The authors present a case study questioning the use of the AAIR mode when programming an ICD. They also examine current clinical trial data to see how to best minimize ventricular pacing.

Case Report
       A 67-year-old male with ischemic cardiomyopathy and a history of syncope underwent a dual chamber defibrillator (ICD) implant. After a device recall, it was elected to replace the device locally with a new ICD. Device interrogation revealed high atrial pacing thresholds (3.5 volts @ 0.5 ms). A new ICD was implanted locally and programmed AAIR with high atrial outputs. Several weeks later, the patient presented to our hospital with recurrent syncope. On device interrogation, atrial pacing thresholds exceeded 1.6 msec at 3 Volts and there were no episodes of ventricular tachycardia or ventricular fibrillation logged. Interrogation of the counters over two months indicated 20% atrial sensing, 80% atrial pacing and 100% ventricular sensing. The bradycardia pacing mode was AAIR (60–90 beats/minute). Post-shock pacing was programmed AAIR. On the monitor, the patient had episodes of Wenckebach AV block and 3º AV block in the emergency room with associated symptoms.

Discussion
       Based on the recent interest to minimize right ventricular pacing in ICDs, various programmed modes have been used, including the AAI mode. There is an ongoing, randomized, controlled clinical trial comparing AAI pacing to VVI pacing (DAVID II). This particular case highlights potential issues with regard to AAI pacing in an ICD:
1. There is no detection of ventricular activity during bradycardia and no capability of back-up ventricular pacing;
2. Transient AV block can and does occur in this ICD population. This can cause syncope and collapse;
3. With AAI pacing in an ICD, there already is a ventricular lead and the capability of pacing in the ventricle. It is potentially egregious not to have a method to pace the ventricle under the worse case scenario. While it may be dangerous to pace the ventricle continuously, occasional pacing during AV block would be indicated.
4. There is no post-shock bradycardia pacing at a time when there is a great risk for complete heart block, and if there are pauses, these may lead to long short episodes causing re-initiation of ventricular tachycardia or fibrillation.

       Newer and better methods are coming to the forefront with regards to minimizing ventricular pacing yet offering the capability to pace the ventricle as needed. Various approaches now exist to minimize ventricular pacing including AV Search Hysteresis. Presently there is an ongoing clinical trial — the Intrinsic RV Trial — which is comparing DDDR (with AV Search Hysteresis) versus VVI pacing in ICD recipients; it is evaluating heart failure hospitalizations and total mortality. Another trial, the MVP trial, is comparing a MVP mode (AAI to DDD automatic mode switch as needed) to VVI programming with outcome measures similar to the INTRINSIC RV study.

Conclusion
       Until these trials (including the DAVID II trial) are completed, it makes sense not to program an ICD to the AAIR mode due to potential risks to the patient.