**2.3 Patient-specific cardiac reshaping net (PS-CRN) with less constraint on RV side**

Based on the above experimental results, we developed a patient-specific cardiac reshaping net (PS-CRN) with less constraint on the right ventricle (RV) by making a hole on the right ventricle side (**Figure 1**) [11]. PS-CRN is made of polyester fiber thread and is prefabricated using computer-controlled knitting machine. The size and shape are designed base on the patient's cardiac images (MRI or contrasted CT) and left ventricular end-diastolic pressure. The relationship between left ventricular end-diastolic pressure and volume is calculated using the Klotz equation [12], The PS-CRN is then designed to achieve 5 to 10 mmHg contact pressure on the LV.

	- i.Cardiac imaging (MRI T2 Whole Heart or Contrast CT).
	- ii.Cardiac catheterization: Measurement of LVEDP.
	- iii. 3D modeling of LV & RV cavity, whole ventricle surface from cardiac images
	- iv. (Converted to STL data).
	- v.Create design paper for computer-controlled knitting machine from 3D cardiac model (custom software), setting contact pressure 5 to 10 mmHg on LV surface at LVEDP 30 mmHg).
	- vi.Prefabrication of the PS-CRN by computer-controlled knitting machine.
	- vii.Washing and packaging with insertion tool and sterilization.

In heart failure patients, cardiac output decreases above a certain LVEDP point (usually LVEDP >30 mmHg), known as the descending slope of the Frank-Starling curve. Therefore, it is reasonable to set sufficient compression pressure (close to 10 mmHg) above the point of the descending slope of the Frank-Starling curve to

#### **Figure 2.**

*a. LV ESPVR & EDPVR. Typical recordings of LV Emax (ESPVR) and EDPVR. Wearing a smaller cardiac net tended to increase Emax. LV EDPVR shifted upward and to the left by wearing a smaller cardiac net. b. RV Emax & EDPVR. Typical recordings of RV Emax (ESPVR) and EDPVR. Wearing a smaller cardiac net tended to increase Emax. RV EDPVR shifted upward and to the left by wearing a smaller cardiac net. Under the 90% size of the cardiac net, RVEDP increased sharply (red dotted line), which meant that RVEDV did not increase further by rapid fluid infusion.*

prevent further dilation. This is another important consideration in the design of the PS-CRN in addition to less right ventricular constraint.

The PS-CRN is designed to set the surface pressure on the left ventricular side to 5 to 10 mmHg at LVEDP 30 mmHg. This requires prediction of the LV end-diastolic

#### **Figure 3.**

*a. Relative changes in Emax (end-systolic pressure-volume relation) of LV & RV. The CHF swine model was created by rapid atrial pacing at 200 bpm for 3 weeks. LVEF decreased from 60.6 ± 12.5 to 20.8 ± 8.5% after 3 weeks of rapid pacing. Conductance and pressure catheters were placed in both ventricles and pressure and volume were recorded using the Sigma5 system. A venous cannula was inserted into the right atrium and connected to a bag filled with 1 L of lactated Ringer's solution. Blood and lactated Ringer's solution were equilibrated by infusing and draining the bag fluid. Emax and end-diastolic pressure-volume relation (EDPVR) were measured by rapid fluid infusion. Each Emax value without cardiac net (control) was set at 100% and compared with Emax data with cardiac net. Emax tended to increase with decreasing cardiac mesh size. In both LV and RV, statistical significance was reached at 85% of the cardiac net size. b. Relative change in diastolic stiffness constant (k) of EDPVR. Each EDPVR data point was fitted to the curve with the equation: EDP = a(ekEDV-1). Each diastolic constant k value without cardiac net (control) was set to 100% and compared to the k value with cardiac net. Diastolic constant k value tended to increase with decreasing cardiac net size. RV k-value reached statistical significance at 90% size, while LV k-value reached statistical significance at 85% size. The increase of k value in RV was greater than that in LV at 85% cardiac net size.*

### **Figure 4.**

*Design and prefabrication process of patient- specific cardiac reshaping net (PS-CRN). (1) cardiac imaging (MRI T2 whole heart or contrast CT). (2) cardiac catheterization: Measurement of LVEDP. (3) 3D modeling of LV & RV cavity, whole ventricle surface from cardiac images (converted to STL data). (4) create design paper for computer-controlled knitting machine from 3D heart. Model (custom software: Set contact pressure 5 to 10 mmHg between LV surface and heart mesh at LVEDP 30 mmHg). (5) prefabrication of PS-CRN by computerized knitting machine.*

pressure-volume ratio (LVEDPVR) in individual cases. We use the LVEDPVR estimation formula reported by Klotz et al. [12]. The concept is based on studies showing that the LVEDPVR of the human heart is almost identical when LVEDV at LVEDP 30 mmHg is normalized to 1 (normalized volume), regardless of whether the heart is normal or diseased (DCM, ICM, and HCM).

> ( ( )) <sup>β</sup> α α β LVEDPVR is expressed as LVEDP = x EDV = 27.78 mmHg, = 2.76 unitless , LVEDV is normalized to 1 at LVEDP 30 mmHg. (1)

In other words, if LVEDV is set to 100% at LVEDP 30 mmHg (more precisely, at 27.78 mmHg), then LVEDP 10 mmHg is approximately 70%, 20 mmHg is approximately 80%, and 25 mmHg is 95% of the size of LVEDV at LVEDP 30 mmHg. With respect to LVEDV, the LVEDP/LVEDV ratio increases sharply as LVEDP approaches 30 mmHg. Similarly, because the PS-CRN is a knitted product, the pressure remains low just above near-unloaded sizes, but when the mesh is fully stretched, the pressure increases rapidly relative to the change in volume. The PS-CRN can be designed to prevent the heart from expanding beyond the LVEDV at LVEDP 30 mmHg by designing the mesh to be nearly fully stretched at the LVEDV size at LVEDP 30 mmHg.

Using a computer-controlled knitting machine that can produce 3D-shaped knitted products without sewing (Shima-Seiki Co. Ltd.), PS-CRN is designed and manufactured in advance for each heart failure patient according to the above concept. It is possible to safely exert sufficient pressure (>5 mmHg) on the left ventricle to improve cardiac function while avoiding right ventricular dilatation failure by making a hole in the right ventricular part.

#### **Figure 5.**

*Operative image of PS-CRN implantation with insertion tool. Insertion of the PS-CRN took approximately 1 minute.*


**Table 2.** *Patient demographics.*

3.Implant procedure (**Figure 5**)


The device can be inserted through either a median sternotomy (**Figure 5**) or a left anterior thoracotomy (minimally invasive approach (MICS). Since the PS-CRN is designed and prefabricated prior to surgery, no intraoperative adjustments are required.
