Effect of a Combined Aspiration and Core Biopsy Technique on Quality of Core Bone Marrow Specimens

Introduction

The indications for bone marrow evaluation in small animal patients are well described.^1–9 For each patient requiring bone marrow evaluation, a clinical decision must be made whether to obtain an aspirate, a core biopsy, or both. Bone marrow aspiration is the procedure most commonly performed, providing cytology smears that can be examined shortly after collection of the marrow. Those smears are ideal for the enumeration of the myeloid/erythroid ratio (M/E ratio) and evaluation of individual cell morphology. Bone marrow core biopsy is performed to obtain marrow for histologic examination. Bone marrow cores are preferred for the evaluation of marrow cellularity and trabecular bone architecture and are indicated when aspirations are either unsuccessful or nondiagnostic (e.g., hypocellularity, “dry taps”) or when there is a suspicion of focal lesions (e.g., myelofibrosis, granulomas, metastatic disease). Core biopsies are technically more difficult to collect and require longer processing times than aspirations, which may cause a delay in obtaining results.

Collection of both a bone marrow aspirate and a core biopsy is desirable for the evaluation of some patients either because of the differential diagnoses under consideration or to provide insurance against the possibility of a nondiagnostic aspirate. In that group of patients, the next clinical decision to be made is whether to collect each specimen directly from a distinctly different site or whether to collect both the aspirate and the core biopsy from the same site, using the same needle (i.e., a combined technique).^8,10–14 The combined technique has the important advantages of decreasing procedural time, material costs, and patient discomfort compared with collection from different sites; however, those advantages must be weighed against a potential loss in quality of the core biopsy.

Studies comparing the direct and combined techniques in humans have provided mixed results. Wolff et al. (1983) concluded that preaspiration did not significantly interfere with interpretation of cellularity in core biopsies from 14 cancer patients, and mean lengths of preaspirated (the combined technique) and nonaspirated (direct) core biopsies were not significantly different.¹² Islam (2007), in contrast, concluded from an observational study that core biopsy following aspiration was technically cumbersome, resulted in a large component of clotted blood in the core specimen, induced more pronounced tissue damage, and depleted intertrabecular spaces of marrow cells.¹¹ Douglas and Risdall (1984) found areas of aspiration-induced intramedullary hemorrhage and early fibrin deposition in all 38 preaspirated (combined) core biopsies they studied.¹³ In three of those cases, the hemorrhagic artifact precluded accurate evaluation of cellularity and hematopoietic elements. To the authors’ knowledge, there are no published studies regarding the combined technique in veterinary medicine.

This study was undertaken to investigate the loss in quality of a bone marrow core biopsy when collected immediately following bone marrow aspiration, using the same anatomic site and needle, to provide objective information for clinicians who are considering a combined technique. Euthanized dogs were used for this study to allow each dog to serve as its own control and to include a larger population than would have been feasible using clinical patients during a similar study period. A preliminary study was performed to ensure that bone marrow core biopsy quality was maintained long enough postmortem for specimen collection to occur. The hypothesis was that bone marrow cores obtained by the combined technique would be histologically inferior to core samples obtained directly.

Materials and Methods

Between Jul 14, 2008 and Aug 13, 2009, bone marrow biopsies were performed on 31 adult dog cadavers at an animal control facility. Five dogs were included in the preliminary study conducted to confirm the postmortem stability of the bone marrow for histologic evaluation. The remaining 26 dogs were included to compare core samples obtained by the combined and direct techniques. The dogs were scheduled for humane euthanasia according to the animal control facility’s policy, and no dogs were euthanized for the purpose of this study. Euthanasia was performed by IV injection of sodium pentobarbital^a. A small number of dogs were administered intramuscular injections of tiletamine hydrochloride and zolazepam hydrochloride^b for sedation prior to the administration of the sodium pentobarbital. Age and breed of the dogs were estimated, gender was ascertained by genital examination, and neuter status was recorded for male dogs.

To verify the postmortem stability of bone marrow core biopsies, two direct biopsies were performed on five dogs. One core was obtained 2–5 min following euthanasia. The second was obtained from the opposite humerus 20 min following euthanasia. The order of humeri sampled (left or right) was alternated. The cutting end of the biopsy instruments had been premarked with hatch marks at 1 cm increments, and the depth of penetration into the first humerus was duplicated as closely as possible during biopsy of the opposite humerus. Two clinical pathologists (C.G. and M.C.) independently evaluated each pair of cores and subjectively assessed relative overall quality. Results from those dogs were not included in the comparisons between the direct and combined techniques.

