Abstract
Aim: Analysis of risk factors for survival in long-term follow-up of children treated at a single pediatric center in Poland. Patients and Methods: Out of 623 children diagnosed with cancer between 1995-2005, 110 were treated for brain tumors and followed-up, with a mean survival of 11.4 years. Results: Overall 5-year survival in the whole cohort was 60.9±4.7%, while 10-year survival was 58.2±4.7%. No relapse, progression or death occurred after six years from initial diagnosis. Survival was 48.1±9.6% for patients with medulloblastoma and primitive neuroectodermal tumors; 83.3±6.2% for low-grade astrocytoma; 56.6±16.6% for ependymoma, while 0% at 72 months for high-grade glioma. Patients with cerebellar tumors had a survival rate 69.0±7.1% at 10 years. Multivariate analysis showed that factors predicting poor outcome were: grade III-IV tumor, incomplete surgical resection, and complications after surgical resection, while diagnosis of low-grade glioma was the only factor predicting good outcome. Progression of the disease during therapy was an additional independent adverse risk factor for survival. Conclusion: Long-term survival was achieved by 58% of children with brain tumors. Advanced tumor stage, incomplete surgical resection, complications of surgical treatment, and progression of the disease during treatment predicted poor outcome.
Central nervous system (CNS) tumors are the second most common type of cancer among children and the most frequent solid tumor in childhood in Europe and in North America (1, 2). They affect approximately 16-27% of all children with malignant diseases, and the annual incidence is 25 cases per one million children (1, 2). Worldwide, they account for 8-15% of pediatric cancer diagnoses and the third most frequent type of childhood cancer, after leukemia (30%) and lymphoma (15%) (3). One third of CNS tumors are diagnosed before three years of age. The incidence of this type of tumor has been progressively increasing, and survival has improved less than for other types of cancer (1-3). After the first year of life, CNS tumors are regarded the most common cause of death due to cancer in adolescence (2). Over the last two decades, adjuvant radiotherapy and chemotherapy were introduced into the therapy of brain tumors in children.
The objective of this study was the analysis of risk factors for survival in long-term follow-up of children treated in a single pediatric center.
Patients and Methods
This retrospective cross-sectional study was performed for the Kujawy-Pomerania region in Poland, which includes 2.1 million inhabitants. Data were collected for all 623 consecutive pediatric patients diagnosed with neoplastic disease between 1995-2005. Diagnosis of brain tumor was made according to the World Health Organization (WHO) classification (4). Patients with secondary or metastatic CNS tumors were excluded from the study. The frequency of brain tumors in the Department was determined based on data reported to the National Cancer Registry over a period of 11 years. The incidence rate of brain tumors in the Kujawy-Pomerania region was determined with regard to the size of population provided by the National Statistic Office in Bydgoszcz. Patients were followed-up until July 2013.
Statistical analysis. Descriptive statistics are used to show the patients' general characteristics. Percentages are reported for categorical variables, median and ranges for continuous variables; overall survival (OS) was calculated from the date of diagnosis to the date of death from any cause, or to the date of the latest follow-up. Death due to any cause was considered as an event. The probabilities of OS were estimated by the Kaplan–Meier method using the log-rank test. The impact of the following variables was analyzed: age, tumor localization, clinical stage, histological diagnosis, degree of surgical resection, post-operative complications (i.e. intracranial hematoma, cerebral edema, meningitis, consciousness clouding, mutism), compliance to treatment protocol, and response to therapy (i.e. remission, stable disease, progression/relapse, death). Cox's proportional-hazards regression model was used to correlate each potential prognostic factor with survival in univariate analysis. The factors that appeared to be important were then fitted together in multivariate analysis and dropped one at a time in a backward stepwise manner using the likelihood ratio test at a 0.05 level until all factors in the model were significant. A final check was made to ensure that no excluded factors would improve the fit. A p-value below 0.05 was regarded as statistically significant.
Results
Demographic data. Between 1995 and 2005, a total of 623 children aged up to 18 years were diagnosed and treated in our Department. Primary brain tumors were diagnosed in 110 (17.7%) patients, including 49 females and 61 males, aged 0-18 (median 9.8) years. According to the WHO classification, these included the following diagnoses: astrocytic tumors, low-grade glioma (LGG) in 36 (32.7%) patients and high-grade glioma (HGG) in nine cases (8.2%); embryonal tumors [medulloblastoma (MBL) and primitive neuroectodermal tumor, (PNET)] in 27 (24.5%) cases; ependymoma in 9 (8.2%) cases; in 8 (7.3%) cases pineal (n=3) and germinal (n=5) tumors (Table I). Other diagnoses were made in 19.1%, including non-malignant tumors or tumors non-verified histologically (e.g. brain stem tumors).
Morbidity expressed as incidence rate (IR) of brain tumors in the area of Kujawy and Pomerania was 19.3/106, and varied in age groups from 22.3/106 (5-9 years), to 18.3/106 (10-14, and 0-4 years), and 17.9/106 (15-18 years).
