Introduction

Percutaneous needle core biopsy has been shown to be a safe and effective method for diagnosing bone and soft tissue lesions [1, 2, 3]. Many bone biopsy needles are available on the market today. The choice of which needle to use is based on many factors including availability, familiarity, cost, ease of use and expected sample quality. During clinical use, we have noted varied quality of specimens obtained with different bone biopsy needles. In general, specimen quality is judged based on achieved versus expected size, specimen integrity and degree of crush or distortion in the sampled tissue. Inadequate tissue sample volume or excess fragmentation of samples, resulting from any of these factors, can lead to repeat percutaneous biopsy or open biopsy. In the literature, soft tissue biopsy needles have been shown to vary in the quality of sample provided [4]. Our hypothesis is that bone biopsy needles also vary in ease of use, ease of sample retrieval, sample size and quality of biopsy tissue they provide. We sought to evaluate a range of commonly used bone biopsy needles on a standard sample to assess whether there are significant differences.

Materials and methods

Eight styles and sizes of bone biopsy needles were evaluated (Fig. 1). These were the Cook Elson/Ackerman 14 g trocar tip [2.03 mm diameter bore] (Cook, Bloomington, Ind.), RADI Bonopty 15 g diamond tip [1.7 mm bore] (RADI Medical Systems, Uppsala, Sweden), Bard Ostycut 15 g diamond tip [1.8 mm bore] (Bard Biopsy Systems, Tempe, Ariz.), Bard Ostycut 16 g diamond tip [1.6 mm bore] (Bard Biopsy Systems, Tempe, Ariz.), MD Tech 11 g with Trap-Lok bevel tip [3.4 mm bore] (Medical Device Technologies, Gainesville, Fla.), MD Tech 11 g without Trap-Lok bevel tip [3.4 mm bore] (Medical Device Technologies, Gainesville, Fla.), Kendall Monoject J-type 11 g bevel tip [3.4 mm bore] (Tyco Healthcare Group, Mansfield, Mass.) and Parallax Core-Assure 11 g diamond tip [3.4 mm bore] (Parallax Medical, Scotts Valley, Calif.). None of the authors has a financial interest in any of the manufacturing companies.

Fig. 1
figure 1

The eight bone biopsy needles evaluated: Cook Elson/Ackerman 14 g (A), RADI Bonopty 15 g (B), Bard Ostycut 15 g (C), Bard Ostycut 16 g (D), MD Tech 11 g with and without the yellow Trap-Lok (E, F), Parallax Core-Assure 11 g (G) and Kendall Monoject J-type 11 g (H)

Full size image

Each single needle was used to obtain five consecutive 1 cm depth samples from a normal cadaveric lamb lumbar vertebral body pedicle (Fig. 2). A single fellowship-trained musculoskeletal radiologist performed all biopsies using each manufacturer’s recommended technique. For coaxial systems, which included a penetration and biopsy set, both sets were used to obtain samples, as would be done in a clinical setting. Imaging guidance was not utilized, as there was no overlying soft tissue to obscure the location of the pedicle.

Fig. 2
figure 2

Each biopsy needle was used to obtain five consecutive samples from a lamb vertebral body pedicle

Full size image

Subjective assessment of ease of needle use and of ease of sample removal from the needle was graded on a 5-point scale, ranging from easy to difficult. The subjective ease of needle use encompassed variables including the physical hand comfort when applying stress to the needle, the overall stability of the needle in gaining purchase to the bone surface and the fatigue produced from the force needed to drill the needle into bone. The ease of sample removal from the needle was based on the force required to dislodge the sample and any difficulty in aligning the removal stylet with the biopsy needle in the manner directed by the manufacturer.

The number of times the needle was inserted into the bone and then removed to unplug any impacted bone in the needle tip (“number of passes”) in order to reach a 1 cm biopsy tract depth from the bone surface was recorded. If the needle tip was plugged before it could be advanced to the 1 cm level, the needle was removed from the tract, unplugged by forcibly passing the stylet through the needle shaft and then placed back into the same partially drilled needle tract to finish drilling to the 1 cm depth. If more than one pass was necessary, then all of the tissue obtained from the different passes was combined for measurement. The operator immediately measured each fresh biopsy sample with vernier calipers (Fig. 3). The same operator graded the degree of gross sample fragmentation on a 5-point scale: cohesive core (1), friable core (2), large tissue fragments (3), small tissue fragments (4), no sample obtained (5). Each sample was then immediately deposited into a pathology cassette, surrounded by thin sponges so that no tissue could be lost, and placed in fixation solution.

Fig. 3
figure 3

Each sample was measured with calipers in the gross state

Full size image

After standard fixation in neutral buffered formalin (4% formaldehyde solution) and 30 min of decalcification (Surgipath, Decalcifier II solution [hydrochloric acid, EDTA, and water], Richmond, Ill.), the samples were evaluated microscopically with an H&E stain, using a 5-point scale ranging from excellent to unacceptable quality. Quality was judged on degree of fragmentation, crush artifact and trabecular distortion. The width and length of each biopsy sample were measured with a microscope micrometer. A single pathologist experienced in evaluation of bone biopsy samples, who was masked to the needle type, analyzed all post-fixation samples.

