P1776Identifying the ideal MitraClip candidate - a COAPT-derived, retrospective subgroup stratification of a large real-life cohort

Background/Introduction

Recently published, diverging data of the COAPT and Mitra.fr trials yielded discussion on the impact of the magnitude of mitral valve insufficiency and of left ventricular end-diastolic volume index (LVEDVI) in patients with functional mitral regurgitation (FMR) prior to MitraClip intervention on outcome measures.

Purpose

We sought to evaluate the leverage of the effective orifice area (EROA) and LVEDVI on long-term outcome in a real-life cohort.

Methods

We stratified 394 patients (74.4±8.6 years, 60.9% male) that had been treated by MitraClip from 09/2008 to 01/2018 into four subgroups: I (FMR, EROA ≤30mm2 [n=76]), II (FMR, EROA 30–40mm2 [n=87]), III (FMR, EROA >40mm2 [n=105]) and IV (degenerative MR [DMR], n=126). Follow-up was conducted at on-site visits 6 months and annually after MC procedure up to 60 months.

Results

At baseline, patients of subgroup IV were oldest (p=0.0011), while subgroups I to III demonstrated significantly higher rates of any cardiomyopathy (I: 78.4%, II: 93.1%, III: 86.7%, IV: 47.6%, p<0.001). The left ventricular ejection fraction (LVEF) differed significantly (I: 38±13%, II: 35±14%, III: 34±14%, IV: 54±12%, p<0.001), as did EuroSCORE values (median, I: 20.4%, II: 24%, III: 23%, IV: 18.4%, p<0.001). Procedural success (placement of ≥1 Clip, residual MR ≤2+) differed significantly (I: 96.1%, II: 96.6%, III: 89.5%, IV: 87.3%, p=0.039).

Up to 60 month follow-up lasting success was noted in 87.5% (I), 81.8% (II), 87.5% (III) and 77.8% (IV) and all subgroups demonstrated a significant improvement in New York Heart Failure Association classification.

In further stratification according to LVEDVI (A: ≤96ml/m2 and B: >96ml/m2), no differences were noted in Kaplan-Meier estimates for death (A: p=0.36 and B: p=53) or in post-hoc comparison of each group (I-IV). Likewise, the combined endpoint of death and cardiac rehospitalization (A: p=0.18, B: p=0.94), MACE (A: p=0.37, B: p=0.54) and MACCE (A: p=0.16, B: p=0.49) and post-hoc comparisons of each group (I-IV) yielded no significant differences in outcome measures.

Conclusions

Despite distinct differences in baseline characteristics of each subgroup, we observed high procedural success rates, long-lasting reductions of MR and beneficial clinical outcome in all patients. In this retrospective analysis of a real-life cohort, EROA and LVEDVI did not influence long-term outcome measures. These results indicate no limitation for MitraClip treatment based on advanced stages of FMR and its underlying pathology, yet mark the necessity for further pre-procedural stratification.

Bilobar spreading of colorectal liver metastases does not significantly affect survival after R0 resection in the era of interdisciplinary multimodal treatment

PURPOSE

Bilobar colorectal liver metastases (CRLM) are often considered incurable or associated with poor prognosis even after R0 resection. In this single-center study, we evaluate the impact of CRLM spreading on recurrence-free survival (RFS) and cancer-specific overall survival (CSS) after R0 resection of CRLM with respect to multimodal treatment strategies including perioperative chemotherapy and multistep resections.

METHODS

Between January 2001 and December 2010, R0 resection could be achieved in 70 patients with bilobar and 100 with unilobar CRLM. Extent of disease, perioperative chemotherapy, surgical procedures, adjuvant treatment, histopathological workup, RFS, and CSS were compared between both cohorts.

RESULTS

Forty-six (66 %) patients with bilobar and 26 (26 %) patients with unilobar CRLM received preoperative chemotherapy (p < 0.001). For bilobar CRLM, more extended and multistep resection including portal vein occlusion were performed (29 % versus 3 %; p < 0.001). Morbidity (39 % versus 28 %, p = 0.183) and mortality (1 % versus 3 %, p = 0.644) rates were comparable in both patients' cohorts. Postoperative therapy was applied in adjuvant intent to 42 (60 %) versus 51 (51 %) patients (p = 0.275). The 5-year RFS and CSS rates were 24 % versus 31 % (p = 0.169) and 42 % versus 55 % (p = 0.131), respectively.

CONCLUSIONS

To our single-center experience, there is no significant effect of CRLM spreading (bilobar versus unilobar) on RFS and CSS rates. Bilobar CRLM are more likely to require extended multimodal efforts to achieve R0 resection.