Safety and efficacy of the p48 MW and p64 flow modulation devices: a systematic review and meta-analysis

OBJECTIVE Flow diverters (FDs) have demonstrated increasing safety and efficacy in treating various types of in - tracranial aneurysms.

ment techniques.Remarkably, given the wide variety of devices available and the growing experience with their use, it is now understood that devices with more wires have a greater and consistent flow-diverting effect across vessels with different anatomical characteristics and dimensions. 8he p64 flow modulation device (p64, Phenox) consists of 64 nitinol wires and 2 radiopaque platinum wires that allow visibility in vivo.The device is compatible with a 0.027-inch inner diameter microcatheter and is the only FD that is mechanically detached, which allows it to be completely unsheathed and resheathed. 8The p48 movable wire flow modulation device (p48 MW, Phenox) has 48 interwoven nitinol wires filled with a platinum core to ensure visibility under fluoroscopy.It is designed for vessels with smaller diameters (1.75-3.5 mm).Deployment can be performed through a 0.021-inch inner diameter microcatheter, and unlike the p64, it is not mechanically detached (a radiopaque marker identifies the point at which the device can be resheathed). 9Additionally, it has an independent nitinol movable wire with an atraumatic distal tip that can be placed distal or proximal for additional support or avoidance of specific vascular territories during deployment.Both devices have been continually improved up to the current generation, which comes with an incorporated, newly developed, glycan-based multilayer hydrophilic polymeric coating (HPC, Phenox) that reduces thrombogenicity when applied to nitinol surfaces. 102][13][14] Hence, we performed a systematic review and meta-analysis of the literature evaluating the safety and efficacy of the p48 MW and p64 in the treatment of intracranial aneurysms.

Protocol and Guidance
This systematic review used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement to report the search results. 15The protocol for this systematic review was registered on the International Prospective Register of Systematic Reviews (PROSPERO) with the registration no.CRD42022381104 (https://www.crd.york.ac.uk/prospero/).

Eligibility Criteria
Original studies reporting on at least 10 patients with intracranial aneurysms treated using the p64 and p48 MW FDs were deemed eligible for inclusion.The exclusion criteria were 1) irrelevant articles; 2) articles not in English, German, Spanish, or Greek; 3) studies with patients treated with other FDs; 4) studies in which FDs were not used for the treatment of intracranial aneurysms; 5) in vitro, animal, and cadaveric studies; 6) narrative or systematic reviews and meta-analyses; 7) editorials, letters to the editor, and comments that did not provide primary patient data; and 8) conference abstracts or abstracts with no published full text.No search filters were used.All eligible studies were assessed for overlap based on the authors, center, and country.In overlapping cohorts, the one with the largest number of participants was included in this review.

Search Strategy and Study Selection
Two independent reviewers (J.V.S. and M.M.C.) searched through MEDLINE (via PubMed), Scopus, Embase, Web of Science, and Cochrane Central Register of Controlled Trials databases to identify eligible studies; the end-of-search date was November 24, 2022.Covidence reference and article manager software was used for all search and study selection stages. 16The search strategies were developed and structured for each of the curated databases.Search terms targeting our top results were refined with field labels and search rules for each database, such as stemming, wild cards, or truncation.The complete search strategy can be found in Supplemental Table 1.A third reviewer (G.S.S.) resolved any disagreements.

Data Collection Process
Two reviewers (J.V.S. and G.S.S.) independently extracted the data from the eligible studies included using a standardized, prepiloted electronic form.Any disagreements were resolved through discussion between the two reviewers, and a third reviewer (M.M.S.) became involved when no consensus was reached.We extracted the following data from each study: 1) study characteristics (first author, publication year, design, country, study period, and number of patients), 2) patient characteristics (age, sex, and pretreatment modified Rankin Scale [mRS] score), 3) aneurysm characteristics (number of ruptured and unruptured aneurysms, location, morphology, and dimensions), 4) procedure characteristics (type of FD implanted, success of implantation, use of adjunctive coils, and technical events experienced), 5) efficacy outcomes (complete aneurysm occlusion and retreatment), and 6) safety outcomes (periprocedural ischemic/hemorrhagic complications and mortality).

