Sir, we would like to congratulate Y. Issa et al. for their paper in the recent issue of the BDJ: Oral lichenoid lesions related to dental materials (BDJ 2005; 198: 361). It provides further evidence of the role played by dental materials, and particularly amalgam, in some oral lichenoid lesions and confirms the value of replacing restorations that are in direct contact with the oral mucosa in such cases.

It is clear that amalgam replacement can be beneficial in patients with amalgam associated lichenoid lesions (OLL) but is unhelpful in cases of idiopathic oral lichen planus (OLP). The dilemma for most clinicians, however, is in clearly distinguishing these two conditions. Where there is only one amalgam restoration in contact with the oral mucosa and one adjacent lichenoid lesion then, in practice, the close topographical relationship between the lesion and restoration may make the decision to replace the one restoration straightforward. However, the real difficulty arises in patients with lots of amalgam restorations and a pattern of lesions that would be consistent with either diagnosis, a not uncommon situation. In such situations, anything that will improve the reliability of the diagnosis is important given the tissue destruction, high financial cost, time and inconvenience to the patient of replacing restorations, particularly if it is of no benefit.

The authors make reference to a study1 in which we concluded that 'A strong clinical association between lesions and amalgam restorations, plus a positive patch test, was a good predictor of lesion improvement on amalgam replacement' ie help distinguish cases of OLP from OLL. Although they agree with our conclusion about the importance of the topographical relationship, they criticise our conclusion about the value of patch testing. However, their criticism appears to be based on a misinterpretation of our results and requires a reply.

Our conclusion was based on the finding that only 3.9% of patients, whose lesions had little or no association with their restorations (ie more likely to be OLP) were patch test positive, compared to 70% of patients whose lesions had a strong or very strong association between their lesions and their restorations (ie more likely to be OLL). This difference was statistically significant (p<0.0001). The paper by Issa et al. does not appear to have looked at the value of patch testing in distinguishing OLP from OLL.

What their study did indicate was that there can be a significant false negative rate when patients with apparent OLL are patch tested. In future, this is something that might presumably be improved with better knowledge of the allergens involved and better patch testing protocols.

We continue to believe that our data support the view that patch testing, in conjunction with the topographical relationship between the lesions and any amalgam restorations, combine in helping to distinguish these two conditions. Indeed, our data showed that the combination of a positive patch test and a strong/very strong clinical association between the lesions and the restorations was a better predictor of lesion improvement following amalgam replacement (with an NNT = 1.1 [the number of patients with a positive test result that would need to be treated to result in one successful treatment outcome]) than either a positive patch test (NNT = 2) or a strong/very strong clinical association alone (NNT = I.46). Since the study by Issa et al. did not look at the value of patch testing in distinguishing OLL from OLP, we do not feel that their conclusion that patch testing is of limited value, or their criticism of our findings, is justified. Interested readers may like to read both papers and draw their own conclusions.

The authors of the paper referred to did not provide a response to this letter.