In colorectal cancer, tumor budding is associated with tumor progression and represents an additional prognostic factor in the TNM classification. Tumor buds can be found at the invasive front (peritumoral budding; PTB) and in the tumor center (intratumoral budding; ITB) of primary tumors. Previous studies have shown that tumor buds are also present in colorectal liver metastases (CRLM). Data on the prognostic and predictive role in this clinical context are still sparse and no standardized approach to evaluate budding in CRLM has been published so far. This study aimed to analyze and correlate perimetastatic (PMB) and intrametastatic budding (IMB) on H&E and pancytokeratin staining, compare it to budding results in corresponding primary tumors and to propose a standardized scoring system in CRLM as the basis for future studies. Tumor tissue of 81 primary tumors and 139 corresponding CRLM was used for ngTMA construction. For each primary tumor and metastasis, two punches from the center and two punches from the periphery from areas with highest tumor budding density were included. TMA slides were stained for H&E and pancytokeratin (Pan-CK). PTB, ITB, PMB, and IMB were analyzed and classified as bd1, bd2, and bd3 according to ITBCC guidelines. ITB and PTB as well as IMB and PMB showed significant correlation on H&E and Pan-CK staining. No correlation was found for tumor bud counts in primary tumors and corresponding metastases. The agreement for categorized tumor bud counts showed fair to good agreement for metastases and poor agreement for primary tumors between different classes on H&E and Pan-CK staining. Based on our results, tumor budding in primary tumors and CRLM seems to be different processes which might be the results of differing surrounding microenvironments. The evaluation of tumor budding in CRLM is challenging in cases without desmoplastic stroma reaction or intense perimetastatic ductular reaction. We therefore propose to evaluate tumor budding only in metastases with desmoplastic stroma reaction based on H&E staining since important morphological features are obscured on Pan-CK staining.

Older patients with venous thromboembolism (VTE) are underrepresented in clinical anticoagulation trials. We examined to which extent elderly patients with VTE would be excluded from such trials and compared the bleeding risk between hypothetically excluded and enrolled patients. We studied 991 patients aged ≥65 years with acute VTE in a prospective multicenter cohort. We identified 12 landmark VTE oral anticoagulation trials from the eighth and updated ninth American College of Chest Physician Guidelines. For each trial, we abstracted the exclusion criteria and calculated the proportion of our study patients who would have been excluded from trial participation. We examined the association between five common exclusion criteria (hemodynamic instability, high bleeding risk, comorbidity, co-medication, and invasive treatments) and major bleeding (MB) within 36 months using competing risk regression, adjusting for age, sex, and periods of anticoagulation. A median of 31% (range: 20–52%) of our patients would have been excluded from participation in the landmark trials. Hemodynamic instability (sub-hazard ratio [SHR]: 2.2, 95% CI: 1.1–4.7), comorbidity (SHR: 1.5, 95% CI: 1.1–2.2), and co-medication (SHR: 1.5, 95% CI: 1.0–2.3) were associated with MB. Compared to eligible patients, those with ≥2 exclusion criteria had a twofold (SHR: 2.16, 95% CI: 1.38–3.39) increased risk of MB. Overall, about one-third of older patients would not be eligible for participation in guideline-defining VTE anticoagulation trials. The bleeding risk increases significantly with the number of exclusion criteria present. Thus, results from such trials may not be generalizable to older, multimorbid, and co-medicated patients.

No abstract