Snoring cessation through radiofrequency treatment of the soft palate:
a preliminary analysis of 119 cases
A New Zealand experience
Neil A Hutchison, FRNZCGP and Andrew D Murley, FRNZCGP, Family Practitioners, Auckland, New Zealand
Aim: To assess the efficacy of radiofrequency tissue volume reduction (RTVR) treatment to the soft palate in reducing snoring (using “Surgitron” equipment) as a simple office procedure.
Methods: A non-randomised, prospective study, where 119 patients underwent some simple clinical testing to best categorise them into:
As seemed appropriate, the options for management were:
Conclusions: These data suggest that the minor surgical procedure of radiofrequency is a valuable tool in the management of the snoring patient across the range of BMIs. There may be an advantage in treating severely congested turbinates concurrently. The success rates, minimal pain, and economy of the RFVTR to the palate make this office minor surgery an attractive alternative for the patient.
Snoring is a very common problem. There is a large industry developing with remedies for this problem. These include sprays for throat and nose, and nasal and mandibular advancement splints. However, there is now a minor treatment with an excellent chance of reducing snoring. Radiofrequency tissue volume reduction (RTVR) has been shown to be tolerable, effective and safe option for treating snoring 1-8. RFVTR equipment is available from two sources: “Somnoplasty”, and the “Ellman Surgitron” alternative we call “Snore-op”. We have used both. These are similar in effect, the former being much more “hightech” (and much more expensive), and the latter, being the much less complex but very adequate “Surgitron”. On the horizon are the cryogenic and sclerosant injection methods, also attempting to effect tightening of the palate through scar tissue. All these methods require local injected anaesthetic, which is the main challenge for most clients.
A review of the literature shows papers extending back to about 1997. Numbers of patients in these trials tend to be quite low (averaging 34 patients, ranging from 12 to 113), and assess radiofrequency to the palate from a number of points of view. These include: The percentage satisfied after 3 treatments 9, single burn results compared with multiple burns 10, comparing the effectiveness of laser-assisted uvulopalatoplasty (LAUP) with the radiofrequency technique 11, and assessing the value of objectivity vs subjectivity in assessing snoring 3. There is also a study of the longevity of the reduction 4 12 13, methods of evaluating pain scores 5 8 14, and assessment of the incidence of mucosal damage, seen and unseen 15. Using RFVTR to successfully reduce turbinate volume has also been recorded9 16-18.
One hundred and nineteen consecutive patients presented for radiofrequency treatment of the palate after the initial clinical assessment. They had been recruited through public advertising, and by referral from other family medical practitioners in our city. One had a previous laser-assisted uvulopalatoplasy (LAUP), which initially helped snoring, but had now “loosened”, and one had a previous uvulopalatoplasy (UPPP), which had not been successful.
Clinical Assessment: History taking involved questions on duration of snoring, snoring position, tiredness on waking and during the day, the presence of a very dry throat in the mornings, smoking, aggravating factors, and history of weight gain. Questions about exercise levels, and what other methods
tried so far were also asked and recorded.
At this assessment visit, a pre-operative snore-score sheet was completed, including a numeric comment from the bedmate. This seemed an effective way to quantify the snoring problem at the beginning and after treatment. (This scoring method was deficient for some who commented that the snoring was now not continuous, but occasional and the same loudness.). This scoring method has been employed in other research 1 3 13.
Scoring was rated zero to ten, where ten indicated one of the partnership leaves the room. One, two and three related to slight snoring not bothersome to the bed partner. Four, five and six indicated snoring loud enough to bother others in the same bed or room, and seven, eight and nine indicated it was bothersome to others in the house. The score sheet also included weight and height and BMI data. Informed consent was obtained.
In the sitting position, examination of the patient included the production of a snore noise with the mouth open and closed, then with the jaw protracted. The palate was examined with “ah” and “ee” to assess for laxity. The uvula was observed for size, oedema, and inflammation. The nose was assessed for blockage from septal, turbinate or other causes. Fibre optic assessment of palatal vibration was not included. We encouraged the patients to make a separate appointment for the actual procedure to allow them to read the details we provided.
After local anaesthetic (1 ml of Xylocaine 1% with adrenaline each side), the palate received two radiofrequency treatments in paramedian positions, with the probe tip carefully placed so as to be deep to avoid surface ulceration. The 20-second burns are not visible. They were placed in a tandem configuration (upper left and lower right) to allow a second procedure in new tissue if required later in the alternate position . The radiofrequency apparatus used was the “Ellman Surgitron”. The settings used were “Partially Rectified”, and at power setting of 3.5 for palate use, for 20 seconds each.
