41. Reoperative Aortic Valve Replacement- Clinical Scenarios

J Hunter Mehaffey, MD, and Kenan W Yount, MD

Adapted from 1^st edition chapter written by Leora Yarboro, MD, and John A. Kern, MD

Concept

• Discuss the common indications and preoperative workup for Redo AVR
• Operative Planning including Surgical vs. Transcatheter Approach
• Potential pitfalls and management

Chief complaint

“A 65-year-old man with history of prior aortic valve replacement now presents with dyspnea on exertion.”

Differential

The differential in this patient includes structural valve disease, endocarditis, valve thrombus or pannus formation, coronary artery disease, and primary pulmonary disorders such as progressive COPD. Certain conditions such as bicuspid aortic valve, rheumatic disease, young age at the time of replacement with a bioprosthetic valve, and endocarditis can lead to early native valve failure. Stented pericardial valves tend to fail by aortic stenosis of the biologic prosthesis and tend to last longer than porcine valve. Smaller valves resulting in patient-prosthesis mismatch (size < 23 mm or EOAI < 0.85) show trends of earlier re-stenosis and limit future options for valve-in-valve transcatheter approaches. Stentless valves (e.g., Toronto valve) have largely fallen out of favor because of early deterioration; in these cases, the mechanism is usually insufficiency rather than stenosis. Although structural valve deterioration is rare for mechanical valves, mechanical valves in the aortic position may still fail due to endocarditis, pannus formation, stroke, or intolerance of anticoagulation. Although mechanical valves have generally been favored for patients < 55 years-old, they do limit a patient’s lifestyle and future TAVR options.

History and physical

A focused cardiopulmonary history and exam is important. Focus on surgical scars and prior vein harvest sites. Listen for new murmurs, JVD and assess femoral and pedal pulses as this may be important in determining alternative cannulation strategies. Determining a patient’s risk and life goals are important given that the mean 30-day mortality after redo AVR approaches 7.1%. Compared to TAVR, redo AVR is associated with a lower incidence of vascular complications and perivalvular leak at the expense of a higher short-term stroke rate, atrial fibrillation burden, AKI, and major bleeding. As transcatheter technologies develop, the risk-benefit analysis associated with this decision is likely to change.

Tests

• TTE (May consider TEE if concern for endocarditis)
• Coronary imaging

• Look for coronary artery disease that may require CABG; also look for any ostial left or right coronary disease)
• Fluoroscopy study if concern for mechanical valve thrombus/pannus
• CTA chest/abdomen/pelvis

• Looking for transcatheter options, root size, STJ width, coronary heights, and access options. These measurements may also help carefully plan a redo open AVR if necessary.
• For patients going on for redo open AVR, important to access femoral access in case cardiopulmonary bypass is required prior to redo sternotomy or emergently. Also assess for any concomitant aneurysmal disease.
• Standard review of anatomic proximity to bone for redo (innominate vein, aorta, RA, RV).
• For patients with endocarditis, can help clarify the extent of a root abscess (determining whether bovine patch or homograft may be required) or aorto-mitral curtain involvement.
• Carotid duplex
• Pulmonary function testing (PFTs)
• Vein mapping (if previous CABG)
• U/A & blood cultures
• Lab: CBC, BMP, LDH (hemolysis), renal insufficiency, T&C
• Obtain previous operative note
  • When was the last operation?
  • What was the manufacturer and size of the last valve?
  • What exactly was the procedure? Was there a root enlargement or other root procedure performed?
  • History of previous CABG? What were the conduits? Do the conduits cross the sternal midline?

Index scenario (additional information)

“Patient has isolated aortic regurgitation on TTE with no significant coronary artery disease.”

The next step is to decide whether the patient would best be suited for valve-in-valve TAVR or redo AVR. Early complication rates from valve-in-valve (ViV) TAVR are acceptable and lower than what would be predicted for redo AVR.  The durability of ViV TAVR remains to be determined.  The above testing and a heart team evaluation will help make the decision.  Endocarditis is an absolute contraindication to TAVR. Mechanical valves preclude ViV TAVR.

