The use of needle holders in CTF guided biopsies as a dose reduction tool

Abstract Purpose The purpose of this study was to evaluate the efficacy of needle holders in reducing staff hand exposure during biopsies guided by computed tomography fluoroscopy (CTF), through the analysis of data acquired during a detailed monitoring study, undertaken in parallel with an ongoing optimization process to reduce hand irradiation. Methods Hand monitoring was performed with 11 extremity detectors, two per finger (base and tip) and one on the back of the wrist, for the left (dominant) hand, during two series of biopsies with comparable characteristics. The first series (47 biopsies) were performed with only quick‐check method (QC) and occasional side‐handle (SH) manipulation of the needle. The second series (63 biopsies) were performed after introducing needle holders (NH) in the course of an optimization process. Results Choice of technique (QC, QC + NH, QC + SH) by the interventional radiologist (IR) was related to biopsy difficulty. Measured hand exposure was low (< 1 mSv) for all QC‐only procedures, and for most of the QC + NH procedures. Occasional side‐handle manipulation still occurred during challenging biopsies, so that 8% of biopsies in the second series accounted for ~70% of total fingertip dose (~90 mSv). The methodology used allowed a detailed insight into the dose reduction achievable with needle holders during real procedures, without the limitations of phantom measurements. Conclusions Needle holders proved effective in reducing mean hand exposure during clinical procedures where real‐time manipulation was necessary. Occasional side‐handle manipulation was found to contribute disproportionately to hand exposure. This highlights the importance of individual hand monitoring during CTF guided procedures.


| INTRODUCTION
Computed tomography (CT) is a useful imaging technique to guide interventional radiology procedures, allowing good visualization of small lesions and neighboring critical structures, as well as the planned needle path. The possibility of in room real-time CT imaging (also known as CT fluoroscopy, CT fluoro or CTF) is an additional advantage, particularly useful in lung biopsies where respiratory motion causes lesion displacement. 1-3 CTF guidance for lung biopsies provides high diagnostic accuracy with fewer complications. 3 Some newer CT scanners offer the possibility of multislice CT guidance (MS-CT guidance), which proved effective in reducing radiation doses to patient and staff. 4 But CTF guidance is still essential when lesions are subject to major respiratory movements. 4 The main concern with CTF is radiation exposure. The designation CTF is adopted in this work, because the name "CT fluoroscopy" can be misleading. As pointed out by Miller et al, CTF is not fluoroscopy at all: it is different from conventional fluoroscopy in both equipment and technique. 5 In CTF guided biopsies, the needle advancement occurs in the imaging/irradiation plane, so direct manipulation of the needle during irradiation would put the hands in the direct beam, where the dose rate may be as high as 4 mGy/s. 6 At this dose rate, the occupational limit of 500 mSv per year for the hands 7,8 would be exceeded in less than 3 min (or 6 procedures considering the mean CTF times in this study). The quick-check (QC) method proposed by Silverman et al prevents direct hand irradiation by using intermittent imaging to check needle position, while needle advancement occurs during beam-off. 9 Alternatively, needle holders (NH) may be used to avoid direct manipulation of the needle during continuous viewing. 1,2,10,11 Dedicated needle holders have been developed 1,10-12 but many authors prefer metallic sponge forceps or towel clamps due to their widespread availability, lightweight, strength, ease of sterilization and relatively low cost. 2,9,13,14 If a towel clamp is used to grasp and manipulate the needle, the hand may be kept 10-20 cm away from the irradiation plane during needle advancement. CTF dose rates drop very rapidly with distance from the scan plane. 6 Other protective devices can be used in combination with needle holders, such as radiation attenuation gloves, lead drapes placed on the patient, and angular beam modulation (ABM). ABM means the CT scanner automatically turns off the irradiation beam during a pre-selected part of the tube rotation. 15 Not all manufacturers offer an ABM option. Nevertheless, it is important to remember that CTF has the potential for very high hand exposures, and operators should be mindful of this when performing these procedures.
All these protective options have been tested on phantoms, 1,12,[15][16][17] but phantom studies are limited: hand movements cannot be reproduced, and it is not possible to take into account the complexity of the procedures (including the difficulties introduced by protective measures themselves). Intermittent imaging has obvious limitations when dealing with respiratory motion, while needle holders have been associated to decreased tactile feedback and grip, as well as bending of thinner biopsy needles. 2,9,13,14,18,19 Because of such limitations, it is likely that interventional radiologists (IRs) will resort occasionally to manual manipulation of the side handle (SH) of the co-axial guiding needle (side-handle manipulation), as described by Buls et al. 20 This puts the IR's hand very close to the irradiation beam, and may lead to high exposures near the fingertips. 21 It is difficult to perform measurements of hand exposure during actual CTF guided procedures. Moreover, to evaluate the effect of protective measures like using needle holders, it would be necessary to compare success rates for similar biopsies performed with different techniques. IR's prefer real-time visualization for smaller lesions, particularly when respiratory motion is a concern (e.g., lung biopsies).
