Basic Research—Technology
Evaluation of the Efficiency of a New File Removal
System in Comparison With Two Conventional Systems
Yoshitsugu Terauchi, DDS,* Le O’Leary, DDS,† Izumi Kikuchi, DDS, PhD,*
Mami Asanagi, DDS,* Takatomo Yoshioka, DDS, PhD, Chihiro Kobayashi, DDS, PhD,* and
Hideaki Suda, DDS, PhD*
Abstract
A novel file-removal system (FRS) was designed to
address weak points of conventional file-removal meth-
ods. The purpose of this study was to compare file-
removal time and dentin removal rates among the FRS,
the Masserann kit (Micro-Mega, Besancon, France),
and an ultrasonic file-removal method. Ninety extracted
mandibular incisors with separated nickel titanium files
were divided into 3 groups of 30 teeth each. Groups 1,
2, and 3 had file-removal attempts made by using the
Masserann kit, a CPR-7 titanium ultrasonic tip (Obtura-
Spartan Corp., Fenton, MO), and the FRS, respectively.
Each group had three operators removing the separated
files. Pre-/postoperative digital radiographs were down-
loaded into image analyzing software that calculated
the amount of dentin removed. The FRS needed less
time and had less dentin loss than the others (p ⬍
0.05). There were statistical differences between the
experienced operator and less experienced operators
regarding the file-removal time and the dentin removal
rates (p ⬍ 0.05). (J Endod 2007;33:585–588)
Key Words
Dentin-removal rate, file removal, file-removal time,
masserann kit, separated nickel titanium file, ultrasonic
tip
*From Pulp Biology and Endodontics Section, Tokyo Med-
ical and Dental University, Tokyo, Japan; and †Advanced Mi-
croendodontics, P.C., Plano, TX.
Address requests for reprints to Dr. Yoshitsugu Terauchi,
1557-4 Shimotsuruma Yamato city, Kanagawa prefecture,
Japan 242-0001. E-mail address: terauti@tk.usen.ne.jp.
0099-2399/$0 - see front matter
Copyright © 2007 by the American Association of
Endodontists.
doi:10.1016/j.joen.2006.12.018
JOE — Volume 33, Number 5, May 2007
N ickel titanium (NiTi) rotary files are now widely used for cleaning and shaping of
the canal system. The inherent characteristics of greater elasticity and resistance to
torsional fracture of the NiTi alloy have allowed clinicians to efficiently obtain predict-
able results with nonsurgical endodontic treatment (1). One of the most dreaded
complications with NiTi files is the separation of the instrument during use. In general,
instruments used in rotary motion separate by two distinct modes: torsional and flexural
(2). Various factors have been associated with the following iatrogenic mishaps: oper-
ator experience (3), rotational speed (4), canal curvature (5), instrument design and
technique (6, 7), torque (8), manufacturing process (9), repeated use (10), and
absence of glide path (11).
Although there is a higher demand for removal of NiTi separated files, to date no
standardized procedure for successful instrument removal has been established (3).
For many years, the traditional method of retrieving instrument fragments among cli-
nicians was to use the Masserann kit (Micro-Mega, Besancon, France) (5– 8). This
system is very effective in the retrieval of instrument fragments located within the straight
part of the canal (5– 8). It cannot be applied to cases involving fragments located in the
midroot or apical third of the root or in severely curved canals because the technique
involves the removal of considerable amounts of sound dentin, which can weaken the
root structure and increase the risk of perforation (9, 10).
Most recently, the use of ultrasonic tips has been found to be the most effective
method for removing separated instruments from root canals without sacrificing a great
deal of sound dentin (11). However, in the use of ultrasonic tips, difficult cases are
occasionally encountered in which the separated file cannot be retrieved from the canal
even though it is seen to be loose. The challenge is that the separated part of the file tends
to engage on the outside wall of a curved canal. These cases ultimately require longer
treatment time, and the process of troughing around the separated instrument is re-
peated until either an iatrogenic error occurs, such as strip perforation, canal trans-
portation, zipping, or ledging, or the fragment is finally extruded out of the canal. Nagai
et al. (12) reported that the time required for ultrasonic file removal varied from 3 to
40 minutes, whereas the Masserann’s method varied from 20 minutes to several hours,
depending on canal shapes. In addition, it has been observed that removal of NiTi
instruments with ultrasonics is more difficult than removal of stainless-steel instruments
because there is a greater tendency for NiTi instruments to fracture repeatedly (13).
