To the Editor: We thank Drs. Gstottner and Arnolder for their thoughtful comments regarding our article and for sharing their approach with regard to surgical implantation and fitting of electroacoustic stimulation (EAS) devices. We do have some comments to add to this scholarly exchange. First, we agree that Greenwood's equation (1) is a valuable tool for estimating which region of the basilar membrane is maximally displaced by a given acoustic frequency at soft levels. However, there is no guarantee that an electrode located in a specific region of the basilar membrane will actually stimulate neurons tuned to the exact same frequency that causes maximal displacement of that basilar membrane location. This happens for at least 3 reasons. First, in the normally functioning cochlea, the place of stimulation varies with stimulus intensity (2), raising the possibility that Greenwood's equation may only be appropriate for threshold-level stimuli. Second, there is no ideal way to predict neural survival for any cochlear region, suggesting that electrical stimulation of a particular location of the cochlea may actually stimulate neurons tuned to other frequency regions. Finally, recent data suggest that a frequency map based on spiral ganglion stimulation differs from that based on the place of excitation in the organ of Corti (3,4), and the place of excitation with a cochlear implant is likely to be predominantly in the spiral ganglion rather than in the cochlea. Drs. Gstottner and Arnolder have identified the crucial issue in EAS: the trade-off between possible improvements in speech perception outcomes and possible loss of hearing. It seems that both of these factors may be influenced by insertion depth, but in opposite directions. On one hand, deeper insertions may result in improved speech perception outcomes, as shown by same-subject comparisons (5) and by across-subject comparisons (6). The latter data are particularly compelling, showing huge differences in speech perception outcomes between 3 subjects who received a 10-mm electrode and 3 others who received a 6-mm electrode. The addition of electrical stimulation to bilateral acoustic hearing aids allowed 10-mm users to increase their consonant recognition by 40 percentage points, and their consonant-nucleus-consonant words scores from approximately 30 to 90%, whereas the 6-mm users only improved their consonant recognition by a meager 10 percentage points (consonant-nucleus-consonant word scores were not reported). A later report with a much greater number of subjects (7) confirmed the substantial gains obtained by users of the 10-mm device and also indicated that the percepts produced by the 6-mm device were reported as "unpleasant." On the other hand, deeper cochlear insertion may increase the risk of a patient losing his or her residual hearing. Studies measuring insertion trauma along the outer cochlear wall in human temporal bones have shown a higher incidence of trauma with deeper insertion depths (8-10), whereas atraumatic insertion of the 10-mm array has been noted (11). An alternate possibility would be the development of an appropriate perimodiolar approach for users of electroacoustic hearing because this approach may help reduce the degree of insertion trauma (12). We are glad to hear that Drs. Gstoettner and Arnolder are achieving a high rate of hearing preservation with the new electrode they are using, and we look forward to the publication (and replication) of those results in the peer-reviewed literature. Finally, it should be kept in mind that devices from different manufacturers may require different fitting procedures, and such factors can also affect outcomes in patients with electroacoustic hearing. In any event, we are very appreciative for this recent commentary from Gstoettner and Arnolder as to how they handle instances in which residual hearing is lost in EAS patients. Clearly, this is a troublesome issue for a small subset of patients, and we welcome as much information as possible with regard to how to best treat these individuals. Matthew B. Fitzgerald, Ph.D. Elad Sagi, Ph.D. Michael Jackson, M.S. William H. Shapiro, M.S. J. Thomas Roland, Jr., M.D. Susan B. Waltzman, Ph.D. Mario A. Svirsky, Ph.D. Department of Otolaryngology New York University Medical Center New York, New York, U.S.A.

Objective: To assess word recognition and pitch-scaling abilities of cochlear implant users first implanted with a Nucleus 10-mm Hybrid electrode array and then reimplanted with a full length Nucleus Freedom array after loss of residual hearing. Background: Although electroacoustic stimulation is a promising treatment for patients with residual low-frequency hearing,a small subset of them lose that residual hearing. It is not clear whether these patients would be better served by leaving in the 10-mm array and providing electric stimulation through it, or by replacing it with a standard full-length array. Methods: Word recognition and pitch-scaling abilities were measured in 2 users of hybrid cochlear implants who lost their residual hearing in the implanted ear after a few months. Tests were repeated over several months, first with a 10-mm array, and after, these patients were reimplanted with a full array. The word recognition task consisted of 2 50-word consonant nucleus consonant (CNC) lists. In the pitch-scaling task, 6 electrodes were stimulated in pseudorandom order, and patients assigned a pitch value to the sensation elicited by each electrode. Results: Shortly after reimplantation with the full electrode array, speech understanding was much better than with the 10-mm array. Patients improved their ability to perform the pitch-scaling task over time with the full array, although their performance on that task was variable, and the improvements were often small. Conclusion: 1) Short electrode arrays may help preserve residual hearing but may also provide less benefit than traditional cochlear implants for some patients. 2) Pitch percepts in response to electric stimulation may be modified by experience.