Reparación endovascular versus reparación abierta del aneurisma asintomático de la arteria poplítea
Resumen
Antecedentes
El aneurisma de la arteria poplítea (AAP) es una dilatación y debilitamiento focal de la arteria poplítea. Si no se trata, el aneurisma puede producir trombosis o rotura, o el coágulo dentro del aneurisma puede producir embolia y causar una morbilidad grave. El AAP puede tratarse de forma quirúrgica mediante la realización de una derivación desde el segmento arterial proximal al aneurisma hasta el segmento arterial debajo del aneurisma, lo cual elimina el aneurisma de la circulación. También puede tratarse con un injerto de stent que se inserta de forma percutánea o a través un corte pequeño en la ingle. El éxito del procedimiento se mide por la capacidad del injerto de seguir presentando permeabilidad durante un periodo prolongado. Aunque generalmente en una emergencia se prefiere el tratamiento quirúrgico, hay evidencia incierta sobre el tratamiento de primera línea en un contexto no urgente. Esta es una actualización de una revisión publicada por primera vez en 2014.
Objetivos
Evaluar la efectividad de un injerto de stent endovascular versus cirugía abierta convencional para el tratamiento de los aneurismas de la arteria poplítea (AAP) asintomáticos en cuanto las tasas de permeabilidad primaria y asistida, la estancia hospitalaria, la duración del procedimiento y las complicaciones locales.
Métodos de búsqueda
El Especialista en Información del Grupo Cochrane Vascular realizó búsquedas en las bases de datos del Registro Especializado del Grupo Cochrane Vascular, CENTRAL, MEDLINE, Embase y CINAHL, en la International Clinical Trials Registry Platform de la Organización Mundial de la Salud y en el registro de ensayos de ClinicalTrials.gov hasta el 29 de enero de 2019.
Criterios de selección
Se incluyeron todos los ensayos controlados aleatorizados (ECA) que comparaban el injerto de stent endovascular versus reparación quirúrgica abierta convencional en pacientes sometidos a la reparación profiláctica unilateral o bilateral de AAP asintomáticos.
Obtención y análisis de los datos
Se recopilaron datos sobre las tasas de permeabilidad primaria y primaria asistida (variables principales de evaluación) así como la duración de la cirugía, la duración de la estancia hospitalaria, la recuperación de la extremidad y las complicaciones locales de la herida (variables de evaluación secundarias). Los resultados se presentaron como riesgos relativos o diferencias de medias con intervalos de confianza del 95% y se evaluó la certeza de la evidencia con criterios GRADE.
Resultados principales
No se identificaron nuevos estudios para esta actualización. Un único ECA con un total de 30 AAP cumplió los criterios de inclusión. Hubo un bajo riesgo de sesgo de selección y sesgo de detección. Sin embargo, los riesgos de sesgo de realización, de sesgo de deserción y de sesgo de informe del estudio fueron inciertos. A pesar de ser un ECA, el nivel de certeza de la evidencia se disminuyó a moderado debido al tamaño de la muestra pequeño, lo cual dio lugar a intervalos de confianza (IC) amplios; sólo 30 AAP fueron asignados al azar durante un periodo de cinco años (15 AAP en cada uno de los grupos sometidos al injerto de stent endovascular y sometidos a la cirugía abierta convencional). La tasa de permeabilidad primaria un año más tarde fue del 93.3% en el grupo endovascular y del 100% en el grupo de cirugía (RR 0,94; IC del 95%: 0,78 a 1,12; evidencia de certeza moderada). La tasa de permeabilidad asistida un año más tarde fue similar en ambos grupos (RR 1,00, IC del 95%: 0,88 a 1,13, moderada certeza de la evidencia). No hubo evidencia clara de una diferencia entre los dos grupos en las tasas de permeabilidad primaria o asistida a los cuatro años (13 injertos presentaron permeabilidad de 15 tratamientos del AAP en cada grupo; RR 1,00; IC del 95%: 0,76 a 1,32; evidencia de certeza moderada); los efectos fueron imprecisos y compatibles con un beneficio del injerto de stent endovascular o de la cirugía o no hubo diferencias. La estancia hospitalaria media fue más corta en el grupo de reparación endovascular (4,3 días para el grupo de reparación endovascular versus 7,7 días para el grupo de cirugía; diferencia de medias (DM) 3,40 días; IC del 95%: ‐4,42 a 2,38; P < 0,001; evidencia de certeza moderada). La duración media de la cirugía también se redujo en el grupo de reparación endovascular (75,4 minutos en el grupo de reparación endovascular versus 195,3 minutos en el grupo de cirugía; DM ‐119,90 minutos, IC del 95%: ‐137,71 a ‐102,09; P < 0,001; evidencia de certeza moderada). La recuperación de la extremidad fue de un 100% en ambos grupos. Los datos sobre las complicaciones locales de la herida no se publicaron en el informe del ensayo.
