3.1. Carina system.

The usual Carina system implantation is performed through a post-auricular incision with a posterior small atticotomy (about 2 cm wide) [18] to expose the body of the incus and the head of the malleus. The arm of the mounting bracket of the device can be modified to place the device on the incus and is fitted to the mastoid cortex using bone screws. Bone beds for the device and the microphone must be drilled so that the electronics capsule and the microphone can be positioned and secured [9], [19].

There are 3 convenient microphone placement locations: anterior and superior to the external auditory canal (temporalis region), posterior to the external auditory canal (retro-auricular region), and on the mastoid tip. It is noteworthy that the microphone is very sensitive to changes in the tissue thickness over time, resulting in feedback. Thus, it may be placed in a region of minimal tissue thickening changes during head and neck movements, which is not the case of the mastoid tip. It is also necessary to avoid the contraction effects of the sternocleidomastoid muscle [19]. Moreover, there is no consensus regarding the optimal placement of the microphone [9].

The tip of the transducer is advanced into the hole on the incus and the positioning is evaluated using software specifically developed by Otologics (Transducer Loading Assistant) to ensure correct placement of the device [17].

The tip of Carinás transducer can be crimped to different tips such as stapes head, stapes superstructure, stapes footplate or round window [9]. The round window may be used in patients for whom multiple ossiculoplasty procedures have been unsuccessful and particularly when the stapes footplate is fixed or no longer accessible [9], [17]. This transducer tip can also be extended by applying a small titanium ball and placed on the body of the incus [9].

The implantation of Carina to stimulate the round window is performed through a post-auricular incision with a posterior tympanotomy to expose the round window niche. The round window membrane movement is checked by the mobilization of the stapes or the long process of the incus. The transducer is placed on the round window with or without placement of the incus. The tip of transducer is adapted by clipping modified total ossicular replacement prosthesis (TORP) to the end of the transducer or can be put in contact with the staples footplate or even coupled to a stapedotomy piston. The other end of the TORP is placed on a fascia graft protecting the round window membrane. This surgical technique allows the implantation of device despite of a nonfunctioning ossicular chain and an abnormal middle ear anatomy as long as round window membrane is present to receive the tip of the transducer [6], [9].

3.2. Esteem Device.

The ear with the poorest functional hearing is selected for implantation. If both ears are equal in performance, the candidate can choose the side to be implanted. The procedure is performed under general anesthesia [1].

A post-auricular incision is made and a bone recess is fashioned posterior to the mastoid to house the sound processor. A tympanomastoidectomy is performed widely exposing the facial recess to accommodate the driver. The chorda tympani nerve is sacrified in about 60% of the cases [13.14]. The intact ossicular motion can be measured using a laser doppler vibrometer. The incus and stapes are disarticulated and the distal 1 to 3 mm of the long process of the incus is gently removed using either malleus nipper or a cutting laser to prevent a mechanic feedback. Transducers are contained in the mastoid cavity with hydroxyapatite cement so their piezoelectric crystals are positioned. The sensor is interfaced with the incus using glass ionomeric cement and the driver is cemented to the stapes [1], [13].

In a study to detect the site of maximum ossicular motion that would be optimal for attachment of the sensor portion of the protesis, Chung et al. [2] used a laser doppler vibrometer to measure the vibrational responses at 7 locations on the middle ear ossicles. They observed that maximum vibrational motion of the middle ear is deliverable to the piezoelectric transducer of Esteem through the superior part of the malleus head, on the lateral part of the incus body, and on the superior part of the incus body near the incudomalleal joint.

After implant placement, the entire system is tested and postoperative functional gain is estimated. If gain is deemed inadequate, the implant is repositioned to improve performance.

Both devices, Carina and Esteem are turned on in about 6 to 8 weeks after implantation [6], . There is an implant programming called “commander” that is used by audiologist for follow-up to program each patient’s device. The patient himself can also modify the filtering of background noise levels, adjust volume, and place the device on stand-by mode using a remote control component called “personal programmer” [1], [22], [23].

5.1. Fully implantable devices compared with hearing aids.

Chen et al. [1] reported that the functional gain for the Esteem implanted in was similar to HAs at 0.5, 1, 2 and 3 KHz and this gain decreased at 3 kHz. Monini et al. [4] found that a mean gain difference of 13 dB favorable to the Esteem device, compared to HAs in both groups moderate-to-severe and severe-to-profound sensorineural hearing loss (SNHL), but with no statistical significance. Otherwise, Memari et al. [20] (n = 10) showed that average gain in 5 frequencies ranged from no gain in 1 patient to 20 dB in another 1. Barbara et al. [11] referred improvement from 70 dB to 48 dB in the whole sample. Shohet et al. [15] reported a functional gain of 22 dB with this same device.

In the study by Lefebvre et al. [6], the 6 patients who underwent implantation of Carina transducing sound via the round window showed essentially the same thresholds for frequencies above 3 kHz comparing to conventional HAs.

The average functional gain at 500, 1 k and 2 k Hz was found to be 35.6 dB and 35.0 dB in 6 patients with Carina and the conventional HA, respectively, revealing an insignificant difference according to Kam et al. [22] Gain on frequencies above 3 kHz is generally limited but residual hearing at such frequencies can still be maintained or slightly improved [14].

