Eur Arch Otorhinolaryngol DOI 10.1007/s00405-014-2932-7
Otology
Stability and survival of bone‑anchored hearing aid implant systems in post‑irradiated patients Mark D. Wilkie · Kathryn A. Lightbody · Ali A. Salamat · Kalyan M. Chakravarthy · David A. Luff · Robert H. Temple
Received: 3 December 2013 / Accepted: 4 February 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Bone-anchored hearing aids (BAHAs) are based on the principle of osseointegration, which is fundamental to implant stability and survival. Previous exposure to ionising radiation may compromise this, as evidenced in relation to dental and craniofacial implants. There is a dearth of data, however, regarding BAHA implant systems in patients with previously irradiated implant sites. We sought, therefore, to investigate implant stability and survival in such patients. Patients were identified retrospectively from our electronic BAHA database. Hospital records were reviewed for demographics; operative technique; complications; and details regarding previous irradiation. Implant stability was assessed by resonance frequency analysis (RFA), generating a numerical value— implant stability quotient (ISQ). Extrapolating from dental studies, successfully loaded implants typically have ISQs of ≥60. Readings were, therefore, interpreted with respect to this. Seven patients were identified for inclusion. Mean time between irradiation and implant insertion was 33 months (range 16–72 months), and mean time from implant insertion to RFA measurement was 41 months (range 3–96 months). Operatively, all patients underwent single-stage procedures under local anaesthesia. One patient suffered a Holger’s grade 2 skin reaction, while two suffered significant skin flap failure, requiring revision procedures. The implant survival rate was 100 %. All ISQ values were >60, with a mean of 66.9 (95 % confidence interval 63.1–70.6). Our data support sufficient osseointegration of BAHA implant systems in post-irradiated patients, but
M. D. Wilkie (*) · K. A. Lightbody · A. A. Salamat · K. M. Chakravarthy · D. A. Luff · R. H. Temple Department of Otorhinolaryngology, Head and Neck Surgery, Countess of Chester Hospital, Chester CH2 1UL, UK e-mail: [email protected]
highlight issues with wound healing. Contemporary soft tissue preservation operative techniques will likely overcome this, facilitating safe and efficacious BAHA insertion in this ever-increasing group of patients. Keywords Bone-anchored hearing aid · Osseointegration · Radiotherapy · Resonance frequency analysis · Implant stability · Implant survival
Introduction Bone-anchored hearing aid (BAHA) implant systems are based on the principle of osseointegration, which is fundamental to implant stability and survival [1]. Osseointegration involves a dynamic process of bone regeneration and remodelling at the bone–implant interface, which is dependent on implant surface characteristics, bone quality, and patient healing behaviour [1]. Ionising radiation may, therefore, compromise this process [2–4], substantiated by the reduced survival rates of osseointegrated dental and craniofacial implants observed in previously irradiated bone in head and neck cancer patients [4–8]. Extrapolating these findings, the stability and survival of BAHA implants may be adversely affected if patients have undergone previous irradiation involving the designated BAHA fixture site. Indeed, this poses a pertinent question as the BAHA surgeon may not infrequently encounter such patients: radiation therapy is a mainstay of management in many head and neck, skull base, and brain malignancies, during which the temporoparietal region is frequently and unavoidably exposed and is associated with aural sequelae amenable to BAHA auditory rehabilitation in a significant proportion of patients [9]. Furthermore, associated skin reactions involving the pinna, external auditory canal, and
13
periauricular region may render the patient unable to wear a conventional air-conduction aid [9, 10]. Despite this, no previous studies have assessed the stability, and only one the survival [10], of BAHA implant systems in patients with post-irradiated implant sites. We sought, therefore, to investigate both objective implant stability and survival in such patients, with the aims of making recommendations about the feasibility of BAHA implantation in this patient population and enhancing our ability to counsel patients regarding expected outcomes.
