Miles Conrad, University of California San Francisco (UCSF), Interventional Radiology section and co-director of the UCSF HHT Centre of Excellence, spoke to Interventional News about a new device being investigated to embolize pulmonary arteriovenous malformations (PAVMs).
What is the prevalence of PAVMs, and what does any treatment seek to achieve in patients with this condition?
The prevalence is rare in the general population. However it is fairly common in patients who have hereditary haemorrhagic telangiectasia (HHT). In those patients who have this autosomal dominant genetic disorder (one in 5,000), that affects blood vessels all over the body, anywhere between 30–50% have pulmonary vascular malformations. For these patients, treatment is seeking to achieve either the resolution of symptoms related to having a physiologic right-to-left shunt (i.e. to occlude pulmonary malformation that results in hypoxia due to the right-left shunt) or to prevent future paradoxical embolus through the right-to-left shunt, which would be manifested as stroke or brain abscess and migraines.
Before the nineties, treatment was universally done with surgical resection, but Bob White at Yale and others developed techniques of feeding artery embolization. (He also established the idea of having an HHT centre, where patients with PAVMs could be sent to have their embolization procedure as it is a difficult procedure with potential treatment failure or non-target embolization of the device or thrombi to systemic arteries). Devices used to treat PAVMs have evolved from the early experience using detachable balloons and non-detachable 0.035 inch coils. Today, many options are available to the interventionalist which include also include detachable microcoils and plugs. Detachable balloons are no longer on the market in the USA. Embolization is this scenario is frequently successful at resolving shunt related symptoms and greatly reducing the incidence of future paradoxical embolic events. After embolization, we do follow up surveillance, using CAT scans to examine the PAVM for persistence. This would be related to recanalisation of the treated feeding artery segment (around or through device), failure to recognise and treat additional feeding arteries which supply the PAVM (complex PAVM), and development of pulmonary artery or systemic collaterals. Endovascular techniques are so effective these days that there is rarely a need for surgery.
How has the embolization landscape for PAVMs evolved since the early days?
Initially surgical resection or lobectomy was the standard of care, but it was a morbid procedure particularly in patients with multiple PAVMs who might require multiple operations. Interventional radiologists then started to look at ways to adequately treat PAVMs using an endovascular approach. One of the big concerns with embolization was that if you have a high flow shunt with an artery feeding a sac or draining varux, it is conceivable that if the coil migrates, it would sail into the pulmonary vein causing stroke, visceral arteries embolus, etc. Techniques were then developed to make this procedure safe using different coils and detachable balloons. While the latter usually stayed in place, they could lose volume over time and result in treatment failure. The arrival of the Amplatzer products in the US reduced fears of device migration and to many these became commonplace for the treatment of PAVMs. The Amplatzer plug (AVP1, St Jude Medical) came around in 2005, the AVP2 in 2007 and the AVP4 in 2012. With the first two plugs it was noticed that patients were starting to have recurrence because of recanalisation through the bare metal device, so interventionalists started depositing coils on the proximal side of the plug to back it up. The plug is generally safe because it doesn’t migrate if appropriately upsized and is detachable, but it has to be delivered through a minimum 4F sheath or 5F guide catheter. For these reasons, at times, it was delivered considerably proximal to the sac. The AVP4 came out and could go through smaller catheters (0.038 inch diagnostic catheters), but there are not many long-term studies looking at using the AVP4 without the back-up of using coils, so many of us still back it up with coils to prevent recanalisation through the plug. One of the big factors associated with PAVMs recurrence is proximity of the embolic device from the sac. For example, embolization performed greater than 1cm from the sac is a risk factor for recurrence. In many situations either due to tortuosity or size, you cannot get a 0.038 catheter as close to the sac as a microcatheter. We like the detachable plug concept, so we became interested in using the MicroVascular Plug (Covidien/Medtronic). This device came on the market recently in the USA in several different sizes, and it is wrapped in goretex on the proximal portion and is also a detachable plug, which reduced our concern for intraprocedural migration. Hopefully there will be no recanalisation because of the goretex coating. It is also microcatheter compatible and it will cause immediate stasis of the pulmonary arteriovenous malformation despite procedural anticoagulation when there is good wall apposition. This new product has potential benefits, but has not been evaluated on a large scale. We want to know if these MVP microvascular plugs migrate or if they fracture over time as they are made of nitinol and the lungs are quite mobile. The device seems safe if properly sized and deployed but it needs to be evaluated in the long term. (MVP has US FDA approval for peripheral embolization.)
