Correspondence

Coronary-Artery Stenting in Acute Myocardial Infarction

N Engl J Med 2000; 342:1447-1448May 11, 2000

Article

To the Editor:

Grines et al. (Dec. 23 issue)1 found that the benefit of coronary stenting after myocardial infarction was attributable only to the marked decrease in the rate of revascularization at six months. However, the unblinded study design may have biased decisions about revascularization in a way that could have favored the stent group. Did the independent clinical-events committee assess the appropriateness of the revascularization procedures used? Were the patients who did not undergo revascularization assessed for potential underuse of these procedures? Before worrying about the ethics of withholding stents from patients with acute infarction, the investigators should first address the possibility that their results may have been confounded by diagnostic-suspicion bias.2

Richard M. Hoffman, M.D., M.P.H.
Laura Allen, M.D.
Albuquerque Veterans Affairs Medical Center, Albuquerque, NM 87108

2 References

1
Grines CL, Cox DA, Stone GW, et al. Coronary angioplasty with or without stent implantation for acute myocardial infarction. N Engl J Med 1999;341:1949-1956
Full Text | Web of Science | Medline

2
Sackett DL. Bias in analytic research. J Chronic Dis 1979;32:51-63
CrossRef | Medline

To the Editor:

There are a few worrisome aspects of the study by Grines et al. First, the exclusion from randomization of patients “if the operator decided that the pa-tient . . . had a high likelihood of requiring a stent” seems inappropriately circular and vague in a study that is specifically designed to assess the value of stenting in patients with acute myocardial infarction. The fact that this was among the most common reasons for excluding patients will leave thoughtful interventionalists wondering about to whom these results apply.

Second, despite an impressive improvement in the combined end point of death, reinfarction, disabling stroke, or target-vessel revascularization for the treatment of ischemia among the patients who underwent stenting, as compared with those who underwent angioplasty alone, the apparent increase in mortality of 1.7 percent at one month and 1.5 percent at six months in patients who received stents is troubling. This result, which translates into eight and seven excess deaths in the stent group, respectively, cannot simply be explained away as “[statistically] nonsignificant” since the study did not have power to address this issue. The appropriate question would be, if this increase in mortality were real, would it be clinically significant and worthy of concern? If so, further studies, with sufficient power to answer this question statistically, would be required before there could be widespread endorsement of stenting in acute myocardial infarction.

The scant reference to the issue of increased mortality made by the authors is problematic and may reflect misunderstanding of the distinction between clinical and statistical significance.

I am also concerned about the potential role of conflict of interest, especially in the field of interventional cardiology, in which clinical investigators commonly double as promoters for and investors in companies that manufacture devices.1 A trend toward an increase in mortality resulting from the use of stents is an inconvenient finding for the underwriter of the study, Johnson & Johnson, and its investors. The relative disregard for this potential finding and the partial implication that, on the basis of the results of this trial, it may be “unethical to withhold stenting from patients with myocardial infarction” leave the reader wondering about the influence of corporate sponsorship on this analysis.

Mark J. Baumel, M.D.
Care Management Consulting, Kennett Square, PA 19348

1 References

1
Eichenwald K, Kolata G. Hidden interest: when physicians double as entrepreneurs. New York Times. November 30, 1999:A1, C16-C17.

To the Editor:

Grines et al. found that stenting provided better angiographic results and a significant reduction in the combined clinical end point as compared with angioplasty alone. Some methodologic aspects, however, require further consideration. First, although a rigorous quantitative angiographic analysis was performed, no data were provided about the final balloon-to-artery ratio in each group. This information is important, because the better results with coronary stenting could have been explained, at least in part, by more aggressive dilation of the vessel with that procedure. Second, the authors speculate about whether extrusion of thrombus through the stented struts may have caused distal embolization that, in turn, could explain the disturbing trend toward a lower grade of blood flow (according to the classification system of the Thrombolysis in Myocardial Infarction [TIMI] trial) in this group. Again, it would be interesting to know whether the patients with abnormal TIMI flow who received stents had more aggressive dilation (larger balloon-to-artery ratios or higher maximal balloon pressures) than the patients who did not. Third, the number of patients with angiographic evidence of thrombus, and whether this important finding influenced results, was not specified. Finally, Grines et al. suggest that the heparin-coated stent may have contributed to the low rate of subacute thrombosis (0.9 percent). I believe that this is difficult to substantiate, because low rates of subacute occlusion have also been obtained with commercially available stents.1,2

Fernando Alfonso, M.D., Ph.D.
San Carlos University Hospital, 28040 Madrid, Spain

2 References

1
Alfonso F, Rodriguez P, Phillips P, et al. Clinical and angiographic implications of coronary stenting in thrombus-containing lesions. J Am Coll Cardiol 1997;29:725-733
CrossRef | Web of Science | Medline

2
Suryapranata H, van't Hof AWJ, Hoorntje JCA, de Boer MJ, Zijlstra F. Randomized comparison of coronary stenting with balloon angioplasty in selected patients with acute myocardial infarction. Circulation 1998;97:2502-2506
Web of Science | Medline

Author/Editor Response

Dr. Grines replies:

To the Editor: In response to Drs. Hoffman and Allen: it is impossible to blind a trial comparing coronary stenting with angioplasty. However, as noted in the Methods section of our article, each end point (including target-vessel revascularization for the treatment of ischemia) was adjudicated by the clinical-events committee. Before the follow-up angiography, the investigator determined the severity of angina, obtained the results of the stress test, and ascertained whether angiography was clinically indicated. Moreover, the clinical benefits of stenting were confirmed, with a reduced incidence of restenosis and reocclusion.

Dr. Baumel states that it is “inappropriately circular and vague” to exclude patients from randomization who had a high likelihood of requiring a stent. On the contrary, our steering committee believed that it was unethical to withhold stenting if the operator believed that the patient required a stent because of extensive dissection or inadequate flow after the initial balloon inflations. In fact, no trial requires investigators to randomize patients for whom they strongly believe that another therapy is more appropriate. Dr. Baumel implies that corporate sponsorship may have influenced the analysis and decision not to emphasize “differences” in mortality (4.2 percent vs. 2.7 percent, P=0.27). We clearly stated in the Discussion section that “a nonsignificantly higher mortality rate was observed in the stent group, perhaps accounted for in part by the trend toward lower rates of TIMI grade 3 flow in the stent group.” Corporate sponsorship played no part in the analysis, which was performed by an independent data-management company with blinded adjudication of clinical events. In addition, the final sentence of the article states “additional trials should be performed and long-term cost effectiveness determined before routine stenting can be recommended as the standard of care for patients with myocardial infarction.”

In response to Dr. Alfonso: Patients with stents who had abnormal TIMI flow (grades 0 to 2) had more frequent angiographic evidence of thrombus but did not appear to have more aggressive dilation (Table 1Table 1Characteristics of Vessels with Abnormal (TIMI Grades 0 to 2) and Normal (TIMI Grade 3) Flow after Stenting.). We agree with Dr. Alfonso and stated in the article that “this trial did not allow us to assess the importance of the heparin coating of the stent relative to that of its metal scaffolding.” The heparin coating gave the investigators the confidence to discontinue the heparin immediately, despite having no glycoprotein IIb/IIIa receptor antagonist on board.