Materials and methods Patients and setting Eligible patients were identified retrospectively from our department’s electronic BAHA record, which includes all patients undergoing BAHA insertion at our institution from 2006 to 2013. This data was cross-referenced with operative registers to ensure no patients were inadvertently excluded. The indications for BAHA insertion were examined for each case and patients included only if they had undergone previous radiation therapy with fields that would have likely included the implant site. Hospital records were subsequently reviewed in more detail and the following data extracted: patient demographics; operative technique; complications; and details and chronology regarding previous radiation therapy. Our department is the regional tertiary referral centre for adult BAHA implantation in Merseyside and Cheshire, UK, thereby serving a population of approximately 2.5 million in this regard. Operative techniques All implants were performed as single-stage procedures under local anaesthetic by two consultant surgeons who lead the regional BAHA programme (DAL and RHT). Fixtures were positioned 55 mm from the external auditory meatus at an angle 45 degrees postero-superior to the horizontal plane. Until February 2013, both surgeons used the same surgical technique, which involved raising an anteriorly based skin flap using an electrical dermatome and undertaking liberal soft tissue reduction, prior to implant insertion using Cochlear’s Osscora handpiece and foot controller (Cochlear Europe Ltd, Addlestone, UK). Subsequent to February 2013, one surgeon (RHT) amended his surgical technique so as to use the new Dermalock BA400 implant system (Cochlear Europe Ltd, Addlestone, UK), designed specifically for soft tissue preservation. Employing this technique, no skin flap was raised and no soft tissue reduction undertaken: a 2–3 cm linear incision was made to gain
13
Eur Arch Otorhinolaryngol
access and extended to a small cruciate incision around the designated implant site. Outcomes and analysis Implant status was evaluated for the duration of post-operative follow-up in each patient, and any case of implant loss treated as an implant failure: an implant survival rate was calculated on this basis. Implant stability was assessed by resonance frequency analysis (RFA) using the Osstell instrument (Osstell, Gothenburg, Sweden). Measurements were performed in accordance with the manufacturer’s instructions (available at http://www.osstell.com/all-osstell-isq-downloads). The Osstell device translates measurements into a numerical value relating to stability–implant stability quotient (ISQ). ISQ values range from 1 to 100: the higher the ISQ, the greater is the stability. All measurements were performed in triplicate and the mean value used for analysis if there were any inter-measurement discrepancies. Although there are few published reports on RFA for BAHA implant systems, RFA is a widely utilised tool within the dental literature. Extrapolating values from the dental literature, an ISQ of 47–59 may be considered stable for loading, but most studies have demonstrated values of ≥60 in successfully loaded implants. Readings obtained in this study were, therefore, interpreted with respect to these values (see discussion). All data were collated in, and descriptive statistics and 95 % confidence intervals calculated using, Excel for Mac 2011 (Microsoft Corp, Redmond, USA). Ethical considerations All data collection and analysis were carried out in accordance with our institution’s clinical information and audit department regulations and approval.
Results Seven patients (five females, two males) were identified as eligible for inclusion. Mean patient age at implantation was 64 years (range 46–74 years). Mean time between radiation therapy and implant insertion was 33 months (range 16– 72 months), and mean time from implantation to RFA measurement and most recent follow-up was 41 months (range 3–96 months). Demographics and details regarding previous radiation therapy and chronology for individual patients are depicted in Table 1. Patients had undergone radiation therapy either as single modality treatment (n = 6) or as adjuvant therapy (n = 1). Prior to radiotherapy, patient 5 underwent a lateral temporal bone resection and total pinnectomy
Eur Arch Otorhinolaryngol Table 1 Clinicopathological characteristics and chronology of previous radiation therapy for all individual patients Patient
Age (years)
Pathology necessitating RT
Time interval from RT to implantation (months)
Time interval from implantation to RFA (months)
1 2 3 4 5 6
73 74 67 46 60 66
VS Glomus tumour VS NPC SCC temporal bone VS
16 26 24 72 19 36
63 9 12 86 15 96
7
62
VS
38
3
RT = radiation therapy, RFA = resonance frequency analysis, VS = vestibular schwannoma, NPC = nasopharyngeal carcinoma, SCC = squamous cell carcinoma
for squamous cell carcinoma (SCC) of the temporal bone. All patients treated for vestibular schwannoma (VS) (n = 4) had undergone stereotactic radiosurgery. In four cases BAHA implantation was indicated for single-sided deafness, with the remainder performed for mixed hearing loss with a significant conductive component. Of the latter, one patient preferred a BAHA to a conventional air-conduction aid; one suffered from intermittent otorrhoea while using a conventional aid; and as a consequence of total pinnectomy patient 5 was unable to wear an air-conduction aid. One patient suffered from diabetes mellitus, but no other recognised risk factors for implant failure were identified in our study population. Operatively, patients 1–6 underwent BAHA implantation using a traditional technique with soft tissue reduction and split-skin grafting, while patient 7 was implanted with the new Dermalock BA400 system (Cochlear Europe Ltd, Addlestone, UK) using a soft tissue preservation technique (see “Materials and methods” for details). One patient (patient 4) suffered a moderate skin reaction at the implant site (Holger’s grade 2), which resolved with conservative measures. Two patients (patients 5 and 6), however, suffered significant skin flap failure requiring revision procedures. In the case of patient 6, who had undergone stereotactic radiosurgery for VS, this was managed successfully with two skin revision procedures, while patient 5, who was treated with high dose adjuvant radiation for temporal bone SCC, still has ongoing issues with wound healing and exposed bone despite two revision procedures, which unfortunately has precluded wearing of the sound processor. No other complications were observed. There were no implant losses in our study cohort during the follow-up period, giving an implant survival rate of 100 %. All ISQ values obtained were ≥60, with a mean value of 66.9 (95 % confidence interval 63.1–70.6). Figure 1 shows ISQ values for each individual patient with respect to the timing of the radiation therapy in relation to their RFA measurement.
