Novel Techniques for the Identification of the Site of Origin of Ventricular Tachycardia: Part II Magdi M. Saba, MD St. George’s Hospital and University.

Novel Techniques for the Identification of the Site of Origin of Ventricular Tachycardia:

Part II

Magdi M. Saba, MD

St. George’s Hospital and University of London, UK

University of Maryland, Baltimore, MD, USA

• When there is no match, there is no current process for determining where to stimulate next

• Prolonged procedures when faced with multiple unstable VTs

Pace-Mapping

Existing Matching Technology

Paced QRS

VT QRS template

Matching software – provides a numerical output based on degree of morphologic similarity between the VT QRS template and the paced QRS; does not take into account the 3D location of pacing site

Concept

• ECG data obtained from pacing at a series of known locations is used to infer the unknown VT SO

• By knowing the locations (input) of multiple paced vectors (output), then we should be able to derive the VT SO (input) when we supply the system with the induced VT template/vector (output)

Step 1: acquire 3D EAM

(or pre-procedural image)ventricle

scar

Step 2: Pace at 4 to 5 locations, record a surface ECG from each site

Step 3: Induce VT and record an ECG

Step 4: Integrate data from PM sites (3D location and QRS vector) with VT QRS vector – receive output as a “VTSO area”

Step 5: repeat Steps 2,3 and 4 on a limited scale around/in “VTSO area”

Step 6: arrive at VTSO, ablate

0 0.1 0.2

V6

V5

V4

V3

V2

V1

avF

avL

avR

III

II

I

12-Lead ECG Plot

Time (sec)

EC

G C

ha

nn

els

Pace 21

VT R12 = 1.4864 mV

0 0.1 0.2

V6

V5

V4

V3

V2

V1

avF

avL

avR

III

II

I

12-Lead ECG Plot

Time (sec)

EC

G C

ha

nn

els

Pace 9

VTR12 = 9.1754 mV

E12 = 1.4864 mV E12 = 9.1754 mV

Patient Chamber R2 Points, n1 RV 0.8209 16

LV 0.6181 152 LV 0.6813 73 LV 0.8160 74 RV 0.7611 8

LV 0.8251 55 LV 0.6269 96 LV 0.7344 87 LV 0.7282 78 LV 0.9562 79 LV 0.4440 23

10 LV 0.5869 1011 RV 0.7432 612 RV 0.9127 11

Correlation: distance / E12

mean 0.73, in a mixed group of patients and pacing from voltage map-defined normal tissue

Data from first 12 patients

Results

• We applied this to a data set with 7 PM points in the LV

• A single PM was arbitrarily set as the Test PM point (serving as a surrogate for a VT SO) to test for feasibility and accuracy

• Sets of 3 PM points were used to determine the physical 3D location of the Test PM point

Prediction of the 3D location of a PM point using only its ECG data and ECG + Location data from other points

Saba M, et al. Prediction of the 3D Location of a Pacing Site from Other Pacing Sites: Experimental Method to Identify the Site of Origin of Ventricular Tachycardia. Heart Rhythm 2009;6(5S):S353.

The Test PM point is not used in the prediction model, but its location is successfully predicted by its ECG output relative to the other known Reference PM sites

Testing all the PM points as reference points in all 20 combinations , the PM reference point fell within or was on the boundary of the predicted target region 15.71 ± 1.38 times (78.6% first iteration predictive accuracy)

PM 10

PM 3

PM 11

3D location of PM 9 Predicted

Test Case Step 1

Select 3 widely spaced points

Spherical Method Surface Distance Method

Third point on other side (not visible)

Test PM

Test Case Step 2

Select new PM point closer to red and replace point furthest from red

Spherical Method Surface Distance Method

Using Identical PM Points

Test PM

Test Case Step 3

• Repeat step 2 - Select new PM point closer to red and replace point furthest from red

Conclusions

• Higher detail in our understanding of substrate and a more rapid method of identifying the likely VT SO or exit site will lead to improved control of ventricular arrhythmias

• Issues to resolve with pacemapping: size of electrode tip, HPS capture, scar (what defines it)

• Prospective, comparative clinical data examining the utility of these novel techniques are required

With ThanksThe Fischell Department of Bioengineering

Keith Herold

Jerry Wierwille

Joe Davis

Stephen Shorofsky

Timm Dickfeld

Martha McLane