Increased incidence of second primary malignancy in patients with malignant astrocytoma: a population-based study

We identified patients diagnosed with malignant astrocytoma (MA) as the first of two or more primary malignancies between 1973 and 2015 from Surveillance, Epidemiology and End Results (SEER) database. Multiple primaries-standardized incidence ratio (MP-SIR) was calculated to quantitate the risk of second primary malignancy (SPM). We further identified the risk factors of developing SPM and factors affecting overall survival (OS) in MA patients with SPM. Our results revealed that overall risk of SPM among MA patients was significantly higher than that in general population (SIR: 1.09, 95% confidence interval (CI): 1–1.18, P<0.05). Specific sites where the risk of SPM increased included salivary gland, bone and joints, soft tissue including heart, brain, cranial nerves other nervous system, thyroid, acute non-lymphocytic leukemia and acute myeloid leukemia. Overall risk of SPM in patients aged ≤29 and 30–59 years significantly increased (4.34- and 1.41-fold respectively). Whereas patients aged ≥60 years had a significantly decreased risk of SPM. Patients in the group of latency at 36–59, 60–119 and ≥120 months carried significantly increased overall risk of SPM. Multivariate analysis revealed that age, race, marital status, WHO grade, differentiated grade of cancer tissues, latency was independent predictor of OS in MA patients with SPM, which were all selected into the nomogram. The calibration curve for probability of survival showed good agreement between prediction by nomogram and actual observation. In conclusion, MA survivors should be advised of their increased risk for developing certain cancers in their lifetime. Our study had clinical implications for the surveillance of MA survivors at risk of developing SPM.


Introduction
Cancer survivors have been increasing due to the improvement of diagnostic modalities and treatment of cancer. Approximately 18 million cancer survivors are expected by 2022 [1]. It is important to figure out the long-term effects of cancer and its treatment among cancer survivors. Second primary malignancy (SPM) is one of the most serious long-term complications in the population of cancer survivors. Previous studies have verified significantly increased risk of SPM after treatment of several primary cancers [2][3][4][5].
The incidence of primary brain tumors has been increasing over the past 30 years, especially in elderly people [6]. Astrocytoma, derived from astrocytic glial cells, accounts for more than 60% of all primary brain tumors [7,8]. According to the current WHO grading system of astrocytoma, which is based on the histopathological characteristics of the underlying tumor tissue, diffuse infiltrating astrocytoma is graded as low-grade diffuse astrocytoma (DA, grade II), anaplastic astrocytoma (AA, grade III) and glioblastoma (GBM, grade IV) in the order of increasing malignancy [9]. The median overall survival (OS) for patients with DA, AA and GBA is 6-8 years, 2-3 years and 10-20 months, respectively [10][11][12][13][14][15].
Multidisciplinary treatment modalities combined surgery, chemotherapy, radiation and targeted therapy have yielded increased survival in patients with astrocytoma [16].
The number of cancer survivor patients with astrocytoma is increasing. Additionally, study on the risk of developing SPM in astrocytoma survivors is still lacking. The current study aimed to evaluate the overall and site-specific risk of developing SPM in patients with malignant astrocytoma (MA). And to investigate the factors affecting OS in MA patients with SPM by using the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) database.

Methods Population
The study population was derived from the SEER database. Dataset of incidence-SEER 9 Regs Research Data, November 2017 Sub  <Katrina/Rita population adjustment> was used for the analysis of multiple primaries-standardized incidence ratio (MP-SIR). Case listing was based on dataset of incidence-SEER 18 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, November 2017 Sub (1973-2015 varying). All patients in the present study were diagnosed with MA by positive histology between 1973 and 2015. Only cases where MA was the first of the two or more primary malignancies were selected. Patients were excluded if the diagnosis of MA was made at autopsy or in the death certificate.

Definition of SPM
SEER definition of SPM took the following factors into account: the site, the behavior, histology and date of diagnosis of the second malignancy. As well as the laterality if it was a paired organ [17]. According to the SEER definition of SPM, the criteria for patients with SPM after the diagnosis of MA were listed as following: (i) a new tumor that contained the same histology as the previous cancer and was diagnosed within 2 months of the previous cancer was considered as a single primary. (ii) Simultaneous multiple tumors with the same histologic type within the brain were regarded as a single tumor, and even when different tumors had different behavior codes. (iii) Multiple tumors with the same histologic type appearing in the brain and in a different site were regarded as multiple primary malignancies unless stated as metastatic tumors. (iv) Multiple tumors of different histologic types within the brain and in a different site were counted as multiple primary malignancies regardless of time.

Statistical analysis
The risk of SPM was presented as standardized incidence ratio (SIR) and its 95% confidence intervals (CI). SIR was calculated by dividing the observed number of SPM cases by the expected number based on general population rates. Risks were considered significant when corresponding 95% CI did not include the null value. Age-adjusted incidence rates were used for comparison to exclude the potential confounding of the differences in age distributions. The analysis was performed in SEER*Stat version 8.3.5 (http://seer.cancer.gov/seerstat/). Multivariable Cox regression was performed to identify factors affected OS in patients with SPM after diagnosis of MA. A P-value <0.05 was considered significant. Analysis was performed using SPSS (Release 23.0, SPSS Inc, IL, U.S.A.). The prognostic nomogram that included all significant independent factors for OS in MA patients with SPM was presented in nomogram, which was performed in R (version 3.5.1; R Foundation).