Have you published your case series?
There is a publication from our experience with 20 PAVMs that has been submitted to the Journal of Vascular and Interventional Radiology (JVIR) that is in press. We used two sizes of plug, the MVP3 and MVP5 and we deployed electrolytic and mechanical detachment devices. We found in our initial series that there were no procedural device migrations. However, we later discovered there was a delayed migration with an MVP5 that was placed in a 5mm feeding artery. Although the device was rated for 5mm and the unconstrained diameter is 6.5 mm, it still migrated in the patient, who was young male labourer, who exerted himself. We discovered this when he came back for a follow-up CT at three months. The plug had migrated into and lodged in the pulmonary vein. Based on the follow up of our remaining patients in this series and others who have undergone treatment since, the MVP seems safe if it is appropriately sized. I believe it should be upsized more than current manufacturer recommendations to fit the feeding artery. We feel it is a good option for treating PAVMs and hopefully it will reduce the instances of recanalisation and ultimately persistence of PAVMs. There are two larger sizes that have recently come on to the market—the MVP7 and the MVP9, with nominal diameters of 9.2 and 13 mm respectively.5mm. Although the device was rated for 5mm, it still migrated in the patient, who was young male labourer, who exerted himself. He demonstrated a delayed migration when he came back for a follow-up CT at three months. The plug had migrated into the pulmonary vein.
So the device seems to be safe if it is appropriately sized and I believe it should be upsized more than current manufacturer recommendations to fit the feeding artery. Apart from the one case of migration, we feel it is a good option for treating PVMs and we will continue to use it and hopefully it will reduce the instances of recanalisation and ultimately persistent PAVMs. There are two larger sizes that have recently come on to the market—the MVP7 and the MVP9, those are ratings for the size of the vessel that you can put it in and not for the nominal diameter of the plug, which is larger.
Before the nineties, treatment was universally done with surgical resection, but Bob White at Yale and others developed techniques of feeding artery embolization. (He also established the idea of having an HHT centre, where patients with PAVMs could be sent to have their embolization procedure as it is a difficult procedure with potential treatment failure or non-target embolization of the device or thrombi to systemic arteries). Devices used to treat PAVMs have evolved from the early experience using detachable balloons and non-detachable 0.035 inch coils. Today, many options are available to the interventionalist which include also include detachable microcoils and plugs. Detachable balloons are no longer on the market in the USA. Embolization is this scenario is frequently successful at resolving shunt related symptoms and greatly reducing the incidence of future paradoxical embolic events. After embolization, we do follow up surveillance, using CAT scans to examine the PAVM for persistence. This would be related to recanalisation of the treated feeding artery segment (around or through device), failure to recognise and treat additional feeding arteries which supply the PAVM (complex PAVM), and development of pulmonary artery or systemic collaterals. Endovascular techniques are so effective these days that there is rarely a need for surgery.
How has the embolization landscape for PAVMs evolved since the early days?
