Graft flow predictive equation in distal bypass grafting for critical limb ischemia

• Published in: Journal of vascular surgery 2019-10, Vol.70 (4), p.1192-1203.e2
• Main Authors: Miyake, Keisuke, Kikuchi, Shinsuke, Okuda, Hiroko, Koya, Atsuhiro, Sawa, Yoshiki, Azuma, Nobuyoshi
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Amputation; Angiography, Digital Subtraction; Blood Flow Velocity; Bypass; Critical Illness; Female; Flowmeter; Graft flow; Graft Occlusion, Vascular - diagnostic imaging; Graft Occlusion, Vascular - etiology; Graft Occlusion, Vascular - physiopathology; Graft Occlusion, Vascular - surgery; Humans; Ischemia - diagnostic imaging; Ischemia - physiopathology; Ischemia - surgery; Limb ischemia; Limb Salvage; Lower Extremity - blood supply; Male; Middle Aged; Models, Cardiovascular; Peripheral Arterial Disease - diagnostic imaging; Peripheral Arterial Disease - physiopathology; Peripheral Arterial Disease - surgery; Regional Blood Flow; Retrospective Studies; Rheology; Risk Factors; Transit time; Treatment Outcome; Ultrasonography, Doppler, Duplex; Vascular Grafting - adverse effects; Vascular Patency; Veins - diagnostic imaging; Veins - physiopathology; Veins - transplantation
• ISSN: 0741-5214; 1097-6809
• DOI: 10.1016/j.jvs.2018.12.057

Graft flow (GF) seems to be an important prognostic predictor in distal bypass for critical limb ischemia, but previous studies have failed to clarify the association between GF and the graft prognosis. GF differs significantly among grafts, and each graft seems to have an optimal GF depending on various factors. We hypothesized that comparison between the measured GF (mGF) and optimal estimated GF (eGF) would be important in predicting graft prognosis. Herein, we aimed to develop a GF predictive equation by assessing GF determinants and to validate the equation against a clinical dataset. A total of 198 distal bypasses with vein grafts for critical limb ischemia from 2011 to 2016 were enrolled. Of these grafts, 135 normal grafts without any abnormalities on early postoperative ultrasound examination were used to develop and validate the equation. Various anatomic and patient-related factors were analyzed to detect GF determinants with stepwise selection, and the GF predictive equation was developed with multiple linear regression analysis. After developing the equation, all 198 grafts were categorized into two groups according to the equation developed based on data from the 135 normal grafts as follows: optimal flow grafts (OFGs), in which mGF > eGF – 14.6, and suboptimal flow grafts (SFGs), in which mGF < eGF – 14.6. The cutoff value of 14.6 was determined using receiver operating characteristic curves to detect graft abnormalities. By comparing OFGs and SFGs, the efficacy of the equation in predicting bypass abnormalities and graft prognosis was assessed. The GF determinants were runoff, hemodialysis (HD), diabetes mellitus (DM), and graft quality (GQ). The predictive equation was estimated as follows: GF(ml/min)=(32.9×run-off)+(9.9×GQ)−(13.0×DM)−(35.1×HD)+12.1 (R2 = 0.71, coefficient: runoff and GQ, 3 [good], 2 [fair], 1 [poor]; DM and HD, 1 [yes], 0 [no]). In the efficacy assessment of the equation, SFGs showed a significantly higher rate of bypass abnormalities (64.0% vs 12.2%; P < .0001), graft intermediate stenosis (10.7% vs 1.6%; P = .0071), graft critical stenosis (28.0% vs 3.2%; P < .0001), and early graft occlusion (17.3% vs 4.3%; P = .0037) than OFGs and were associated with a higher rate of revision surgery within 2 years after surgery (50.7% vs 34.2%; P = .026). SFGs also showed significantly lower primary patency rates (P < .0001) and secondary patency rates (P = .0005). GF was well-estimated with runoff, GQ, and the presence of DM and HD.