Conclusiones de los autores
La evidencia para determinar la efectividad del injerto de stent endovascular versus cirugía abierta convencional para el tratamiento de los AAP asintomáticos se limita a los datos de un estudio pequeño. Al año hay evidencia de certeza moderada de que la permeabilidad primaria puede mejorar en el grupo de cirugía, pero las tasas de permeabilidad primaria asistida fueron similares entre los grupos. A los cuatro años no había ningún beneficio claro del injerto de stent endovascular o de la cirugía de permeabilidad primaria o asistida (evidencia de certeza moderada). Como tanto el tiempo de operación como la estancia hospitalaria se redujeron en el grupo endovascular (evidencia de certeza moderada), puede representar una alternativa viable a la reparación abierta del AAP. Un ECA más amplio y multicéntrico podría proporcionar más información en el futuro. Sin embargo, es probable que haya dificultades para reclutar suficientes pacientes, a menos que se trate de una colaboración internacional que incluya una serie de centros vasculares de gran volumen.
PICO
Resumen en términos sencillos
Tratamiento mínimamente invasivo versus tratamiento quirúrgico del aneurisma de la arteria poplítea
Antecedentes
La arteria poplítea es un vaso sanguíneo situado detrás de la articulación de la rodilla. A veces se debilita y se expande como un globo, lo cual se conoce como aneurisma. Si no se trata, el coágulo sanguíneo dentro del aneurisma puede migrar o el aneurisma puede romperse o bloquearse. Cualquiera de estas complicaciones puede dar lugar a la pérdida de la extremidad o incluso la muerte. Tradicionalmente, el aneurisma de la arteria poplítea (AAP) ha sido tratado de forma quirúrgica. Sin embargo, también es posible tratar el trastorno mediante una técnica mínimamente invasiva en la que se coloca una malla metálica forrada de tela (injerto de stent) a lo largo del aneurisma mediante una pequeña punción en la ingle. El injerto de stant forma un sello dentro de la arteria. El éxito de la intervención viene determinado por la capacidad del injerto aplicado de forma quirúrgica o del injerto de stent de continuar funcionando y no presentar bloqueos (conocido como permeabilidad). El bloqueo del injerto reduce la circulación de la pierna, lo cual puede requerir cirugía de urgencia. Algunos consideran que la técnica quirúrgica convencional es el estándar de referencia, aunque la factibilidad de la técnica mínimamente invasiva ha sido bien documentada en muchos estudios en los que los datos se analizaron de forma retrospectiva. Sin embargo, la evidencia de calidad alta solo se obtienen al comparar dos técnicas entre sí en un ensayo clínico en el que los datos se recogen de manera prospectiva, también conocido como ensayo controlado aleatorizado (ECA).
La finalidad de esta revisión fue combinar los datos de todos los ECA realizados hasta la fecha que compararon la técnica quirúrgica versus técnica mínimamente invasiva (endovascular) para el tratamiento del AAP no urgente.
Características de los estudios y resultados clave
Se realizó una búsqueda extensiva en las bases de datos de bibliografía médica (actualizadas hasta el 29 de enero de 2019). Se encontró solo un ECA completado. En el ECA publicado, se trataron 30 AAP (15 con la técnica endovascular y 15 con la técnica quirúrgica). Se realizó el seguimiento de cada caso durante un mínimo de cuatro años. En el grupo de pacientes tratados con la técnica endovascular hubo dos bloqueos. En un caso se realizó una nueva colocación de stent y el otro caso requirió una derivación quirúrgica. En el grupo de cirugía también hubo dos bloqueos, los cuales no requirieron tratamiento. No hubo pérdidas de la extremidad. El tiempo hasta completar el procedimiento y la duración de la estancia hospitalaria fueron más cortos en el grupo de reparación endovascular. El informe del ensayo no proporcionó información sobre las complicaciones de la herida.
Fiabilidad de la evidencia
La limitación principal de este estudio fue que hubo sólo 15 AAP en cada grupo, por lo que la certeza de la evidencia se redujo de alta a moderada. Debido a las limitaciones de la evidencia actual, no es posible determinar los efectos de un injerto de stent endovascular versus cirugía abierta convencional para el tratamiento de los AAP asintomáticos. Se requiere un ensayo clínico multicéntrico más grande para que pueda haber más confianza en los hallazgos. No es posible decir si hubo un claro beneficio general en la permeabilidad de cualquiera de los dos grupos (evidencia de certeza moderada). Como tanto el tiempo de operación como la estancia hospitalaria se redujeron en el grupo endovascular (evidencia de certeza moderada), puede representar una alternativa viable a la reparación abierta del AAP.
