Needle Biopsy Adequacy in the Era of Precision Medicine and Value-Based Health Care

An ideal needle biopsy yields an adequate sample for the intended diagnostic assessment. Although this criterion is highly functional, it is retrospective and provides no technical or quantitative information on the sample characteristics required for adequate needle biopsy. More comprehensively, an ideal adequate biopsy provides a sample containing the most reliable diagnostic information representative of the lesion in the timeliest fashion, with the least burden to the patient, the biopsy operator, and health system costs. It is important to recognize that biopsy adequacy is much more than a binary concept. Conceptually, needle biopsies can be classified as inadequate, somewhat adequate, or fully adequate for the intended diagnostic purpose. The characteristics for an ideal needle biopsy sample are presented in Table 1 and discussed further below.

Needle biopsies are aimed to extract a sample directly from a specific lesion or even a specific tissue domain within a lesion. The rationale is that the lesion is the location where the most representative and most concentrated sample is found. Recently, liquid biopsy, extraction of representative circulating DNA or circulating tumor cells from blood, has been developed.^14–16  Liquid biopsy methods have clinical utility where sampling by needle biopsy is too difficult or where the sample is deliberately intended to represent multiple discrete lesions. However, liquid biopsy offers no localization information, and requires ultrasensitive analytical methods that result in lower sensitivity for most clinical purposes.

In contrast, for solitary or diffuse tissue lesions, needle biopsies provide precise localization and are much more likely to yield representative samples cost-effectively with a similar safety profile.

Adequate material in a sample refers to cell number (cell mass), proportion of cells that are potentially diagnostic (eg, tumor cells), and/or the amount of RNA, DNA, or protein markers present in representative lesional cells. In clinical practice, the range of diagnostic techniques to be considered may be known only when the sample is examined immediately after retrieval. Sample adequacy is then determined by the cytotechnologist-cytopathologist team. Portions of the sample are allocated for particular studies and preserved in appropriate media as determined by the provisional assessment made by the pathologist. Critical to this assessment are the pathologist's knowledge of the clinical condition, the diagnostic questions, and their examination priority in the presence of a small sample, as well as the laboratory professionals' knowledge and capability to initiate appropriate processing for ancillary diagnostic techniques. For some needle biopsy applications, where cytologic phenotype is known previously (eg, for selection of therapy), immunocytochemistry or genomic markers are often considered as the needle biopsy's primary intended use rather than as an ancillary technique.^17–21 

Typically, FNAB yields enough cells for 1 or 2 slides. For presence of cells diagnostic for tumor, this may be sufficient for cellular assessment, even though tissue architecture is not seen. However, samples for immunocytochemistry or genomic assessment usually require additional passes. Preanalytical preparation technique is a critical factor in sample adequacy. Liquid-based cytology preparations, which have advantages for morphologic analysis, are also suitable under appropriate circumstances for immunocytochemistry and genomic ancillary techniques.²² 

Core needle biopsy may yield sufficient material with histologic architecture preserved. For immunohistochemistry, additional sections can be prepared from these CNB samples. However, for genomic analysis it is usual that a portion of sample be removed and immediately fixed in an RNA preservative. For some probes, formalin-fixed, paraffin-embedded samples can be used, but, related to variability of formalin fixation and subsequent processing, a single CNB sample may not be enough. Nassar²  and VanderLaan⁵  have reviewed the comparative benefits and drawbacks of FNAB and CNB. Both technologies can have substantial sampling adequacy limitations. A rigorous approach to defining needle biopsy adequacy requires quantification in terms of cell number²³  or mass yield^24,25  for each type of analysis as well as the proportion of specific lesional cells such as tumor cells. This approach is most useful for biopsy instrument design and for estimating the number of needle passes required to obtain a clinically adequate biopsy. To date, quantification during the clinical biopsy procedure is seldom performed. Rather, adequacy is usually assessed visually and quantitated by manual estimation only.