Discussion Hearing loss can be a significant cause for morbidity following radiation therapy to the head and neck region, and as such auditory rehabilitation is of utmost importance [9]. Radiation-induced fibrosis and/or destruction of middle ear components, tympanic mucositis, and cochleitis from radiation vasculitis may result in a wide spectrum of conductive, sensorineural, or mixed losses [9, 10]. This, together with the fact that post-irradiated patients are frequently unable to wear air-conduction aids secondary to skin and soft tissue reactions and/or disrupted anatomy, proffers BAHA as an ideal option for such auditory rehabilitation [9, 10]. Knowledge of the behaviour and retention of BAHA implant systems in these patients is therefore paramount in making informed decisions about the feasibility of the procedure, and indeed for counselling patients on expected outcomes and risks. At present, however, this knowledge is based on one small-scale study examining implant survival, but not stability, in post-irradiated nasopharyngeal carcinoma (NPC) patients [10], and inferences from the dental and craniofacial literature. The lack of published data examining BAHA implants in post-irradiated patients would imply that such patients are a rarity in clinical practice. However, middle ear side effects occur in up to 40 % of cases following head and neck irradiation and approximately one-third of patients develop significant sensorineural hearing loss [9]. A lack of assessment of hearing outcomes and awareness of BAHA as an auditory rehabilitation solution among oncologists may explicate this discrepancy, a perception supported by the general underevaluation of radiation-induced ototoxicity within the oncology literature [9]. Notwithstanding, the number of such patients is only set to increase, with the upsurge in use of radiation therapy, particularly for VS [11]. The principal finding of this study is the demonstration of successful osseointegration, as assessed both by objective implant stability and survival, of BAHA implant systems in post-irradiated patients. Although our study
13
Eur Arch Otorhinolaryngol
Fig. 1 Implant stability quotient (ISQ) values for each individual patient in relation to the time interval between radiation therapy and resonance frequency analysis (RFA) measurement. ISQ values of
≥60 are typically observed in successfully loaded implants, with the dashed line in the figure representing this critical value
is limited to a degree by a small study cohort, which may raise questions over reliability and precludes meaningful statistical analysis to identify factors that may influence implant stability and survival, our survival rate of 100 % is consistent with that of the only study published previously, in which no implant losses were observed in post-irradiated NPC patients [10]. Curiously, this is at odds with the general consensus from the dental and craniofacial literature. Several investigators have reported significantly increased failure rates of osseointegrated implants used to reconstruct craniofacial defects when placed in irradiated bone, with a relative risk as high as 12 [4, 6]. Similarly, although more contentious, observed survival rates of dental implants tend to be reduced in irradiated bone [3, 12, 13]: a recent review identified three studies that demonstrated a statistically significant difference, all reporting a two- to threefold greater risk of implant failure in irradiated bone [4]. Such disparity in findings would not appear to be an artefact of insufficient patient follow-up. Mean patient follow-up in our study and in that of Soo and colleagues [10] was 41 and 39.8 months, respectively, while implant losses in irradiated bone have typically been observed during the first three years after implant surgery, most commonly in the first year [6, 8]. Whilst it is plausible that implant failure may occur in our study cohort at a later date, indeed two of our patients have been followed for less than a year at the time of writing, our mean follow-up period captures the time of most frequent implant loss. Also from a chronological standpoint, the time interval from irradiation to implant surgery has been implicated in influencing implant survival [7]. Radiobiologically, the optimal