This study was conducted as a series following our previously published case
report (14). We developed a new file removal system (FRS) with the intended goal of
minimizing both the dentin removal rate and the time required to remove the separated
instruments. The purpose of this study was to compare the file-removal time and the
dentin removal rates among the FRS, the Masserann kit, and the ultrasonic tip, and
among operators with different experience levels.
Materials and Methods
Ninety extracted human mandibular incisors with straight roots were used in the
study. It was visually recognized that all the experimental teeth had completely devel-
oped root apices. The clinical crowns were removed, and conventional access cavities
were prepared with high-speed diamond burs. The working length was established by
inserting a #15 K file into the canal until the tip was visualized at the apex. The root
canals were prepared to the canal terminus by using 25-mm long greater taper rotary
New File-Removal System 585
Basic Research—Technology
Figure 1. Magnified views of each instrument. (a) CBA. (b) CBB. (c) Ultrasonic
tip. (d) Loop device. Vertical bar on the right side corresponds to 1 mm in
length.
nickel-titanium files with #20/.06 taper (Dentsply/TulsaDental, Tulsa,
OK) on an Endo-Mate TC electronic motor (NSK Inc., Tochigi, Japan). A
separated instrument was then created in the canal by notching a greater
taper rotary file with #20/0.10 taper to a depth of half the instrument
thickness and at 3 mm from the tip by using a low-speed diamond disk.
The notched instrument was then used in the canal at 250 rpm until it
separated in the apical one third of the canal. The coronal end of the
separated instrument measured approximately 0.5 mm at 3 mm from
the tip.
The teeth were then divided into 3 groups of 30 samples each:
groups 1, 2, and 3 for separated file removal with the Masserann kit, a
CPR-7 titanium ultrasonic tip (Obtura-Spartan Corp., Fenton, MO), and
the FRS, respectively. The CPR titanium ultrasonic instruments are
known for their long lengths and thin diameters, allowing them to work
well in severely restricted space. The CPR-7 (0.4 mm in diameter) was
selected because it was the first instrument from the CPR series (from
largest to smallest) that had adequate length and diameter to passively
reach the separated files under the condition of this study. These exper-
iments were performed by three operators. Operators 1 and 2 had 8 to
9 years of clinical endodontic experience at the time of the study, and
they both had no previous experience with the instruments used in this
study. Operator 3, the first author of this study, had 14 years of clinical
endodontic experience and 5 years of familiarity with all the file-removal
instruments involved in this study. File-removal attempts continued until
the entire separated file was removed, regardless of secondary fracture
occurring during the procedure. If the separated file was extruded out
of the apex or the canal was perforated, the attempt was cancelled. This
procedure was continued until each operator successfully removed 10
separated files in each group. File-removal attempts were also cancelled
when the procedure exceeded 90 minutes. The irrigation protocol for
each test group consisted of rinsing the canals intermittently with 17%
TABLE 1. Number of Unsuccessful Cases Because of Perforation (P), Overtime
Attempt (T), Overtime Attempt With Secondary Separated File Fracture (S),
and a Separated File Pushed out of the Apex (A)
Group 1 Group 2 Group 3
Operator 1 P,P P,S — Operator 2
Operator 2 P P,T,A — Operator 3
Operator 3 — — — a
586 Terauchi et al.
TABLE 2. File-Removal Time in Minutes (Mean⫾Standard Deviation) With
Multiple Comparisons Among the Operators and Among the Experimental
Groups
Group 1 Group 2 Group 3
Operator 1 8.1 ⫾ 4.4 20.5 ⫾ 18.7 5.5 ⫾ 3.5 a
Operator 2 16.1 ⫾ 9.3 19.2 ⫾ 12.0 4.6 ⫾ 1.6 b
Operator 3 24.0 ⫾ 16.8 29.1 ⫾ 20.4 11.1 ⫾ 3.1a,b
c c c
EDTA solution, followed by 5.25% sodium hypochlorite, after which the
canals were dried with paper points.