The oldest and most common application of needle biopsies is the detection and characterization of cancer cells from a tumor. Operator variability and lesion heterogeneity are particular concerns for FNAB. However, as definitive diagnosis can often be made by examination of only a few representative cells, proposed quantitative criteria for cytomorphology have not been definitive. In 1997, a national conference on FNAB of breast lesions²⁶  did make recommendations regarding adequate FNAB cell yield for cancer diagnosis, namely that with a specimen deemed representative and adequate by the pathologist, the clinician deems that the cytologic results are consistent with the clinical findings. Although functional, these remain as retrospective and subjective criteria. Nonetheless, the committee recommended that the amount of cells present be estimated on a visual analog scale as follows: few (occasional clusters), moderate (clusters easy to find), or abundant (epithelial cells in almost every field). In contrast, the Canadian Society of Cytology in 2003 did attempt to record minimum cell requirements for thyroid FNAB from the literature, noting number of cell clusters observed and number of cells in each cluster as visually estimated.²⁷  It should be noted that if few diagnostic cells are present from an FNAB or if the cells are distorted by processing, adequacy is jeopardized. Incremental assessment time may be required by the cytotechnologist and cytopathologist, and less specific diagnoses than ideal, such as suspicious for malignancy, may be rendered.²⁸  A higher cell yield and presence of undistorted cells raises the likelihood of more definitive assessment. Although technically feasible, and capable of providing extensive objective information, cytomorphometry is rarely applied in diagnostic cytopathology and no standards for cytomorphometric quantitation for routine cytopathology have been developed.

Theoretically, FNAB has advantages over CNB. With FNAB immunocytochemistry, entire cells can be observed with minimal postbiopsy processing.^29,30  This advantage applies only if immunocytochemistry is performed directly on the cell aspirates. If a cell block is prepared, then the technology is similar to immunohistochemistry.³¹  The cell aspiration immunocytochemical technique was developed in 1999 for estrogen and progesterone³²  and in 2003 for human epidermal growth factor receptor 2 (HER2) assessment from breast cancer FNAB.³⁰  However, related to variably inadequate FNAB yield, current standard of care recommends CNB or open biopsies for assessment of breast cancer biomarkers such as estrogen receptor, progesterone receptor, and HER2. This remains a problem, as it is desirable to obtain these biomarker assessments in metastatic lesions that are often accessible by FNAB but rarely by CNB. An emerging application with similar considerations is tumor or immune cell analysis for immunologic markers such as programmed death ligand-1 (PD-L1).^33–35 

Fine-needle aspiration biopsy washout fluid can be used for biomarker assessment to assist the diagnosis of particular entities—for example, medullary thyroid carcinoma by calcitonin analysis³⁶  and metastatic thyroid cancer in lymph node by thyroglobulin analysis.^37,38 

Using FNAB for these techniques should be performed only after rigorous validation in individual laboratories, including validation of fixatives different from 10% formalin commonly used for immunohistochemistry.

The assessment of genomic markers, particularly for mutations and gene rearrangements in tumors,^20,21,39  is evolving quickly, with preparations ranging from DNA isolation⁴⁰  and RNA probes^41,42  to next-generation sequencing.^43,44  The amount of sample required varies with the genomic technology. For DNA extraction from CNB histologic preparations, 1 μg of DNA is required. For epithelial tumors, currently, this requires two 1-cm-long CNBs from an 18-gauge needle.⁴⁰  In the practice of a major cancer center, 20% of CNBs were found to yield inadequate DNA.⁴⁵  For RNA probes using quantitative polymerase chain reaction, only about 100 cells are required, but immediately after sample acquisition these cells must be fixed promptly in RNA fixative.^41,46  With FNAB, the entire cell nucleus is present within each cell,⁴⁷  and accordingly, DNA yield from FNAB is superior to DNA extracted from CNB formalin-fixed blocks.⁴⁸  With CNB, RNA preservation may be compromised, as RNA undergoes autolytic degradation faster than RNA fixative can penetrate a tissue core. Also, the high-velocity insertion forces from CNB guns may compromise RNA integrity.⁴⁹  For next-generation sequencing, either cell blocks or direct smears from FNAB can be used, depending on cellularity. Cellularity is classified on smears as follows: high, at least 1 or 2 slides with more than 5000 cells; moderate, 1000 to 5000 cells; and low, fewer than 1000 cells.⁴³  Even with this relatively high cellular requirement, next-generation sequencing is successful only 50% of the time.^43,50  Further, even high cellularity as defined above may not be adequate with tumors containing intratumor mutational heterogeneity; it is likely that higher tumor cell yields would facilitate better diagnostic results.