Authors' conclusions
Summary of findings
| Endovascular stent graft treatment compared with surgical treatment for asymptomatic PAA | ||||||
| Patient or population: patients older than 50 years with asymptomatic PAA Settings: hospital (non‐emergency) Intervention: endovascular treatment Comparison: open surgical treatment | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | No of PAA | Certainty of the evidence | Comments | |
| Risk with open surgical treatment | Risk with endovascular treatment | |||||
| Primary patency rate at 4 yearsa | Low risk populationb | RR 1.00 (0.76 to 1.32) | 30 (1) | ⊕⊕⊕⊝ | Although, the long‐term follow‐up data were available there was a limited number of cases: two occlusions in each group | |
| 867 per 1000 | 867 per 1000 | |||||
| Assisted primary patency rate at 4 years | Low risk population | RR 1.00 (0.76 to 1.32) | 30 (1) | ⊕⊕⊕⊝ | Although, the long‐term follow‐up data were available there was a limited number of cases | |
| 867 per 1000 | 867 per 1000 | |||||
| Hospital stay in days | The mean hospital stay was 7.7 days | The mean hospital stay was 3.4 days less (4.4 days less to 2.4 days less) | ‐ | 30 (1) | ⊕⊕⊕⊝ | Favours endovascular repair |
| Operating time in mins | The mean operating time was 195.3 mins | The mean operating time was 119.9 mins less (137.7 mins less to 102 mins less) | ‐ | 30 (1) | ⊕⊕⊕⊝ | Favours endovascular repair |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a The risk of an occlusion is an inverse event with respect to the primary outcome of patency. The occluded grafts are equal to the total procedures minus the number of grafts/stent grafts patent at 4 years | ||||||
Background
Description of the condition
An aneurysm is defined as a localised dilatation of the arterial wall. The popliteal artery is considered aneurysmal when it is more than 1.5 times the diameter of the adjacent normal arterial segment (Callum 1974). Popliteal artery aneurysm (PAA) accounts for 70% of all peripheral arterial aneurysms, although aneurysms of the aorta are more common (Szilagyi 1981). The overall incidence of PAAs is 0.1% and about half of these cases have bilateral disease and an association with aortic or femoral aneurysms (Dawson 1991; Dawson 1997). Males between the age of 60 and 70 years are most commonly affected by PAA. The incidence of PAA in females is extremely rare. In a retrospective study of 485 patients undergoing PAA treatment in a single institution over 28 years, only 8 (1.6%) were females (Peeran 2016). The precise etiopathogenesis is unknown, however, it is likely to be a combination of atherosclerosis, proteolytic vessel wall degradation, genetic predisposition and mechanical stress (Dawson 1997). Increasing numbers of PAA are now recognized at an asymptomatic stage, either incidentally or in screening programmes that are aimed at detecting abdominal aortic aneurysms (Björck 2014; Wright 2004). A common complication of PAA is chronic limb ischaemia, which occurs secondarily to distal embolisation of fragments of thrombus contained within the sac of the aneurysm. Acute limb ischaemia may also result from sudden thrombotic occlusion or rupture of the aneurysm sac. Indirect symptoms may be caused by compression of the popliteal vein or the sciatic nerve, leading to leg oedema and pain respectively (Lowell 1994).
Ischaemic complications occur in up to 45% of people with PAA, and major amputation (above the ankle) in 4% of those patients (Roggo 1993). However, major amputation rates that are as high as 48% have been reported (Gifford 1953). Hence prophylactic treatment of PAA is indicated when the aneurysm sac measures more than 2 to 3 cm. The need is greater if significant thrombotic load or occlusion of the tibial arteries (anterior tibial, posterior tibial and peroneal arteries) from chronic thromboembolism (Galland 2008).
Description of the intervention
The treatment of PAA has evolved considerably over the years. Currently the two main techniques are the conventional open surgical approach and covered endoluminal stent graft repair. The surgical repair of PAA involves ligation of the popliteal artery both above and below the PAA and bypassing the excluded segment using either a prosthetic graft or a reversed autologous vein graft. The latter technique is generally preferred (Huang 2007). This allows exclusion of the aneurysm from the limb circulation and provides adequate revascularisation. The site of the proximal and the distal anastomoses of the bypass conduit varies with the quality of the non‐aneurysmal popliteal artery and the superficial femoral artery below and above the aneurysm respectively. Sometimes the proximal anastomosis may have to be made on the common femoral artery in the groin and the tibio‐peroneal trunk below the knee. More distal anastomoses are rare, except in symptomatic aneurysms. An alternative surgical technique is aneurysmotomy or aneurysmectomy followed by transposition of a tubular prosthetic graft between the proximal and distal normal segments of the popliteal artery. The latter technique is similar to abdominal aortic aneurysm surgery, however the former approach is more common (Wain 2007).