Even in abdominal biopsies, preference for the quick-check method is associated with larger sized lesions. 9 So far, the most comprehensive study is that of Irie et al, which compared finger doses per procedure (measured with ring dosimeters) using needle holders of different lengths, with and without a protective lead plate, for 55 procedures (mostly in the chest area). 10 In the studies of Carlson et al 14 20 However, finger doses may be 20 times higher than at the back of the hand when side-handle (SH) manipulation is used. 21 Other recent studies have focused on patient doses, 4,23 or assessment of recently introduced scanner features, like MS-CT 4 and iterative reconstruction. 24 Because the total length irradiated during CTF is very small, patient effective doses associated with CTF are lower than those of a diagnostic CT scan. 6 CTF used for needle positioning typically accounts for~15% of the total DLP (dose length product) of a CTF guided procedure. 23,25 Peri-interventional acquisitions (such as the initial helical CT scan acquired to determine the optimal access, and the lesion examination after the intervention) account for~85% of the total DLP in CTF guided biopsies. 23 On the other hand, since irradiation involves multiple rotations at the same position, patient skin doses may be a concern with CTF. 6 The purpose of this work was to assess and improve occupational exposure, particularly the hand exposure of the IR, which was monitored in detail during CTF guided biopsies, in parallel with an ongoing optimization process to reduce hand irradiation. The routine use of needle holders (NH) was introduced as part of this optimization process, and therefore hand exposure was measured for two series of biopsies, before (before NH) and after (after NH) the intro-

| MATERIALS AND METHODS
Detailed hand monitoring was performed using the methodology described by Pereira et al. 26 The location of the in room viewing monitor favors left-handed needle manipulation, therefore the left hand is usually the most exposed. A thin plastic glove was prepared for the left hand, containing casings for the placement of 11 extremity detectors, two per finger (base and tip) and one on the back of the wrist, as shown in Fig. 1(a). Another glove was prepared for the right hand, with only six casings for detectors, one per finger (at the base) and the 6th at the back of the wrist, as shown in Fig. 1b).
These gloves were tested before use, and found not to reduce hand mobility, dexterity or sensitivity in any way. 26 A secondary objective of this detailed monitoring was to charac- The WB dosimeters consist of the Harshaw 8814 card and holder containing two LiF:Mg,Ti (TLD-100) detectors, with adequate filtration for the measurement of H p (10) and H p (0.07). The operational quantities H p (10) and H p (0.07) are recommended for assessment of effective dose and equivalent dose to local skin, respectively, defined as the dose equivalent to soft tissue at a depth of 10 mm and 0.07 mm below a specified point on the body. 7,27 The extremity detectors used in the gloves were of the Ext-Rad type with LiF:Mg,Cu,P (TLD-100H), for the measurement of H p (0.07) in routine individual monitoring.
All detectors were calibrated and read out by an approved dosimetry service (ADS, as defined in report RP160 of the European Commission 27 ) which is also a provider of individual monitoring services. This allows direct comparison with ring dosimeter measurements from routine individual monitoring. Dosimeters were read using two Harshaw 6600 readers. The extremity dosimeters were calibrated in terms of H p (0.07) using a N120 X-ray beam incident on an ISO rod phantom, and the WB dosimeters were calibrated in terms H p (10) and H p (0.07) using a 137 Cs beam incident on a ISO water phantom. [28][29][30][31] The minimum limit of detection is 0.02 mSv in terms of H p (10) and H p (0.07) for WB detectors, and 0.07 mSv in terms of H p (0.07) for the extremity detectors. All biopsies were performed by the same experienced interventional radiologist (IR), using a Toshiba Asteion four-slice scanner and 120 V, 0.75 s rotation time and 8 mm beam collimation. 40 mA was used for most biopsies, and increased to 50 mA whenever necessary (usually for abdominal biopsies). The IR wears a wrap-around lead apron (0.35 mm lead equivalent at the back, 0.70 mm at the front) and a thyroid shield (0.5 mm lead equivalent). As specified by the hospital's radiation safety program, both hands are normally monitored with ring dosimeters, and a whole body dosimeter is worn under the lead apron. These three dosimeters are read monthly, and are separate from this study.
