First Results of the Drug Eluting Balloon-AMI Trial
First Results of the Drug Eluting Balloon-AMI Trial
The main findings of this randomized, multicenter study are: 1) DIOR DEB failed to demonstrate angiographic superiority over BMS, with similar late-luminal loss and binary restenosis rates; 2) DES showed significantly better angiographic and clinical results compared with both DEB and BMS; and 3) DEB had significantly more combined uncovered and malapposed struts compared with BMS, but less compared with the DES group.
DES were developed with the knowledge that the process of restenosis after stent implantation is gradual and progressive. Therefore, drug release from the stent was deliberately prolonged with the use of polymer coatings, providing a long-term and sustained drug release. On the other hand, laboratory results have shown that even short contact between taxol compounds with vascular smooth muscle cells can inhibit the proliferation of these cells for a long period, suggesting that stent-based sustained drug release may not be necessary. With this knowledge, a DEB was developed and tested in trials, which confirmed this inhibitory effect on neointimal hyperplasia. This effect was demonstrated for the currently used DIOR DEB, and for the Sequent Please paclitaxel-coated balloon (Braun Melsungen AG, Melsungen, Germany).
An appealing extension of the use of DEB appeared to be the treatment of STEMI, combining DEB with a BMS. Specific advantages of this approach might theoretically be: 1) homogeneous administration of the drug to the vessel wall, especially at the area of the culprit plaque, whereas the DES delivers the drug only in the proximity of its struts; 2) better angiographic results, and hence less need for TLR; 3) less malapposition, with potentially less stent thrombosis with respect to DES; 4) preservation of endothelial function with respect to DES; and 5) possibly less prone to the potential clinical consequences in case of shortened dual antiplatelet duration, or in patients incapable of adhering to 12-month dual antiplatelet therapy.
Notwithstanding these potential advantages, the DEB used in this study failed to prove superior angiographic outcomes. We were not able to demonstrate a beneficial effect of pre-treatment with DEB in STEMI. Interestingly, however, the percentage of uncovered and malapposed struts as seen on OCT suggest that there is a drug effect induced by DEB that shows morphological changes compared with BMS alone. The DES group showed even more pronounced morphological changes. These findings are in line with a recent OCT study in STEMI patients. These results may suggest that the DEB did induce some effects on neointimal proliferation as demonstrated by OCT; however, they were insufficient to cause enough inhibition of the process to reduce late-luminal loss as compared with the BMS group. As has been previously demonstrated for DES in OCT studies, effective inhibition of late-luminal loss seems to be accompanied by specific morphological changes, seen as uncovered and malapposed struts. These delayed healing processes may contribute to an increased risk of stent thrombosis. Also, the acetylcholine testing findings in the present study point toward a drug effect in DEB-treated patients. After incremental acetylcholine infusions, paradoxical vasoconstriction occurred in the DEB- and DES-treated patients, with nonsignificantly more pronounced vasoconstriction in DEB compared with DES. By contrast, endothelial function in the BMS group was stable after incremental acetylcholine concentrations.
A possible explanation of the findings relies on the fact that the currently used DEB may have failed to warrant sufficient bioavailability of paclitaxel at the lesion site. A clarification for this might be the excipient used, which consists of shellac. Recently, a side-by-side comparison in a porcine model of various DEB relying on different excipients demonstrated differences in late-luminal loss in one DEB over the other. In that same study, fibrin deposition and inflammatory response were more pronounced in the most effective DEB in comparison to a normal angioplasty balloon and the less effective DEB. Although the currently used DEB was not specifically used in that study, the study itself shows the importance of the excipient or drug carrier in DEB technology.
A second justification might be the fact that although mandatory per protocol, only 60% of patients in the DEB group underwent pre-dilation with a regular balloon. Pre-dilation before using a DEB should improve drug uptake by the vessel wall due to the creation of micro-dissections and thus facilitate drug transport through the intima and media. Third, in case of calcified lesions, pre-dilation will facilitate lesion crossing with the usually more bulky DEB and prevent potential scrape-off of the drug. Therefore, in order to understand the results obtained, we performed a series of post hoc "hypothesis-generating" analyses focused on the DEB group. Of 25 patients who had pre-dilation, the late-luminal loss was 0.49 ± 0.52 mm versus 0.85 ± 0.56 mm in the 17 patients without pre-dilation in the same DEB group (p = 0.04). Finally, in 10 patients in whom more than 1 stent was placed at the lesion site, a protocol-mandated DEB dilation was not performed in the segment where the additional BMS was placed. In these 10 patients with additional BMS and without an extra DEB dilation, the late-luminal loss was 1.01 ± 0.75 mm versus 0.52 ± 0.44 mm in the 32 patients without an additional stent (i.e., no geographical mismatch) (p = 0.01). Noticeably, 5 of these 10 patients (50%) had a TLR.