Each set of cores was evaluated before (gross assessment) and after fixation (microscopic assessment) by comparing length, width and specimen quality. The differences were assessed for statistical significance using an analysis of variance (ANOVA) test among groups.

Results

The RADI Bonopty 15 g and Kendall Monoject J-type 11 g needles were rated the easiest to use. The Parallax Core-Assure 11 g and the Bard Ostycut 16 g were rated the most difficult to use. The RADI Bonopty, Bard Ostycut 15 g, MD Tech without Trap-Lok and Monoject needles had the easiest sample removal. The Cook Elson/Ackerman and the Parallax Core-Assure had the most difficult sample removal. The Cook Elson/Ackerman was the only needle that jammed with impacted bone at the needle tip, requiring it to be removed from the bone and unplugged before it could be drilled to a 1 cm depth (Table 1).

Table 1 Results for eight bone biopsy needles in acquisition of five standard samples drilled to a 1 cm depth from a lamb vertebra: No. of passes required, ease of sample acquisition and ease of sample removal from needle (rated 1 = very easy to 5 = very difficult) (g gauge)
Full size table

There was a significant difference in pre-fixation sample length between needles (P<0.0001), despite acquiring all cores to a standard 1 cm depth (Table 2), which was also present in the post-fixation sample lengths (P<0.0001) without a change in trend among the needle types. Average sample length was greatest using the Cook Elson/Ackerman (8.2 mm, combined tissue from two passes), MD Tech with Trap-Lok (6.9 mm, single pass) and Kendall Monoject (7.1 mm, single pass) needles. Shortest cores were acquired using the Bard Ostycut 16 g needle (4.0 mm, single pass). Sample width is expected to be dependent on needle gauge. As expected, the smaller-gauge needles yielded thinner cores (Table 3).

Table 2 Results for eight bone biopsy needles in acquisition of five standard samples drilled to a 1 cm depth from a lamb vertebra: core length before and after fixation (g gauge)
Full size table
Table 3 Results for eight bone biopsy needles in acquisition of five standard samples drilled to a 1 cm depth from a lamb vertebra: core width before and after fixation (g gauge)
Full size table

The sample length and width consistently dropped from the gross to the fixed state. The gross sample length ranged from 1.6 to 9.9 mm with an average for all needles of 5.8 mm (ideal = 10 mm). The sample length in the fixed state ranged from 0 to 7.7 mm, average 4.2 mm, representing an average decrease of 28% after fixation (Table 2). This was not significantly different among the groups although the Cook Elson/Ackerman 14 g had the smallest decrease (13%) and the Bard Ostycut 16 g had the largest decrease (60%). The remaining needles were in the 20–40% decrease range. The gross sample width ranged from 1.6 to 2.4 mm, averaging 1.9 mm. The sample width in the fixed state ranged from 0 to 2.0 mm, averaging 1.1 mm, for an average decrease of 42% after fixation (Table 3). There were significant differences among the group. The Bard Ostycut 16 g showed the largest decrease (about 80%). The RADI Bonopty 15 g, Bard Ostycut 15 g, Kendall Monoject J-type 11 g and Parallax Core-Assure 11 g showed an intermediate decrease (40–50%). The Cook Elson/Ackerman 14 g, MD Tech with Trap-Lok 11 g and MD Tech without Trap-Lok 11 g showed the least decrease (about 25%).

Parallax Core-Assure 11 g and Kendall Monoject J-type 11 g had the highest quality (most cohesive) specimen in the gross state; the Cook Elson/Ackerman 14 g and Bard Ostycut 16 g yielded the lowest quality (most fragmented). The MD Tech without Trap-Lok 11 g had the highest quality core after fixation, while the Bard Ostycut 16 g had the lowest quality (Table 4).

Table 4 Results for eight bone biopsy needles in acquisition of five standard samples drilled to a 1 cm depth from a lamb vertebra: quality of specimen (g gauge)
Full size table

The needles ranged in cost from $19.37 to $98 each. Overall results, including relative cost, are summarized in Table 5.

Table 5 Summary of needle performance (g gauge)
Full size table

Discussion

Many factors play a role in bone biopsy needle selection in clinical practice. Coaxial systems are extremely useful, especially for sampling lesions that are deep or require precise localization. Smaller-gauge needles may be required for sampling lesions near vital structures. The eccentric drill system is useful for accessing lesions several centimeters deep to the cortex or for sampling sclerotic lesions. However, radiologists are occasionally left to choose needles based only on needle familiarity, availability or perception of effectiveness. Detailed knowledge of the strengths and weaknesses of different needles would help users make an appropriate selection, but to our knowledge there no standardized comparison of needles is currently available for bone biopsy.