timing for implant surgery would coincide with the abatement of acute tissue reactions and the commencement of established healing, which, when accounting for the risk for severe tissue reactions caused by surgical trauma in the irradiated tissue, would likely fall 6–12 months following irradiation [7]. The high survival rate observed in our study may be attributable to the relatively long interval between irradiation and implantation in our patient group: mean interval was 33 months and range 16–72 months. The influence of anatomical location on implant survival may also go some way to reconcile the contrariety in findings. Several studies have shown that osseointegrated implants are not uniformly successful in the craniofacial region, with failure rates varying between sites [4, 14]. Implant survival rates noted in the literature for several locations in non-irradiated bone are as follows: mastoid >95 %, orbital 35 %–91 %, and nasal 71–81 % [4]. Interestingly, most investigators have observed highest survival rates in implants placed in the temporal region [4, 6, 14], which may account for our high survival rates and those reported by Soo et al. [10]. Whilst it is not entirely clear why such differences between anatomical sites may exist, it is likely, and indeed intuitive, that this would relate to the structure and volume of the bone, particularly the proportion of compact relative to trabeculated bone [15]. The fact that craniofacial implant survival tends to be lowest in the nasal floor and medial orbital region, where bone is thinnest and consists mainly of a loose trabecular structure, lends credence to this notion [4, 6]. A further factor which may account for the high implant survival rate observed in our study is radiation dosimetry.
13
Eur Arch Otorhinolaryngol
Although in craniofacial applications the importance of radiation dose in determining implant survival has been difficult to assess, owing to small study cohorts [4], a dosedependent relationship has been demonstrated [5]. Specifically, cumulative radiation effect (≥30) as a measure of radiation dose was significantly related to implant failure [5]. Similarly, Visch et al. reported a greater than twofold reduction in long-term survival of dental implants in oral cancer patients who had received radiation doses in excess of 50 Gy [16], while a systematic review of 19 studies of dental implants in irradiated bone noted no failures with radiation doses below 45 Gy [8]. Although unfortunately we were unable to obtain data on specific radiation doses for our study population, based on the typical pathologyspecific radiotherapy regimes, it is presumed that most patients included in our study would have received relatively low radiation doses. Whilst doses for the treatment of NPC and temporal bone SCC (patients 4 and 5 respectively) would have been in the order of 60 Gy [17, 18], patients treated with stereotactic radiosurgery for VS or glomus tumours (71 % of patients in this study) would typically have been exposed to marginal doses of approximately 15 Gy [19, 20]. In addition to implant survival, objective implant stability was also assessed in our patient group, which to the best of our knowledge is the first report of such for BAHA implant systems in post-irradiated patients. RFA is a method to assess the stability of the bone–implant complex at any point in time and provides a numerical value relating to stability, and was the chosen method in our study owing to its non-invasive nature and well-documented use within the dental literature. Although there are few published reports of RFA values in human temporal bone for comparison, RFA measurement was first introduced into implant dentistry in the 1990s and has been used extensively to assess dental implant stability in mandibular and maxillary bone and guide timing of implant loading [21, 22]. Whilst it has been suggested that the trend in RFA values over time, rather than an absolute RFA value, is more relevant in determining long-term implant success and lifetime stability [23], the comprehensive data from the dental literature implies that ISQs of ≥60 are found in the majority of successfully loaded implants, while ISQs as low as 47–59 may be considered stable for loading [24, 25]. On this basis, comparing values obtained in the present study with respect to these seems entirely tenable, particularly as the objective of our study was to determine whether implants had attained sufficient stability in irradiated bone rather than using these values to predict future implant behaviour. In further support of this, findings from studies of BAHA implant systems examining changes in RFA values with time indicate that a plateau of stability is reached and maintained at approximately 1 month post-implantation
[1], which when applied to our study population, with intervals from implantation to RFA measurement ranging from three to 96 months, would insinuate that in all cases a steady state of stability had been achieved. Importantly, ISQ values were >60 in all patients studied, implying that commensurable stability is attained in post-irradiated patients. Examining prospective stability measurements in such patients, however, presents an interesting area for future research to characterise the process and chronology of osseointegration in post-irradiated bone. Compromised wound healing is a well-recognised consequence of radiation therapy and is a common and challenging clinical problem. Consonant with this, two patients in our study (28.6 %) suffered significant skin flap failure leaving exposed bone around the abutment site, a considerably higher rate than the