If the patient had a prior bioprosthetic valve, it is important to know what the initial gradient was leaving the OR and what the gradients were during the first 1-2 years after the first AVR. If the gradients were always high, then it indicates the patient may have had too small of a valve placed during the index AVR and that they would not be an optimal candidate for ViV TAVR; instead, they may need a redo AVR with a root enlargement. Other flags for caution for ViV TAVR may increase the risk of possible coronary obstruction include inadequate or effaced sinuses, low coronary heights (e.g., < 10 mm), or a prior surgical valve in which the leaflets are on the outside of the frame rather than the inside of the frame (e.g., St. Jude Trifecta). While there is technology to reduce the risk of coronary obstruction in these scenarios, patients at high risk for coronary obstruction need to have a compelling reason not to proceed with redo AVR.

Prior bioprosthetic valves < 23 mm are challenging in that they may require the smallest size TAVR valve to be placed inside of them or they require fracturing of the surgical valve (via high pressure balloon dilation). Again, while such maneuvers have been performed successfully, if you are considering putting in the smallest available TAVR valve or fracturing a prior SAVR valve, then the patient must have a compelling reason not to proceed with redo AVR. Similar technical challenges for ViV TAVR include a narrow STJ or a small root.

One important footnote is that the labelled size of surgical bioprosthetic valve refer to the external diameter of the sewing ring—thus the landing zone of the native valve annulus measured by the surgeon at the time of initial AVR. However, for ViV TAVR, the internal diameter of the sewing ring, the struts, and the leaflets is the more relevant dimension because the TAVR valve is being landed inside (rather than outside) the old surgical valve. The internal diameters of prior SAVR valves can be found from various industry sources and mobile apps, but prior SAVR internal diameter must always be confirmed by CTA or TEE; operative notes notoriously have errors. Fortunately, many patients carry a card in their wallet with the surgical valve information.

Finally, if the prior bioprosthetic valve has failed due to AI rather than AS, then there needs to be a compelling reason not to proceed with redo AVR as currently approved TAVR systems are designed for AS rather than primary AI, although some systems that target AI are under investigation. Finally, other considerations for deciding on ViV TAVR vs redo AVR include the patient’s access options, concomitant pathology (e.g., co-existing aneurysm > 5.5 cm, multivessel CAD better treated with CABG), age and life-expectancy. All patients being evaluating for ViV TAVR vs redo AVR deserve careful consideration by a multidisciplinary heart team.

By the same token, there are several risk factors that should lead on to strongly consider ViV TAVR over redo AVR if the former is feasible. These include older age (e.g., > 75-80 years-old) prior CABG with patent LIMA (especially if near the sternum), heart failure with reduced ejection fraction (HFrEF), poor conditioning, calcified ascending aorta, and renal failure. Such risk factors can severely affect the early and long-term survival and recovery from redo AVR.

It is debated whether self-expanding valves or balloon expandable valves are best for ViV TAVR, although there is some bias in real world practice for self-expanding valves in ViV scenarios. First, self-expanding valves have slightly more favorable post-TAVR gradients, which is important given that the valve area is being reduced by putting a valve inside a valve. Second, the higher pacemaker risk of self-expanding TAVR valves is less of a concern for ViV procedures since the conduction system is somewhat protected by the old bioprosthetic valve. Third, the aorta may be calcified in redo settings and thus the self-expanding valve may be less likely to traumatize a root that is calcified; this is also important if there is any concomitant aneurysm or if a root enlargement procedure was performed during the index TAVR. Fourth, self-expanding valves do not require rapid pacing for deployment, which may be relevant if the patient has a sicker ventricle due to recurrent AS. One caveat to these biases is that the taller nitinol stent in the current iterations of self-expanding valves (e.g., Evolut R & Evolut Pro) may complicate future coronary access if the patient has CAD.

