The following was part of a listserve dialog on left atrial tear. It was published as a VIN page to facilitate rapid access to images.

Hello again Dan,

It looks like you are running into some interesting cases. I too would advise against pericardiectomy in a case of left atrial hemorrhage but will suggest another procedure. In addition, I thought I should share some information and illustrations for the LA "tear" dialog based on my experience with the condition. I summarized 22 cases in 1964 and 30 cases were included in my thesis in 1967. I will include the reference below (Ref 1) since it pre-dated medline and you might have difficulty finding it.

1. Buchanan JW, and Kelly AM: Endocardial Splitting of the Left Atrium in Dogs with Hemorrhage and Hemopericardium. J. Am. Vet. Radiol. Soc. 5:28-39, 1964.

First, I will give my reasoning for calling the condition one of endocardial or endomyocardial splitting rather than a left atrial tear. The word "tear" brings to my mind ripping and tearing with shredded edges. That feature does not describe nearly all the many (probably hundreds of) left atrial splits that I have seen. Splits can be fresh or healed separations of just the endocardium (Fig. 1) or they can extend variable depths into the myocardium with or without perforating the epicardium. I believe the syndrome of left atrial rupture in dogs is a combination of endocardial degeneration and elevated left atrial pressure - most often from mitral regurgitation. Most dogs with splits have one or more ruptured chordae tendineae. Atrial splitting is not a disease syndrome recognized by human cardiologists and I only found one similar case report in a man during my literature search many years ago. I have also seen multiple splits in a young dog with PDA and a large left atrium so I think the dog endocardium must be inherently weaker than in man

Figure 1. Small healed split

The largest fresh split I saw (Fig 2, 8y m Dachshund with MR for 5y) did not penetrate the epicardium but there was external evidence of it in the form of myocardial hemorrhage in the caudal wall of the left atrium (Fig 3). A small healed split was also present in this dog (Fig 4). This case was recently superceded by the mother of all splits, a healed one that almost encircled the entire left atrium in a 9y m Cavalier KCS with no evidence of hemopericardium (Fig. 5). 

Figure 2. Large fresh split

Figure 3. subepicardial hemorrhage

Figure 4. Healed split histology

Figure 5. Mother of all LA splits (S)

Often there were healed, endothelialized splits as well as one or more fresh, thrombus covered splits in the same heart. The thrombus over most splits was minimal and seldom large enough to be recognizable in an echocardiogram. About one third of dogs with splits had hemopericardium and about 10 % had acquired atrial septal defects. Those causing hemopericardium were usually (always?) in the dorsocaudal wall of the left atrium or the LA appendage. Fresh splits are obvious as exposed myocardium covered with a layer of thrombus. Healed splits can be distinguished from jet lesions by their shape, depth and location. Most splits are parallel to the annulus and have tapered ends. Healed splits are smooth surfaced, endothelialized and usually depressed below the endocardial surface (Figs 1, 4, 5). Histologically, the endocardium is separated (note separated elastic fibers in Figure 4) and replaced by fibrous tissue which extends variably into the myocardium where there had been previous dissecting hemorrhage. Jet lesions (impact lesions from MR jets) are proliferative reactions to trauma, rise above the intact endocardium, are stacked in radiating fashion above a regurgitant site and often resemble ridges left in the sand by waves (Figs. 6, 7). 

Figure 6. Jet lesions from MR

Figure 7. Jet lesion histology

One dog with healed and fresh endocardial splits had hemopericardium of unknown duration. Its tense pericardium was thin and transparent (Fig. 8) but the entire epicardium had an opaque, granular surface which I presume was reactive mesothelial cells and/or organizing fibrin (Fig. 9). It might have become restrictive over a period of time but I didn't think it was then. According to the literature at that time, blood in the pericardial sac was said to cause reactions in the mesothelium. The liver was classically enlarged with rounded edges so the dog had been in right heart failure for some time. 