To determine whether the combined technique resulted in core biopsies that were histologically inferior to those obtained directly, paired cores were collected from alternate humeri within 15 min of euthanasia from the remaining 26 dogs. One core was obtained directly and the other by the combined technique. The order in which the direct and combined techniques were performed and the order of the humerus used first (i.e., left or right) were alternated. Depth of penetration into the first humerus was duplicated as closely as possible during biopsy of the opposite humerus. All core samples were obtained using either a direct or combined technique by a single investigator. That investigator (J.R.) did not subsequently assess the samples.

Direct core biopsies were performed using 11-gauge Jamshidi bone marrow biopsy needles^c that had been primed with 3% calcium disodium ethylenediaminetetraacetic acid (EDTA). The hair was clipped over the proximal humerus, and a stab incision was made through the skin using a No. 11 scalpel blade. The cutting end of the Jamshidi instrument was placed on the intertubercular groove, and pressure was applied with 30°, back-and-forth axial rotation to seat the tip firmly in the cortical bone. The stylet was removed, and 90° axial rotation was used to advance the Jamshidi instrument into the medullary cavity, maintaining the shaft as parallel as possible to the long axis of the humerus. When maximal penetration was reached, the instrument was briskly rotated in repetitive clockwise and counterclockwise directions. A 6 cc syringe was attached to the Jamshidi aspiration port, and gentle traction was applied to the plunger to create a slight vacuum. The Jamshidi was then removed from the bone. The bone marrow core biopsy was extruded from the aspiration port with a shepherd’s hook (i.e., a blunt stylet) onto filter paper and placed in 10% formalin.

Core samples obtained by the combined technique were performed using identical procedures as described for direct core biopsies (above), except that after seating the Jamshidi tip into cortical bone and removing the stylet, 0.25–0.5 mL of marrow was aspirated into an EDTA-primed syringe (Figure 1). Aspirated marrow was expressed into a watch glass containing EDTA and was later evaluated cytologically for the presence of marrow elements. Failure to identify marrow elements in the aspirated marrow resulted in exclusion of that dog from the study.

View Figure

• Download as Powerpoint
• Download Figure

All core biopsies were decalcified and embedded in paraffin. Slides were made from a single slice of each core that was cut along the long axis and stained routinely with hematoxylin and eosin. Slides were randomized, and the identification key was concealed from the two clinical pathologists (C.G. and M.P.) who independently evaluated all cores. Each slide was evaluated twice. The first analysis included all slides presented in a random order and the second analysis involved assessing the two slides from each dog paired. Randomly ordered slides were measured, assessed, and scored. The scoring system (Table 1) allowed for quantitation of specific features for each specimen based on the clinical experience of the each pathologist. The terms “diagnostic,” “length of diagnostic marrow,” and “diagnostic quality” indicated the presence of sufficient marrow elements for accurate histologic assessment. Length of marrow (mm) was measured by ocular micrometer. The presence of artifacts (such as hemorrhage, cell drop-out, crush artifact, and cortical bone) was also noted. Slides were then organized in pairs by a third party and re-presented to each clinical pathologist for the comparative assessment of relative overall quality.

Table 1 Numerical Scoring of Histologic Properties of Bone Marrow Core Samples

View Fullscreen

Results

Bone marrow core biopsies were performed on 31 dogs (16 males). The mean estimated age was 3.25 yr (range, 0.6–10 yr), and the mean estimated body weight was 20.2 kg (range, 5–40 kg). Breeds represented were Pit bull terriers (n = 11), mixed-breed dogs (n = 11), Australian shepherds (n = 2), and one each of the following: Doberman pinscher, rottweiler, German shepherd dog, dachshund, border collie, Labrador retriever, and beagle.