Group 1
The Masserann kit was used according to the manufacturer’s in-
structions. The Peeso reamer (Mani, Inc., Tochigi, Japan) was chosen
over the Gates Glidden (Mani, Inc.) drill for straight-line access based
on the assumption that the parallel configuration of the bur would help
maintain a straight pathway down the canal and possibly conserve the
dentin during the process. A staging platform was created with a #4
Peeso reamer, 1.3 mm in diameter, which would make room for the
entry of the smallest (1.2 mm in diameter) extractor down the canal.
After the access, the trephine bur (1.1 mm in diameter) was worked in
a counter-clockwise direction to further expose the coronal portion of
the fragment. Once an adequate amount of file was exposed and con-
firmed under a microscope (Opmi 111; Zeiss, Oberkochen, Germany),
the Masserann extractor was used to retrieve the separated instrument.
Group 2
Straight-line radicular access was achieved with a # 2 Peeso reamer
(0.9 mm in diameter). The CPR-7 was then activated at the lowest power
setting (Spartan Ultrasonic Endo J15; Obtura-Spartan Corp., Fenton, MO)
and trephined circumferentially in a counter-clockwise direction to expose
the coronal aspect of the file in dry conditions as recommended by Ruddle
(15). This procedure was continued until the obstruction was freed and
floated out of the canal by the ultrasonic tip.
Group 3
The FRS involved three steps comprising three different techniques
and four newly designed instruments (14) (Fig. 1). Each step was
performed sequentially until the separated file was removed. The details
of the techniques can be obtained from our previously referenced case
report (14).
The procedure was timed from the moment the file-removal pro-
cess began up to the point when the separated file was removed from the
canal, and it was rounded off to the nearest whole minute.
A standard urethane jig was designed to hold an X-ray tube head at
a fixed distance from a DEXIS digital radiographic sensor (Provision
Dental Systems, Palo Alto, CA). The head of the X-ray tube was fixed in
the aperture of the urethane jig, and the digital radiographic sensor was
placed in the wall on the opposite side perpendicular to the X-ray tube.
Mesiodistal (MD) and labiolingual (LL) planes were created on the
TABLE 3. Dentin Removal Rates (%) in the Mesiodistal (MD) Direction
(Mean⫾Standard Deviation) With Multiple Comparisons Among the
Operators and Among the Experimental Groups
Group 1 Group 2 Group 3
Operator 1 162.7 ⫾ 17.6 147.8 ⫾ 25.6 130.2 ⫾ 15.0
164.9 ⫾ 23.0 138.2 ⫾ 28.1 117.6 ⫾ 14.5
158.8 ⫾ 27.0 135.7 ⫾ 17.3 125.7 ⫾ 13.2
a a
JOE — Volume 33, Number 5, May 2007

TABLE 4. Dentin Removal Rates (%) in the Labiolingual (LL) Direction
(Mean⫾Standard Deviation) With Multiple Comparisons Among the
Operators and Among the Experimental Groups
Group 1 Group 2 Group 3
Operator 1 162.8 ⫾ 33.0 154.1 ⫾ 28.7 144.8 ⫾ 12.9 a
Operator 2 176.6 ⫾ 19.0 156.4 ⫾ 40.8 128.8 ⫾ 21.3 b
Operator 3 164.5 ⫾ 18.1 121.3 ⫾ 14. 113.6 ⫾ 9.6 a,b
c c c the canal during step 2, it was highly expected that the fragment would
sample roots with quick self-curing acrylic resin (Unifast; GC Dental
Products Corp., Aichi, Japan) to stabilize the sample roots for radio-
graphic superimposition. Cross lines were drawn on both planes of the
sample roots and the plate of the digital radiographic sensor holder with
a black oil-based marker so that they could be placed back in the same
positions for pre- and postoperative radiographs in both the MD and the
LL directions. The standardized digital radiographs were downloaded in
JPEG format from the DEXIS digital radiographic system and imported
into the image analyzing software (Photoshop 6.0, Adobe Systems, Inc.,
San Jose, CA) to compare the pre- and postoperative radiographs. This
was accomplished by outlining the periphery of the entire canal on the
image, and the number of the pixels within this outline representing the
canal space was then recorded. The determination of the dentin re-
moval rate for each case was calculated by dividing the value of the canal
space on the preoperative pixels into that of the postoperative pixels.