The typical biopsy procedure is very similar to that described by Buls et al. 20 The patient is brought into the CT room, the procedure is explained and a consent form signed. The patient is then positioned for a preliminary CT scan, restricted to the region where the lesion was previously detected. Once the lesion is located, a plane (slice position) is chosen and the couch moved to the appropriate position. The position indicated by the laser lights is marked with a radio-opaque marker, and a single axial scan is acquired. The distance from the marker to the lesion is measured on the console, and a needle course is plotted. A sterilized drape is placed on the patient, the biopsy region is sterilized and local anaesthesia is administered.
Only the IR remains in the CT room during irradiation. The radiology technologist operates the CT scanner from the main console outside the CT room, with communication via the audio system.
The optimization process started with the acquisition parameters, to reduce the tube current to the lowest possible value. All the data presented here were obtained with optimized acquisition parameters. Initially, the IR wore attenuation gloves and, whenever the quick-check (QC) method proved insufficient, the biopsy needle was grasped by the side handle (SH), as described by Buls et al. 20 A total of 47 biopsies were performed with this methodologythese constitute the first series, named "before needle holder" ("before NH").
In a second series, named "after needle holder" ("after NH"), an

| RESULTS
Patient statistics and biopsy type were similar for both series of measurements, as summarized in Tables 1 and 2, respectively. Maximum H p (0.07) reading for the second series was 42.89 mSv for the left hand, and 0.87 mSv for the right hand. This confirms that the left hand is dominant for needle manipulation, independent of the location of the IR relative to the CT couch ( Table 3). Interestingly, the highest value of H p (0.07)max is approximately the same for both series, as indicated by an arrow in Fig. 4. This is consistent with the fact that side-handle manipulation (SH) was considered necessary in a small number of biopsies, even after the introduction of needle holders. SH is clearly associated with high hand exposure (Fig. 4, Table 4).
The effect of introducing needle holders is clearly seen in Fig. 4 | 253 the beam. This is reflected also in the variation in exposure across the hand ( Table 4). The greatest range of variation is associated with side-handle manipulation (SH), where hand exposure is strongly influenced by spatial location and finger positioning.

| DISCUSSION
This study has some limitations, because hand exposure was assessed for only one IR, one CT scanner, a specific patient population, and certain types of CTF guided procedures. However, this also  Table 2). The selection process was the same for both series: hand exposure was assessed for all biopsies on randomly chosen days. Therefore, in this scenario, it seems reasonable to assume that the distribution of biopsy difficulty was similar for both series, and that hand exposures can be compared.
As shown in Fig. 2   the mean hand exposure observed in the first series of biopsies, the same limit of 500 mSv would be reached after only 100 procedures.
The introduction of needle holders was associated with a 9 s increase in the mean value of CTF beam-on time ( Table 2). Assessment of patient doses is outside the scope of this paper, which focuses on the hand exposure of the IR and aims to prevent it from exceeding the regulatory limits. Nevertheless, for a complete analysis it is important to estimate the impact of this increase (in CTF time) on patient skin doses. Considering the dose rate estimated by Keat of 4 mGy/s in the direct beam, 6 the maximum patient skin dose associated with CTF in this work was 408 mGy or 0.4 Gy (corresponding to the maximum beam on time of 102 s in Table 2). This is well below the threshold for deterministic effects (2 Gy for transient erythema 32 ). A 9 s increase in CTF beam-on time results in an estimated increase in patient skin dose of 36 mGy (~0.04 Gy), which is not a cause for concern. Moreover, it may be a transitory effect resulting from lack of familiarity with needle holders, which had just been introduced. The 9 s increase in CTF beam-on time did not noticeably affect mean procedure time (which includes prepping the patient and collecting the biopsy sample).
Evaluation of biopsy success rates is equally outside the scope of this work, because the use of needle holders remained optional during the second series of biopsies. As shown in Table 4, the IR chose to use QC + SH during the more challenging biopsies in the second series (~14%). Therefore, these biopsies were performed with the same technique used in the first series. In this situation, biopsy success rates (e.g., accuracy of advancement and needle placement) are assumed equal for both series. If the IR had been constrained to use only QC or QC + NH in the second series, than the success rate for these 14% biopsies would have had to be assessed. Without this constraint, the percentage of QC + SH procedures in the second series of biopsies gives some indication of how frequently this IR felt the needle holders to be a limitation.
The data presented in Fig. 4 suggests that, during the second series of biopsies, needle holders replaced side-handle manipulation in most procedures where real-time manipulation was necessary. As expected, most QC + NH procedures involve low hand exposure, so the high number of procedures performed with needle holders instead of side-handle manipulation resulted in the significant reduction in hand exposure seen in Fig. 2. As seen in Fig. 4, nearly half the procedures in the first series of biopsies resulted in H p (0.07)max > 5 mSv, compared to less than 20% of procedures in the second series. Some QC + NH biopsies were still associated with H p (0.07)max > 5 mSv, but this occurred mostly in the first month of the second series of measurements, and may have been associated with inexperience in the use of needle holders.