Marin et al first described endovascular repair of PAAs in 1994 (Marin 1994). It is usually performed percutaneously or via small incision in the groin to access the ipsilateral common femoral artery. A covered stent graft is deployed within the lumen of the PAA under radiological guidance. A seal is formed at the proximal and the distal landing zones (non‐aneurysmal artery where the proximal and distal ends of the graft rest). This helps in excluding the aneurysm sac from the circulation (Cina 2010). Occasionally, the landing zones may not be adequate, due to the sural branches emanating from the main popliteal artery. Embolisation of these branches may be required prior to deployment of the main stent graft to avoid blood leaking back into the sac, from reverse flow in the branches (type‐2 endoleak). The length of the parallel walled landing zones above and below the aneurysm and the lack of mismatch in size between these two landing zones determine the anatomical suitability for endovascular repair. If adequate landing zones are not attained, either a primary surgical approach or conversion to the conventional surgical approach may be necessary (Antonello 2005a).
How the intervention might work
While successful repair of PAA is possible by both the endovascular and open surgical approaches, there is no consensus on the preferred technique in the various clinical settings. Both techniques are able to exclude the thrombus that is inside the aneurysm from the limb circulation, hence removing the possibility of sudden rupture and distal embolisation of the fragments of thrombus. For a patient with an aneurysm presenting with symptoms including acute ischaemia and those resulting from pressure, conventional surgical treatment is generally preferred (Hupalo 2009), although some reports have indicated that endovascular repair is feasible in some cases (Gerasimidis 2003). For asymptomatic aneurysms the choice of technique often reflects the available local resources, including expertise.
Unfortunately most grafts, whether prosthetic, native or endoluminal, eventually get blocked. The durability of the graft varies but many factors interplay to determine the longevity. These include the graft material, patient factors and medication such as antiplatelet therapy.
Many proprietary covered stent grafts are available for this procedure (Beregi 1999; Kudelko 1998; Tielliu 2010). Most of them are based on the same principle, a self‐expanding tubular mesh made out of a malleable metal alloy and lined with a thin synthetic membrane (such as polytetrafluoroethylene (PTFE)) that simulates the endothelium of the native artery. Recently, a multilayer flow modulator stent (without any fabric lining) has also been successfully tested in 19 patients (Ucci 2018). This works by flow diversion and eventual spontaneous thrombosis of the extraluminal aneurysm sac, a principle that is commonly used to treat wide‐necked cerebral aneurysms. Therefore, it is possible that with further improvements in the elastic properties of the metal alloys and the overall design of stent grafts, the tribulations of stent fracture and kinking will be overcome.
Why it is important to do this review
There is no clear consensus on the treatment of PAA. The proponents of the open surgical technique argue that surgical repair offers better long‐term patency rates of the bypass conduit. A North American case series that consisted of a large cohort of patients undergoing PAA repair (583 open operations in 537 patients) showed a limb salvage rate of 99% at 30 days, 97.6% at 1 year, and 96.2% at 2 years (Johnson 2008). However, consistent developments in stent graft technology and the relative success of endovascular aortic aneurysm repair have inspired the increasing popularity of endovascular treatment of PAA. Nevertheless, compared to the aorta the popliteal artery is much smaller and is subject to greater external mechanical stress from frequent movement at the knee joint. This increases the risk of stent kinking, fracture or migration, which may require secondary endovascular or surgical intervention. It may also lead to graft failure (Antonello 2005a). Despite the challenges, endovascular stent grafting is more appealing than surgery because it may carry minimal morbidity and lead to shorter recovery times (Midy 2010). Indeed, now there is an increasing trend towards endovascular treatment for both symptomatic and even asymptomatic PAA (Wrede 2018). Recent studies have suggested that it can offer acceptable patency rates. For example, in a study of 73 patients with PAAs that were managed with endovascular stent grafts the five‐year primary patency rate was 70% and the assisted patency rate was 76%. The five‐year primary patency rate improved to 80% with enhanced technical expertise and the addition of clopidogrel to postoperative patient management (Tielliu 2007). One of the arguments in favour of endovascular first approach is that, the autologous vein conduit is still available for possible future surgical intervention. However, performing a salvage surgical bypass in a patient with failed stent graft can be more challenging, especially if the site of distal anastomosis has been used for landing the distal end of the stent graft. This review brings together trial evidence to evaluate the advantages and disadvantages of endovascular stent grafts compared to conventional open surgical repair of PAA. This will help the clinicians to decide on the optimum treatment strategy for PAA.
Objectives
To assess the effectiveness of an endovascular stent graft versus conventional open surgery for the treatment of asymptomatic popliteal artery aneurysms (PAA) on primary and assisted patency rates, hospital stay, length of the procedure and local complications.
Methods
Criteria for considering studies for this review
Types of studies
We included randomised controlled trials (RCTs).
Types of participants
We included participants undergoing unilateral or bilateral prophylactic repair for asymptomatic PAA.
Types of interventions
We included RCTs comparing endovascular stent grafting versus conventional open surgical repair of asymptomatic PAA.