Otology
Stability and survival of bone‑anchored hearing aid implant systems in post‑irradiated patients Mark D. Wilkie · Kathryn A. Lightbody · Ali A. Salamat · Kalyan M. Chakravarthy · David A. Luff · Robert H. Temple
Received: 3 December 2013 / Accepted: 4 February 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Bone-anchored hearing aids (BAHAs) are based on the principle of osseointegration, which is fundamental to implant stability and survival. Previous exposure to ionising radiation may compromise this, as evidenced in relation to dental and craniofacial implants. There is a dearth of data, however, regarding BAHA implant systems in patients with previously irradiated implant sites. We sought, therefore, to investigate implant stability and survival in such patients. Patients were identified retrospectively from our electronic BAHA database. Hospital records were reviewed for demographics; operative technique; complications; and details regarding previous irradiation. Implant stability was assessed by resonance frequency analysis (RFA), generating a numerical value— implant stability quotient (ISQ). Extrapolating from dental studies, successfully loaded implants typically have ISQs of ≥60. Readings were, therefore, interpreted with respect to this. Seven patients were identified for inclusion. Mean time between irradiation and implant insertion was 33 months (range 16–72 months), and mean time from implant insertion to RFA measurement was 41 months (range 3–96 months). Operatively, all patients underwent single-stage procedures under local anaesthesia. One patient suffered a Holger’s grade 2 skin reaction, while two suffered significant skin flap failure, requiring revision procedures. The implant survival rate was 100 %. All ISQ values were >60, with a mean of 66.9 (95 % confidence interval 63.1–70.6). Our data support sufficient osseointegration of BAHA implant systems in post-irradiated patients, but
M. D. Wilkie (*) · K. A. Lightbody · A. A. Salamat · K. M. Chakravarthy · D. A. Luff · R. H. Temple Department of Otorhinolaryngology, Head and Neck Surgery, Countess of Chester Hospital, Chester CH2 1UL, UK e-mail: [email protected]
highlight issues with wound healing. Contemporary soft tissue preservation operative techniques will likely overcome this, facilitating safe and efficacious BAHA insertion in this ever-increasing group of patients. Keywords Bone-anchored hearing aid · Osseointegration · Radiotherapy · Resonance frequency analysis · Implant stability · Implant survival
Introduction Bone-anchored hearing aid (BAHA) implant systems are based on the principle of osseointegration, which is fundamental to implant stability and survival [1]. Osseointegration involves a dynamic process of bone regeneration and remodelling at the bone–implant interface, which is dependent on implant surface characteristics, bone quality, and patient healing behaviour [1]. Ionising radiation may, therefore, compromise this process [2–4], substantiated by the reduced survival rates of osseointegrated dental and craniofacial implants observed in previously irradiated bone in head and neck cancer patients [4–8]. Extrapolating these findings, the stability and survival of BAHA implants may be adversely affected if patients have undergone previous irradiation involving the designated BAHA fixture site. Indeed, this poses a pertinent question as the BAHA surgeon may not infrequently encounter such patients: radiation therapy is a mainstay of management in many head and neck, skull base, and brain malignancies, during which the temporoparietal region is frequently and unavoidably exposed and is associated with aural sequelae amenable to BAHA auditory rehabilitation in a significant proportion of patients [9]. Furthermore, associated skin reactions involving the pinna, external auditory canal, and
13
periauricular region may render the patient unable to wear a conventional air-conduction aid [9, 10]. Despite this, no previous studies have assessed the stability, and only one the survival [10], of BAHA implant systems in patients with post-irradiated implant sites. We sought, therefore, to investigate both objective implant stability and survival in such patients, with the aims of making recommendations about the feasibility of BAHA implantation in this patient population and enhancing our ability to counsel patients regarding expected outcomes.
