Page 82 - New epidemiological and PSMA-expression based paradigms in salivary gland tumors
P. 82

80
Chapter 5
person-time data. Statistical analysis was performed using STATA 13 (StataCorp LP, College Station, USA).
Results
Our study included 3,650 women: 1,567 women with SGC and 2,083 women with SGPA, translating into a yearly mean of 60 and 417 women respectively (Table 1; Supplementary table 2). The overall median age at diagnosis of the salivary gland tumor was 56 years (IQR 45-68). The median follow-up was 7 years (IQR 2-12) after SGC. The median follow-up after an SGPA diagnosis was 6 years (IQR 2-16). During follow-up 52 SGC patients developed BC, of whom 46 (88%) had invasive and 6 (12%) in situ BC (Table 2; Supplementary table 3). Of the SGPA patients 74 developed BC, of whom 68 (92%) had invasive and 6 (8%) in situ BC. Overall, the median age at BC diagnosis was 63 years (IQR 51-74). The median interval between salivary gland tumor and BC diagnosis was 6 years in SGC (range 0-24; IQR 3-9) and 7 years (range 0-20; IQR 3-11) in SGPA. In SGC patients, the 5- and 10-year survival rate after invasive BC was respectively 59.8% (95%CI 42.7-73.4%) and 42.0% (95%CI 24.9-58.1%). For patients younger than 65 years, this was 78.9% (95%CI 52.4-91.7%) and 70.1% (95%CI 40.5-87.0%). The BC receptorstatus could be determined in most cases (Table 2).
Comparison with the general population: The cumulative risk of BC in SGC patients was 5.3% (95% CI 3.8-7.3%) at 10 and 8.2% (95% CI 5.8-11.5%) at 20 years, respectively (Table 3). The comparison to the expected risk based on the general population of 2.9% at 10 years and 6.0% at 20 years (dotted line) is made in figure 1A. The cumulative risk of BC among SGPA patients was 3.5% (95% CI 2.6- 4.6%) at 10 years and 7.2% (95% CI 5.4-9.4%) at 20 years, respectively (Table 3). In comparison, the expected cumulative BC risk, based on age-matched incidence in the general population, was 2.2% at 10 and 5.2% at 20 years, respectively (Figure 1B). Overall, among SGC patients the incidence of BC was 1.59 times (95% CI 1.19-2.09) higher than expected based on general population rates. BC incidence was 1.48 times (95% CI 1.16-1.86) higher than expected among SGPA patients. The SIRs when risk of DCIS and invasive BC in the SGC-cohort were estimated separately were similar (SIR 1.86 for in situ BC; 95% CI 0.68-4.04 and SIR 1.57 for invasive BC; 95% CI 1.14-2.09, respectively). SIRs for BC (invasive or in situ)





























































































   80   81   82   83   84