Introduction. An ideal biopsy needle consistently recovers representative specimens of adequate length with minimal dependence on patient characteristics. We have shown that bone marrow biopsy needle gauge influences the length and quantity of recovered specimens. A needle that performs independently of patient characteristics should yield specimens demonstrating a normal distribution of specimen lengths. We evaluated the influence of age and needle gauge on the specimen length distribution curves of patients undergoing bone marrow biopsies using SNARECOIL needles.

Methods. 88 bone marrow core specimens were recovered from 72 patients using 11G SNARECOIL bone marrow biopsy needles. The mean age of the patients was 61.4 and the m/f ratio was 45/27. 53 specimens were recovered from 39 patients (mean age = 47.1, m/f = 27/12) ≤ 64 and 35 specimens were recovered from 33 patients (mean age = 78.6, m/f = 18/15) ≥ 65. 106 patients underwent 127 bone marrow biopsy procedures using 8G SNARECOIL needles. The m/f ratio was 56/50 and the mean age of the patients was 63.1 years. 72 specimens were recovered from 56 patients (mean age = 51.4, m/f = 30/26) ≤ 64 years old and 55 specimens were recovered from 49 patients (mean age = 76.1, m/f = 26/23) ≥ 65. 40 additional specimens were prospectively recovered from 39 patients ≥ 65 (mean age = 74.3, m/f = 19/20) with 8G SNARECOIL needles.

Results. The mean specimen lengths of the 11G and 8G specimens were statistically the same (mean±SEM, 1.97±0.07 cm vs. 1.99±0.05 cm, respectively, p = 0.8). However the 11G specimen length frequency distribution curve deviated markedly from a normal distribution (skewness(skw) = 0.52) while the 8G specimen distribution was nearly normal (skw = 0.04). While the frequency distribution curves of older patients (≥ 65) biopsied with 8G needles showed minimal deviation from normality (skw = 0.12,) the distribution curve of older patients (≥65) biopsied with 11G needles deviated markedly from a normal curve (skw = 0.64). 40 specimens prospectively obtained from patients ≥ 65 demonstrated a mean specimen length of 1.77±0.09 cm with a near normal specimen length frequency distribution curve (skw = 0.07).

Conclusions. 1. Although commonly reported, mean specimen lengths do not completely characterize the performance of biopsy needles. 2.0 Specimen length frequency distribution curves provide added characterization of needle performance in defined patient cohorts. 3.0 In the older patient population, near-normal 8G specimen frequency distributions and skewed 11G distributions suggest that 8G specimens may provide more representative sampling.

Author notes

Corresponding author