Core samples obtained 20 min after euthanasia to verify postmortem stability were of similar overall histologic quality as those obtained within 5 min of euthanasia, and no postmortem artifact was noted in any specimen. Both clinical pathologists’ assessments of relative overall quality were identical for all five pairs. Core samples collected in 20 min were of lesser quality in two pairs, of greater quality in two pairs, and of equal quality in one pair of samples. Recorded reasons for assessing one core as relatively inferior were lower cellularity, shorter length of core, and greater crush artifact. No autolysis or aberration of cellular morphology was observed in any specimen.

For evaluation of the effect of the combined technique on core quality, core samples were successfully obtained by both the direct and combined techniques in 22 of 26 dogs (85%). In four dogs (15%), either one or both biopsy techniques failed to yield a core biopsy sample. In all four of those dogs, the combined technique was unsuccessful. In one of those four dogs, the direct method was also unsuccessful. Only the 44 cores from the 22 dogs with paired specimens were analyzed further. Cytology of all aspirates obtained with the combined technique confirmed aspiration of marrow.

Based on histologic evaluation of individual, randomized slides, the only significant differences found between core samples obtained by the combined technique and those obtained directly were length of diagnostic marrow and presence of the artifact of hemorrhage (Table 2). The length of marrow measured by both clinical pathologists was shorter in core samples obtained by the combined technique compared with cores obtained directly (C.G., P = 0.011; M.C., P = 0.028). The mean difference in length of core sample between methods measured by C.G. was 3.89 mm (core samples obtained by combined technique were 28% shorter); the mean difference measured by M.C. was 2.36 mm (combined technique cores were 18% shorter). Hemorrhage was noted as an artifact more often in core samples obtained by the combined technique compared with core samples obtained directly (C.G., P < 0.001; M.C., P < 0.001). No other histologic property or specific artifact was found to be significantly different between collection methods as assessed by either clinical pathologist, including megakaryocyte numbers, cellularity, M/E ratio, iron stores, or diagnostic quality.

Table 2 Comparison of Bone Marrow Core Samples Collected by the Direct and Combined Techniques Assessed Independently by Two Clinical Pathologists

*

Mean ± standard deviation

†

Median (25th percentile, 75th percentile)

M/E ratio, myeloid/erythroid ratio.

Pathologist 1 (C.G.), pathologist 2 (M.C.).

View Fullscreen

Core biopsies obtained by the combined technique were subjectively assessed to be of lesser overall quality than core biopsies obtained directly when compared in pairs, but statistical significance was not achieved (P = 0.286). Assessments by the two clinical pathologists were identical for each pair of cores. In 14 of 22 pairs (64%), the core biopsies obtained by the combined technique were of lesser quality. Only 6 of 22 pairs (27%) of core biopsies obtained directly were of lesser quality. No difference in quality was found in two pairs. The histologic features that contributed to the clinical pathologists’ assessment of lesser quality recorded for each pair included the following: shorter core length, the presence of intramedullary hemorrhage, crush or fragmentation of the specimen, zones of decreased or absent cellularity, poorer quality of staining, alteration of cellular morphology, and increased zones of cortical bone or cartilage.

Discussion

Bone marrow core biopsies collected following the aspiration of marrow through the same biopsy needle at the same anatomic site from recently euthanized dogs were altered; however, the overall quality was not greatly affected. Core biopsies collected using the combined technique were significantly shorter and more often had hemorrhage noted as an artifact than core biopsies collected directly.

Length of core can affect diagnostic accuracy of marrow specimens in certain cases, with longer specimens increasing the sensitivity for detecting focal disease. Studies in humans have shown that the use of long-core and bilateral trephine biopsy increases the rate of detection of discrete neoplastic lesions. Brunning et al. (1975) reported that a neoplastic lesion was found in only one of two bilateral core biopsies obtained from the posterior iliac spine in 11 of 50 patients (22%) with non-Hodgkin’s lymphoma, 3 of 7 patients (43%) with Hodgkin’s lymphoma, and 5 of 14 patients (36%) with metastatic marrow tumors.¹⁶ Islam and Henderson (1988) demonstrated the value of long-core (18–20 mm) over short-core (8–10 mm) marrow biopsies in certain individuals when 5 of 256 (2%) specimens obtained from leukemia patients had a single, focal malignant lesion at the innermost extent of long-core biopsies.¹⁷ Although the decrease in length of cores obtained by the combined technique in the current study was slight (18–28%), it is possible that this difference could affect diagnostic accuracy in patients with focal marrow disease.

