November 07, 2016

Another trial on maxillary protraction shows it works: but does it?

Another trial on maxillary protraction shows it works: but does it?

A few weeks ago I discussed a trial that showed early intervention with protraction headgear could correct class III malocclusion. This new trial on investigates the effectiveness of skeletal anchored maxillary protraction. The authors concluded that this was effective. But is it?

I think that one of the most potentially important recent innovations, in orthodontics, is the introduction of bone anchored protraction of the maxilla. Hugo De Clerck developed this new technique and he has shown some very impressive case series. In effect, patients wear class III elastics from zygomatic mini plates to symphyseal mini plates in the mandible and this protracts the maxilla. This recent study evaluated this technique and was published in the AJO – DDO.

screen-shot-2016-11-04-at-17-18-10Comparative evaluation of 2 skeletally anchored maxillary protraction protocols

Mohammed H. Elnagar et al

Am J Orthod Dentofacial Orthop 2016;150:751-62 DOI:

What did they do?

This was a two centre study. The authors set out to measure the effects of skeletal anchored facemask protraction or Class III elastics to skeletal anchors.

They divided up 30 growing patients with class III malocclusion using random numbers to 3 groups of 10. These were:

  • Facemasks attached to mini plates anchored into the zygomatic buttress
  • Class III elastics extending from mini plates in the maxilla to mini plates in the mandible
  • Untreated control

They used cephalometric measurements as their primary outcome

They took cephalometric radiographs before and after maxillary protection or observation, digitised the radiographs and then analysed a large amount of cephalometric variables.

What did they find? 

There were no differences between the three groups at the start of treatment. They observed the participants for approximately nine months.

They presented their data in a very detailed table of multiple cephalometric variables. I found this very difficult to interpret. They also carried out multiple statistical testing with the risk of finding significant differences just by chance. I decided that the best way to discuss this data was to select a few relevant variables. I have put these in this table here;

CorticotomyPiezocisionControl (merged)
Mean total tooth movement (mm)4.744.032.5

In general, it appeared that the maxilla was displaced forward significantly in the treatment groups when compared to the control group. There was very little difference between the two interventions. The cephalometric changes are clinically and statistically significant and interestingly the confidence intervals are narrow and again clinically meaningful.  The authors went into a large amount of detail in the discussion and showed some nice case reports. Overall they concluded that the bone anchored maxillary protracted treatment and achieved favourable outcomes in the short term.

What did I think?

I thought that this was an interesting paper in which the investigators attempted to answer and very relevant clinical question. But I was not sure  whether this was a trial.  When I first read the paper, I thought that it was. However, this was not clear in either the abstract or the main paper. This was a surprise to me because the AJO has high standards in the way that they report trials.

Putting this aside, I could not find a sample size calculation, method of randomisation and concealment. I was not sure on the number of study drop outs and it looks as though the treatment was complete for all patients. This is usually reported for a trial. As a result, I am not sure about the nature of this study.

But perhaps most importantly, I could not see any evidence of blinding at data collection. This is crucial in any study because of biases that may occur, particularly with cephalometric data recording. When I looked at the cephalograms that they showed in the paper, I could clearly see the bone plates. As a result, I can only conclude that there was no blinding in the data recording. There is a high probability that this would lead to bias in the study.

My other concern was that they did not provide any information on how they made sure that the mandible was not postured forwards when the initial cephalograms were taken. This again is an important issue that I would like to have seen clarified.

I wonder if I have been a little critical about this study, but they do suggest that this treatment is effective. In view of these issues in terms of methodology, data recording and potential bias I’m afraid I’m not so confident. My concern is particularly relevant when I consider that this treatment is very invasive and involves a moderate surgical procedure. I certainly would like to see a more robust study before I adopt this treatment.

I do wonder if early interceptive treatment with face masks is a better option than this method of Class III correction?

Related Posts

Have your say!

  1. Hi Kevin . I do agree that the early facemask therapy which you have referred to and also described by Benedetti and I 2000 were more convincing.

  2. Of particular interest to me would be the specific ages of the patients in the study and the amount of skeletal anchorage failure found.
    Dr Peter Ngan, chair of the Department of Orthodontics at West Virginia has done extensive work with skeletal anchorage based Class III treatment and I have utilized his treatment protocols extensively enough to have identified a problem with skeletal anchorage failure, especially in the maxilla. Annectdotally, this failure in maxillary anchorage seems age related (higher failure rates in younger patients, possibly due to less dense tuberosity bone?). We have utilized Bollard plates and these were placed by several different very competent oral Surgeons. So, from a clinical efficacy standpoint, system failure rates and drop out rates are of utmost importance in determining clinical usefulness.

