The real incidence of secondary leukemias in these reports is likely underestimated, as these studies focused on SMNs that occurred after 5 years of follow up. However, a number of case reports and even large series have described instances of secondary leukemias that occurred within a shorter interval. In a French series of 61 cases of secondary leukemias detected at 0.8–12.8 years (median, 3.4) from an initial diagnosis of solid tumors, nine occurred in patients with neuroblastoma. Three of these nine patients, in whom leukemia occurred at 16, 17 and 27 months from diagnosis, respectively, had intermediate risk factors and had been treated with standard-dose chemotherapy ( ). In addition, Kushner reported eight cases of secondary leukemia that occurred at 12–50 months (median, 16) from the initial diagnosis in a group of 184 high-risk neuroblastoma patients treated with dose-intensive induction chemotherapy ( ). Finally, at least two cases of simultaneous occurrence of neuroblastoma and acute myeloid leukemia have been reported ( ).

In the experience of the Italian Neuroblastoma Registry (Garaventa, unpublished), based on more than 3000 children and adolescents diagnosed with neuroblastoma between 1979 and 2011, a total of 27 patients developed an SMN, including one who also developed a third tumor ( Table 6.1 ). SMN or myelodysplasia occurred at a median interval of 10 years from the primary diagnosis. Thyroid carcinoma was the most frequently observed SMN (n=11). A possible explanation for this high incidence may be the aggressive treatment we administered to these patients and, in particular, to the stage 4 patients that were diagnosed after 1985 and who underwent hematopoietic stem cell transplantation (HSCT) and total body irradiation (TBI). A second reason could be related to the different attitudes in the screening policies that were adopted to detect SMNs in these patients. For example, in several Italian centers, the thyroid gland is evaluated by ultrasonography every year and nodules>1 cm in diameter are routinely biopsied thus making it possible to detect some tumors that would otherwise never be detected. Finally, we did not observe any SMNs of the gastrointestinal or genitourinary tracts, likely because a number of patients were lost to follow up after entering adulthood.

Incidence of Selected SMNs