Materials and methods Patients and setting Eligible patients were identified retrospectively from our department’s electronic BAHA record, which includes all patients undergoing BAHA insertion at our institution from 2006 to 2013. This data was cross-referenced with operative registers to ensure no patients were inadvertently excluded. The indications for BAHA insertion were examined for each case and patients included only if they had undergone previous radiation therapy with fields that would have likely included the implant site. Hospital records were subsequently reviewed in more detail and the following data extracted: patient demographics; operative technique; complications; and details and chronology regarding previous radiation therapy. Our department is the regional tertiary referral centre for adult BAHA implantation in Merseyside and Cheshire, UK, thereby serving a population of approximately 2.5 million in this regard. Operative techniques All implants were performed as single-stage procedures under local anaesthetic by two consultant surgeons who lead the regional BAHA programme (DAL and RHT). Fixtures were positioned 55 mm from the external auditory meatus at an angle 45 degrees postero-superior to the horizontal plane. Until February 2013, both surgeons used the same surgical technique, which involved raising an anteriorly based skin flap using an electrical dermatome and undertaking liberal soft tissue reduction, prior to implant insertion using Cochlear’s Osscora handpiece and foot controller (Cochlear Europe Ltd, Addlestone, UK). Subsequent to February 2013, one surgeon (RHT) amended his surgical technique so as to use the new Dermalock BA400 implant system (Cochlear Europe Ltd, Addlestone, UK), designed specifically for soft tissue preservation. Employing this technique, no skin flap was raised and no soft tissue reduction undertaken: a 2–3 cm linear incision was made to gain
13
Eur Arch Otorhinolaryngol
access and extended to a small cruciate incision around the designated implant site. Outcomes and analysis Implant status was evaluated for the duration of post-operative follow-up in each patient, and any case of implant loss treated as an implant failure: an implant survival rate was calculated on this basis. Implant stability was assessed by resonance frequency analysis (RFA) using the Osstell instrument (Osstell, Gothenburg, Sweden). Measurements were performed in accordance with the manufacturer’s instructions (available at http://www.osstell.com/all-osstell-isq-downloads). The Osstell device translates measurements into a numerical value relating to stability–implant stability quotient (ISQ). ISQ values range from 1 to 100: the higher the ISQ, the greater is the stability. All measurements were performed in triplicate and the mean value used for analysis if there were any inter-measurement discrepancies. Although there are few published reports on RFA for BAHA implant systems, RFA is a widely utilised tool within the dental literature. Extrapolating values from the dental literature, an ISQ of 47–59 may be considered stable for loading, but most studies have demonstrated values of ≥60 in successfully loaded implants. Readings obtained in this study were, therefore, interpreted with respect to these values (see discussion). All data were collated in, and descriptive statistics and 95 % confidence intervals calculated using, Excel for Mac 2011 (Microsoft Corp, Redmond, USA). Ethical considerations All data collection and analysis were carried out in accordance with our institution’s clinical information and audit department regulations and approval.
Results Seven patients (five females, two males) were identified as eligible for inclusion. Mean patient age at implantation was 64 years (range 46–74 years). Mean time between radiation therapy and implant insertion was 33 months (range 16– 72 months), and mean time from implantation to RFA measurement and most recent follow-up was 41 months (range 3–96 months). Demographics and details regarding previous radiation therapy and chronology for individual patients are depicted in Table 1. Patients had undergone radiation therapy either as single modality treatment (n = 6) or as adjuvant therapy (n = 1). Prior to radiotherapy, patient 5 underwent a lateral temporal bone resection and total pinnectomy
Eur Arch Otorhinolaryngol Table 1 Clinicopathological characteristics and chronology of previous radiation therapy for all individual patients Patient
Age (years)
Pathology necessitating RT
Time interval from RT to implantation (months)
Time interval from implantation to RFA (months)
1 2 3 4 5 6
73 74 67 46 60 66
VS Glomus tumour VS NPC SCC temporal bone VS
16 26 24 72 19 36
63 9 12 86 15 96
7
62
VS
38
3
RT = radiation therapy, RFA = resonance frequency analysis, VS = vestibular schwannoma, NPC = nasopharyngeal carcinoma, SCC = squamous cell carcinoma
for squamous cell carcinoma (SCC) of the temporal bone. All patients treated for vestibular schwannoma (VS) (n = 4) had undergone stereotactic radiosurgery. In four cases BAHA implantation was indicated for single-sided deafness, with the remainder performed for mixed hearing loss with a significant conductive component. Of the latter, one patient preferred a BAHA to a conventional air-conduction aid; one suffered from intermittent otorrhoea while using a conventional aid; and as a consequence of total pinnectomy patient 5 was unable to wear an air-conduction aid. One patient suffered from diabetes mellitus, but no other recognised risk factors for implant failure were identified in our study population. Operatively, patients 1–6 underwent BAHA implantation using a traditional technique with soft tissue reduction and split-skin grafting, while patient 7 was implanted with the new Dermalock BA400 system (Cochlear Europe Ltd, Addlestone, UK) using a soft tissue preservation technique (see “Materials and methods” for details). One patient (patient 4) suffered a moderate skin reaction at the implant site (Holger’s grade 2), which resolved with conservative measures. Two patients (patients 5 and 6), however, suffered significant skin flap failure requiring revision procedures. In the case of patient 6, who had undergone stereotactic radiosurgery for VS, this was managed successfully with two skin revision procedures, while patient 5, who was treated with high dose adjuvant radiation for temporal bone SCC, still has ongoing issues with wound healing and exposed bone despite two revision procedures, which unfortunately has precluded wearing of the sound processor. No other complications were observed. There were no implant losses in our study cohort during the follow-up period, giving an implant survival rate of 100 %. All ISQ values obtained were ≥60, with a mean value of 66.9 (95 % confidence interval 63.1–70.6). Figure 1 shows ISQ values for each individual patient with respect to the timing of the radiation therapy in relation to their RFA measurement.