Types of outcome measures
Primary outcomes
-
Primary patency rate at one and four years. According to the Trans‐Atlantic Inter‐Society Consensus (TASC) document, primary patency indicates uninterrupted patency following initial graft deployment (Dormandy 2000).
-
Assisted primary patency rate at one and four years. This conveys the cases in which a revision or a revascularisation method is applied to prevent impending occlusion or progression of stenosis (Dormandy 2000).
Secondary outcomes
-
Length of hospital stay.
-
Operating time or time taken for the procedure to be completed.
-
Limb salvage.
-
Wound complications (bleeding, haematoma, seroma and infection).
Search methods for identification of studies
Electronic searches
The Cochrane Vascular Information Specialist conducted systematic searches of the following databases for randomised controlled trials and controlled clinical trials without language, publication year or publication status restrictions:
-
the Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web searched on 29 January 2019);
-
the Cochrane Central Register of Controlled Trials (CENTRAL) Cochrane Register of Studies Online (CRSO 2018, issue 12);
-
MEDLINE (Ovid MEDLINE® Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE®) (searched from 1 January 2017 to 29 January 2019);
-
Embase Ovid (searched from 1 January 2017 to 29 January 2019);
-
CINAHL Ebsco (searched from 1 January 2017 to 29 January 2019):
-
AMED Ovid (searched from 1 January 2017 to 29 January 2019).
The Information Specialist modelled search strategies for other databases on the search strategy designed for CENTRAL. Where appropriate, they were combined with adaptations of the highly sensitive search strategy designed by the Cochrane Collaboration for identifying randomised controlled trials and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions Chapter 6, Lefebvre 2011). Search strategies for major databases are provided in Appendix 1.
The Information Specialist searched the following trials registries on 29 January 2019:
-
the World Health Organization International Clinical Trials Registry Platform (who.int/trialsearch);
-
ClinicalTrials.gov (clinicaltrials.gov).
Searching other resources
We checked the reference lists of relevant articles for any additional possible studies.
Data collection and analysis
Selection of studies
Two review authors (DJ and YG) independently assessed the titles and abstracts of reports identified by the search and obtained the full texts for those which met the criteria listed in Criteria for considering studies for this review. Two authors (DJ and CM) independently assessed these reports for inclusion. Any disagreements were resolved by discussion or arbitration by the third author (BG).
Data extraction and management
The review authors (DJ and CM) extracted data independently using proformas designed by Cochrane Vascular. Discrepancies were settled by discussion with the senior author (BG). If required, authors of individual studies were contacted to obtain more information such as missing data.
Assessment of risk of bias in included studies
We used Cochrane's 'Risk of bias' tool for assessing bias (Higgins 2011), independently by two authors (DJ and YG). Discrepancies between the authors' assessments were resolved through discussion with the third author (BG). The results of the assessment of bias of each study were presented in a 'Risk of bias' table and 'Risk of bias' graph (Higgins 2011).
Measures of treatment effect
We used Review Manager 2014 to perform statistical analysis according to Cochrane Vascular's statistical guidelines for review authors. We analysed the time‐to‐event data for the primary outcomes (primary and assisted patency rates of the graft) as dichotomous data because the status of all participants in a study are likely to be known at a fixed time point. Hence, the patency rates were expressed as risk ratios (RR) with 95% confidence intervals (CI).
The secondary outcomes were measured as follows:
-
The length of hospital stay and the time taken for the procedure to be completed: the methods for measures of continuous outcomes were employed and the difference in the mean number of days spent in the hospital and the mean time taken for the procedure to be completed were measured amongst the two groups of participants. The published report of the included RCT provided the range but did not include the standard deviation (SD) as the measure of variance. An approximate SD was derived from the formula SD = range/4 (Taylor 2014).
-
Wound complications and limb salvage: we treated the presence or absence of wound complications and limb salvage as dichotomous data, and expressed the outcomes as RRs.
Unit of analysis issues
The unit of analysis in this review is the PAA rather than the participant. Therefore, if a participant had bilateral popliteal artery aneurysms and both aneurysms were randomised and operated on independently of each other then they were counted as two events rather than one. We ensured that the studies followed a simple parallel group design for clinical trials, that is the number of observations in the final analysis matched the number of units that were randomised. We planned to make appropriate corrections if this was not the case. However, it was not necessary on this occasion.
Dealing with missing data
There were no missing data reported in the included study. If there are any missing data in the studies included in the future, we will ignore data 'missing at random' and analyse only the available data as the missing data are unlikely to make any significant change in the outcome. For the data that are deemed to be 'not missing at random' we will contact the original investigators. If no clear answer is available, we will allocate replacement values equal to the mean of the available values. We will perform a sensitivity analysis to assess how sensitive the results are to the reasonable changes in the assumptions that were made. The potential impact will be addressed in the Discussion section.