Risk of developing SPM in MA patients
The overall risk of SPM among MA patients was significantly higher than that in the general population (SIR: 1.09, 95% CI: 1-1.18, P<0.05). Specific sites where the risk of SPM increased included salivary gland, bone and joints, soft tissues including heart, brain, cranial nerves other nervous system, thyroid, acute non-lymphocytic leukemia and acute myeloid leukemia. Additionally, the risk of SPM in liver, breast and prostate was significantly decreased ( Table  1).

Effect of age at diagnosis, gender and latency on the site-specific risk of SPM in MA patients
We further validated the effect of age at diagnosis, gender and latency (interval from the diagnosis of MA to that of SPM) on risk of developing SPM among MA patients. Based on the age at MA diagnosis, patients were divided into three groups: 0-29, 30-59 and ≥60 years. Overall risk of SPM in patients aged ≤29 and 30-59 years significantly increased (4.34-and 1.41-fold respectively). Whereas patients aged ≥60 years had a significantly decreased risk of SPM. As stratified by gender, the overall risk of SPM in males and females did not change among MA patients. The detailed site-specific risk of SPM in each group is shown in Table 2.
Median latency and corresponding interquartile range (IQR) were 22 months and 80 months, respectively. According to latency, patients were divided into five groups: ≤11, 12-35, 36-59, 60-119 and ≥120 months. Patients in the group of latency at 36-59, 60-119 and ≥120 months carried significantly increased overall risk of SPM. Patients in the group of latency at ≤11 and 12-35 months did not have significant change in the overall risk of SPM. Site-specific risk of SPM in each group is shown in Table 3.

Characteristics of MA patients with SPM
A total of 1181 patients with SPM after diagnosis of MA were recorded in SEERStat by case listing function. Characteristics of these patients are shown in Table 4

Factors affecting the OS in MA patients with SPM
As shown in Table 5
at 1, 3 and 5 years after diagnosis showed an optimal agreement between the prediction by nomogram and actual observation ( Figure 2).

Discussion
Given the elevated incidence and survival of MA patients [18], there is an increasing focus on long-term effects of cancer and its related treatment, especially the risk of developing SPM. This population-based cohort study of MA patients strongly indicated that there was a significantly elevated risk of tumor in brain (SIR: 12.36, 95% CI: 10.05-15.03) and cranial nerves (SIR: 16.53, 95% CI: 7.14-32.58). One explanation could be a known consequence of cranial irradiation [19,20]. Another consideration is whether these brain and cranial nerves tumor diagnoses are metastases from the primary malignancy. As stated in the 'Methods' section, we used SEER definition to identify patients with SPM, which could exclude metastatic and recurrent tumors from SPM counts. Previous study evaluated the validity of SPM in public database with a revaluation of diagnoses, which showed that 98% of the second primary diagnoses were correct [21]. Therefore, we can almost get rid of the concern. It was also notable that patients with astrocytoma were at a significantly higher risk of developing malignancy in bones and joints (SIR: 8.3, 95% CI: 3.98-15.26), soft tissue including heart (SIR: 3.35, 95% CI: 1.73-5.86), leukemia (SIR: 2.04, 95% CI: 1.38-2.92) and thyroid (SIR: 1.72, 95% CI: 1.1-2.55). Explanations for why patients with astrocytoma had higher risk of developing above-mentioned malignancies are not quite clear. It is usually attributed to treatment-related factors including radiation and chemotherapy. Temozolomide, widely used for treating high-grade   The present study carried several strengths. First, it was the first study which reported the risk of SPM in MA patients. Second, it was based on a population with large sample size and long-term follow-up. Third, it provided strong evidence for the development of guidelines of surveillance of MA survivors. We analyzed the risk of developing SPM in subgroup of patients with different age, gender and interval from diagnosis of MA to that of SPM, which provided detailed information regarding patients' prognosis and focusing clinical follow-up in each subgroup. For example, our study showed that patients aged ≤29 years had significantly increased risk of developing colon cancer. If a patient understands his/her risk for developing the aforementioned cancer, he/she may choose to alter her treatment to be more aggressive in hopes of reducing his/her risk. He/she may also need to be counselled regarding preventable risk factors, such as weight gain and poor dietary habits. Lack of variables such as radiation, chemotherapy and genetic make-up was the main limitation of the study.

Conclusion
In conclusion, we observed that patients survived from MA were at a significantly higher overall risk of developing SPM. Specific sites where the risk of SPM was increased including brain, cranial nerves thyroid, acute non-lymphocytic leukemia and acute myeloid leukemia. Additionally, excess risks varied with age, gender and interval from diagnosis of MA to that of SPM. Age, race, marital status, WHO grade, differentiated grade of cancer tissues, latency were independent predictors of OS in MA patients with SPM. The results of our study suggested the need for closer surveillance of at-risk MA patients for the development of SPM and for further research into the molecular pathways underlying the occurrence of SPM.