The decrease in diagnostic core biopsy length in samples obtained by the combined technique could have been influenced by the intramedullary hemorrhage noted more often in those core biopsies. The measurement of diagnostic core length excluded regions where marrow elements were absent due to artifacts such as hemorrhage and cell drop out. This explanation is consistent with findings of previous studies in which length of interpretable marrow in preaspirated specimens was decreased by areas of histologic distortion.^11–13 The depth of the biopsy needle within the bone at the beginning of trephination for the core and the subsequent aspiration of marrow may also have been contributing factors to shorter core sample length by decreasing the marrow collected from the proximal (outer) region of the bone.

The combined technique did not result in a significant difference in the scored properties of megakaryocyte numbers, cellularity, M/E ratio, iron stores, diagnostic quality, and overall quality of the core biopsy based on paired comparisons. The combined technique failed to yield a core biopsy more often than the direct technique, but this difference did not reach statistical significance. The failure to find a significant difference for some of the above-mentioned criteria could be the result of an insufficient number of dogs included in this study. It is also impossible to predict whether differences would be found in populations of dogs with specific bone marrow diseases. Further, testing a population of dogs without known bone marrow disease precludes any ability to draw conclusions about diagnostic sensitivity.

Another limitation of this study was the use of euthanized dogs. Hemorrhage artifact may have been less in core biopsies obtained from either technique in the absence of circulating blood. Dogs were able to serve as their own controls, minimizing concerns about unknown disease status and lack of concurrent complete blood cell counts. Further, no bone marrow pathology was noted in any specimen. Preservation of the histologic integrity of core biopsies for up to 20 min postmortem was demonstrated in the initial phase of this study, and a previous study also demonstrated the postmortem preservation of marrow.¹⁸ The order of collection technique used in each dog was alternated to minimize any potential effect of elapsed time. By using euthanized dogs, the study was strengthened by the ability to have each dog serve as its own control and to have a larger population than would have been possible using clinical cases during the same time period. Concerns about unnecessary procedures and discomfort for living dogs were avoided. All dogs were scheduled for euthanasia without consideration for this study.

Variability may have been introduced into the study either by the clinician collecting the bone marrow or the clinical pathologists evaluating the slides. To minimize variability in collection technique, a single clinician obtained all specimens and a strict, written protocol was followed. The procedure used was alternated between left and right limbs to minimize differences introduced by the relationship between humeral anatomy and the dominant hand of the operator.

There was not complete agreement between clinical pathologists with respect to individual measurements and scores. The purpose of this study was not to determine consistent criteria or evaluate scoring systems for the assessment of bone marrow core biopsies. Instead, this study was developed to address the important question of whether the quality of core biopsies was affected by using a combined technique. The study design provided two separate methods for identifying differences in quality: the measuring and scoring of randomized slides and direct comparison of paired slides from the same dog. Slides and pairs were randomized to avoid any bias. The findings that both clinical pathologists found core biopsies collected by different methods to be significantly different for the same two criteria (i.e., length and hemorrhage), and in all cases agreed in the assessment of relative quality of paired slides, provide strength to the results of this study.

Finally, different results may have been obtained by using an alternative technique or an anatomically different site of collection. Alternatives to the combined technique used in this study include changing the angle of penetration of the biopsy needle following aspiration, replacing the stylet and penetrating deep to the aspirated zone before biopsy collection, and collecting the aspirate after deep penetration into the bone. A smaller volume of marrow could also be aspirated. In this study, marrow was aspirated in sufficient volume for suspension in EDTA prior to particle isolation and smear.^1,2,5,7 In the authors’ practice, such larger volume specimens of 0.25–0.5mL are routinely collected because of subjectively better slide quality and the frequent need for marrow for either flow cytometry or polymerase chain reaction testing for infectious organisms or round cell neoplasia.

Conclusion

This study confirmed that bone marrow core biopsies collected following the aspiration of marrow through the same biopsy needle at the same site are altered by the procedure. Direct core biopsy remains the ideal technique; however, the compromise to core biopsies collected by the combined technique in this population of dogs without marrow pathology was small. In clinical patients, the advantages of the combined technique include shorter procedure time, decreased costs, and less patient discomfort. Those advantages should continue to be weighed on a case-by-case basis against a potential loss of diagnostic sensitivity.