gms | German Medical Science

Introduction: The Cancer Registry of Rhineland-Palatinate has been recording cancers epidemiologically since 1998 and clinically and epidemiologically since 2016.

In 2023, the WiZen Comparative Cohort Study [1] published an evaluation of overall survival of cancer patients treated initially at certified centers versus other hospitals. Using data from 4 clinical cancer registries (2006-2017) in eastern and southern Germany and health insurance data, the results often showed better survival with treatment in certified centers. In this study, the cancer registry wanted to examine the transferability of the cancer registry results to Rhineland-Palatinate and therefore reproduced similar multivariate analyses for overall survival for patients initially treated in this federal state. Preliminary results are presented.

Methods: Definition for treatment in certified centers: patient was treated at least 1 year after start of certification; patients treated in center during first year of certification were excluded from analysis; initial operation (OP) was performed at the center or (if not reported), the diagnosis was done at the center. The comparison group consists of all other patients with OP or diagnosis outside certified centers. Multivariate survival analyses (Cox-regression) were calculated for patients with colorectal cancer (ICD-10: C18/C19/C20, n=14.298), prostate cancer (ICD-10: C61, n=17.003) and breast cancer (ICD-10: D05/C50, n=22.125). An analysis for lung cancer is planned (C34). These are the most common cancers in Rhineland-Palatinate. Hazard ratio (HR) effects were estimated for center vs. non-center (and covariates). As done by WiZen, no model selection was made, but full models with all covariates were used for calculation. Sensitivity analyses of subpopulations using Cox models and stratification were performed to check for heterogeneity (i.e. nonproportional effects). Covariates where ICD-10, UICC-stage, age, grade/gleason-grade (C61), sex (C18-C20), TNM_v, TNM_l (i.e. lymph-node- and vene-invasion), time-effect (2016-2018/2019-2022).

Results: Cox model analysis for colorectal cancer showed significant decreased center effects (<1), i.e. initial treatment in centers could be beneficial (HR:0.96, 95%-CI:[0.87;0.98]). For prostate cancer, HR for centers where non-significant (HR:1.02, 95%-CI:[0.92;1.12]), for breast cancer the estimates were significantly larger than 1 (HR: 1.147, 95%-CI:[1.06;1.25]). In most cases, the full models showed some non-proportional effects, but Cox-regression in subpopulations with proportionality showed analogue results. Some sub analyses showed more beneficial effects of initial treatment in certified centers for subpopulations of breast and prostate cancer.

Conclusions/outlook: This initial evaluation indicates similar results for survival after colorectal cancer here and in WiZen study, but different results were achieved for breast and prostate cancer in Rhineland-Palatinate. One reason for these differences could be that prostate and breast cancers have a good prognosis in lower (most common) stages. Especially considering that differences in treatment of cancer patients in certified centers vs. other facilities should occur at intermediate stages (UICC II and III), longer follow-up may be necessary to detect differences in overall survival of prostate and breast cancer patients. Clinical cancer registration started at different times in the federal states, there may be differences in data collection or treatment. Further analysis in different regions should be considered to further investigate the observed differences.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.