Ellman have other probes for different palatal work (which were not used), and for the treatment of the turbinates. The latter is a double-pronged device that is autoclavable.
We have modified the palatal tip by straightening it to 5 degrees to accomplish this task, and results seem most satisfactory for many patients. Topical 20% Benzocaine is placed on the turbinate, and then injected with Xylocaine 1% with adrenaline 1 ml each side. This achieves excellent analgesia. The surgitron is set on “Partially Rectified” on a power setting of 2 for the turbinates. About twelve applications to each inferior turbinate produced a visible change in contour. Haemostasis was achieved by applying a burn on exiting.
In our previous pilot study of 50 cases, post-operative pain was assessed on a visual analogue scale (V.A.S.). The mean pain rating on a scale of 0 to 10 (where 0 = no pain and 10 = maximum pain) was 2 (range: 0 – 4). One out of four patients required mild analgesia. Pain was not assessed in the current study.
In this study of 119 cases, there were 97 males (81.5%) and 22 females (18.5%). The age range was 19 to 79 years of age. Mean improvement of males and females was similar, being 3.31 and 3.45 respectively (figure 1).
Scores for snoring have been quoted using the partner scores, not the patient scores, because often patients could not offer an opinion on their own scores. Partner scores were consistently higher than patient scores. (Table 1)
Comparison of patient’s account of snoring compared with partner’s account
| Mean score | Patient | Partner |
| Preoperative score | 8.2 | 9.0 |
| Postoperative score after one treatment | 5.5 | 5.8 |
| Postoperative score after second treatment | 4.3 | 4.7 |
After the first procedure, 79.6 % of patients had improved, the mean improvement being 4.65 on the snoring scale. After the second procedure (n = 35), 62 % of patients had a further improvement, the mean being 2.36 on the snoring scale. After the first treatment, 20.4% showed no improvement.
Based on age groupings, reduction in snoring occurred over all ages with the first treatment. In the under 40-year-old group (n = 15), the mean improvement in snoring scale was 2.3. The mean improvement in snoring scale for the 40 to 49 year olds was 3.3 (n = 37). The mean for the 50 to 59 year olds was 3.5 (n = 43), and the 60 to 69 year olds was 3.1 (n = 18). Although the over 70’s group had the best improvement score at 4.45, the numbers were low (n = 5). (Figure 2)
When grouping by Body Mass Index (BMI), all groups had average improvements of more than 3. Categories were BMI of 20 to 25 (n = 26) where the mean improvement was 3.9 on the snoring score. With a BMI of 25 to 30 (n = 59), the improvement was 3.1. With the BMI of 30 to 35 (n = 24), the mean improvement was 3.0, and the very overweight BMI of 35-plus (n = 6), the mean improvement was 3.8. (Figure 3).
RFVTR application to the palate only, and RFVTR to both palate and turbinates in the different BMI groupings were compared. In most cases, the mean reduction in snoring score was greater where the combined procedure was done. With the BMI of 20 to 25, the mean score in the palate group was 5.34, and 5.23 for those receiving palate and turbinate radiofrequency. When the BMI range of 25 to 30 was compared, the mean postoperative scores were 5.21 and 4.74 respectively. With BMI 30 to35, the scores were 4.67 and 4.88, and with the BMI over 35, scores were 4.5 and 4.0 (Table 2)
BMI Preoperative score Postoperative Palate Palate & Turbinates
| BMI range | Preoperative score | Postoperative score after first treatment | Postoperative score after second treatment |
|---|---|---|---|
| 20-25 (n=33) | 9.24 | 5.34 | 5.23 |
| 25-30 (n=39) | 9.23 | 5.21 | 4.74 |
| 30-35 (n = 15) | 8.26 | 4.67 | 4.88 |
| 35 plus (n = 5) | 8.8 | 4.50 | 4.00 |
The incidence of adverse reactions was very low. Three patients felt faint, and needed to be reclined. There were 11 mucosal burns, most of which were trivial, and not noticed by the patients. Two of these were considered moderate, and healed without problem by six weeks. There were no palate perforations in this series, and no one needed narcotic analgesia. Paracetamol sufficed when needed.
