Secondary Logo

Journal Logo

Letters

Incidence Shifts Within Central Nervous System Malignancies

Kopel, Eran

Author Information
doi: 10.1097/EDE.0b013e3182625a6e
  • Free

To the Editor:

Studies of central nervous system (CNS) malignancies in relation to mobile phone use are at the heart of a growing public and scientific debate.15 Cancer registries are common sources of data used by these studies; time trends of anatomic site- or tumor-specific incidence rates, however, could potentially bias results.

For instance, the distribution of tumors could shift over time because of improved imaging techniques that increase the accuracy of tumor diagnosis in specific locations and decrease the number diagnosed in ill-defined locations. Such artifacts of improved imaging could be mistaken for real changes in disease pattern. I describe the dynamics of site- and tumor-specific incidence trends over time in a large cancer registry.

I examined primary CNS malignancies in the US Surveillance, Epidemiology, and End Results (SEER) data6 for time trends of anatomic site-specific incidence between 2000 and 2008. Incidence was defined as the annual number of CNS malignancies in each specific anatomic site, divided by all CNS malignancies in that year, and presented as an annual rate per 1000 CNS malignancies. Tumor-specific incidences were similarly calculated for frequently occurring malignant gliomas. I assessed trends over time and correlations between trends with Spearman's rho. ICD-O-3 is routinely used in the SEER registries for the classification of malignancies by their primary anatomic site (topography codes) and histologic grading (morphology codes), both of which were abstracted for this study under a data-use agreement. The statistical analyses were performed with IBM SPSS version 19.0 (Chicago, IL).

Of the 3,098,562 malignancies registered between 2000 and 2008, 44,089 (≈1%) were of CNS origin: 24% occurred in the frontal lobe, 18% in the temporal lobe, 12% in the parietal lobe, 6% in the cerebellum, 4% in the brainstem, and 3% in the occipital lobe. The poorly specified malignancies approached 25%, with 14% occurring in overlapping anatomic sites (topography code C71.8) and 11% being classified as origin not otherwise specified (topography code C71.9).

Annual site-specific incidence trends (Figure) were positively correlated with time (2000–2008) for frontal lobe tumors (rs = 0.95) and temporal lobe tumors (rs = 0.88), or negatively correlated for overlapping tumors (rs = −0.97). The trend for overlapping tumors for 2000–2008 was negatively correlated with both frontal (rs = −0.88) and temporal (rs = −0.93) lobes tumors.

Figure
Figure:
Annual site-specific incidences per 1000 central nervous system malignancies during 2000–2008 in 17 registries of the United States Surveillance, Epidemiology and End Results (SEER).

The tumor-specific incidence trend of anaplastic and low-grade astrocytomas combined (morphology codes 9400/3 and 9401/3, respectively, mean annual incidence 117.6 per 1000 CNS malignancies), was negatively correlated with the trend of glioblastoma multiforme (morphology code 9440/3, mean annual incidence 453.4 per 1000 CNS malignancies) (rs = −0.87). The trend for glioblastoma multiforme increased with time (rs = 0.93), whereas the trend for anaplastic and low-grade astrocytomas declined with time (rs = −0.87). Mixed and unspecified malignant gliomas (morphology code 9380/3) had a mean annual incidence of 57.8 per 1000 CNS malignancies, without substantial correlations with time or with either of the other groups of gliomas.

The interchangeable dynamics in the site- and tumor-specific incidence rates of malignant gliomas over time could be derived, for example, from higher diagnostic sensitivity, accompanied by lower specificity and higher false-positive rate, of the imaging and tumor-classification tools used over time to diagnose high-grade gliomas in the temporal and frontal lobes. Such phenomena might introduce into epidemiologic studies a nondifferential misclassification bias of site- and tumor-specific outcomes that could distort the measured association and thus should be accounted for.7

Eran Kopel

Tel Aviv, Israel [email protected]

REFERENCES

1. Ahlbom A, Feychting M, Green A, Kheifets L, Savitz DA, Swerdlow AJ. Epidemiologic evidence on mobile phones and tumor risk: a review. Epidemiology. 2009;20:639–652.
2. INTERPHONE Study Group. Brain tumour risk in relation to mobile telephone use: results of the INTERPHONE international case-control study. Int J Epidemiol. 2010;39:675–694.
3. Hardell L, Carlberg M, Hansson Mild K. Re-analysis of risk for glioma in relation to mobile telephone use: comparison with the results of the Interphone international case-control study. Int J Epidemiol. 2011;40:1126–1128.
4. Frei P, Poulsen AH, Johansen C, Olsen JH, Steding-Jessen M, Schüz J. Use of mobile phones and risk of brain tumours: update of Danish cohort study. BMJ. 2011;343:d6387.
5. Deltour I, Auvinen A, Feychting M, et al.. Mobile phone use and incidence of glioma in the Nordic countries 1979–2008: consistency check. Epidemiology. 2012;23:301–307.
6. Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER 17 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2010 Sub (1973–2008 varying)—Linked To County Attributes—Total U.S., 1969–2009 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2011 (updated 10/28/2011), based on the November 2010 submission. Accessed December 27, 2011.
7. Greenland S, Lash TL. Bias analysis. In: Rothman KJ, Greenland S, Lash TL eds. Modern Epidemiology. 3rd ed. Philadelphia, PA: Lippincott Williams and Wilkins; 2008:345–380.
© 2012 Lippincott Williams & Wilkins, Inc.