Zenner and Rodriguez [24] (n = 50) reported that the average functional gain varied by frequency between 25 and 30 dB for audiometric test frequencies of 5 and 6 KHz.

The word recognition (WR) improvement for Esteem comparing to conventional HA’s varied among the studies. While Chen et al. [1] found an index of improvement with Esteem of 17% compared to HAs, Kraus et al. [13] found that 62% of their subjects (n = 52) had improvement, 27% were the same, and 11% were worse. Neither studies presented information of statistical significance. Monini et al. [4], showed that WR raised to 55% with conventional HA and to 66% with the Esteem in the group with moderate-to-severe SNHL. In the group with severe-to-profound SNHL the improvement was to 46% with conventional HA and to 57% with the Esteem. However the difference was not statistically significant. Carina results in WR were conflicting in all articles studied. Zenner and Rodriguez [24] (n = 50) reported that a significant improvement, up to 82% correct, in speech discrimination scores was obtained. The case report of Deveze et al. [26] showed improvement from 40% to 80% with Carina device at 65 dB, comparing to the patient’s own HAs. Nevertheless, Kam et al. [22] found insignificant difference between conventional HAs and the Carina in terms of WR both in quiet and in noise.

The APHAB (Abbreviated Profile of Hearing Aid Benefit) scale was applied before and after implantation in some studies [1], [10], [11], [13], [18], [19], and it consists of questionnaires in four areas: EC (Ease of Communication), BN (Background Noise), RV (Reverberation) and AV (Aversiveness) [1], [10], [19]. According to Instruction for Manual Scoring of the APHAB, a significant benefit has occurred if a difference of 22% is obtained for the EC, RV, or BN score [1]. Jenkins et al. [19] found that patients preferred the Carina over their own HAs for all questionnaires. According to Kam et al. [22], the APHAB scale for Carina and HA were 84.9 and 37.2, respectively.

5.2. Fully implantable devices compared with unaided.

Comparing the outcomes of Carina implantation to unaided patients, the mean functional gain was 29 dB and 24 dB (p  = 0.0004) in Martin et al. [9] and Bruschini et al. [17] respectively. Tringali et al. [7] showed a mean improvement of 39 dB with no information about statistical significance. Bruschini et al. [18] reported a mean functional gain of 26.4 dB (p = 0.0000001).

Five patients suffering from congenital auricular atresia submitted to Carina had functional gain of 36 dBHL in 4 frequencies by pure tone audiometry (1, 2, 3 e 4 kHz) as shown by Siegert et al. [8].

Kraus et al. [13] (n = 57) studied the 12-month results of the Esteem and found that the functional gain was 27 dB for 48 patients and 4 of them were stable at ±10 dB. Gerard et al. [14] found that the mean gain was 25±11 dB and the best results were obtained at frequencies between 500 and 3 k Hz.

According to Gerard et al. [14], the Esteem device, when compared to unaidaded, showed a mean WR gain of 64±33% at 50 dB SPL. At 50 dB, the WR improved from 10% at unaided condition and 23% with HA to 78% with Esteem according to Shohet et al. [15] A great improvement was reported by Barbara et al. [11] (42% to 79% in a group with moderate hearing loss and 30% to 72% in a group with severe hearing loss). Murali et al. [22] showed that postoperative WR on an average for all their 3 patients for closed set and open set were 100% and 95% respectively. All Carina results in WR comparing to unaided condition showed improvement. WR mean melioration according to the studies of Bruschini et al.[17], [18] were 18% to 58% [17] and from 32.5% to 68.75% [18].

When the APHAB scale was applied, Martin et al. [9] showed significant benefit for Carina over unaided conditions for EC (from 49.8 to 19.9%) and RV (from 57.7 to 44.8%). AV increased (25.8 to 38.6%). Chen et al. [1] found that the average score for AV was −33 and −6 for HAs and Esteem compared to unaided condition, respectively, what means a 27% improvement of Esteem over the HAs. On this same comparison, Kraus et al. [13] showed that the mean difference on the global scale was 8.9±2.6 (p<0.01). The APHAB questionnaire revealed 85% of satisfaction improvement with Esteem compared to HAs in the 4 subscales in the study of Gerard et al. [14] Their 2 dissatisfied patients underwent revision surgery for poor functional results.

5.3. Fully implantable devices activated and inactivated.

The middle and inner ear conditions were evaluated by some authors. Tringali et al. [21], Martin et al. [9] and Bruschini et al. [18] found no significant changes postoperatively, indicating minimal surgical trauma during Carina implantation. Jenkins et al. [19] observed no pre- nor post implantation differences for bone conduction and slight differences in pre- and post implant for air conduction.

As Esteem implantation induces an additional conductive hearing loss, Monini et al. [4] observed a conductive threshold shift of 35 dB on average over the whole frequency range. Barbara et al. [11], [25] showed a bone conduction threshold worsening from baseline after Esteem implantation. For Kraus et al. [13] the average change was mean –0.8±1.1 dB. At 12 months only one patient had a threshold shift from 55 to 75 dB at 4 kHz. In the other hand, Chen et al. [1] and Gerard et al. [14] showed no significant changes of cochlear function by comparing bone conduction threshold before and after implantation of Esteem.