One patient who presented had LAUP several years ago had resumed snoring, and wanted to try RFTVR to reduce his snoring. One patient who had previously had UPPP without initial success, also presented. The former patient had what he considered a successful outcome to radiofrequency treatment to the palate, with an initial snoring score of 10, and a reduction to 8 with one procedure, and to 5 with a second one. The spacing was 8 weeks between these procedures. The latter patient did not improve, scoring 10 preoperatively, and 10 at eight weeks, and 10 two months after a second attempt.
Snoring is a distressing problem for the victims and the perpetrators, and is very prevalent. This prospective non-randomised preliminary study provides more data on some areas reported on previously by others in smaller samples, and offers some new data in some areas of radiofrequency tissue volume reduction (RFTVR). Our data differ from some other published studies, inviting discussion and further investigation.
Our snoring score scale has been used by many investigators 1-5, 8, 10, 13, 19, 20 and seems to be a reasonable method to quantitate preoperative disturbance levels. As patients improve, however, some develop intermittent snoring, and this scoring method is sometimes not adequate to accurately quantitate these reductions. Embarrassed snorers are often quite reluctant or unable to put a score to their snoring volume. It is however a good, simple, and reproducible scoring system when scored by the spouse 4. A reduction in snoring, as scored by the partner, in 79.6% of patients after one treatment compares well with other studies 1-5, 7, 8, 10, 19,21, and the mean improvement of 4.65 on the snoring scale represents an excellent result. A second treatment, given when the patient wished for a further reduction, showed a further reduction in 62% of these patients by a mean of 2.36 on the scale. Complete data is not yet available for recipients of a third procedure. The percentage having no significant improvement with the first treatment was 20.4%, and it remains to be seen whether this group would improve with further treatments.
Response to treatment was evaluated looking at different age groups. All age groups responded with improvements in their mean snore scores. This varied from a mean reduction of 2.3 in the under 40 year olds (n = 14) to 4.4 in the over 70 year olds (n = 5). This may be an aberration because of the smaller numbers represented at either end of the graph, but does suggest that RFVTR treatment of snoring is worthwhile at all ages. This data has not been reported before in the literature.
Previous data in a small series has suggested that response to RFTVR of the palate is closely related to the BMI, 2. Our results showed a consistent improvement over all ranges of BMI, including the very obese BMI of >35. This improvement suggests that this procedure is generally successful in the full range of body weights, and that one is not necessarily disadvantaged by being obese. Further trials of larger numbers are needed to resolve this question.
Using data from one operator, a comparison has been made between patients who have had RFVTR to soft palate alone, and to a combination of palate and turbinates, using the same apparatus at the same visit (Table 2). This data was further broken down into BMI groupings. Patients were selected on details of
history and clinical impression as to whether the nasal obstruction was relevant to their snoring or not. In the history, waking during the night to sip water, or waking with an excessively dry mouth suggested mouth breathing during the night. When not associated with a lot of sneezing, this was considered to be
due to chronic turbinate congestion, rather than acute allergy. Clinically, when the above turbinates largely occlude the nasal airway, and they are accessible, treatment was offered with RFTVR. When reviewed at 6 weeks, tissue reduction is usually still obvious in the turbinate contours. We have observed
this phenomenon for as long as 2 years postoperatively. Utley et al 9, in their series of 10 cases, have quoted 100% success using similar methods for volume reduction of turbinates. In a number of ways, Utley et al endeavoured to quantitate nasal patency. They concluded that: “submucosal ablation of the inferior turbinates with radiofrequency is effective in reducing nasal obstruction secondary to turbinate hypertrophy”. A more efficient way of equating nasal patency would be helpful in evaluating this treatment.
Adverse reactions to RFVTR are minimal and well tolerated, though caution in cases of patients with cardiac pacemakers is advised. Pain is usually minimal, and this is an advantage when the inevitable loosening of tissues occurs perhaps one to 3 years later13, as is common in all mechanical tightening procedures. Acceptance of re-treatments is high, 20.
Whether RFVTR can be appropriately used to augment a diminishing effect after LAUP and UPPP needs more research, as the much more minor operation of RFVTR has been shown to be equally effective and better tolerated when compared with LAUP for snoring reduction 11.
RFVTR is a safe and effective method for the reduction of snoring of all age groups, and all categories of BMI patients. It may also have a place in concomitantly reducing the volume of the turbinates in patients where this is relevant.
The authors wish to acknowledge the expert assistance of Paul Prosée and Paul Hutchison in collating and interpreting data.