RESUMEN
The aim of this study was to investigate how breast cancer follow-up in the Netherlands changed during the COVID-19 pandemic, compared to 2018-2019, and to what extent follow-up during the pandemic corresponded to the patient risk of recurrence. During the early phase of the pandemic the Dutch Society for Surgical Oncology (NVCO) issued a report with recommendations on how follow-up could be postponed, as a guidance for the pandemic, based on a low, intermediate or high risk of recurrence. In this study we investigated to what extent this advice was followed. A dataset of 33160 women diagnosed with primary invasive breast cancer between January of 2017 and July of 2021 was selected from the Netherlands Cancer Registry (NCR) and Dutch Hospital Data (DHD). The pandemic, 2020 and weeks 1-32 of 2021, was divided into six periods (A to F), based on the number of hospitalized COVID patients in the Netherlands. The five-year risk of locoregional recurrence (LRR) was determined for each patient with the INFLUENCE nomogram. The LRR risk was compared to the risk groups from the NVCO report with a Kruskal-Wallis test. The percentage of patients who received a mammogram during period A to F was compared to the same periods of 2018-2019 with a chisquared test. Correlation between the LRR risk, and if patients had a mammogram, was investigated with logistic regression. This analysis was repeated separately for the risk groups. Correlation between the LRR risk, and time intervals between surgery and the first and second mammogram was analyzed using cox proportional hazard models, this was also repeated for the risk groups. There was a significant difference in LRR risk between the NVCO risk groups. In the low-risk group (n=7673), 86 patients (1.1%) had a risk >5%. In the intermediate risk group (n=19197), 18364 patients (95.7%) had a risk of < 5%, and 65 patients (0.34%) had a risk of >10%. In the high-risk group (n=2674), 2365 patients (88.4%) had a risk < 10%. The percentage of patients who received a mammogram was significantly lower in periods B to F of the pandemic. Logistic regression showed a negative correlation between the risk of LRR and if patients had a mammogram in 2020 (OR 0.93) and 2021 (OR 0.93). There was also a negative correlation between the risk groups and mammography in 2020 (OR 0.92 for intermediate and 0.80 for high), and for the risk groups and mammography in 2021 (OR 0.98 for intermediate and 0.95 for high). There was no significant impact of LRR risk, or risk group, on time intervals between mammograms. During the pandemic, patients with a higher LRR risk, or a higher risk according to NVCO advice, had lower odds of having a mammogram. If the advice would have been followed, in 0.5% of the patients scheduled for follow-up, the recommendation was to postpone in contrast to a high estimation of the individual risk. For 62.7%, a follow-up was recommended, despite a low estimated individuals risk. Because the number of high-risk patients is relatively low, individual risk prediction could be supportive, in case of future restrictions. This way the high-risk patients can be identified and prioritized for follow-up, and can also be encouraged to come to the hospital.
RESUMEN
Introduction: The burden of the COVID-19 pandemic forced the Dutch health care services to discontinue their national screening programs on 16 March 2020 (week 12). For breast cancer, the program invites women aged 50-74 years for biennial screening mammography. From mid-June 2020 (week 25) the breast cancer screening program was resumed, albeit with reduced capacity (max 60%). We aimed to investigate the impact of resuming the screening program on incidence, tumor-, and T-stage of screen- and non-screen-detected ductal carcinoma in situ (DCIS) and invasive breast cancer (IBC) in the Netherlands. Methods: Women, 50-74 years of age, diagnosed with DCIS or IBC in weeks 2-35 of 2018, 2019 and 2020 were selected from the Netherlands Cancer Registry. Weekly incidence of screen- and non-screen-detected tumors was calculated and expressed per 1 million women aged 50-74 years living in the Netherlands. Weeks 2-35 were divided in seven periods, representing separate phases of the pandemic. For each period, incidence of screen- and non-screen-detected tumors was stratified by clinical tumor stage (TNM) and clinical T-stage. Incidence in each period of 2020 was compared with the incidence in the same period of 2018/2019 (averaged). Results: In weeks 2-35, 7,250 patients were diagnosed in 2018, 7,299 in 2019, and 5,300 in 2020. During weeks 2-12 2020 the weekly average incidence of screen-detected tumors was 42/1 million women aged 50-74. In week 13 incidence dropped to 19, and was almost zero during weeks 14-25. Incidence increased to 4 in week 26, when the screening was gradually restarted, and increased to an weekly average of 24 tumors/1 million in weeks 30-35, when screening had restarted in most of the Netherlands. During weeks 14-16, 17-25, 26-29, and 30-35 2020 incidence of screen detected DCIS, stage I-II tumors and T0-2 tumors was lower than in the same period of 2018/2019. Distribution of tumor- and T-stage in newly diagnosed tumors did not differ between the periods (24% DCIS, 57% stage I, 18% stage II, 1% stage III, 1% stage IV;24% T0, 59% T1, 15% T2, 2% T3, 0% T4). During weeks 2-11 2020 the weekly average incidence of non-screen-detected tumors was 38/1 million women aged 50-74. Incidence dropped to 16 in week 14, and increased to an average of 37 tumors per week in weeks 17-35. During weeks 17-25, 26-29, and 30-35 the incidence and distribution of non-screen-detected DCIS, stage I-IV and T0-4 tumors was comparable with 2018/2019 (weekly incidence (distribution): 3 DCIS (9%), 15 stage I (40%), 13 stage II (36%), 3 stage III (7%), 3 stage IV (8%);3 T0 (9%), 17 T1 (45%), 12 T2 (34%), 3 T3 (8%), 2 T4 (5%)). Conclusion: The temporary suspension of the breast cancer screening program reduced the incidence of breast cancer. After resuming screening the incidence did not raise above the incidence in 2018/2019, therefore it is expected that the incidence will rise during the next months. The results to date did not show a shift towards a higher tumor stage or T-stage.