Assessment of heterogeneity
We did not perform tests for heterogeneity as only one trial was included. If sufficient trials are available for analysis in future updates, we will assess the degree of heterogeneity amongst the trials by using the I2 statistic according to the formula I2 = 100% × (Q ‐ degrees of freedom)/Q, where Q is the Chi2 statistic (Higgins 2011). If significant heterogeneity is present, we will calculate a summary statistic for each outcome using a random‐effects model.
Assessment of reporting biases
As there was only one study available, the publication bias was not assessed.
Data synthesis
Where possible, we will perform a fixed‐effect model meta‐analysis if sufficient trials become available in the future. This is based on the assumption that each study will estimate the primary and assisted patency rates of the grafts as primary endpoints. We will use the standard error of the summary (pooled) intervention effect to obtain the following:
-
the confidence interval (CI), which will estimate the accuracy (or ambiguity) of the summary estimate;
-
the P value, which will reveal the strength of the evidence against the null hypothesis of no significant difference between endovascular repair and conventional open repair of asymptomatic PAA.
Subgroup analysis and investigation of heterogeneity
We did not perform subgroup analysis as there was only one trial included in this review. In future, if there is an adequate number of trials to merit subgroup analysis, we may consider some groups of participants separately. For example, subgroups can be created if one trial uses a different kind of stent graft to the others. Due to the limited range of commercially‐available stent grafts, however, it seems unlikely that an additional subgroup will be required.
Sensitivity analysis
We did not perform sensitivity analysis as there was only one trial included. If significant heterogeneity is present in future updates, we will perform sensitivity analyses by repeating the analysis after omitting the trials which have low scores on individual quality items.
Summary of findings and assessment of the certainty of the evidence
We presented the main findings of this review in a 'Summary of findings' table (see summary of findings Table for the main comparison). We created a table for the comparison of endovascular treatment versus surgical treatment using GRADEpro GDT 2015, and we included the outcomes primary patency at four years, assisted primary patency at four years, length of hospital stay and operating time. We assessed the certainty of the body of evidence using the specific evidence grading system developed by the GRADE working group (GRADE Working Group 2004). We made judgements on the certainty of the evidence transparent by using footnotes in the 'Summary of findings' table.
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies
Results of the search
See: Figure 1
Study flow diagram.
Included studies
No new studies were identified for this 2019 update. The only included study (Antonello 2005), is described below in the table Characteristics of included studies.
Excluded studies
One previous ongoing study has now been excluded as it has been terminated due to unsuccessful trial accrual (NCT01817660).
Risk of bias in included studies
The only included study was a randomised, controlled single‐centre trial. An in‐depth description of the methods is provided in the original article (Antonello 2005). The relevant regulatory bodies of the institution approved and monitored the study. The details of the risk of bias assessment are described in Figure 2.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
The randomisation was performed by using sealed opaque envelopes containing the allocation to group A or group B. The envelopes, put into a container in blocks of 10 (five group A, five group B) were extracted by the study controller a day before the operation. If a participant did not wish to be included in the study, he or she was excluded from the trial (low risk of bias).
Blinding
There was no blinding of the trial participants or personnel as it would have been impractical to do so (unclear risk of bias). The trial outcome is unlikely to have been affected by this. The follow‐up of the graft or the stent graft was performed independently and in an objective fashion: a stenosis of greater then 50% was considered an indication for re‐intervention to prevent an occlusion (low risk of bias).
Incomplete outcome data
None of the trial participants were lost to follow‐up. However, any wound complications including infection, haematoma and seroma were not documented (unclear risk of bias).
Selective reporting
The initial trial protocol was not available, hence it was not possible to comment if there were any outcomes that were not reported in the completed trial (unclear risk of bias).
Other potential sources of bias
The main limitation of the trial was that it included a lower number of participants than it initially calculated in order to achieve a statistically significant difference at one‐year follow‐up and the study was therefore underpowered. However, PAA is a relatively uncommon condition and it was not possible to enrol a large number of participants from the same centre (unclear risk of bias).
Effects of interventions
Primary patency rates at one and four years
The effects of the intervention have been derived from a single RCT with 15 PAA randomised to each group (30 total). The initial trial period took place over five years (Antonello 2005); hence the mean follow‐up period was 45.9 months (range 12 to 65 months) for the endovascular group and 46.1 months (range 12 to 72 months) for the surgical group. In the endovascular group, two stents occluded: one in the early post‐operative period and the other one in the late post‐operative period. The cause of the blockages was attributed to more than 20% oversizing of the stent with respect to the popliteal artery (some oversizing of the stent graft is essential to obtain a snug fit). The early post‐operative blockage was treated by an endovascular approach while the late blockage required a surgical bypass. In the surgical group, there were two late occlusions of the bypass grafts that were detected on the surveillance ultrasound scan, performed at 24 months and 48 months. The participants only had intermittent claudication hence no treatment was initiated. The one‐year primary patency rate of the endovascular treatment versus surgical treatment was 93.3% versus 100%; RR 0.94, 95% CI 0.78 to 1.12; P = 0.47; moderate‐certainty evidence; Analysis 1.1), the four‐year primary patency rates of both the endovascular and the surgical groups were similar: two occlusions each in the 15 PAAs treated (86.6%; RR 1.00, 95% CI 0.76 to 1.32; P = 1.00; moderate‐certainty evidence; Analysis 1.2).