Outcomes and Prioritization
The primary efficacy outcome was the proportion of complete angiographic occlusion at the last follow-up.Complete occlusion was defined as Raymond-Roy class 1 and O'Kelly-Marotta grade D, according to the selfadjudicated or core-lab adjudicated reports by the authors. 17,18Secondary efficacy outcomes were the rate of technical success, the rate of adjunctive coil usage, and the retreatment rate.The primary safety outcomes were the composite safety rate (including intra-and postprocedural complications) and the all-cause mortality rate.Intra-and postprocedural complications included ischemic events (thromboembolic events and vessel occlusions) and hemorrhagic events (any intracranial bleeding event) that caused permanent morbidity.

Quality of Evidence Assessment
Two reviewers (M.M.C. and G.S.S.) used the Newcastle-Ottawa Quality Assessment Scale (NOS) for cohort studies to assess the methodological quality of the studies included. 19Because there were no studies comparing cohorts, the items "selection of the nonexposed cohort" and "comparability" were excluded from the scale, resulting in a maximum score of 6 points.In the items assessing follow-up, the cutoffs were set a priori at ≥ 12 months and > 80%.

Data Synthesis
Continuous variables were summarized as means ± standard deviation (SD), whereas categorical variables were summarized as percentages and frequencies.We estimated all relative rates based on available data for the variables of interest, and available data were handled according to the principles of the Cochrane Handbook. 20 random-effects meta-analysis of proportions was performed with a generalized linear mixed model to calculate pooled rates and 95% confidence intervals (CIs) for each outcome.Statistical heterogeneity across studies was assessed with the I 2 test (> 50% implies significant heterogeneity), while heterogeneity between subgroups was assessed using Cochran's Q test. 21Furthermore, we performed a sensitivity analysis for the population treated exclusively by the devices with surface modification (p64 HPC, p64 MW HPC, and p48 MW HPC).All analyses and plots were generated using R statistical software (version 4.1.3)and RStudio.

Study Selection and Characteristics
Three hundred nineteen documents were identified, and 165 duplicates were removed.As a result of the initial screening by title and abstract, there were 102 potentially eligible documents.Of these, 1 document could not be retrieved.In the full-text evaluation, 32 documents were excluded for the following reasons: ineligible outcomes, ineligible population, insufficient data available, ineligible intervention, ineligible study type, and ineligible language (Supplemental Table 2).Ultimately, 20 studies from the final systematic search were included (Fig. 1).

Patient, Aneurysm, and Treatment Characteristics
A total of 1781 patients with 1957 aneurysms were included in our aggregate meta-analysis; of these, 149 patients harboring 156 aneurysms were treated with the p48 MW FD.Patient ages ranged from 20 to 89 years, and 78.7% were female.The mean number of FDs implanted per aneurysm ranged between 0.8 and 1.2.In some cases, 1 FD was implanted to treat more than 1 aneurysm in the same vessel.Only 141 aneurysms (7.2%) were ruptured.Nonsaccular morphology, including fusiform, blister-like, dissecting, and segmental disease, was present in 3.2%.Aneurysm size ranged between 0.8 and 50 mm, and neck size ranged between 1 and 20 mm.Most aneurysms were located in the anterior circulation (93.1%).Previous treatment with coils, clips, or a combination of endovascular techniques (assisted or intrasaccular techniques) was performed in 14.9% of the aneurysms.
All procedures were performed under general anesthesia with perioperative heparinization.The use of bi-and tri-axial catheter systems was variable.Most procedures were performed uneventfully regarding technical events.The intraprocedural technical events reported included detachment difficulties, excessive friction advancing the device, opening issues, foreshortening, fishmouthing, migration, twisting, and failure to deliver.These events were encountered in 4% (n = 54) of the procedures.Additionally, balloon-assisted angioplasty was performed in 2%.Details about the technical events during the procedures can be found in Table 1.
In most studies, patients were maintained on dual antiplatelet therapy (DAPT) before the procedure and continued the same regimen between 3 and 12 months after the procedure.However, in 2 studies, patients received single antiplatelet therapy (SAPT) with 10 mg of prasugrel daily preprocedure. 26,31For these patients, the same regimen was maintained after the procedure for at least 6 months, after which they were switched to 100 mg of aspirin (ASA) daily.Details regarding antiplatelet regimens can be found in Table 2.

