The Match Trial: Disappointments and surprises

The negative results of the Management of Atherothrombosis with Clopidogrel in High-risk patients (MATCH) Trial¹ are disappointing and surprising, but perhaps this can be said about the conduct of the trial itself. This is especially true because the standard practice for the secondary prevention of stroke is not clopidogrel (Plavix), but aspirin, at doses ranging from 81 mg to 325 mg per day. The rationale for MATCH is more logical if one believes that clopidogrel is superior to aspirin for secondary prevention in stroke and that this superiority warrants the use of clopidogrel as a standard therapy (rather than aspirin), in spite of its relatively great cost.

According to the Clopidogrel versus As­pirin in Patients at Risk of Ischemic Events (CAPRIE) Trial,² this is not true for recurrent stroke alone (7.15% clopidogrel vs 7.71% as­pirin) and is only marginally true for a composite of annual vascular outcomes (5.32% clopidogrel vs 5.83% aspirin, P = .043; number needed to treat = 197). Therefore, the results of the MATCH trial, which adds aspirin or placebo to clopidogrel as a standard therapy for transient ischemic attack (TIA) or mild stroke, are greeted by many of us with disappointment and some frustration at this clinical non sequitur.

Another surprising feature of MATCH was the delay it allowed in beginning trial therapy. The trial enrolled 7599 patients within 3 months of a TIA or stroke and with at least 1 of a list of certain risk factors for vascular disease; all received clopidogrel, 75 mg per day, without a loading dose. The patients were randomly assigned 1:1 to placebo or aspirin, 75 mg per day, and the trial continued for 18 months. The results showed a nonsignificant relative risk reduction of 6.4% for a composite primary outcome of ischemic stroke, myocardial infarction, vascular death, death from hemorrhage, and hospitalization for any ischemic event including TIA or angina over the 18-month period of observation.

These results seem to contrast with those of the Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial,³ in which the addition of clopidogrel to aspirin reduced by 20% the relative risk of a similar composite outcome to that of MATCH over 12 months. However, the CURE patients were started on the trial therapy within 24 hours, and clopidogrel was given as a 300-mg loading dose before beginning the daily 75-mg dose. The standard approach to TIA or mild stroke is to begin secondary preventive therapy within 48 hours, rather than wait 3 months, and this ap­proach is amply supported by literature, which demonstrates that the highest risk of recurrent stroke is within the first 30 days of stroke or TIA.^4,5 Therefore, because the median time of recruitment to MATCH was 15 days following the qualifying event, the trial cohort was “diluted” with patients at lower risk of outcome, at least in terms of recurrent stroke.

Another disappointment with the conduct of MATCH is the heterogeneity of its participants, ie a composite “income” as well as outcome. The etiology for the qualifying TIA or stroke was classified as atherothrombosis using the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria⁶ in only 34% of MATCH participants, while 53% of these events were classified as due to penetrating artery disease. However, if atherosclerotic vascular disease was the cause of only a minority of the acute (or rather, subacute) events, is it still useful to consider a composite of atherosclerotic vascular disease manifestations as a primary outcome? In addition, whether or not one believes that penetrating artery disease is strongly associated with atherosclerotic vascular disease, the 2 conditions differ significantly with respect to outcome. The risk of recurrent stroke, TIA, or stroke death is far lower with penetrating artery disease than with other stroke etiologies,⁷ so that the inclusion of acute stroke patients with this etiology serves, in effect, to dilute the effect of the active arm in a secondary prevention trial. I believe that MATCH shares this weakness with many other clinical trials and observational studies of stroke, which tend to consider ischemic stroke as an homo­geneous manifestation of atherosclerotic vascular disease, rather than what it really is: a heterogeneous syndrome of multiple vascular etiologies.⁸

Despite the disappointments and surprises, I believe there is useful information from the MATCH trial overall. It effectively casts doubt on the assumption that the addition of as­pirin to clopidogrel is uniformly more effective for secondary prevention in stroke or TIA than the use of either agent alone, particularly for a composite outcome of vascular disease, and when considered with the results of CAPRIE. It also demonstrates an important increase in the risk of life-threatening bleeding us­ing the combination over the 18-month trial time period (1.9% versus 0.6% with clopidogrel alone). It therefore seems logical, based on the results of MATCH, to limit the use of a clopidogrel/aspirin combination over an indefinite time frame, and to limit its use to stroke due to penetrating artery disease. Further studies are needed, however, to support or refute the practice of using clopidogrel with a loading dose and aspirin in the acute setting of TIA or mild stroke due to a demonstrated atherothrombotic vascular lesion.