Acute ischemic mitral regurgitation (MR) is seen in patients with myocardial infarction (MI), and is associated with increased morbidity and mortality. Severe MR may occur with a frequency ranging from 0.3 to 3 % [1, 2] and is more related to the location of the infarction (more commonly seen with inferior MI) rather than the degree of infarction [3–7]. Left ventricular remodeling, tethering of the mitral valve leaflets, apical displacement of the papillary muscle, and less commonly papillary muscle rupture, contribute to the MR [2, 8–10]. It is generally accepted that patients with acute ischemic MR due to papillary muscle rupture need emergent mitral valve surgery given the exceedingly high mortality rate (up to 80 %) without intervention [11, 12]. However, in patients with acute ischemic MR but without papillary muscle rupture, there is no general consensus whether mitral valve surgery is warranted or beneficial as opposed to medical therapy alone. Subgroup analysis of the SHOCK registry (Should we revascularize Occluded Coronaries for cardiogenic shocK) showed that early mitral valve replacement was associated with better survival when compared to medical therapy alone [6]. However, these were observational data with selection bias. Hence, the aim of this study was to conduct a systematic review of the medical literature to assess the relative benefits and harms of mitral valve surgery with medical therapy versus medical therapy alone for patients with acute ischemic MR.

We published the research protocol of this systematic review online on PROSPERO systematic review database prior to the literature search and data extraction. There were no modifications between the planned protocol and the actual conduct of the review.

Reference URL for published protocol: http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42013005843. Search strategy: http://www.crd.york.ac.uk/PROSPEROFILES/5843_PROTOCOL_20131106.pdf. (PROSPERO registration number CRD42013005843).

The search strategy identified 887 citations (137 were duplicates and removed). Of the 750 titles, 709 were excluded (519 were non-relevant and 190 were review articles and case reports). Of the 41 remaining abstracts, 37 were retrospective cohorts and four excluded acute MR, leaving no eligible study for analysis (Additional file 2: Figure S1). Based on our search strategy, we did not find any published randomized controlled trial comparing outcomes of medical versus surgical treatment of acute ischemic MR. However, we found one registered ongoing randomized controlled trial that is being conducted at Southern Illinois University entitled by “Medical Versus Surgical Management of Patients With Moderate Mitral Regurgitation Following Percutaneous Coronary Intervention for Myocardial Infarction: A Pilot Prospective Randomized Trial” (ClinicalTrials.gov Identifier: NCT01156441). Unfortunately, the study has been just withdrawn in June 2014 with no enrollment after 4 years.

Our systematic review serves as the first one in the medical literature that attempts to summarize the evidence from randomized controlled trials comparing medical to surgical management of at least acute moderate ischemic MR. We found no published randomized controlled clinical trials to answer our inquiry.

The current review is an empty review given no studies met the eligibility criteria. Empty reviews are not uncommon and account for 8.7–12 % of Cochrane database of systematic reviews [14, 15]. One reason for this review being empty is that we limited inclusion to randomized clinical trials. The intention was to exclude data from non-randomized studies given they are unlikely to be reasonably informative for either guideline development or clinical decisions.

Importantly, our systematic review shows the gap in the knowledge about what is the best intervention in the treatment of MR following acute myocardial infarction. It is our hope that this empty systematic review will help stimulate appropriate research, whether well-designed prospective observational studies or, ideally, well-designed and adequately powered, randomized clinical trials.

There were several retrospective data on the subject as summarized in Table 1. Of the observational data that we encountered, one was an analysis of the SHOCK registry comparing mortality outcome of severe acute ischemic MR treated by mitral valve annuloplasty repair/replacement versus no valve intervention [6]. Between April 1993 and August 1997, the SHOCK registry enrolled 1190 patients with suspected cardiogenic shock complicating acute MI. A subgroup analysis of this registry included 94 patients with severe MR. Almost half of the cohort (N = 43) underwent concomitant coronary artery bypass graft with mitral valve surgery (valve replacement was done in most and valve repair was done in six patients only), while the remaining patients (N = 51) underwent coronary artery bypass grafting alone with medical management of the MR. The in-hospital mortality rates for valve surgery and non-valve surgery groups were 40 and 71 % respectively [6]. The lower mortality in the group that underwent valve surgery however, was predominately driven by a significantly higher left ventricular ejection fraction (40 versus 29 %, p = 0.04), causing a selection bias. These observational data that were identified non-systematically showed some potential benefit for early surgery given the high mortality rate of un-intervened severe MR following acute MI [6]. These results must be interpreted with caution however Another multicenter retrospective study evaluated the postoperative outcome of emergency surgery for acute severe MR [21]. Of 279 total patients enrolled, 126 (45 %) had an acute MI and did worse than those without MI. On multivariable analysis, associated coronary artery disease and acute MI were independent predictors of increased short-term and long-term mortality. The type of mitral valve surgery (repair versus replacement), however did not influence the outcomes [21].