Figure 8. Hemopericardium

Figure 9 Shaggy epicardium

Another case to mention was one we operated on (11y, 16 lb f Poodle) (Ref 2). Since many readers of this forum may not have seen pericardial linear clots, I will attach her echocardiograms (Fig 10). Because perforating splits are usually in the caudolateral wall of the LA, I anticipated that it would be accessible for surgery through the left 6th intercostal space. Upon opening the pericardium, there was a large clot plus 150 ml of blood in the pericardial sac. The 10 mm LA split was at the juncture of the LA appendage and the main body of the LA (Fig 11). It had stopped bleeding by the time we opened the pericardium but it was identifiable as a thrombus covered line at the appendage juncture. Unsupported sutures tore a bigger hole and released many red cells but we were able to clamp it shut and close it with buttressed mattress sutures (Fig 11). I had toyed with the idea of puncturing the interatrial septum with a probe through a purse string controlled LA appendage approach but I chickened out due to the friability of the appendage and LA wall. She died suddenly 12hrs after surgery apparently from ventricular tachycardia. At necropsy, the vertical split at the juncture of the LA appendage was near but separate from the lateral edge of a much larger split (Fig. 12).

Figure 10. Right parasternal long-axis view of a two-dimensional echocardiogram

Figure 11. Diagram of left atrial rupture site

Figure 12. Large horizontal split and vertical perforating split

2. Sadanaga KK, MacDonald MJ, Buchanan JW: Echocardiography and Surgery in a Dog with Left Atrial Rupture and Hemopericardium. JVIM 4:216-221, 1990.

The reason for making a small ASD was (and is) the fact that spontaneous acquired ASD's from splits through the atrial septum almost always are otherwise healed and endothelialized along with the other non-perforating splits that are usually present (Fig. 13a & b, Boxer). The ASD splits (D) are usually located in the thin area of the septum primum but distinct from the fossa ovalis (FO) when viewed from the right side (Fig. 14, 10y-m Italian Greyhound). I think the ASD split decompresses the left atrium and allows splits to heal and the dogs to survive for sometime. 

Figure 13a. ASD from large healed split	Figure 13b. Close up of 13a

Figure 14. RA view of acquired ASD

I treated a 7 yr-old Cocker spaniel for 21 months that progressed from left-sided CHF to right-sided failure, probably when he developed an acquired ASD confirmed by angiography and necropsy when he was 9yrs-old (3). The acquired ASD was separate from the fossa ovalis ridge in the left atrium (FO in Fig 15). He developed refractory ascites the last 6 months but was a bright, alert, happy dog with the help of occasional abdomenocentesis. The ascites might be controlled better now with vasodilators and better diuretics than we had 40 years ago.

Figure 15. LA view of ASD and fossa ovalis ridge

3. Buchanan JW: Spontaneous left atrial rupture in dogs. . Advances in Exp. Med.and Biol., 22:315-324, 1972

An ASD is relatively simple to make via a right thoracotomy and might be considered in a relatively stable case such as yours. I would plan to do it with inflow venous occlusion and a Satinsky clamp, basically as demonstrated in my cardiac surgery website in the section on pulmonic stenosis. http://cal.vet.upenn.edu/projects/cardiosf/index.htm but performed through the right 4th or 5th ICS where it is easy to tourniquet the venae cavae and azygos vein.

Another approach to consider would be a fossa ovalis balloon tear via a femoral vein. A transeptal puncture is pretty safe and easy when the septum is approached from the caudal vena cava. With fluoroscopic guidance, a curved catheter takes a sudden dip and drops into the fossa ovalis when withdrawn along the atrial septum and the tip slips over the intercaval ridge in the medial wall of the right atrium (IR in Fig. 16). A sharpened, shielded guide wire can then be advanced through the thin area of the septum primum and exchanged for a balloon catheter. Ordinarily, puncture ASD's heal over quickly and a piece of the septum must be removed to assure a continuing left to right shunt in experimental dogs. In spontaneous splits however, I think the high LA pressure keeps the shunt open and left to right shunting continues. The raised endocardial lesion (J?) in the right atrium (Fig 16) below the Boxer's acquired ASD (D) also shown in Figure 13 could be a jet lesion but it is hard to imagine a left to right atrial shunt causing that much of a jet effect.

Figure 16. RA view of ASD, jet lesion and intercaval ridge (IR)

Good luck with your case. Keep us ''posted" (hopefully in a few years)

James W. Buchanan, DVM, M Med Sci
Emeritus Professor of Cardiology
Sch Vet Med, Univ of Penn,
Philadelphia PA 19014