Discussion Hearing loss can be a significant cause for morbidity following radiation therapy to the head and neck region, and as such auditory rehabilitation is of utmost importance [9]. Radiation-induced fibrosis and/or destruction of middle ear components, tympanic mucositis, and cochleitis from radiation vasculitis may result in a wide spectrum of conductive, sensorineural, or mixed losses [9, 10]. This, together with the fact that post-irradiated patients are frequently unable to wear air-conduction aids secondary to skin and soft tissue reactions and/or disrupted anatomy, proffers BAHA as an ideal option for such auditory rehabilitation [9, 10]. Knowledge of the behaviour and retention of BAHA implant systems in these patients is therefore paramount in making informed decisions about the feasibility of the procedure, and indeed for counselling patients on expected outcomes and risks. At present, however, this knowledge is based on one small-scale study examining implant survival, but not stability, in post-irradiated nasopharyngeal carcinoma (NPC) patients [10], and inferences from the dental and craniofacial literature. The lack of published data examining BAHA implants in post-irradiated patients would imply that such patients are a rarity in clinical practice. However, middle ear side effects occur in up to 40 % of cases following head and neck irradiation and approximately one-third of patients develop significant sensorineural hearing loss [9]. A lack of assessment of hearing outcomes and awareness of BAHA as an auditory rehabilitation solution among oncologists may explicate this discrepancy, a perception supported by the general underevaluation of radiation-induced ototoxicity within the oncology literature [9]. Notwithstanding, the number of such patients is only set to increase, with the upsurge in use of radiation therapy, particularly for VS [11]. The principal finding of this study is the demonstration of successful osseointegration, as assessed both by objective implant stability and survival, of BAHA implant systems in post-irradiated patients. Although our study
13
Eur Arch Otorhinolaryngol
Fig. 1 Implant stability quotient (ISQ) values for each individual patient in relation to the time interval between radiation therapy and resonance frequency analysis (RFA) measurement. ISQ values of
≥60 are typically observed in successfully loaded implants, with the dashed line in the figure representing this critical value
is limited to a degree by a small study cohort, which may raise questions over reliability and precludes meaningful statistical analysis to identify factors that may influence implant stability and survival, our survival rate of 100 % is consistent with that of the only study published previously, in which no implant losses were observed in post-irradiated NPC patients [10]. Curiously, this is at odds with the general consensus from the dental and craniofacial literature. Several investigators have reported significantly increased failure rates of osseointegrated implants used to reconstruct craniofacial defects when placed in irradiated bone, with a relative risk as high as 12 [4, 6]. Similarly, although more contentious, observed survival rates of dental implants tend to be reduced in irradiated bone [3, 12, 13]: a recent review identified three studies that demonstrated a statistically significant difference, all reporting a two- to threefold greater risk of implant failure in irradiated bone [4]. Such disparity in findings would not appear to be an artefact of insufficient patient follow-up. Mean patient follow-up in our study and in that of Soo and colleagues [10] was 41 and 39.8 months, respectively, while implant losses in irradiated bone have typically been observed during the first three years after implant surgery, most commonly in the first year [6, 8]. Whilst it is plausible that implant failure may occur in our study cohort at a later date, indeed two of our patients have been followed for less than a year at the time of writing, our mean follow-up period captures the time of most frequent implant loss. Also from a chronological standpoint, the time interval from irradiation to implant surgery has been implicated in influencing implant survival [7]. Radiobiologically, the optimal