When looking at the data of the non-randomized studies (Table 1), one could notice the mixed cohorts that included patients with papillary muscle rupture versus leaflet tethering. These are totally different cohorts; yet, the reported outcomes are combined and pooled for all patients. Furthermore, the studies were of small sample size, most of them involving only one treatment arm (most often mitral valve replacement) which makes any possibility of fair comparison of treatment strategy to outcomes almost impossible. In addition, the cohorts included a very high prevalence of patients with cardiogenic shock and already depressed left ventricular systolic function. Pooling the numbers of all the studies did show a significantly lower in-hospital (<30 days) and 1–12 months mortality for the registry of patients that underwent MVR (68/353 versus 51/101 for in-hospital mortality, and 9/102 versus 25/50 for 1–12 months mortality, respectively) (Table 1). However these numbers must be interpreted very cautiously due to the inhomogeneity of the cohorts, study designs and selection bias.

In absence of robust data, the management of acute ischemic MR mostly remains based on expert opinion. Nowadays at least, the standards of management involve the early identification of MR during acute MI, and the differentiation of whether there is papillary or chordal disruption or not. It is clinically relevant to differentiate papillary muscle ischemia from papillary muscle/chordal rupture, since the former might be corrected by revascularization alone, hence sparing the patient the added risk of mitral valve repair or replacement. Indeed, observational data suggested that acute ischemic MR without papillary muscle rupture may be treated with coronary angioplasty; the 3 year survival of coronary intervention group was 70.2 versus 45.6 % in medical group [23]. Figure 1 is a proposed algorithm that summarizes a potential approach in the management of patients with acute ischemic MR.

Papillary muscle rupture, on the other hand, is an unusual cause of ischemic MR occurring in less than 0.1 % of patients [24]. When present, it carries a poor prognosis with mortality rates reaching 95 % following MI if left without surgical intervention, and with 70 % of death occurring within the first 24 h [11, 25]. Other observational data studies showed that the mortality rate is high regardless of valve surgery; it appears to be that these patients are at an already advanced stage where intervention may be too late whenever shock occurs [7]. Data from a larger and more recent study, however, showed that there has been considerable decrease in perioperative mortality across decades (after 1990 as compared to prior to 1990) for patients with acute MR secondary to papillary muscle rupture undergoing mitral valve surgery, particularly when concomitant coronary artery bypass graft is performed (odds ratio 0.18; 95 % confidence interval 0.04–0.83) [20]. Also, the long term survival was similar to those who had an MI without papillary muscle rupture [20].

Furthermore, Chevalier et al. evaluated the predictors of increased perioperative mortality of 55 consecutive patients with acute MR (45 % papillary muscle rupture, 22 % partial rupture, and 33 % papillary muscle dysfunction) post MI undergoing mitral valve surgery (94 % valve replacement). Although the perioperative mortality was high, it was significantly reduced when concomitant coronary artery bypass graft was performed (34 versus 9 %, p = 0.02) [26].

In patients presenting with acute ischemic MR, those with papillary muscle rupture warrant emergent mitral valve repair/replacement, while those with papillary muscle dysfunction or leaflet tethering might have significant improvement in the degree of MR after coronary revascularization and optimal medical therapy (Fig. 1). The role of intra-operative transesophageal echocardiogram in guiding whether MVR is needed as an adjunct to coronary revascularization in such situation seems justified, although there is no consensus on such approach that varies with different clinical practices.

Based on our systematic review we conclude that there is a lack of randomized data comparing the clinical outcomes of medical management versus mitral valve surgery in patients with acute MR of at least moderate severity. In absence of randomized controlled trials, it is hard to make a solid informed decision about the optimal management of acute severe ischemic MR in the absence of papillary muscle rupture, and the treatment remains to be based on expert opinion and physician expertise. Medical management in such situation might be reasonable if the degree of MR is less than severe or if there high suspicion of improvement and reversibility of the MR with coronary revascularization. The withdrawal of the randomized trial “Medical Versus Surgical Management of Patients With Moderate Mitral Regurgitation Following Percutaneous Coronary Intervention for Myocardial Infarction” due to failure to enroll patients was disappointing. It is our hope that this empty systematic review will help stimulate appropriate research, whether well-designed prospective observational studies or, ideally, well-designed and adequately powered, randomized clinical trials.