Survival analysis. A total of 64/110 (58.2%) patients were alive at the end of the follow-up, with a mean survival of 11.4 years (95% confidence intervals, CI=9.8-13.0 years); 37 had died due to disease relapse/progression, while 9 died due to therapy complications. Overall 5-year survival for the whole cohort was 60.9±4.7%, while 10-year survival was 58.2±4.7%. No relapse, progression or death occurred after six years from initial diagnosis. Survival was 48.1±9.6% for patients with medulloblastoma and PNET; 83.3±6.2% for those with low-grade astrocytoma; 56.6±16.6% for those with ependymoma (Figure 1). We found a lower survival rate for patients with high-grade gliomas: 55.6±16% at 12 months, 11.1±10% at 24 months and 0% at 72 months. Patients with brain stem tumors also had a low survival rate, 36.8±11.1% at 24 months. Patients with cerebellar tumors, in contrast, had a survival rate of about 73.8±6.8% at 5 years and 69.0±7.1% at 10 years.
Risk factor analysis of survival. By univariate analysis, factors predicting poor outcome were: brain stem localization, HGG diagnosis, grade III-IV tumors according to the WHO classification, incomplete surgical resection, complications after surgical resection, and progression of the disease during therapy, while diagnosis of LGG was the only factor predicting for good outcome (Table II). Other factors such as age<3 years, diagnosis of medulloblastoma or PNET, cerebellar localization, presence of metastases or therapy modifications had no prognostic impact on overall survival. By multivariate analysis, factors predicting poor outcome were: grade III-IV tumor according to the WHO classification, incomplete surgical resection, complications after surgical resection, and progression of the disease during therapy, while diagnosis of LGG was the only factor predicting for good outcome (Table II). When progression during therapy was taken as a separate factor in the univariate analysis, it strongly predicted therapy failure (hazard ratio, HR=7.3; 95% CI=3.5-15.4; p<0.001). When the additional prognostic model was constructed with this factor, progression during therapy was the strongest independent adverse risk factor for OS (HR=5.6; 95% CI=2.5-12.3; p<0.001).
Patients' characteristics.
Discussion
This study was aimed analyze factors influencing OS in children with brain tumors diagnosed and treated at a single pediatric Polish center at long-term follow-up. The main independent adverse prognostic factors for the primary outcome for the whole group were tumor grade III-IV, incomplete surgical resection and complications after surgical treatment. Diagnosis of HGG and brain stem tumors were significant adverse risk factors only in univariate analysis. Another finding of our study is that progression of the disease during therapy (chemotherapy or radiotherapy) had a highly negative impact on survival. This observation would seem to be an obvious factor, however its value was not well-documented, and rather presented as ‘salvage therapy’ (5-7). Diagnosis of LGG was a positive prognostic factor, but other variables did not have any statistically significant effect on survival of the cohort overall.
Probability of overall survival with respect to diagnosis.
Risk factor analysis for overall survival.
Tumor grade, site and histopathological type usually have an important impact on surgical resection, and further therapy. In the study of 6,212 patients younger than 20 years at diagnosis of glioma between 1973-2005 in the USA, the survival of children with gliomas was influenced by tumor grade, age <3 years, histological subtype (i.e. HGG), and extent of resection (8).
Surgery other than gross total resection was an independent adverse prognostic factor in our study, in agreement with other reports in the literature (8-11). The complications after surgical treatment were usually not analyzed by other study groups. We found that intracranial hematoma, cerebral edema, meningitis, consciousness clouding, or mutism have negative impact on outcome due to delay of further treatment, necessity for therapy modifications, or permanent disabilities. Impact of these complications has been shown only in a few studies (12-15).
We did not analyze impact of chemotherapy and radiotherapy as these modalities are commonly used depending on histological diagnosis in most national protocols (16, 17). Radiotherapy is associated with better survival of children (17, 18). It affects the survival of patients with medulloblastoma and PNET or with brain stem tumors, but not of those with low-grade astrocytoma. The indication for radiotherapy in cases of low-grade astrocytoma is usually limited to patients that did not undergo complete resection or had recurrent disease, which limits the evaluation of the role of radiotherapy (3). The role of chemotherapy for pediatric brain tumors is already well-defined, and its positive effect is well-known, especially in medulloblastoma and PNET (17, 19-21).
In conclusion, we have shown that 10-year survival for children with brain tumors was 58%; it was the best for LGG, and the worst for HGG patients. No relapse, progression or death was observed after six years from initial diagnosis. We found that advanced tumor stage, incomplete surgical resection, complications of surgical treatment and progression of the disease during treatment predicted poor outcome.
Footnotes
-
↵* These Authors contributed equally to this study.
- Received October 12, 2013.
- Revision received November 13, 2013.
- Accepted November 15, 2013.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.