October 06, 2014

New evidence on the treatment of palatally displaced canines!

Should we extract primary canines to prevent palatal canine impactions?

This post is on the most recent evidence on the effectiveness of removing primary canines in the hope that this encourages the eruption of palatally displaced permanent canines (PDC). As we all know, this is an established procedure that is practiced widely. However, a recent Cochrane review highlighted that there is limited evidence underpinning this practice. I have reviewed this previously. In the review they pointed out that previous studies have not included control groups or were not clear in their design and reporting. As a result, they concluded that there was no evidence to support the removal of primary canines to intercept palatal impaction of the permanent canine.

In my recent presentation to the British Orthodontic Conference, I discussed the effect of clinical uncertainty on treatment decisions. This was an interactive presentation and I asked the delegates to vote on whether they would remove primary canines in cases of PDC, in the absence of scientific evidence. It was really interesting to find that 78% stated that they would remove the primary canine and 81% were very certain about their decision! There is no doubt that this is a controversial area!

I was then very interested to see that a new trial has been published on this question in the European Journal of Orthodontics. I am going to review this in the same way that I would for a Cochrane review and concentrate on issues concerned with design and the risk of bias.

Extraction of the deciduous canine as an interceptive treatment in children with palatal displaced canines—part I: shall we extract the deciduous canine or not?

Julia Naoumova, Juri Kurol and Heidrun Kjellberg

European Journal of Orthodontics: Epub in advance of print.

doi: 10.1093/ejo/cju040

This study was done by a well known research team and they aimed to evaluate the effect of interceptive extraction of the primary canine in children with palatally displaced canines.

They also wanted to analyse radiographic changes in the PDC and measure any root resorption. These were secondary outcomes and I was more interested in whether the permanent canines erupted or not, so I will not discuss these outcomes.

I felt that this was a well carried out trial. They randomised Children with PDC  to either have the extraction of primary canines or no extractions. The inclusion criteria were clearly set out. I thought that it was interesting that one exclusion criteria was maxillary crowding greater than 2mm.   I could not understand the rationale for this step, as it must have excluded a fairly large number of children. It also means that the results of this study are only relevant to children with minimal incisor crowding.

They included 67 patients in the trial with a total number of 89 PDCs. because some of the patients had bilateral impactions.

Randomisation was clearly outlined and concealment was by previously prepared sealed envelopes. All the participants had CBCT radiographs taken at the start, 6 months and 12 months. The last two radiographs were not taken if the permanent canine had erupted.

The primary outcome was the eruption of the permanent canine. But the sample size calculation was not based on this outcome, the authors based it on change in angulation of the PDC. I thought that this could result in the study being at risk of not having a large enough sample size to detect differences in the primary outcome.

 What did they find?

They found that 69% of the PDCs erupted in the extraction group compared with 38% in the non-extraction. This was statistically significant and I would say that this was a clinically significant finding.

I felt that it would have been useful if they had presented other relevant data, for example, the odds ratio, risk ratio and numbers needed to treat. I have outlined this in a previous blog. So I reached for my odds ratio calculator and worked these out myself.

This revealed the following

VariableMean Change/Year(Degrees)Confidence Interval
ANB-1.14(-1.52, -0.77)

 What does this all mean?

From this interesting paper, I can conclude that following the extraction of the primary canine when the permanent canine is palatally displaced.

  •  There is a 40% reduction in permanent canine impaction
  • The risk ratio suggests that the chance of eruption of the permanent canine was 1.7 times greater than in the extraction group.
  • Numbers needed to treat reveals that we need to extract 4 primary canines to prevent the impaction of one permanent canine.

These figures are encouraging. But we must take care in interpreting them and we must also look at the 95% confidence intervals. These are all wide and this suggests that there is still a degree of uncertainty in the data.

 What will I do?

So, we have to consider that we have one trial that provides a moderate level of evidence with a degree of uncertainty in the data. We also need to evaluate the risks of removing the primary canine and these are low. As a result, in terms of risk benefit considerations, I have decided to start extracting primary canines.

