Dentine hypersensitivity: development and evaluation ofamodel in situ to study tubulepatency

Lesions of dentine hypersensitivity have numerous tubules open at the dentine surface as opposed to non-sensitive dentine where tubules are mostly covered by a smear layer. The present two studies were designed to model both states in situ and evaluate the effects of agents on the model.

Etched (sensitive) and smeared (non-sensitive) dentine specimens prepared from human third molar teeth were retained in lower buccal acrylic appliances. The results of Study 1 led to the development of a method to ensure tubules were sectioned at right angles. Study 1 was a 5-day period, seven treatment regimens randomised, part blind cross over design involving five subjects. Treatments were 2x day application of a desensitising (SA) or non-desensitising toothpastes (F) or chlorhexidine (CHX) mouthwash with or without drinking orange juice (OJ)(1 l/day). A no treatment group (P) allowed plaque to accumulate. The evaluation of effects was observational by scanning electron microscopy (SEM). Study 2 involved 1 subject, 4 treatments applied once and studied after 0, 6, and 12 h by SEM with image analysis. The 12 h groups were studied with and without imbibing water or OJ. Treatments were two desensitising (SA & SC) and one non-desensitising (F) toothpastes and an in office product (DS).

Study 1: Treatment SA resulted in occlusion of tubules an outcome little changed by OJ. Treatment P produced a bio-film, which covered the dentine surface. CHX produced some tubule occlusion in three of the five subjects. F+/-OJ and CHX+OJ had little effect and tubules remained open. For smeared specimens toothpastes and OJ removed the smear layer but SA+/-OJ blocked tubules. P and CHX had no effect on the smear layer. Study 2: At 0 h, tubule occlusion was in order of magnitude DS>or=SA>SC>F. After 6 and 12 h with SA, SC and F some loss of occlusion occurred but not DS. Water and OJ by 12 h decreased occlusion for SA, SC and particularly F. Water and OJ removed virtually all of DS.

The aims of Study 1 were achieved and effect of treatments was not inconsistent with data in vitro. The need for more standardisation of specimens was appreciated and applied in Study 2 to allow image analysis to quantitatively record data. Further use of the model in randomised controlled clinical trials is envisaged.