Quality of Evidence Assessment
The mean modified NOS score was 5.4 ± 0.7.Only 1 study was prospective. 8The detailed quality assessment for each included study is shown in Supplemental Table 3. Visual assessment of the funnel plots identified publication bias (Supplemental Fig. 1, Fig. 2).The results from Egger's tests did not show publication bias in the associated outcomes (Supplemental Table 4).
The rate of the composite safety outcome was 2% (95% CI 1%-4%; Fig. 3C).Fischer et al. reported two thromboembolic events (pontine ischemic infarcts) due to DAPT interruption against medical advice.Both events occurred during the first 30 days postimplantation. 22Briganti et al. reported one permanent ischemic event due to in-stent thrombosis. 34Morais et al. reported one ischemic event found on postprocedure day 7; that patient experienced central facial palsy and dysarthria that required invasive management. 30guilar Perez et al. reported two hemorrhagic events that produced clinical deterioration. 11In one case, wire perforation of the pericallosal artery produced massive intracerebral bleeding and death; in the other, the retrieval of a migrated coil into the superior branch of the left middle cerebral artery (MCA) caused intracranial bleeding.Additionally, on later follow-up, four in-stent thromboses causing permanent mRS score worsening and one large intracerebral hemorrhage (most likely due to transformation of an ischemic lesion) after treatment of a pericallosal aneurysm were reported.De Beule et al. reported one ischemic event due to acute in-stent thrombosis and one hemorrhagic event due to distal wire perforation, both producing permanent neurological deficits. 12Sirakov et al. reported one delayed rupture of a giant, wide-necked, ophthalmic aneurysm of the right internal carotid artery (ICA).The patient reported a thunderclap headache 9 days after the procedure and the rupture was confirmed with imaging; management was successful, but the patient remained with deficits. 33Hellstern et al. reported four ischemic complications with permanent deficits and one fatal hemorrhagic complication on the day of discharge. 27Bonafe et al. reported eight major ischemic complications that left clinical deficits (seven intraprocedure and one postprocedure), one aortic dissection that led to myocardial infarction and death, and one subdural hematoma (thought to be caused by antiplatelet medications) that caused death. 8Additionally, at follow-up 1 patient showed mRS score worsening after a minor stroke.Winters et al. reported one delayed aneurysm rupture 2.5 weeks after FD implantation that caused a high-flow carotid cavernous fistula that was successfully treated. 28Hellstern et al. reported seven ischemic events that left minor neurological deficits and one intraparenchymal hemorrhage 2 months after the procedure. 26Sensitivity analysis evaluating the studies that used the p64 HPC showed a composite safety outcome rate of 4% (95% CI 1%-17%; Supplemental Fig. 3D).Finally, the overall mortality rate was 0.49% (95% CI 0%-1%).

FIG. 3.
Meta-analysis of studies reporting outcomes after implantation of the p64 FD.Forest plots compare the proportion of patients in each study with a reported complete aneurysm occlusion (among 1462 aneurysms, the pooled occlusion rate was 77%) (A), the proportion of adjunctive coiling use (among 1184 aneurysms, the pooled use of adjunctive coils was 7%) (B), and the proportion of ischemic and hemorrhagic complications associated with the procedure (among 1632 patients, the pooled rate of complications was low [2%]) (C).
The rate of the composite safety outcome was 3% (95% CI 1%-11%; Fig. 4C).AlMatter et al. reported one case of minor ischemia with neurological deficits and one death due to apnea (no intracranial hemorrhage or ischemia was found). 14Pierot et al. reported one ischemic event due to the migration of a previously placed FD; neurological deficits remained in the patient at discharge. 29Van den Bergh et al. reported two microwire vessel perforations (one patient recovered with neurological sequelae and the other died) and one vessel rupture after dilation with a compliant balloon due to incomplete device expansion. 13Sensitivity analysis evaluating the studies that used the p48 MW HPC showed a composite safety outcome rate of 2% (95% CI 0%-6%; Supplemental Fig. 4C).Finally, the overall mortality rate was 2% (95% CI 1%-6%).

Discussion
This systematic review and meta-analysis of 20 stud-ies demonstrated that treatment with the p64 and p48 MW devices is feasible with low rates of periprocedural ischemic and hemorrhagic complications.Complete aneurysm occlusion after implantation of the p64 was found in 77%, while after implantation of the p48 MW the rate was 67%.The rate of ischemic and hemorrhagic complications was 2% for the p64 and 3% for the p48 MW.Finally, the mortality rates were 0.49% and 2% for the p64 and p48 MW, respectively.Given the evolution and emergence of new endovascular devices for the treatment of intracranial aneurysms and the limited availability of certain flow-diverting devices to specific regions of the world, this study offers valuable information about the safety and efficacy of another tool in the armamentarium of the neurovascular surgeon.