Needle biopsy to retrieve marrow stem cells is a long-established application.⁵¹  Needle biopsies are increasingly used to retrieve tumor cells for xenograft diagnostic studies.^52–56  Ongoing adequacy issues include retrieval of enough cells with preserved viability. Fine-needle aspiration biopsy can be used to obtain cells for flow cytometry assessment, for example in lymphoma classification.^57,58  Here, lesion representativeness and sufficient cell number are prominent adequacy considerations. Single cell analysis is an emerging analytical technique to characterize properties of rare cells, to analyze cell expression heterogeneity,^59,60  and even to characterize properties of a single molecule class within a cell.⁶¹  For single-cell analysis, accessing and extracting single cells with minimal perturbation of native expression profiles remains challenging.⁶² 

A representative sample is retrieved from a tissue site, usually from a lesion detected by one or more imaging modalities. Imaging provides contrast to determine size, shape, and texture of target lesion and to confirm that the biopsy needle is indeed inside the lesion. As handheld high-resolution ultrasound imaging can now achieve a resolution of ≈1 mm,⁶³  colocalization of the needle biopsy can be problematic for both FNAB and CNB. For FNAB, typically the pass within the lesion involves not a precise single site, but, using the recommended technique, different internal passes within a lesion volume.⁶⁴  For CNB, the issue is to achieve precise localization within the limitations of the biopsy gun geometry. In addition, large core or vacuum assisted needle biopsies have been known to inadvertently remove most of the tumor, making neoadjuvant therapy or subsequent surgical tumor margin assessment impossible.⁶⁵  The second issue for obtaining a representative sample is to obtain sufficient and representative cell yield from the lesion site. Fine-needle aspiration biopsy is intrinsically limited by the manual skills of the operator and the cellularity of the lesion. Lesions with collagenous connective tissue or microcalcifications may yield limited cells even with the most skilled operators. Irrespective of needle lumen diameter, CNBs have difficulties as well. With CNB, epithelial lesions with abundant connective tissue tend to fragment. The fragments retrieved may or may not be fully representative of the lesion. The limitations of CNB are illustrated by the diagnostic upgrade rates of 18% between breast CNBs yielding atypical hyperplasia and more-aggressive lesions diagnosed at surgical excision.⁶⁶  Conversely, Ki67 cell proliferation index was higher in CNBs than in surgical excisions.⁶⁷  Similarly, CNBs for soft tissue tumors subsequently resulted in upgraded diagnosis in 33% of cases.⁶⁸  What constitutes a representative sample is dependent on the intended clinical use. If the purpose is to detect malignant cells in a mass, or cytopathic effect of virus affecting all cells of an organ, only a very few cells are needed. By contrast, if the purpose is dependent on an extensive sample of tissue architecture (eg, chronic hepatitis grade), then a CNB core at least 20 mm long containing 11 or more portal tracts is recommended.⁶⁹ 

Similarly, if a lesion potentially includes several different cell populations and potential diagnoses with different options for subsequent management, then samples with higher cell yield are required to reduce risk of false-negative results.⁷⁰  This raises the question of tailoring the diagnostic criteria to the sample that can be reliably obtained by needle biopsy. This is the underlying principle of cytology and currently of biomarker identification and next-generation sequencing analysis.