Considering the patients who had pre-dilation with a normal balloon and 1 stent (i.e., no geographical mismatch of DEB and BMS), late-luminal loss was 0.74 ± 0.60 mm (n = 29), 0.43 ± 0.45 mm (n = 19), and 0.19 ± 0.30 (n = 35) in the BMS, DEB, and DES groups, respectively. In this specific subgroup, the DEB had a significantly lower late-luminal loss than the BMS subgroup (Table 6).
It remains difficult to judge the impact of these protocol deviations on the outcomes in the DEB arm. Nevertheless, it is important to consider these results as "hypothesis-generating" and possibly useful when applying DEB in future studies.
Furthermore, whereas in DES the release of paclitaxel is regulated by a polymer coating that ensures a sustained and gradual release over time, DEB are applied by a single short exposure to the vessel. This may still have an impact on the long-term outcomes in patients with de novo lesions. However, it has been demonstrated that with a short exposure to DEB, the amount of paclitaxel in the vessel wall was still in a bio-effective range after 7 days. Therefore, we do not believe that the half-life of paclitaxel significantly contributed to the negative results of this study.
This study was powered for angiographic outcomes and not aimed at detecting clinical differences between groups. Hence, no firm conclusions can be drawn on the safety of DEB. Since this was the first study with DEB in STEMI, no reference late-luminal loss for the DEB group was available for the power calculation. Because late-luminal loss was higher in the DEB group than assumed, the study might have been insufficiently powered to detect smaller differences in angiographic outcomes between the BMS and DEB groups. Therefore, even a reduction in late-luminal loss <50% (as the original assumption of this study was) could be of clinical significance when applied to a larger cohort of patients. The OCT outcomes with a reduction of neointimal hyperplasia in the DEB group seem to suggest that the changes induced by DEB might indeed have clinical significance when applied to an appropriate number of patients. Hence, future larger randomized studies should be performed in order to put the current findings into perspective. Second, the study was single blinded, thus potentially resulting in treatment bias. Third, a risk of selection bias, which however should equally apply to all 3 treatment arms, could not be completely ruled out. Finally, the deviations from the protocol regarding the lack of pre-dilation in the DEB arm, as well as the lack of additional DEB dilation in case of additional BMS implantation, may have influenced the results negatively. Nevertheless, the consequences of these post-hoc analyses should be applied with thought and used as hypothesis-generating outcomes.
Discussion
The main findings of this randomized, multicenter study are: 1) DIOR DEB failed to demonstrate angiographic superiority over BMS, with similar late-luminal loss and binary restenosis rates; 2) DES showed significantly better angiographic and clinical results compared with both DEB and BMS; and 3) DEB had significantly more combined uncovered and malapposed struts compared with BMS, but less compared with the DES group.
DES were developed with the knowledge that the process of restenosis after stent implantation is gradual and progressive. Therefore, drug release from the stent was deliberately prolonged with the use of polymer coatings, providing a long-term and sustained drug release. On the other hand, laboratory results have shown that even short contact between taxol compounds with vascular smooth muscle cells can inhibit the proliferation of these cells for a long period, suggesting that stent-based sustained drug release may not be necessary. With this knowledge, a DEB was developed and tested in trials, which confirmed this inhibitory effect on neointimal hyperplasia. This effect was demonstrated for the currently used DIOR DEB, and for the Sequent Please paclitaxel-coated balloon (Braun Melsungen AG, Melsungen, Germany).
An appealing extension of the use of DEB appeared to be the treatment of STEMI, combining DEB with a BMS. Specific advantages of this approach might theoretically be: 1) homogeneous administration of the drug to the vessel wall, especially at the area of the culprit plaque, whereas the DES delivers the drug only in the proximity of its struts; 2) better angiographic results, and hence less need for TLR; 3) less malapposition, with potentially less stent thrombosis with respect to DES; 4) preservation of endothelial function with respect to DES; and 5) possibly less prone to the potential clinical consequences in case of shortened dual antiplatelet duration, or in patients incapable of adhering to 12-month dual antiplatelet therapy.
Notwithstanding these potential advantages, the DEB used in this study failed to prove superior angiographic outcomes. We were not able to demonstrate a beneficial effect of pre-treatment with DEB in STEMI. Interestingly, however, the percentage of uncovered and malapposed struts as seen on OCT suggest that there is a drug effect induced by DEB that shows morphological changes compared with BMS alone. The DES group showed even more pronounced morphological changes. These findings are in line with a recent OCT study in STEMI patients. These results may suggest that the DEB did induce some effects on neointimal proliferation as demonstrated by OCT; however, they were insufficient to cause enough inhibition of the process to reduce late-luminal loss as compared with the BMS group. As has been previously demonstrated for DES in OCT studies, effective inhibition of late-luminal loss seems to be accompanied by specific morphological changes, seen as uncovered and malapposed struts. These delayed healing processes may contribute to an increased risk of stent thrombosis. Also, the acetylcholine testing findings in the present study point toward a drug effect in DEB-treated patients. After incremental acetylcholine infusions, paradoxical vasoconstriction occurred in the DEB- and DES-treated patients, with nonsignificantly more pronounced vasoconstriction in DEB compared with DES. By contrast, endothelial function in the BMS group was stable after incremental acetylcholine concentrations.