In conclusion, at present, redo AVR remains the treatment of choice for younger patients at low surgical risk with severe structural valve deterioration.

Operative steps

• Perioperative monitoring
  • TEE, PA catheter, cerebral oximetry, Foley with temp probe, arterial line, R2 pads.
  • Review CT scan for areas of concern – proximity of aorta/right ventricle/innominate vein/atrium/grafts. Check for excessive ascending aortic calcification.
  • Prep legs in case of coronary injury and need for vein.
  • As with any AS case, the scrub nurse should be ready to go and the fellow/attending should be in the room and ready to quickly start given hemodynamic instability.
• Cannulation
  • Develop a plan for potential alternate cannulation sites – axillary vs. femoral
    • Be sure to at least prep in the right axillary access site during draping.
    • While it is not always necessary, it is probably safest practice to either have the femoral vessels exposed, or percutaneous wire access in case of emergency prior to opening the sternum. At the very minimum, small (4F) sheaths should be placed in the femoral vessels.
  • Be able to discuss which patients you would go on bypass for prior to opening the sternum (i.e., CT scan demonstrates live grafts in close proximity to sternum or previous aortic injury on entry). A reasonable approach would be the following:
    • Structures a safe distance away: expose the femoral vessels or access percutaneously prior to sternotomy (if you injure something you can heparinize and cannulate).
    • Structures in proximity: Depending on cannulation strategy (axillary vs. femoral), have vessels ready to go (wires in, graft sewn, etc).  This can be done using small doses of heparin. 
    • Critical grafts under the sternum, aorta close to the sternum, high PA pressures with RV close to sternum: go on bypass vein prior to sternotomy (if you injure something you can empty out and even drop flows as needed) – the main drawback to fully heparinizing and cannulating prior to the dissection is bleeding.  One area of discussion/debate is whether to cool here.  Cooling can result in fibrillation (~28-32C) which can be difficult to deal with in a redo with an undissected heart. 
    • Calcified aorta: refer to porcelain aorta chapter.
    • Other reasons to cannulate peripherally electively prior to or after sternotomy deal with “real estate” – usually the ascending aorta can be safely cannulated distal or lateral to the prior cannulation site, but if you do not have enough room because of live grafts or it proves too hazardous to expose then cannulate the axillary artery/fem artery. If the atrium is stuck, then proceed to femoral venous or (if additional drainage needed) bicaval cannulation.
  • Myocardialprotection
    • Prefernces vary but some combination of anterograde, retrograde and ostial cardioplegia. If you abandon a retrograde catheter, you will be limited to repeated handheld doses unless you use longer duration (e.g., del Nido) cardioplegia. If due to adhesions you cannot dissect enough of the heart out to confirm retrograde catheter placement, there is always the option to directly place it (which requires bicaval cannulation).
      • Aortic insufficiency further necessitates both retrograde cardioplegia to obtain arrest and an LV vent to prevent distension. If there is no or minimal AI, an LV vent is not needed, but other options to facilitate a bloodless field include a PA vent or simply using a Ross tip pump sucker across the aortic valve into the LV.
    • Cool to 32-34°C.
• Dissection
  • Dissect out the right atrium and ascending aorta for cannulation
  • Heparinize, cannulate, clamp, arrest.
  • Previous aortotomy site may be calcified. Open just proximal or distal to it. The prior operative note may indicate whether the prior aortotomy was lower or higher than desired for optimal exposure. Excise valve sharply using a Kocher to help stabilize the valve. Remove all sutures with blade. Use a freer-elevator to gently remove valve. Vigilantly identify and remove all pledgets. Examine the root after removal for any inadvertent injury.
• Valve
  • Size annulus after debriding pannus. Determine whether root enlargement or repair is necessary with a bovine suture. If a bovine patch is needed, it may be more prudent to take bites outside to inside the aorta such that any pledgets around the patch end up outside the aorta rather than inside the aorta. Place annular sutures and tie down valve.
• Close aortotomy
• Place pacing wires (consider a-wires if complete heart block)
• Post-op
  • Redo AVRs have significantly longer operative, bypass and cross clamp times. Consequently, they may have variable degrees of vasoplegia and cardiogenic shock requiring resuscitation. They also have higher rates of heart block so make sure you have well placed and reliable wires.