I have also had an afterthought and I wondered whether there were any predictors for the success of treatment. Unfortunately, I could not find any. However, this is part I of a publication, so maybe this was addressed in the second part?

In summary this was a good study that provided useful clinical information on an established practice that up until now had no evidence base!

Naoumova, J., Kurol, J., & Kjellberg, H. (2014). Extraction of the deciduous canine as an interceptive treatment in children with palatal displaced canines–part I: shall we extract the deciduous canine or not? The European Journal of Orthodontics DOI: 10.1093/ejo/cju040

Related Posts

Have your say!

  1. Thanks for these posts, I was really interested in this one as there has been varying evidence. Not scientific and I expect to be shot down but I have long felt that, even with dubious evidence, the upside of extraction, even in hope rather than expectation, is so much greater than the downside that it is hard to resist the extraction in a fully informed patient

  2. Hi Kevin

    I was one of the ones who voted for extraction of primary canines at BOC. Nice to know it was a decent choice. To be honest I think the advent of cone beam radiographs gives us far more information than we’ve ever had before. I know that I’ve certainly looked at cases where I felt that the canine was so displaced that extraction of the primary canine would be of little benefit. I’m also finding it easier to identify the cases where I feel that extraction of the primary canine will almost certainly encourage eruption of an impacted permanent canine. Yet more scope for research?


  3. Is there any evidence that palatal expansion in conjunction with primary cuspid extraction increases its effect?

    • The only studies that suggest this are those carried out by Baccetti, but they were not clear in their write up and this has meant that they were not included in the Cochrane review. I do wonder if there is a case to be made for expanding to create space, but the evidence is currently lacking,

      Best wishes: Kevin

  4. Kevin,

    I appreciated the review. As a Pediatric Dentist I’ve recommended the extraction of Primary canines many times. I too am glad that there is additional evidence to justify this protocol. As a side note… perhaps the next step is to examine the effectiveness of double extraction. The article below proved interesting… I smell a new clinical trial.
    Double vs Single Primary Teeth Extraction Approach as Prevention of Permanent Maxillary Canines Ectopic Eruption
    Alessandri Bonetti Giulio, MD , DDS 1 • Incerti Parenti Serena, DDS 2 • Zanarini Matteo, DDS 3 • Marini Ida, MD , DDS 4
    Pediatric Dentistry v 32 / no 5 SEP / OCT 10

  5. Well blogged Dr Kevin.
    Long awaited evidence on already predominantly practiced clinical outcome.
    Thanks Dr Kevin.

  6. Thanks for the great blog. The results reported in the study demonstrated that 39% of PDC erupted in the control group. This figure is actually inflated as they explained that 5 of the 10 patients at T2 underwent extraction of the non-mobile primary canines in the control group. These patients shouldn’t really be included in the 39% as they had extraction but yet were in the control group. If you only include non-extraction patients in the results 27% of PDC erupted. What are your thoughts?

  7. Dear professor Kevin O’brien,

    Thank you for blogging about our article (Extraction of the deciduous canine as an interceptive treatment in children with palatal displaced canines—part I: shall we extract the deciduous canine or not?). It was really interesting to read your comments.
    You had some comments that I would like to ask you if it would be of value that we include them in part II of this trial, since it still has not been published.
    * The exclusion criteria of crowding more than 2 mm, would you like us to explain why we had that as exclusion in part II? Would it be strange to have an explanation there and not in part I?

    * The number of patients needing treatment is given in table 3 in the article.

    * OR for factors affecting the main outcome will be presented in part II as well as predictors for success of treatment.

    * I liked your comment about the power analysis. We had discussion about this before starting the study and decided in the end to go with the change in angulation of the PDC. Would it be wrong to make a power analysis retrospective to see if our sample was enough to detect differences in the primary outcome?

    Thank you once again for your comments and a very nice and educating blog.

    Best wishes

    Julia Naoumova

    • Thanks for a great blog and a nice review of a very interesting paper. I actually has a question that Julia N might answer: How servere were the impactions? Sectors? It would be nice to know which type of cases were included especially since the rate of bilateral cases were nearly 50%. Perhaps this is coming in part 2?