Safety
The overall complication rate for the p64 FD was 2%, while the overall mortality rate was 0.49%.Although it has been suggested that the denser coverage of the p64 FD FIG. 4. Meta-analysis of studies reporting outcomes after implantation of the p48 MW FD.Forest plots compare the proportion of patients in each study with a reported complete aneurysm occlusion (among 136 aneurysms, the pooled occlusion rate was 67%) (A), the proportion of adjunctive coiling use (among 156 aneurysms, the pooled use of adjunctive coils was 4%) (B), and the proportion of ischemic and hemorrhagic complications associated with the procedure (among 149 patients, the pooled rate of complications was 3%) (C).
may result in more ischemic complications, 11 our findings regarding these complications are similar and, in some cases, showed lower rates than previous studies or metaanalyses for other FDs.Considering the older generation of the PED, a meta-analysis by Texakalidis et al. reported a 6.6% rate of symptomatic thrombotic episodes and a 3% rate of hemorrhagic events, 36 whereas for the PED Flex (Medtronic Neurovascular), Bhatia et al. reported a major complication rate of 1.8% and mortality rate of 0.8% in their meta-analysis. 37A systematic review on the flow redirection endoluminal device (FRED, MicroVention) reported a morbidity rate of 3.9% and mortality rate of 1.4%. 38lso, a prospective study evaluating the performance of the Surpass Streamline FD (Stryker Neurovascular) found permanent morbidity in 6% and mortality in 2.7% of the patients. 39A meta-analysis of the Derivo embolization device (Acandis GmbH & Co.) found that the rate of periprocedural ischemic and hemorrhagic complications was 4.9% and the mortality rate was 2.1%. 40Finally, Florez et al. performed a meta-analysis of the Silk FD (Balt) and found rates of thromboembolic and hemorrhagic complications of 6.1% and 1.6%, respectively, and a 2.8% mortality rate. 41Of note, most studies we evaluated treated aneurysms located in the anterior circulation (93.1%), which might explain the lower incidence of complications in our study, considering that (in general) aneurysms located in the posterior circulation are associated with higher rates of complications. 39,42otably, the complication rate in the p64 with surface modification was slightly higher (4%), which seems counterintuitive considering the inclusion of an antithrombogenic coating to reduce platelet aggregation.Of the 4 studies that reported outcomes with surface-modified devices, Hellstern et al. was the only study that used an SAPT protocol, in which the postprocedural ischemia occurred mainly in patients with MCA or MCA bifurcation aneurysms. 26Both aspects are risk factors for ischemic events and could have contributed to the high rate of complications reported by the authors.However, it is important to emphasize that 3 of the patients who had an ischemic complication were not compliant with the SAPT regimen. 26verall, although we found that the p64 has a safe profile, further studies are needed to evaluate the safety of the newer-generation FDs with surface modification and the potential use of SAPT in these cases.
The overall complication rate for the p48 MW FD was 3%, and the mortality rate was 2%.The literature on the p48 MW FD is still limited, but morbidity and mortality rates were comparable to other low-profile FDs.A systematic review of the Silk Vista Baby FD (Balt) found that major stroke occurred in 1.2% of cases, stent thrombus formation or branch occlusion in 5.5%, and deaths in 2.5% (1.8% procedure-associated). 43A recent meta-analysis evaluating the FRED Jr. (MicroVention) reported a 5.3% stroke rate, 0.4% hemorrhagic event rate, and 0% mortality. 44Furthermore, a meta-analysis evaluating the outcomes after flow diversion in distal anterior circulation aneurysms found a 5.4% morbidity rate with a 2.2% mortality rate. 45Notably, the implantation of FDs in arteries with smaller diameters is technically challenging due to distal navigation, and the risk of treatment-related complications is potentially higher due to the coverage of bifurcation branches and perforators, and in some cases, even in-stent thrombosis. 31onsidering these risks, the impact of adequate antiplatelet therapy is significant, and it has been previously shown that premature discontinuation is associated with a higher rate of complications. 46Finally, although the lower rate of complications (2%) in the studies that used devices with surface modification might suggest a favorable safety profile of this technology, larger studies are needed to confirm these findings and optimize the selection of the antiplatelet regimen.