Potential questions/alternative scenarios

“Patient fibrillates during initial dissection.”

• Two things that cause this include bovie electrocautery and CPB with an incompetent valve. Thus, preoperative planning for this scenario is critical, particularly if you know the patient has at least moderate AI.
  • Have external defibrillator pads on the patient prior to prepping. If you do fibrillate prior to having the heart dissected, shock with external pads at 100-200 Joules.
  • Try to avoid going on CPB until the heart is dissected out. Try to avoid electrocautery near the LV during the initial dissection until you are ready to go on pump and arrest.  Keep cautery low (<45).
  • If the patient becomes hemodynamically unstable, cannulate (centrally or peripherally) and go on pump immediately. If the heart then distends, decompress manually or with an LV vent until you are able to clamp and arrest.
  • Ideally, an LV vent could be placed through the RSPV. However, in an emergent scenario (when the heart distends and defibrillates before you have the heart dissected out) options include (1) stabbing LV apex (if enough exposed) or (2) clamping the aorta and performing an aortotomy combined with direct handheld cardioplegia to achieve arrest.

“Patient with previous CABG.”

Preoperative cardiac catheterization is crucial in these patients.

• If a prior LIMA is patent, have the choice to keep it patent and use frequent antegrade and retrograde cardioplegia and consider cooling the patient to 28-30^oC. If you choose to clamp the LIMA, do not risk injuring it during your dissection. If you absolutely had to reduce LIMA flow for myocardial protection, from the left pleural space you can place a tonsil across all tissue where the LIMA is expected to be and that may be enough to slow down the flow, obviating the need to specifically isolate and clamp it. You must be ready to clamp and arrest prior to attempting patent LIMA dissection because any injury will require immediate CPB support and myocardial protection.
• You will likely need to mobilize previous grafts to allow for aortotomy. If unable to mobilize graft and it is still patent may need to transect and perform bypass. The proximal can be to the aorta or hood of the vein graft and the distal can be to the vein graft or a new distal. Ligate the old graft.

“Unable to access the root and get exposure of the valve through your aortotomy”

Consider aortic transection to enable exposure and allow better access to the old valve.

“Tear in aorta/outflow tract when removing old valve.”

Valve may be incorporated, or previous dissection may have been extensive. Be prepared with bovine pericardium or even homograft for reconstruction if necessary.

“Right Coronary ostia injured during excision of valve.”

Be prepared by prepping in the legs on all redo operations. Perform bypass. Reference cath to be sure that bypass is distal to any native disease.

“You notice significant RV dysfunction coming off bypass.”

Always assume any regional functional problem coming off bypass is a coronary problem. Coronary problems are obviously life-threatening and cannot be ignored. If you’ve proven to yourself that the problem is not air in the coronary, the safest plan is to assume you’ve obstructed to the right coronary with your newer, larger valve and thus proceed with an expeditious vein graft to the right coronary artery. Do not instead try to reopen the aorta and re-replace the valve. This situation also highlights the importance of ensuring you can see the os of both coronary after the valve is lowered into place. This can also be double-checked by giving retrograde cardioplegia and looking for blood return from the os.

“After redo-AVR, you find it difficult to close the aortotomy due to tension on the aorta.”

In redo settings, the aorta can be less compliant and thus a large prosthesis can place undue tension on the suture line. If this is the case, augment your closure with a bovine pericardial patch to minimize tension on the suture line.

“You instead have a 77-year-old patient with a mechanical valve that has failed and 2+ MR with no obvious mitral abnormalities.”