  8. Hi Kevin

    I have just read this article and will look forward to tomorrow morning as it’s the chosen paper for our journal club at Sheffield
    It’s great to have a study using CBCT, definitely an enjoyable read and has helped widen knowledge about this controversial area. However, after my first read of the article, I have a few comments:

    Since over 60% of the sample were between the ages of 10 & 11 at T0, I think it’s a shame that the non-erupted canines were extracted in the control group at T2. This is 12 months after baseline records and so over 60% of the sample would be aged 12 yrs or under (similar age for normal development). I therefore don’t understand the statement about it being ‘unethical’ to leave them any longer.

    I’m curious to know if the exclusion criteria of ‘over 2mm crowding’ includes the difference in size between the primary canine and permanent canine.

    I also wonder how displaced some of the canines were at baseline looking at the values for canine cusp tip –dental arch plane 1.9 +/- 1.2 (extn) and 2.5 +/- 2.0 (control). I suspect this is the most important measurement in terms of palatal displacement. It’s nice to see that only patients were included over the age of 10 yrs but I wonder if we should focus on maturation of the lateral incisor root rather than chronological age? What we know about ‘normal development’ is that the permanent canine is often in a medial position prior to apex closure of the lateral incisor so some of the linear and angular measurements at baseline don’t mean a lot to me.

    Kevin, what do you think to including all canines in the analysis? Obviously the canines taken from the bilateral cases aren’t independent. I’m not sure this matters…

    My conclusion from this study (at first read) is that extraction of the primary canine speeds things up but I’m not convinced it means that more will erupt spontaneously, well not in the numbers quoted. It would be nice to see the numbers in each group that deteriorated in their position rather than grouping them together with those that showed no improvement. Perhaps those that showed no improvement might have improved with time?

    I still don’t have a black or white view of what to do when faced with the dilemma of extracting the primary canine to manage PDC’s in young children. It must depend on the degree of malposition of the permanent canine and other patient factors such as suitability for fixed appliances. It is mentioned in the paper that all extracted primary canines showed no root resorption (or less than a third) and none were mobile. So they might have a good long term prognosis if left in situ & if the permanent canine remained impacted.
    Kevin, I disagree with you with regard to the risks of extraction being low. I think there is a considerable risk for extracting the primary canine and the degree of this risk may depend on motivation/OH/ socio economic group of the patients we are treating. I believe the risks of exposing and aligning PDCs in some patients referred to us at Sheffield and Barnsley are high and that these patients would be better off keeping their primary canine and avoiding decalcification, root resorption , failure to complete etc…

    As an aside, this is a subject that has interested me for years. Baccetti, Leonardi, Olive & Kockich were/are strong believers that space creation is probably more important. Despite there being no strong evidence for this approach, I have been practicing this many years with great success. So my interceptive treatment tends to involve creating space between the premolar and lateral incisor & leaving the primary canine in situ. Having created sufficient space and maintained the space for a period of time the primary canine normally becomes mobile and exfoliates naturally. Have you tried this approach Kevin, rather than going straight for the forceps 🙂

    Best wishes, Nicola

  9. Dear Kevin,
    I found your review of this paper interesting and this is clearly a contentious area. Many years ago when I spoke with Professor Erickson about the original Erickson and Kurol paper he said that their cases vere usually uncrowded. So maybe they didn’t have to exclude as many cases as you might think? This is clearly a very different situation to the majority of cases that we see in this country. Certainly when Mary Short and I looked at our cases who had deciduous canine extractions carried out crowding was a significant factor in determining improvement in the canine position. The cases without clinical crowding, or where active intervention was carried out to increase space in the canine area, showed a more favourable response to the interceptive deciduous canine removal.
    Best wishes


    • Hi Susan, this is also my experience as I think that most of my canine cases are uncrowded. But I have never come across a reference that reinforces this clinical opinion. You are also correct in that id does seem logical that if you make space then a tooth should erupt. There was a suggestion that this is the case in the papers by Leonardi and Baccetti, but these trials were not reported well and so this remains conjecture, even though it logical

      Best wishes: Kevin

  10. Hi Kevin,
    I have been extracting C’s for years – convinced (without robust evidence) that it is an effective and sensible thing to do…. nice to be able to say “I told you so”!!