timing for implant surgery would coincide with the abatement of acute tissue reactions and the commencement of established healing, which, when accounting for the risk for severe tissue reactions caused by surgical trauma in the irradiated tissue, would likely fall 6–12 months following irradiation [7]. The high survival rate observed in our study may be attributable to the relatively long interval between irradiation and implantation in our patient group: mean interval was 33 months and range 16–72 months. The influence of anatomical location on implant survival may also go some way to reconcile the contrariety in findings. Several studies have shown that osseointegrated implants are not uniformly successful in the craniofacial region, with failure rates varying between sites [4, 14]. Implant survival rates noted in the literature for several locations in non-irradiated bone are as follows: mastoid >95 %, orbital 35 %–91 %, and nasal 71–81 % [4]. Interestingly, most investigators have observed highest survival rates in implants placed in the temporal region [4, 6, 14], which may account for our high survival rates and those reported by Soo et al. [10]. Whilst it is not entirely clear why such differences between anatomical sites may exist, it is likely, and indeed intuitive, that this would relate to the structure and volume of the bone, particularly the proportion of compact relative to trabeculated bone [15]. The fact that craniofacial implant survival tends to be lowest in the nasal floor and medial orbital region, where bone is thinnest and consists mainly of a loose trabecular structure, lends credence to this notion [4, 6]. A further factor which may account for the high implant survival rate observed in our study is radiation dosimetry.
13
Eur Arch Otorhinolaryngol
Although in craniofacial applications the importance of radiation dose in determining implant survival has been difficult to assess, owing to small study cohorts [4], a dosedependent relationship has been demonstrated [5]. Specifically, cumulative radiation effect (≥30) as a measure of radiation dose was significantly related to implant failure [5]. Similarly, Visch et al. reported a greater than twofold reduction in long-term survival of dental implants in oral cancer patients who had received radiation doses in excess of 50 Gy [16], while a systematic review of 19 studies of dental implants in irradiated bone noted no failures with radiation doses below 45 Gy [8]. Although unfortunately we were unable to obtain data on specific radiation doses for our study population, based on the typical pathologyspecific radiotherapy regimes, it is presumed that most patients included in our study would have received relatively low radiation doses. Whilst doses for the treatment of NPC and temporal bone SCC (patients 4 and 5 respectively) would have been in the order of 60 Gy [17, 18], patients treated with stereotactic radiosurgery for VS or glomus tumours (71 % of patients in this study) would typically have been exposed to marginal doses of approximately 15 Gy [19, 20]. In addition to implant survival, objective implant stability was also assessed in our patient group, which to the best of our knowledge is the first report of such for BAHA implant systems in post-irradiated patients. RFA is a method to assess the stability of the bone–implant complex at any point in time and provides a numerical value relating to stability, and was the chosen method in our study owing to its non-invasive nature and well-documented use within the dental literature. Although there are few published reports of RFA values in human temporal bone for comparison, RFA measurement was first introduced into implant dentistry in the 1990s and has been used extensively to assess dental implant stability in mandibular and maxillary bone and guide timing of implant loading [21, 22]. Whilst it has been suggested that the trend in RFA values over time, rather than an absolute RFA value, is more relevant in determining long-term implant success and lifetime stability [23], the comprehensive data from the dental literature implies that ISQs of ≥60 are found in the majority of successfully loaded implants, while ISQs as low as 47–59 may be considered stable for loading [24, 25]. On this basis, comparing values obtained in the present study with respect to these seems entirely tenable, particularly as the objective of our study was to determine whether implants had attained sufficient stability in irradiated bone rather than using these values to predict future implant behaviour. In further support of this, findings from studies of BAHA implant systems examining changes in RFA values with time indicate that a plateau of stability is reached and maintained at approximately 1 month post-implantation
[1], which when applied to our study population, with intervals from implantation to RFA measurement ranging from three to 96 months, would insinuate that in all cases a steady state of stability had been achieved. Importantly, ISQ values were >60 in all patients studied, implying that commensurable stability is attained in post-irradiated patients. Examining prospective stability measurements in such patients, however, presents an interesting area for future research to characterise the process and chronology of osseointegration in post-irradiated bone. Compromised wound healing is a well-recognised consequence of radiation therapy and is a common and challenging clinical problem. Consonant with this, two patients in our study (28.6 %) suffered significant skin flap failure leaving exposed bone around the abutment site, a considerably higher rate than the