Lamb vertebral bodies were selected for this experiment due to their ease of procurement and their prior use as a model for the human spine in the literature [5]. The vertebral bodies had been previously frozen, but were biopsied at room temperature. Freezing of bone has been shown to have no effect on the strength of cortical and trabecular bone [6, 7]. Use of a standard biopsy subject helped remove confounding factors from analysis of needle effectiveness. A similar study in available cadaveric human bones would be subject to variation related to wider differences in trabecular density, cortical thickness, location of sampling and underlying disease processes. The lamb vertebral body model simulates a bone biopsy in which normal cortical bone is traversed to sample tissue in the medullary space, in this case normal bone marrow. The lamb vertebral body lacked the overlying soft tissue, which differs from the clinical setting, but this was felt to be of little consequence as bone biopsy needles easily traverse fat and muscle.

We found some needles easier to advance into bone than others, and ease of core removal also varied. This seemed to be multifactorial, related to needle gauge, strength, sharpness, needle tip configuration and handle ergonomics. The Bard Ostycut 16 g needle repeatedly flexed during the drilling process. This flexion makes it difficult to transmit a downward force to the drill tip, making the drilling more difficult. The ergonomic handle of the Kendall Monoject J-type, MD Tech and Parallax Core-Assure facilitated generation of torque over smaller handle configurations such as the Cook Elson/Ackerman and Bard Ostycut. Conversely, the small handle of the RADI Bonopty was more than compensated for by the eccentric drill tip on the trocar, which greatly facilitated cortical penetration. Needles with sharp bevel tips (e.g., Kendall Monoject J-type) also facilitated cortical purchase compared with the serrated edge of the Cook Elson/Ackerman, which tended to slip along the angled bone surface. An overall best needle could not be chosen, as the different qualities assessed (ease of sample acquisition, ease of sample removal, specimen quality and cost) all would vary in importance to different individuals.

The needle price varied widely, with the coaxial systems being the highest priced. Our cost for individual needles cannot be published due to the confidentiality of negotiated pricing. Needle cost may also vary between institutions based on the number of needles purchased. Thus, only the relative needle cost is shown in Table 5. Another variable with needle cost is the potential for reuse. We know from personal communication that some institutions sterilize and reuse the Cook Elson/Ackerman needles on multiple patients. These needles are specifically labeled as being for single use only. We do not participate in or support this off-label multiple use.

This study is limited by several factors. First, a single observer obtained the tissue samples. A single observer was chosen to achieve consistency in the biopsy technique. However, this makes the “ease of use” assessment extremely subjective and this information may be of little help to people choosing biopsy needles. The single observer was familiar with all the needles but had the most practical experience using the Cook Elson/Ackerman needle. Using multiple observers, with differing strength, dexterity, experience and handedness, may have given different results. Second, biopsy samples were obtained in a single biopsy scenario, that being a biopsy of the normal medullary bone contents after healthy bone cortex was traversed. In clinical practice, the bone being biopsied would often be abnormal, unless normal bone had to be traversed to obtain access to a lesion. It is not known how differently the needles would perform if compared against each other using a standardized sample of abnormal bone. This study also does not address the performance of needles in situations where the bone is completely destroyed, partially destroyed or densely sclerotic. Also, some of the needles tested are available in different lengths and gauges. We did not test every needle configuration option or even every bone biopsy needle that is currently available for purchase. We tested only the needles that we had readily available in our institutions. Finally, the needle performance was based on the sole use of that needle, as packaged. In clinical situations, multiple needles or portions of needle sets can be used together, such as using the RADI Bonopty penetration set to drill through bone and act as a sheath, with the biopsies then being taken using a Bard Ostycut needle.

Core width varied relative to needle gauge, as was expected. Less expected was the variation in core length, since all needles were advanced to a standard depth of 1 cm below the bone surface. In the case of the Cook Elson/Ackerman needle, the cortical bone became impacted in the needle, preventing further advancement. After the tip was unplugged, the needle could be advanced. This needle produced the longest specimens when the two specimens, the plug of cortical bone and the subsequent sample, were combined. It was with this needle only that the soft tissue sample was not crushed against the traversed cortical bone. This “crushing” effect may explain variation in length of specimens acquired using the other needles as well; the shorter specimens presumably were caused by more compaction within the needle and thus may have an effect on decalcification and fixation. Specimen compression damages the three-dimensional lattice of the tissue, which could allow more collagen shrinkage between trabecular fragments, thus making a shorter (and narrower) core post processing.

As might be expected based on the general tendency for tissue to shrink during processing, we also found that there was a consistent decrease in size of the sample during the fixation process, averaging a 28% decrease in length and a 42% decrease in width relative to measurement in the gross state. Some of this may be a function of the accuracy of gross versus micrometer measurement. One needle in particular, the Bard Ostycut 16 g, yielded only disorganized tiny fragments after fixation for three of five core samples, making exact measurement impossible. Since fixation reduces sample size, providing the pathologist with samples of adequate size is paramount.

Conclusion

Based on data we have provided on needle performance in a controlled setting, we conclude that bone biopsy needles vary significantly in the size and quality of specimens they generate. We also found that fixation reduces sample size and may affect diagnostic accuracy. Detailed knowledge of the strengths and weaknesses of different bone biopsy needles is important for making an appropriate selection for each individual’s practice.