The percentage of biopsies performed with the quick-check method (QC) alone appears to have remained the same, despite availability of needle holders (Fig. 4). This is good, because QC results in very low hand exposures, so its substitution by QC + NH would confer no advantage in terms of radiological protection. Use of QC alone is probably related to less challenging procedures. Realtime visualization is more important when respiratory motion is a concern, but the need for real-time manipulation is not determined by anatomical area alone (Table 4). Other factors are clearly involved, such as size and accessibility of lesion.
In the second series of biopsies, there were only five values of H p (0.07)max higher than 10 mSv, mostly associated with side-handle manipulation. H p (0.07)max values cannot be added directly, because they occur at different locations for each biopsy. H p (0.07) values were added for each measuring location [see Fig. 1(a)], for the five biopsies with H p (0.07)max higher than 10 mSv. The most exposed locations were the tips of the thumb, middle and ring fingers of the left hand where the mean exposure was~18 mSv/procedure (cumulative total~90 mSv). These five biopsies with H p (0.07)max > 10 mSv correspond to about 8% of the biopsies in the second series, but were responsible for nearly 70% of the total fingertip dose in this series. More importantly, this occurrence of H p (0.07)max > 10 mSv remained approximately constant, at 6%-8% of biopsies (mostly associated with SH), until this study ended, 1 year after the introduction of needle holders into routine practice.
Needle holders are associated with loss of tactile feedback and grip, F I G . 5. Maximum H p (0.07) values plotted as a function of the corresponding CTF time, for needle holder (NH) and sidehandle (SH) manipulation; the CTF times given are estimates for the duration of NH and SH manipulation (T_NH and T_SH), and the fit shown pertains to H p (0.07)max as a function of T_SH.
which may be critical during particularly challenging biopsies, even for an experienced IR well acquainted with their use.
This data highlights the importance of constant individual monitoring of hand exposure. The need for side-handle manipulation in~7% of biopsies is easily overlooked in a busy routine, and left out of dose estimates based on approximate hand distance to the beam. Our results show that these rare occurrences contribute disproportionally to overall hand exposure, and therefore there is always a potential for dose escalation, either through inexperience or overconfidence. Moreover, individual attitudes toward risk vary between individuals, and concerns over patient safety also play a role. But methods of extremity monitoring during CTF guided biopsies need to be improved, because fingertip dose is not correctly assessed by ring dosimeters. 21 The methodology used in this study, comparing hand exposure for two series of biopsies, is different from previous reports of staff dose studies, and allows a more detailed insight into the dose reduction achievable with needle holders in a clinical scenario. In this study, some biopsies could be performed with the quick-check method alone, and for these there was no advantage in using needle holders. When real-time manipulation was necessary, needle holders proved extremely effective at reducing overall hand exposure, by greatly reducing the number of procedures where SH was used. This reduced the mean hand doses per procedure to less than half the values observed for the first series of biopsies (Fig. 2).
Unfortunately, the methodology used in this study requires an ongoing optimization process, so opportunities for such comparisons are rare. Nevertheless, it would be interesting to have similar data for other types of needle holders, different IRs, different biopsy types, and CT scanners with angular beam modulation (ABM). Hopefully, the data presented here will lead to greater awareness of the potential for escalation of hand exposure, and prompt further studies during ongoing optimization processes.

| CONCLUSIONS
Hand exposure was measured for two series of biopsies with comparable characteristics. This allows a detailed insight into the effect of protection measures during real procedures, without the limitations of phantom measurements.
Needle holders proved extremely effective at reducing hand irradiation during CTF guided biopsies when real-time manipulation is necessary. Use of the quick-check method alone leads to even lower exposures-therefore, the quick-check method should be preferred if real-time manipulation is not essential. In this study, the introduction of needle holders did not alter the percentage of procedures performed with the quick-check method alone, nor did it completely prevent side-handle manipulation of the needle. However, availability of needle holders greatly reduced the number of procedures where side-handle manipulation was used, and this lowered the mean hand exposure considerably (Fig. 2).
When needle holders are available, occasional high hand exposure related to side-handle manipulation during challenging biopsies is a rare occurrence (~8% of biopsies in the second series), but contributes disproportionately to hand exposure (nearly 70% of total hand dose in this study). This highlights the importance of constant individual hand monitoring, to avoid dose escalation through inexperience or overconfidence.