Outcomes of each value for both the working time and the dentin re-
moval rates were analyzed by Fisher’s Protected Least Significant Differ-
ence (PLSD) at 0.05 significance level.
Basic Research—Technology
The time needed for file removal in group 3 ranged from 2 to 17
minutes, whereas group 1 varied from 2 to 70 minutes and group 2 from
4 to 80 minutes. The removal time varied as a result of the root canal
shape and the wedging severity of the fragment (12). From an observa-
tional point of view, even when using a microscope, it was difficult to
predict when the separated files in groups 1 and 2 would be dislodged
from the canals. In group 3, when a separated file was seen swaying in
be removed before long using the loop device. In group 3, 23 of 30
separated files were successfully removed with ultrasonics without re-
sorting to the loop device, which indicated that ultrasonic vibration
alone is very effective in removing separated files from straight roots.
Two cases in group 2 had secondary fracture, and they ultimately re-
quired longer time to complete the file-removal process. Overapplica-
tion of ultrasonic vibration to a separated file could cause secondary
fracture, resulting in more complication because extra time is then
required for removal of the remaining fragment. In fact, Hulsmann and
Schinkel (16) reported a shorter fragment was more difficult to retrieve
than a longer fragment.

Results
Ninety of 98 separated files consumed in all the groups were suc-
cessfully removed. The number of unsuccessful cases and causes for
failure are listed in Table 1A. Mean file-removal time in minutes, mean
dentin removal rates (%) of MD and LL directions with standard devi-
ations and multiple comparisons among the operators and among the
experimental groups were determined by using Fisher PLSD and are
listed in Tables 2-4. Representative samples in each group are shown in
Figure 2. The results showed that there were significant differences
between operators 1 and 3 as well as operators 2 and 3 in the file-
removal time and in the LL dentin removal rates (p ⬍ 0.05). Group 3
had the shortest mean file-removal time, whereas group 2 had the
longest. Group 1 had the largest mean dentin removal rate, whereas
group 3 had the least.

Discussion
In the results, there were no statistically significant differences
between operators 1 and 2 in both dentin removal rates and working
time. However, between operator 3 and the other operators, there were
significant differences in the working time and the dentin removal rate
in the LL direction. The results showed operator 3 had the longest
file-removal time and the lowest dentin removal rate among all the
operators in each group, which shows that operator 3 was very careful
not to sacrifice too much dentin during the process. The removal of a Figure 2. (a) Representative sample from group 1. (a1, a2) Pre-and postop-
large amount of dentin loosens the file from the canal wall, thus making erative radiographs from the LL view. a2 sample has an arrow pointing to a
the file-removal attempts much easier. Unfortunately, excessive loss of trenched out area created by a trephine bur. (a3, a4) Pre-and postoperative
dentin does increases the risk of iatrogenic accidents such as canal radiographs from the MD view. (b) Representative sample from group 2. (b1,
b3) Pre- and postoperative radiographs from the LL view. (b2) Radiograph
perforation, as in the case of operators 1 and 2 having five cases for
showing a secondary fracture during file-removal attempts from the LL view.
groups 1 and 2. Even though operator 3 had the longest file-removal (b4, b6) Pre-and postoperative radiographs from the MD view. (b5) Radio-
time, there was more preservation of dentin when compared with the graph showing a secondary fracture during file-removal attempts from the MD
other operators. Operator 3 was successful in removing all separated view. (c) Representative sample from group 3. (c1, c2) Pre- and postoperative
files in all groups without any procedural errors. The success is prob- radiographs from the LL view. (c3, c4) Pre- and postoperative radiographs from
ably because of the operators’ experience in removal of separated files. the MD view.

