For virtually every patient with colorectal cancer (CRC), hematoxylin–eosin (HE)–stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can extract prognosticators directly from these widely available images. We hand-delineated single-tissue regions in 86 CRC tissue slides, yielding more than 100,000 HE image patches, and used these to train a CNN by transfer learning, reaching a nine-class accuracy of >94% in an independent data set of 7,180 images from 25 CRC patients. With this tool, we performed automated tissue decomposition of representative multitissue HE images from 862 HE slides in 500 stage I–IV CRC patients in the The Cancer Genome Atlas (TCGA) cohort, a large international multicenter collection of CRC tissue. Based on the output neuron activations in the CNN, we calculated a “deep stroma score,” which was an independent prognostic factor for overall survival (OS) in a multivariable Cox proportional hazard model (hazard ratio [HR] with 95% confidence interval [CI]: 1.99 [1.27–3.12], p = 0.0028), while in the same cohort, manual quantification of stromal areas and a gene expression signature of cancer-associated fibroblasts (CAFs) were only prognostic in specific tumor stages. We validated these findings in an independent cohort of 409 stage I–IV CRC patients from the “Darmkrebs: Chancen der Verhütung durch Screening” (DACHS) study who were recruited between 2003 and 2007 in multiple institutions in Germany. Again, the score was an independent prognostic factor for OS (HR 1.63 [1.14–2.33], p = 0.008), CRC-specific OS (HR 2.29 [1.5–3.48], p = 0.0004), and relapse-free survival (RFS; HR 1.92 [1.34–2.76], p = 0.0004). A prospective validation is required before this biomarker can be implemented in clinical workflows. In our retrospective study, we show that a CNN can assess the human tumor microenvironment and predict prognosis directly from histopathological images.

Risk of colorectal cancer (CRC) is considerably higher in men compared to women; however, there is inconclusive evidence of sex differences in CRC prognosis. We aimed to assess and explain sex differences in 5-year relative survival using standard and model-based period analysis among 164,996 patients diagnosed with CRC from 1997 to 2006 and reported to 11 German cancer registries covering a population of 33 million inhabitants. Age-adjusted 5-year relative survival was higher in women (64.5% vs. 61.9%, P<0.0001). A substantial survival advantage of women was confirmed in multivariate analysis after adjusting for CRC stage and subsite in subjects under 65 years of age (relative excess risk, RER 0.86, 95% CI 0.82–0.90), but not in older subjects (RER 1.01, 95% CI 0.98–1.04); this pattern was similar in the 1st and in the 2nd to 5th year after diagnosis. The survival advantage of women varied by CRC stage and age and was most pronounced for localized disease (RERs 0.59–0.88 in various age subgroups) and in patients under 45 years of age (RERs 0.59, 0.72 and 0.76 in patients with localized, regional or advanced disease, respectively). On the contrary, sex differences in survival did not vary by location of CRC. In conclusion, our large population-based study confirmed a survival advantage of female compared to male CRC patients, most notably in young and middle aged patients and patients with localized disease. The effect of sex hormones, either endogenous or through hormonal replacement therapy, might be the most plausible explanation for the observed patterns.

Background. Breast cancer survivors may experience adverse effects of cancer and/or treatment years after completion of therapy, which can considerably decrease quality of life (QoL). Little is known about the time course of QoL in breast cancer survivors beyond the fifth year post-diagnosis, when routine follow-up care has usually terminated. We therefore explored in detail whether and to what extent restrictions in breast cancer survivors persist and whether changes or aggravations in QoL occurred over time. Material and methods. QoL was assessed 1, 3, 5, and 10 years post-diagnosis in a population-based cohort of initially 387 female breast cancer patients from Saarland (Germany), using the EORTC QLQ-C30 and QLQ-BR23. Time course of QoL over 10 years post-diagnosis was assessed for survivors and survivors’ QoL was compared cross-sectionally to the German general population after adjustment for age. Results. A total of 182 out of 238 patients alive (76.5%) responded in the 10-year, 160 patients (67.2%) participated in all follow-ups. Although breast cancer survivors and controls reported comparable general health and overall QoL, survivors reported significantly more restrictions on most functioning and symptom scales at each follow-up. Detriments in various QoL dimensions (e.g. physical and social functioning; pain, financial difficulties) aggravated from year 5 to 10. Generally, restrictions were largest for the youngest survivors. Conclusion. Relevant restrictions in QoL persist over years in breast cancer survivors and affect predominantly younger women. The aggravation of restrictions in QoL beyond the fifth year may indicate deficits in health care and psychosocial support of breast cancer patients after completion of routine follow-up care.