  3. Kevin:
    I wonder if the maxillary was “displaced” as you note in your observations. One of the downsides of traditional orthodontic nomenclature is that it sometimes loses meaning in the broader scientific context. In my courses I refer to this as a discrepancy between quantum physics and quantum biology. Is the above use of the term “displacement” robust; think Archimedes. In other words was there translation of the maxilla thru space? Maybe. More likely, IMHO, it’s an example of allometry (size-related-shape change), which is impossible to decompose using traditional 2D cephalometrics. In addition, the authors may have inadvertently targeted the (anterior) cranial base. For example, please see
    Mitani Y, Banabilh SM, Singh GD. Craniofacial changes in patients with Class III malocclusion treated with the RAMPA system. Int J Orthod Milwaukee. 2010;21(2):19-25. Best wishes –

  4. Many thanks for reviewing my article
    Let me clear some points
    For the sample size we made the power analysis, we found many publications Many studies on maxillary protraction used approximately the same size of sample and sometime even less De Clerck HJ et al 2009 , Heymann et al 2010 and Lee NK et al 2012,
    All cases included in the study were selected according to same selection criteria then randomly divided in three groups
    All cases with mandible functional shift were excluded from this study, all the radiographs were de identified, and this is the best to bed done for blinding.

    The two bone anchored maxillary protraction protocols used in this study achieved favorable treatment outcomes for Class III patients with maxillary deficiency at late mixed and early permanent dentition (10-14 y). They could considered as alternatives for patients with severe maxillary deficiency who missed the early treatment and would be a candidate for orthognathic surgery which is more invasive than the skeletally anchored maxillary protraction. These protocols provide a short interval between growth modification phase and fixed appliance phase; hence the miniplates could be used in fixed appliance phase as anchorage to obtain more space by distalization of the maxillary teeth. In addition improvement of facial esthetics in teenagers instead of postponing a surgical approach until completion of growth may have a favorable psychosocial effect.
    We were able to protract the maxilla 6mm as the direct application of force to maxilla or zygomatic bone using bone anchored maxillary protraction achieved more maxillary advancement than facemask/rapid maxillary expansion therapy. Moreover, facemask/rapid maxillary expansion therapy produce greater dental than skeletal changes in contrast bone anchored maxillary protraction produce more skeletal changes and open circummaxillary sutures. In addition facemask is bulky and less easily tolerated than Class III elastics; hence more patient compliance can be obtained with intraoral Class III elastics.

    Based in our study The same rate of maxillary protraction can be achieved by either facemasks anchored with miniplates using heavy interrupted force or by Class III elastics from infrazygomatic miniplates in the maxilla to symphyseal miniplates in the mandible using moderate continuous force without mesialization for the maxillary dentition.
    Class III elastics to miniplates can provide vertical closing of the mandibular plane more than facemasks anchored with miniplates. In addition lower incisors retroclination decreased which makes it the treatment of choice in high angle cases.

    Over all the perfection is not but we do our best for our profession and our patients

    • Thanks for clarifying the points that I raised about your paper. This answers the points that I raised. Can you confirm that this was a randomised trial with blinding of data recording? Were the second cephs taken after the bone anchors were removed. It was not made clear in the paper and the paper was not written according to CONSORT guidelines and this led to my comments. Best wishes

      • yes it was randomized trial with blinding of data recording, for each patient had 2 pre treatment cephs with and without the anchor miniplates and 2 post treatment cephs with and without the miniplates on which the measurements were done. however for more illustration in the article we but images for the cephs with miniplates.

        and again many thanks for for reviewing my article its a great honor for me, there will be more coming puplication for me in skeletal anchorage I hope it will attract your attention as well .


        • Thank you Dr Elnagar for providing these clarifications on your paper.

          Since this turns out indeed to be a randomized trial, I was wondering why this was not reported in the original report from the start

          AJO-DO, as a leader in its field has gone to great length, following the initiative of the late past editor and the assistance of brilliant associate editors, to improve the reporting of randomized trials and systematic reviews in AJODO.

          Although your trial is very interesting it has many issues that are unclear (was the trial registed a priori somewhere? is the initial protocol available? / how was the random sequence generated / was allocation of patients concealed in any way? / was sample size calculated from the start / was blinding indeed performed (contradicting statement-paper data) / were there any protocol deviations (including dropouts) / were there any harms observed? / and why ON EARTH, did you take 4 cephs per patient.

          All these key issues have been in previous years not reported in full detail in orthodontics, and therefore AJODO has introduced special Instruction to Authors for RCTs (see journal site) and has divided submitted RCT papers in different sections to make them more easily/transparently reported. This was one effective measure to tackle the problem of suboptimal reporting of trial details.

          Unfortunately, as your trial was not labeled, I take it, as a randomized trial in its title/abstract/text, it didn’t go the “usual way” RCTs go in AJO-DO (correct me if I’m wrong). For example, most of my questions from above would be automatically answered, if the trial was written according to the CONSORT statement (and a filled CONSORT checklist was submitted with the paper).

          Although, as stated above, your trial deals with a very interesting scientific question, the existing issues in the trial and its report cast doubts about the usefulness of the extracted conclusions.


        • Are there any orthodontists in NYC or the surrounding areas that use this method? Can it be used on 13 year old girl?

Leave a Reply

Your email address will not be published. Required fields are marked *