The patient probably has functional mitral regurgitation that is accentuated by severe AS. The safest option is performing an expeditious redo AVR and seeing if the mitral regurgitation will decrease thereafter. If were to persist several weeks afterwards, there are other options down the road one could consider for functional mitral regurgitation (e.g., MitraClip).

Pearls/pitfalls

• Valve type, valve size, implantation technique, and root enlargement during the first AVR can affect the technical success of either a redo AVR operation or ViV TAVR procedure.
• The differential diagnosis of the failed prior surgical valve must be considered, and it is critical to rule out (1) high gradients as a result of patient prosthesis mismatch (PPM) and (2) endocarditis if the patient is being considered for ViV TAVR.
• Redo AVR is still the treatment of choice for younger patients with favorable risk factors.
• Review echo and cardiac catheterization to ensure no other procedures or evaluation is necessary.
• CTA Chest/Abdomen/Pelvis is critical for planning. If it cannot be obtained due to renal disease (in which case they are likely better suited by ViV TAVR), then non-contrast CT ± TEE at the very minimum.
• Obtain and study previous operative note.
• Discuss mechanical vs. tissue valve preoperatively.
• After reviewing the CTA and patient’s BSA, have an idea of what size valve the patient currently has and what their anatomy can accommodate, as well as what size the patient needs or can tolerate.
• Careful preoperative planning for cannulation and myocardial protection is critical, particularly if re-entry will be difficult or if AI is present.
• Have the femoral vessels exposed or accessed before sternotomy and have cross-matched blood in the room (typically 4 units).
• Injury to coronary arteries à bypass with vein grafts.
• Increased risk of heart block postoperatively à place v-wires and maybe even a-wires.
• Be prepared for complications, e.g., need for aortic root enlargement/replacement.
• Any complication that can occur during primary AVR is more likely to occur in a redo.

Suggested readings

1.           Hirose H, Gill IS, Lytle BW. Redo-aortic valve replacement after previous bilateral internal thoracic artery bypass grafting. Ann Thorac Surg 2004;78(3):782-785.

2.           Potter DD, Sundt TM, 3rd, Zehr KJ et al. Operative risk of reoperative aortic valve replacement. J Thorac Cardiovasc Surg 2005;129(1):94-103.

3.           LaPar DJ, Yang Z, Stukenborg GJ et al. Outcomes of reoperative aortic valve replacement after previous sternotomy. J Thorac Cardiovasc Surg 2010;139(2):263-272.

4.           Kaneko T, Vassileva CM, Englum B et al. Contemporary outcomes of repeat aortic valve replacement: A benchmark for transcatheter valve-in-valve procedures. Ann Thorac Surg 2015;100(4):1298-1304; discussion 1304.

5.           Ejiofor JI, Yammine M, Harloff MT et al. Reoperative surgical aortic valve replacement versus transcatheter valve-in-valve replacement for degenerated bioprosthetic aortic valves. Ann Thorac Surg 2016;102(5):1452-1458.

6.           Lau C, Gaudino M, Mazza A, Munjal M, Girardi LN. Reoperative aortic valve replacement in a previous biologic composite valve graft. Ann Thorac Surg 2016;102(5):e477-e480.

7.           James Edelman J, Khan JM, Rogers T et al. Valve-in-valve tavr: State-of-the-art review. Innovations (Phila) 2019:1556984519858020.

8.           Sedeek AF, Greason KL, Sandhu GS, Dearani JA, Holmes DR, Jr., Schaff HV. Transcatheter valve-in-valve vs surgical replacement of failing stented aortic biological valves. Ann Thorac Surg 2019;108(2):424-430.

9.           Webb JG, Murdoch DJ, Alu MC et al. 3-year outcomes after valve-in-valve transcatheter aortic valve replacement for degenerated bioprostheses: The partner 2 registry. J Am Coll Cardiol 2019;73(21):2647-2655.