Some of the participants in the initial RCT were not followed up for four years. These additional data were obtained from a subsequent publication by the same authors (Antonello 2007), which meant that all participants in the initial group had at least a four‐year follow‐up. In the later trial period all participants from the initial trial were followed up for three more years (Antonello 2007). An additional 12 cases were included in the surgical group and six cases were included in the endovascular group. These additional cases were not randomised hence we have not included them in this review. In this new group of participants there were only two complications, both of which took place within two months of the index procedure. The remaining complications listed in the later study were identical to the complications in the report on the initial RCT (Antonello 2005), hence it can be assumed that there were no further complications in the randomised participants in the completed four‐year follow‐up period.
Assisted patency rate at one and four years
One endovascular stent graft was occluded in the early post‐operative period. It was made patent by performing another endovascular procedure. Two more stent grafts were found to have more than 60% stenosis (but not occlusion) in the late post‐operative period. One of these cases was successfully treated by an endovascular approach, while the other required surgery as described in the previous paragraph. The surgical group had two occlusions in the late post‐operative period; however assisted patency was not attempted in these cases due to minimal symptoms. Hence, no differences were detected in the assisted patency rate at one year or four years in the endovascular versus the surgical group respectively ((RR 1.00, 95% CI 0.88 to 1.13; P = 1; moderate‐certainty evidence; Analysis 1.3) and (RR 1.00, 95% CI 0.76 to 1.32; P = 1; moderate‐certainty evidence; Analysis 1.4)).
Length of hospital stay
The length of hospital stay was reduced in the endovascular group as compared to the surgical group (mean 4.3 days versus 7.7 days; mean difference (MD) ‐3.40 days, 95% CI ‐4.42 to ‐2.38; P < 0.001; moderate‐certainty evidence; Analysis 1.5).
Operating time
The operating time was reduced in the endovascular group compared to the surgical group (75.4 minutes in the endovascular group versus 195.3 minutes in the surgical group; MD ‐119.90 minutes, 95% CI ‐137.71 to ‐102.09; P < 0.001; moderate‐certainty evidence; Analysis 1.6).
Limb salvage
None of the participants in the study underwent an amputation. Hence, there was no difference in the limb salvage rate between the endovascular and the surgical groups (RR 1.00, 95% CI 0.88 to 1.13; P = 1; moderate‐certainty evidence; Analysis 1.7).
Wound complications
Wound complications including bleeding, haematoma, infection and seroma were not published in the report of the included RCT.
Discussion
Summary of main results
An untreated PAA can cause severe morbidity. The treatment may be either endovascular or surgical. There is currently no clear evidence to suggest better outcomes with one treatment over the other in a non‐emergency setting. Conventional surgical repair is still considered by some as the gold standard. On the other hand, several large retrospective case series have established the safety of the endovascular approach. However, there is little level‐1 evidence comparing the two techniques. We found only one RCT that was published in 2005 (Antonello 2005). Unlike abdominal aortic aneurysm (AAA), PAAs are relatively uncommon and the included RCT was conducted in a single centre, which resulted in an underpowered study. See summary of findings Table for the main comparison. Most cases achieved a four‐year follow‐up during the trial, however the data for the completed four‐year follow‐up of the remaining participants were obtained from the authors' subsequent publication (Antonello 2007). Antiplatelet therapy in the form of aspirin was used indefinitely in both groups. Two cases required embolisation of the collateral vessels prior to the endovascular treatment to prevent type‐2 endoleak. There is moderate‐certainty evidence that one‐year primary patency rate was higher in the surgical group, however the four‐year primary patency rates of both the endovascular and the surgical groups were similar (two occlusions each, moderate‐certainty evidence). There was no difference in the assisted patency rate and limb salvage at four years in the endovascular versus the surgical group (moderate certainty evidence). Both occlusions in the surgical bypass group were present in the prosthetic graft category. There were no occlusions in the autologous vein category. The stent graft occlusions were attributed to adverse anatomy and oversizing of the stent graft (> 20%). The operating time and the length of hospital stay were reduced in the endovascular group (moderate‐certainty evidence). The study did not report on wound complications.
Overall completeness and applicability of evidence
Only one small underpowered study was identified for inclusion in the review. Wound complications were not reported by the trialists. The study does, however, demonstrate the safety of endovascular non‐emergency treatment of PAA compared with conventional open repair. Indeed, with adequate surveillance for evolving stenotic lesions the long‐term assisted patency approaches that of conventional open repair.
