Original article
Adult cardiac
Novel Formula to Calculate Three-Dimensional Angle Between Inflow Cannula and Device Body of HeartMate II LVAD

https://doi.org/10.1016/j.athoracsur.2019.04.124Get rights and content

Background

The acute angle between inflow cannula and device body of HeartMate II left ventricular assist device (LVAD) (Abbott, Pleasanton, California) is associated with device thrombosis. However, most studies utilized two-dimensional (2D) angle obtained from chest roentgenogram (CXR), which is unlikely accurate. We aimed to create and validate a formula to estimate actual three-dimensional (3D) angle.

Methods

We retrospectively reviewed the cohort undergoing HeartMate II LVAD implantation between 2008 and 2016. A formula for calculating 3D angles of the LVAD inflow cannula relative to the device body was mathematically derived, using simple 2D measurements from CXR.

Results

The cohort included consecutive 275 patients with HeartMate II (median age: 60 [25% quartile: 51, 75% quartile: 68] years). There was no significant difference between the calculated 3D angles (from formula) and actual 3D angles (from computed tomography) from the subset group with 3D computed tomography (n = 28) (71.7° ± 13.4° vs 71.1° ± 11.5°, P = .858). Among all participants, the calculated 3D angle (from formula) was 74.4° ± 14.2°, which was significantly larger than the 2D projected angle (from CXR) (65.2° ± 11.3°, P < .001). There was no statistical difference in the calculated 3D angles (from formula) between patients with/without device thrombosis, hemorrhagic stroke, ischemic stroke, or mortality (P > .05 for all).

Conclusions

We established a novel formula to mathematically calculate actual 3D angles between inflow cannula and device body of HeartMate II. The formula would help investigators to validate their findings of the relationship between 2D projected angle (from CXR) and device thrombosis.

Section snippets

Patient Selection

Consecutive patients who underwent HMII implantation and subsequent CXRs at 1-3 month(s) after the surgery in our institute between July 2008 and Jan 2016 were retrospectively reviewed. The actual 3D angles between the inflow cannula and the device body were estimated by using the below-described formula in all the patients. Subsequently, the patients undergoing both CXR and 3D CT were reviewed for the validation of the estimation formula. We compared the calculated 3D angles (from formula)

Baseline Characteristics

A total of 275 HMII patients were enrolled (Table 1). Most of the patients were male (80%) with median age of 60 (51, 68) years and a body mass index of 28.4 (24.5, 33.2) kg/m2. All CXRs were obtained at 1-3 month(s) after LVAD implantation.

We performed intra- and inter-observer reliability test to validate the reliability of our measurements. As for intra-observer reliability, ICC (1, 2) = 0.996 for reviewer 1, and ICC (1, 2) = 0.987 for reviewer 2. As for the inter-observer reliability, ICC

Comment

We have developed a novel and “easy-to-calculate” formula with high reliability, which enables us to estimate the actual 3D angle of inflow cannula relative to the device body of HMII using simple measurements of parameters from standard CXRs. The calculated 3D angles (from formula) were statistically larger than the 2D projected angles (from CXR). In other words, the 2D projected angles (from CXR), which is conventionally used to predict clinical outcomes, may underestimate the actual angle,

References (7)

There are more references available in the full text version of this article.
∗

Mr. Yasuyuki Tsushima and Dr Imamura contributed equally to this work.

View full text
© 2020 by The Society of Thoracic Surgeons Published by Elsevier