The preparation of curved root canals has always been a challenge. Many studies [14–16] have showed that canal curvature results in an increased risk of canal transportation and instrument fracture, which leads to the failure of root canal therapy. Unfortunately, around 84% human teeth have clinically noticeable root canal curvatures [17]. It is well-known that NiTi instruments are superior to SS ones in shaping the narrow and curved canals.
However, several of the recent studies [9, 12] indicated that, as a unique rotary system made from SS, GF was more flexible and prepared better for oval root canals than some of the NiTi instruments, and our study showed the GF system was also more centered and less transported than PT in the middle and apical sections of the narrow and curved root canals. This finding is consistent with that of Moreinos [9] who reported that GF files created less vertical force on bent canals.
There is a large volume of published studies [18, 19] describing the threat of root canal transportation. Once canal transportation scores exceed 0.15 mm, the possibility of preparation failure will increase. Excessive transportation results in unreliable canal walls—one side of canals could not be completely cleaned, While the other side remains thin dentin, which will lead to a series of complications such as perforation, root fracture, periapical periodontitis, and so on [20, 21]. The instrumentation of PT F2 in 5–7 mm in our study seemed to be riskier than GF on account of their excessive results at these levels (average transportation value is 0.174,0.305,0.224, respectively), obvious transportation even can be observed in the middle section (Fig. 2).
The results seemed that PT is not suitable for narrow and severely curved canals. In addition to the curvature mentioned above, two other possible reasons are the big taper and brushing movement of PT system. PT F2 is a multi-taper instrument, the maximum taper can reach 0.08, which means that, compared with GF system, more dentin needs to be removed to achieve the same shaping effect and several reports [22, 23] have shown that 0.06 or smaller taper is enough for mechanical debridement. Furthermore, brushing motion [24] has been proved to be more aggressive but safer than pecking. Multiplying the number of unidirectional brushing strokes will lead to significant cutting of the canal wall and result in unnecessary root canal enlargement in one direction [24].
However, in the present study, two cases of instrument separation occurred in GF groups. This outcome was contrary to that of Moreinos [9] who found GF files have greater resistance to cyclic fatigue than NiTi files. It is worth mentioning that resin materials have some limitations in simulating human teeth preparation, especially under high rotation speed. Although plastic blocks make it possible to standardize and visualize samples and the credibility of resin blocks as an experimental model for the analysis of preparation effect has been proved [25], due to the difference of hardness, preparation in simulated root canals would generate much heat, which might soften the resin material and increase the risk of fracture, especially at a speed of nearly 6500 rpm in GF groups. Thus, it is necessary to take further studies in extracted teeth with curved root canals.
Interestingly, there occurred a new mode of instrument separation in the present study. The separated file was just like a ring in the canal, thus it was hollow so that other instruments can insert into the canal and take it out easily (Fig. 3). The reason for this is not clear but it may have something to do with its coiling method of the three SS wires.