Efficacy
40]47 A relevant study to consider is a meta-analysis by Fiorella et al., in which the authors evaluated the outcomes of flow diversion in unruptured small-and medium-sized intracranial aneurysms located in the ICA and found a 74.9%rate of complete occlusion at 12 months. 47Notably, the cumulative occlusion rate reported in this study had a correction factor for the studies that reported self-adjudicated outcomes. 47The findings in this study converge with our results, considering that most of the aneurysm sizes in the studies we included were < 12 mm and were mostly located in the anterior circulation.However, most studies (n = 19) from our analysis presented self-adjudicated outcomes, which can potentially overestimate the rates of complete occlusion. 48Furthermore, the high variability in the follow-up times of the included studies (range 2-13.1 months) has to be considered, given that higher occlusion rates are observed with longer follow-ups.Notably, the rate of adjunctive coiling in the studies we included was low (4%), which favors evaluation of the direct efficacy of the FD as there was limited influence of adjunctive techniques.In comparison, the occlusion rate of the group treated with the surface-modified device (p64 HPC) was lower (65%), but the mean follow-up time ranged from 5.1 to 6.4 months.Similarly, the adjunctive coiling in these patients remained low (3.6%).Overall, the efficacy profile of the p64 is consistent with the results reported after the implantation of other FDs, and our findings confirm that flow diversion is a reliable and reproducible technique.
The overall complete occlusion rate with the p48 MW was 67%, which is comparable to the 66.7% reported in a meta-analysis evaluating the FRED Jr. but lower than the 72.1% reported rate with the Silk Vista Baby FD. 43,44 Furthermore, a large meta-analysis evaluating the performance of FDs in unruptured distal anterior circulation aneurysms found 82.7% complete occlusion at final follow-up.Similar to the p64 studies, the variability in the length of follow-up evaluations for these patients was considerable between studies.Additionally, the rate of adjuvant coiling was low.From the studies included in the pooled rate, AlMatter et al. 14 and Van den Bergh et al. 13 reported low occlusion rates (51% and 53%, respectively).These authors suggested that this finding might be explained by the reduced metal density of this low-profile device (48 vs 64 wires) and/or the extended use of DAPT due to local institutional protocol at one of the centers. 13,14owever, some of the advantages of using low-profile FDs for distal aneurysms are the possibility of using 0.021-inch microcatheters for delivery and avoiding device oversize or overlap.Finally, the occlusion rate with the p48 MW HPC was slightly higher (71%), which seems promising as more experience is gained in treating distal aneurysms and the best antiplatelet regimens for ruptured and unruptured cases are defined.

Antiplatelet Therapy: DAPT and SAPT
The implantation of FDs necessitates the use of DAPT to prevent intrastent thrombosis or other thromboembolic events; 49 however, the response to antiplatelet medications is heterogenous, with some patients hyperresponsive and others hyporesponsive, potentially increasing the risks of hemorrhagic and ischemic complications.As a result, the development of surface modifications such as the HPC seems promising to avoid the need for DAPT.Among the 5 FDs with antithrombogenic surface modifications available, only the p64 and p48 MW HPC include instructions for use with SAPT. 261][52][53][54] Their results confirm feasibility, but safety was not significantly improved under this regimen.In the 2 studies we included, the SAPT regimen consisted of 10 mg of prasugrel daily for 6 months and then a switch to daily ASA. 26,31The authors reported comparable safety results to uncoated FDs with DAPT and mentioned the need for further studies to confirm their outcomes.

Limitations
We acknowledge that our study has limitations.First, the studies were mainly retrospective and single-center experiences with an imbalance across sample sizes (range 11-520 participants), which contributed to the heterogeneity of the results.Second, outcome reporting was heterogenous, with variable definitions of the ischemic and hemorrhagic events and various studies excluding data on follow-up duration and posttreatment morbidity.We note that almost two-thirds of patients can present with ischemic infarcts after the procedure as shown on diffusionweighted imaging without any clinical symptoms; 55 we did not consider those patients for the pooled proportion of the safety composite outcome.Third, the angiographic outcomes were mostly self-adjudicated and measured at different time points.Fourth, the inclusion of studies that treated patients with newer devices with an antithrombotic coating could have introduced additional confounders into our analysis due to the variation in the antiplatelet therapy prescribed.Fifth, regarding the data, we found significant heterogeneity across some outcomes.However, although we report several limitations, our meta-analysis is the first to evaluate the p48 MW and p64 FDs for treating intracranial aneurysms.

Conclusions
This systematic review and meta-analysis found that the p64 and p48 MW have acceptable efficacy and favor-able safety profiles.Both devices, however, have been primarily used in Europe thus far.Although the addition of antithrombotic coatings seems promising, more data are needed to confirm their potential benefits.Furthermore, we suggest future studies are needed to evaluate the efficacy and safety of SAPT after implantation of surfacemodified FDs.