Does LIPUS speed up tooth movement?

Just when I thought that the methods of speeding up tooth movement had disappeared in a web of proper science.  I came across yet another method that claims miracles.  This is called LIPUS. I found two papers by a Canadian based team who promoted this type of care. Unfortunately, these papers were not good quality. Read on..

LIPUS stands for Low-Intensity Pulsed Ultrasound. This is a non-invasive, non-pharmacological method that has been used in medicine for many years.  A dental mouthpiece, covering the teeth, delivers an acoustic pressure wave to the teeth.  Our patients use this for only 20 minutes a day.  LIPUS appears to work on animals, but studies on humans are lacking.  LIPUS is marketed as the Aevo system by SmileSonica, Alberta, Canada.  They claim:

“Regular use of the Aevo System^™during orthodontic treatment will facilitate faster tooth movement and decrease orthodontically induced tooth root resorption”.

The company states that the cost of the AEVO is 20-25% of the cost of braces.

The British Medical Journal published a rapid review of LIPUS in bone healing and they recommended against the use of LIPUS. I thought that this was pretty compelling.

But lets have a look at some recent orthodontic research.

I came across two papers about LIPUS, and I thought that I should have a look at them.  A team from the Dental School at the University of Alberta did the studies.  The Journal of Clinical Medicine published the papers.  One was a retrospective investigation, and one was an RCT.

Let’s start with the retrospective study.

This is an open-access journal, so you can easily have a look at the paper.

What did they ask?

They did the study to:

“Analyse the overall treatment duration and percentage treatment reduction in patients using LIPUS who were undergoing Invisalign treatment and compare this with Invisalign only”.

What did they do?

They did a retrospective study of cases selected from one orthodontic practice.  They picked the first 34 patients who completed their Invisalign treatment with the LIPUS and compared this with a sample of 34 Invisalign patients randomly selected from the clinics completed patients.

The LIPUS patients had used the appliance for 20 minutes a day.

What did they find?

The mean duration of treatment for the LIPUS/Invisalign patients was 541.4 days (SD=192) and for the Invisalign only this was 1061.05 days (SD=455). This was a reduction of 49%.

The average co-operation with the LIPUS device was 66%.

What did I think?

Firstly, I thought that the difference in duration was remarkable. If this is correct, this will change orthodontic treatment forever.  However, you do not need to understand much about research to discover the fatal flaw in this study.  This was that the investigators selected the first 34 cases to finish treatment with LIPUS. This is similar to recording the times of the first 34 finishers in a race and using this data to calculate the average speed of the total number of runners. It is such a fundamental error it is remarkable.  I was also surprised at the duration of the Invisalign only treatment this was 3 years. Indeed this is too long for a mean duration of treatment.

In summary, this was a very poor paper, even if we consider that it was retrospective etc.

Was the randomised trial any better?

Again this was led by the Dental School at Alberta, and the Journal of Clinical Medicine published the paper.

What did they ask?

They wanted to

“Evaluate the effect of LIPUS on the rate of orthodontic tooth movement”.

What did they do?

They did a five site randomised controlled trial with a 1:1 allocation ratio.  The PICO was:

Participants:  Orthodontic patients aged between 12 and 40 years old requiring extraction of first premolars

Intervention: AEVO

Control: Treatment as usual

Outcome: Rate of space closure in mm/week.  Other results were the amount of root resorption.

Randomisation, concealment and sample size calculation were good and clear.  The AEVO device was only active on one side. As a result, this was a split-mouth design.  Blinding was good as the AEVO device cannot be felt.  They entered 47 patients into the trial. They defined a cut-off point of compliance at 67%. If compliance was less than this, they excluded the patients from the trial.

In this paper, they stated that space closure was done using TMA canine retraction T-loop.   They measured the extraction space intra-orally. The method that they used to do this was not too clear.

What did they find?

They entered 47 patients into the trial. As compliance with the device was not good, they excluded 24 participants.  Two others were excluded because of imaging being absent and a non-working device. Finally, a further 4 were excluded because the tooth movements either side of the arch were not comparable. This left a final sample of 17 participants. This is an exclusion rate of 63%.

The mean rate of tooth movement for the LIPUS group was 0.266mm/week, and for the control, it was 0.232mm/week. This was a mean difference of 0.03mm/week. This was statistically significant. The authors also stated that LIPUS increased the rate of tooth movement by 29%.  But I could not get this figure when I did the calculation using the formula that they outlined in the paper.

What did I think?

There are multiple severe problems with this trial. I shall only note a few.

• It was split-mouth, and there must be some carryover of this intervention.
• It was not clear on the number of cases treated by each clinic, so we could not consider any operator effects.
• Some of the patients had been treated as part of a Masters project. When I looked at this thesis, the method of space closure was different.  But I am not sure how many of the Master’s project cases were included in this final sample.
• The effect size was tiny and not clinically significant.  This would make no difference to the overall duration of treatment.
• But most importantly, you cannot exclude 63% of the patients in a trial. Notably, as 24 were excluded because they did not co-operate.  To be honest, I am amazed that the authors did this. If you do a trial on a voluntary intervention, you need to include all the patients regardless of their level of co-operation. This is because you are trying to measure the effect of giving someone an appliance etc.

Finally, we need to consider the Journal.  This is a “pay to publish” Journal and is part of the MDPI publishing group. This was once listed as a predatory journal and appealed and was removed from the list of predatory journals.  Here is a link to the Wikipedia entry on MDPI.

Final thoughts.

I thought that these were two remarkably poor papers. Nevertheless, SmileSonica is using the data in their publicity.  I am not sure why the Journal accepted these papers. I am confident that they would not be accepted in any of the leading orthodontic journals.  The PI in the studies has declared he has no conflict of interest.

Finally, after reading these papers, I cannot help feeling that AEVO lacks evidence to support its use.  I wonder if it is already heading the same way as AcceleDent, MOPS and Piezocision.