Quality of the evidence
A single study has been included in this review, which provided moderate‐certainty evidence based on the GRADE system for evaluating the certainty of the evidence. While the study had a robust design and consistent results, there are concerns with regards to the precision of the results. As shown in the summary of findings Table for the main comparison there were a limited number of participants included in the study, which led to wide confidence intervals for the primary and secondary outcomes of the trial. We downgraded each outcome from high to moderate‐certainty evidence due to imprecision.
Potential biases in the review process
The search for suitable studies was conducted independently by the Cochrane Vascular Information Specialist. The selection of suitable studies was performed independently by two review authors (DJ and RJ). No discrepancies arose in the selection process. The unit of analysis for this review is the PAA instead of the individual. This is a potential selection bias, however we believe that the outcome of treatment of PAA on one side is unlikely to affect the outcome of treatment on the other side. According to the study four individuals had bilateral PAAs, which were randomised independently of each other and received treatment on different occasions. Two participants received endovascular treatment for both PAAs, while the other two participants underwent surgical repair on one side and endovascular repair on the other side. The included study reported number of occlusions and percentage patency rate. We deemed occlusion an inverse event with respect to the primary outcome of patency, i.e. the occluded grafts are equal to the total procedures minus the number of patent grafts/stent grafts.
Agreements and disagreements with other studies or reviews
Our results are consistent with a recent meta‐analysis of 14 studies that included 4880 popliteal artery aneurysm repairs (3915 open surgery and 1210 endovascular) (Leake 2017). The meta‐analysis included randomised as well as non‐randomised studies. It concluded that the open surgery patients were younger and more likely to have worse tibial runoff than endovascular patients. As expected, open surgery patients had increased incidence of wound complications and longer hospital stay, however, fewer re‐interventions and lower thrombotic complications. The primary patency was better for open surgery at one year and three years, however, there was no difference between the groups in assisted primary patency at one year and three years.
Study flow diagram.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 1 Primary patency rate at 1 year.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 2 Primary patency rate at 4 years.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 3 Assisted patency rate at 1 year.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 4 Assisted patency rate at 4 years.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 5 Length of hospital stay.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 6 Operating time.
Comparison 1 Endovascular versus open surgical repair of asymptomatic popliteal artery aneurysm, Outcome 7 Limb salvage at 4 years.
| Endovascular stent graft treatment compared with surgical treatment for asymptomatic PAA | ||||||
| Patient or population: patients older than 50 years with asymptomatic PAA Settings: hospital (non‐emergency) Intervention: endovascular treatment Comparison: open surgical treatment | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | No of PAA | Certainty of the evidence | Comments | |
| Risk with open surgical treatment | Risk with endovascular treatment | |||||
| Primary patency rate at 4 yearsa | Low risk populationb | RR 1.00 (0.76 to 1.32) | 30 (1) | ⊕⊕⊕⊝ | Although, the long‐term follow‐up data were available there was a limited number of cases: two occlusions in each group | |
| 867 per 1000 | 867 per 1000 | |||||
| Assisted primary patency rate at 4 years | Low risk population | RR 1.00 (0.76 to 1.32) | 30 (1) | ⊕⊕⊕⊝ | Although, the long‐term follow‐up data were available there was a limited number of cases | |
| 867 per 1000 | 867 per 1000 | |||||
| Hospital stay in days | The mean hospital stay was 7.7 days | The mean hospital stay was 3.4 days less (4.4 days less to 2.4 days less) | ‐ | 30 (1) | ⊕⊕⊕⊝ | Favours endovascular repair |
| Operating time in mins | The mean operating time was 195.3 mins | The mean operating time was 119.9 mins less (137.7 mins less to 102 mins less) | ‐ | 30 (1) | ⊕⊕⊕⊝ | Favours endovascular repair |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a The risk of an occlusion is an inverse event with respect to the primary outcome of patency. The occluded grafts are equal to the total procedures minus the number of grafts/stent grafts patent at 4 years | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Primary patency rate at 1 year Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.94 [0.78, 1.12] |
| 2 Primary patency rate at 4 years Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.0 [0.76, 1.32] |
| 3 Assisted patency rate at 1 year Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.0 [0.88, 1.13] |
| 4 Assisted patency rate at 4 years Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.0 [0.76, 1.32] |
| 5 Length of hospital stay Show forest plot | 1 | 30 | Mean Difference (IV, Fixed, 95% CI) | ‐3.40 [‐4.42, ‐2.38] |
| 6 Operating time Show forest plot | 1 | 30 | Mean Difference (IV, Fixed, 95% CI) | ‐119.90 [‐137.71, ‐102.09] |
| 7 Limb salvage at 4 years Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.0 [0.88, 1.13] |