JOE — Volume 33, Number 5, May 2007 New File-Removal System 587
Basic Research—Technology
There were significant differences in the dentin removal rates
among all the groups. Group 1 had the largest instruments out of the
three groups for file removal, whereas group 3 had the smallest instru-
ments. The ultrasonic tip used for group 3 was two times thinner than
group 2. The differences between groups 2 and 3 are the diameter of the
tip and its ability to work in tight spaces. The dentin removal rates were
directly in proportion to the instrument size, even though the time
needed for file removal in each group was significantly different.
The allotted working time for removal of the separated instruments
was set at 90 minutes because this parameter would give adequate time
for each system to prove itself and it’s also a reasonable chair time for
treatment. Suter et al. (13) had recommended that removal attempts of
separated instruments from root canals should not exceed 45 to 60
minutes because the success rates may drop with increased treatment
time. They suggested that the lowered success rate could be because of
operator fatigue or overenlargement of the canal, which compromises
the integrity of the tooth and increases the risk of perforation.
With any bench-top study, what we can glean from the findings is
the potential of the system. Therefore, under the condition of this study,
the FRS was found to be very effective in reducing file-removal time and
dentin removal rates. Clinicians inexperienced in separated file removal
tend to sacrifice more dentin than an experienced clinician.
References
1. Walia H, Brantley WA, Gerestein H. An initial investigation of the bending and tor-
sional properties of nitinol root canal files. J Endod 1988;14:346 –51.
2. Cohen S, Hargreaves K. Pathways of the pulp, ed 9. St. Louis: Mosby, 2006.
588 Terauchi et al.
3. Mandel E, Adib-Yazdi M, Benhamou L-M, et al. Rotary NiTi profile systems for pre-
paring curved canals in resin blocks: influence of operator on instrument breakage.
Int Endod J 1999;32:436 – 43.
4. Sano S, Miyake K, Osada T. A clinical study on the removal of the broken instrument
in the root canal using Masserann kit. Kanagawashigaku 1974;9:50 –7.
5. Masserann J. Removal of metallic fragments from the root canal. J Br Endod Soc
1971;5:55–9.
6. Feldman G, Solomon C, Notaro P, Moskowitz E. Retrieving broken endodontic in-
struments. J Am Dent Assoc 1974;88:588 –91.
7. Krell KV, Fuller MW, Scott GL. The conservative retrieval of silver cones in difficult
cases. J Endod 1984;10:269 –73.
8. Yoldas O, Oztunc H, Tinaz C, Alparslan N. Perforation risks associated with the use of
Masserann endodontic kit drills in mandibular molars. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 2004;97:513–7.
9. Friedman S, Stabholtz A, Tamse A. Endodontic treatment: case selection and tech-
niques. J Endod 1990;16:543–9.
10. Yared G. In vitro study of the torsional properties of new and used Profile nickel
titanium rotary files. J Endod 2004;30:410 –2.
11. Ward JR, Parashos P, Messer HH. Evaluation of an ultrasonic technique to remove
fractured rotary nickel-titanium endodontic instruments from root canals: an exper-
imental study. J Endod 2003;29:756 – 63.
12. Nagai O, Tani N, Kayaba S, Osada T. Ultrasonic removal of broken instruments in root
canals. Int Endod J 1986;19:298 –304.
13. Suter B, Lussi A, Sequeira P. Probability of removing fractured instruments from root
canals. Int Endod J 2005;38:112–23.
14. Terauchi Y, O’Leary L, Suda H. Removal of separated files from root canals with a new
file removal system: case reports. J Endod 2006;32:789 –97.
15. Ruddle CJ. Nonsurgical retreatment. J Endod 2004;30:827– 45.
16. Hulsmann M, Schinkel I. Influence of several factors on the success or failure of
removal of fractured instruments from the root canal. Endod Dent Traumatol 1999;
15:252– 8.
JOE — Volume 33, Number 5, May 2007