    • Hi Joe, thanks, I am glad that you are reading the blog, yes it appears to be effective but the evidence is only just beginning to come through. However, we know that orthodontics is all about the interaction between evidence and clinical experience, so may be the clinical experience is working for this problems

      Best wishes: Kevin

  11. I totally agree with the comment posted by Nicola Parkin.
    I personally think that the study has flaws as other previously reported studies on this topic.
    I wouldn’t extract primary canines at ages 12 &13, in particular, when they appear to be intact with no root resorption (inclusion criteria is inappropriate). Of course the degree of displacement should be taken into consideration.

    I hope you reply to the points addressed by Nicola Parkin.

  12. Hi Simon,
    as written in the article, we used intention-to-treat analysis (ITT) and therefore the number of 39 % in the control group is correct.
    Below you can find information about ITT citated from the article,

    ITT analysis includes every subject who is randomized according to randomized treatment assignment. It ignores noncompliance, protocol deviations, withdrawal, and anything that happens after randomization. ITT analysis is usually described as “once randomized, always analyzed”. ITT analysis avoids overoptimistic estimates of the efficacy of an intervention resulting from the removal of non-compliers by accepting that noncompliance and protocol deviations are likely to occur in actual clinical practice.

  13. Hi Sigurd,
    We didn’t used the method from Ericson and Kurol (1988) with sectors, alpha angle, distance to the occlusal plane etc, instead we developed another method that was used on the CBCT images.
    However, we have panoramic radiographs on all included patients and in part III we are planning to compare 2D images with 3D.

    We couldn’t an fortunately find a reason on why we had that many bilateral cases. The patients were collected from 15 different clinics and were diagnosed by three observers: the general dentist, the consultant orthodontist and the treating orthodontist.

    Best wishes, Julia

  14. Firstly, I would like to congratulate Prof. O’Brien on his blog, this is a really good adjunct to my MOrth prep. I am also glad there is some evidence to support this uncertain topic, it will hopefully make my Viva that little bit easier.

    My only concern with any interceptive treatment is the system we work in.

    For example, a patient attends with everything class I, mild crowding and a unilateral impacted canine. We extract the C and alas the canine has come through and we are all patting ourselves on the back, the patient is then re-referred the canine isn’t fully aligned and suddenly the patient no longer qualifies for treatment on the NHS. The IOTN has dipped from a 5i to a 3d.

    From a clinicians point of view, we have saved them a GA and potentially unnecessary treatment, unfortunately from the patients point of view, they are saving for private treatment.

    However, if you used an appliance be it whatever headgear, RME, fixed to create space you don’t have this issue.

    I believe some units would carry out the treatment as they take the initial IOTN, but I doubt this is universal.


  15. Hi Kevin,
    I have always wondered how the un-erupted canine would be influenced by the extraction of the c if there was a solid wall of bone between it and the c. Is it not possible that there are three different conditions: one the failure of the c to exfoliate even though the canine is contacting the root of the deciduous tooth (obviously the extraction of the c would be helpful in these cases) and another condition where the path of eruption of the canine was abnormal and the canine is not contacting the root of the deciduous tooth ( clearly the extraction of the deciduous tooth would be un-helpful) and a third condition where there is crowding.

  16. Hi Kevin
    Doesn’t the data suggest the chance of ” non” eruption was 1.7 times greater than in the extraction group ?

    “The risk ratio suggests that the chance of eruption of the permanent canine was 1.7 times greater than in the extraction group.”

    Thanks for the blog

  17. Kevin
    Thank you for your great blog and invaluable information. Like Nicola Parkin above, I note that extraction of the Cs appears to accelerate the eruption of palatal U3s. It is a shame then that the observation period stops at 24 months as the average for the nonextraction group approaches that period. It reminds me of the first phase of the Tulloch et al “functional appliances” study and a longer period of study may decrease the actual benefit. Still, may reconsider treatment recommendations the next time I see this presentation as you have until a longer study is done.

Leave a Reply

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