Initially surgical resection or lobectomy was the standard of care, but it was a morbid procedure particularly in patients with multiple PAVMs who might require multiple operations. Interventional radiologists then started to look at ways to adequately treat PAVMs using an endovascular approach. One of the big concerns with embolization was that if you have a high flow shunt with an artery feeding a sac or draining varux, it is conceivable that if the coil migrates, it would sail into the pulmonary vein causing stroke, visceral arteries embolus, etc. Techniques were then developed to make this procedure safe using different coils and detachable balloons. While the latter usually stayed in place, they could lose volume over time and result in treatment failure. The arrival of the Amplatzer products in the US reduced fears of device migration and to many these became commonplace for the treatment of PAVMs. The Amplatzer plug (AVP1, St Jude Medical) came around in 2005, the AVP2 in 2007 and the AVP4 in 2012. With the first two plugs it was noticed that patients were starting to have recurrence because of recanalisation through the bare metal device, so interventionalists started depositing coils on the proximal side of the plug to back it up. The plug is generally safe because it doesn’t migrate if appropriately upsized and is detachable, but it has to be delivered through a minimum 4F sheath or 5F guide catheter. For these reasons, at times, it was delivered considerably proximal to the sac. The AVP4 came out and could go through smaller catheters (0.038 inch diagnostic catheters), but there are not many long-term studies looking at using the AVP4 without the back-up of using coils, so many of us still back it up with coils to prevent recanalisation through the plug. One of the big factors associated with PAVMs recurrence is proximity of the embolic device from the sac. For example, embolization performed greater than 1cm from the sac is a risk factor for recurrence. In many situations either due to tortuosity or size, you cannot get a 0.038 catheter as close to the sac as a microcatheter. We like the detachable plug concept, so we became interested in using the MicroVascular Plug (Covidien/Medtronic). This device came on the market recently in the USA in several different sizes, and it is wrapped in goretex on the proximal portion and is also a detachable plug, which reduced our concern for intraprocedural migration. Hopefully there will be no recanalisation because of the goretex coating. It is also microcatheter compatible and it will cause immediate stasis of the pulmonary arteriovenous malformation despite procedural anticoagulation when there is good wall apposition. This new product has potential benefits, but has not been evaluated on a large scale. We want to know if these MVP microvascular plugs migrate or if they fracture over time as they are made of nitinol and the lungs are quite mobile. The device seems safe if properly sized and deployed but it needs to be evaluated in the long term. (MVP has US FDA approval for peripheral embolization.)
Have you published your case series?
There is a publication from our experience with 20 PAVMs that has been submitted to the Journal of Vascular and Interventional Radiology (JVIR) that is in press. We used two sizes of plug, the MVP3 and MVP5 and we deployed electrolytic and mechanical detachment devices. We found in our initial series that there were no procedural device migrations. However, we later discovered there was a delayed migration with an MVP5 that was placed in a 5mm feeding artery. Although the device was rated for 5mm and the unconstrained diameter is 6.5 mm, it still migrated in the patient, who was young male labourer, who exerted himself. We discovered this when he came back for a follow-up CT at three months. The plug had migrated into and lodged in the pulmonary vein. Based on the follow up of our remaining patients in this series and others who have undergone treatment since, the MVP seems safe if it is appropriately sized. I believe it should be upsized more than current manufacturer recommendations to fit the feeding artery. We feel it is a good option for treating PAVMs and hopefully it will reduce the instances of recanalisation and ultimately persistence of PAVMs. There are two larger sizes that have recently come on to the market—the MVP7 and the MVP9, with nominal diameters of 9.2 and 13 mm respectively.5mm. Although the device was rated for 5mm, it still migrated in the patient, who was young male labourer, who exerted himself. He demonstrated a delayed migration when he came back for a follow-up CT at three months. The plug had migrated into the pulmonary vein.
So the device seems to be safe if it is appropriately sized and I believe it should be upsized more than current manufacturer recommendations to fit the feeding artery. Apart from the one case of migration, we feel it is a good option for treating PVMs and we will continue to use it and hopefully it will reduce the instances of recanalisation and ultimately persistent PAVMs. There are two larger sizes that have recently come on to the market—the MVP7 and the MVP9, those are ratings for the size of the vessel that you can put it in and not for the nominal diameter of the plug, which is larger.