A possible explanation of the findings relies on the fact that the currently used DEB may have failed to warrant sufficient bioavailability of paclitaxel at the lesion site. A clarification for this might be the excipient used, which consists of shellac. Recently, a side-by-side comparison in a porcine model of various DEB relying on different excipients demonstrated differences in late-luminal loss in one DEB over the other. In that same study, fibrin deposition and inflammatory response were more pronounced in the most effective DEB in comparison to a normal angioplasty balloon and the less effective DEB. Although the currently used DEB was not specifically used in that study, the study itself shows the importance of the excipient or drug carrier in DEB technology.
A second justification might be the fact that although mandatory per protocol, only 60% of patients in the DEB group underwent pre-dilation with a regular balloon. Pre-dilation before using a DEB should improve drug uptake by the vessel wall due to the creation of micro-dissections and thus facilitate drug transport through the intima and media. Third, in case of calcified lesions, pre-dilation will facilitate lesion crossing with the usually more bulky DEB and prevent potential scrape-off of the drug. Therefore, in order to understand the results obtained, we performed a series of post hoc "hypothesis-generating" analyses focused on the DEB group. Of 25 patients who had pre-dilation, the late-luminal loss was 0.49 ± 0.52 mm versus 0.85 ± 0.56 mm in the 17 patients without pre-dilation in the same DEB group (p = 0.04). Finally, in 10 patients in whom more than 1 stent was placed at the lesion site, a protocol-mandated DEB dilation was not performed in the segment where the additional BMS was placed. In these 10 patients with additional BMS and without an extra DEB dilation, the late-luminal loss was 1.01 ± 0.75 mm versus 0.52 ± 0.44 mm in the 32 patients without an additional stent (i.e., no geographical mismatch) (p = 0.01). Noticeably, 5 of these 10 patients (50%) had a TLR.
Considering the patients who had pre-dilation with a normal balloon and 1 stent (i.e., no geographical mismatch of DEB and BMS), late-luminal loss was 0.74 ± 0.60 mm (n = 29), 0.43 ± 0.45 mm (n = 19), and 0.19 ± 0.30 (n = 35) in the BMS, DEB, and DES groups, respectively. In this specific subgroup, the DEB had a significantly lower late-luminal loss than the BMS subgroup (Table 6).
It remains difficult to judge the impact of these protocol deviations on the outcomes in the DEB arm. Nevertheless, it is important to consider these results as "hypothesis-generating" and possibly useful when applying DEB in future studies.
Furthermore, whereas in DES the release of paclitaxel is regulated by a polymer coating that ensures a sustained and gradual release over time, DEB are applied by a single short exposure to the vessel. This may still have an impact on the long-term outcomes in patients with de novo lesions. However, it has been demonstrated that with a short exposure to DEB, the amount of paclitaxel in the vessel wall was still in a bio-effective range after 7 days. Therefore, we do not believe that the half-life of paclitaxel significantly contributed to the negative results of this study.
Study Limitations
This study was powered for angiographic outcomes and not aimed at detecting clinical differences between groups. Hence, no firm conclusions can be drawn on the safety of DEB. Since this was the first study with DEB in STEMI, no reference late-luminal loss for the DEB group was available for the power calculation. Because late-luminal loss was higher in the DEB group than assumed, the study might have been insufficiently powered to detect smaller differences in angiographic outcomes between the BMS and DEB groups. Therefore, even a reduction in late-luminal loss <50% (as the original assumption of this study was) could be of clinical significance when applied to a larger cohort of patients. The OCT outcomes with a reduction of neointimal hyperplasia in the DEB group seem to suggest that the changes induced by DEB might indeed have clinical significance when applied to an appropriate number of patients. Hence, future larger randomized studies should be performed in order to put the current findings into perspective. Second, the study was single blinded, thus potentially resulting in treatment bias. Third, a risk of selection bias, which however should equally apply to all 3 treatment arms, could not be completely ruled out. Finally, the deviations from the protocol regarding the lack of pre-dilation in the DEB arm, as well as the lack of additional DEB dilation in case of additional BMS implantation, may have influenced the results negatively. Nevertheless, the consequences of these post-hoc analyses should be applied with thought and used as hypothesis-generating outcomes.
Source...