Although samples that are diagnostic can be obtained where lesional tissue is scant, this is done at a risk that the sample is inadequate. Rather, it is highly desirable that the biopsy instrument and procedure be designed to obtain predominantly lesional material.

Typically, CNBs performed with a biopsy gun or without a stylet contain tissue along the entire needle pathway. At best, nonlesional tissue can limit yield, but if it is abnormal, such as with reactive atypia, difficulty can be added to the diagnostic interpretation. For CNB, nonlesional tissue can be minimized by controlling the insertion angle of biopsy. Some FNAB techniques, particularly those designed to access deep tissues, use a stylet to occlude the lumen, excluding cells until the needle is in the lesion. For FNABs with needle gauge 21 and higher, few cells enter the needle without cutting action and low-pressure suction applied via syringe. Therefore, a stylet is not usually required for FNABs of palpable lumps, as the syringe suction pressure can be applied after the needle has entered the lesion. The ideal clinical conditions required to facilitate adequate needle biopsy are shown in Table 2 and discussed further below.

Reliability of a needle biopsy procedure to deliver an adequate representative sample on the first pass or a few passes is a major economic and patient-care concern. Multiple passes are in part a result of varied lesion texture and consistency (eg, presence of fibrosis, necrosis, fat, blood), in part related to requirements for multiple analytical studies, and in part related to the state of present biopsy devices. For FNABs, this is a problem that can be mitigated by observation of adequacy during the biopsy procedure by the cytotechnologist-pathologist team—for example, with rapid on-site evaluation (ROSE) technologies.^71,72  With ROSE, adequacy even with multiple passes may be only 75%, although ROSE may increase adequacy to recommend a fixed number of passes to yield sufficient cells.⁷³  Macroscopic methods such as reflectance spectroscopy to distinguish tumor from nontumor in needle biopsy samples are under development.⁷⁴ 

For CNBs, it is common even for superficial biopsies to obtain 2 or 3 cores, and for molecular studies, up to 10 passes may be required to obtain a 92% adequacy rate for lung biopsies using a 20-gauge side core needle.⁷⁵  Repeat CNBs, although uncommon, have similar adequacy rates to first biopsies.⁷⁶  In both situations, incremental procedure time is required. With a 16-gauge needle it is possible on a single pass to obtain ≈250 ng DNA, which is often sufficient for molecular diagnostic analysis.⁷⁵  However, needles with a gauge this large are used uncommonly, and, because of trauma risks for deep lesions, are usually reserved for superficial biopsies. Ideally, an adequate biopsy should be obtained in the same procedure time or less than a single pass from FNAB.

Presently, FNABs can be performed in a treatment room, but these procedures do require considerable operator manual skill. Core needle biopsies require highly skilled operators (eg, surgeons, radiologists), and often require imaging facilities, as well as specialized biopsy guns and stereotactic equipment. It is highly desirable to reduce the technical complexity of needle biopsy procedures and, for deep sites, to reduce complication rates. From an ergonomic and standardization perspective, it is desirable that the same needle biopsy device be used for a particular tissue irrespective of the texture or composition of the lesion.

Fine-needle aspiration biopsies use needle gauges of 21 or higher, which minimizes bleeding risk. For CNBs, the tendency is to use 18-gauge or larger needles, with corresponding higher risk, principally from bleeding. This limits the use of CNBs with gauges less than 20 for deep lesions and patients with coagulation problems. Core needle biopsies using large-gauge needles incur the slight risk of tumor seeding in the needle track.⁷⁷  For CNBs, adequacy in terms of yield²⁴  or whether biopsies are deemed clinically adequate appears independent of needle gauge used. Needle biopsies of lesions with high vascularity have the risk of bleeding. From an adequacy perspective, particularly for FNAB, aspiration of blood can limit the accession and interpretation of lesional cells. Thyroid and liver biopsies are 2 sites where this consideration is particularly important. The risk for bleeding with FNAB can be reduced by use of needles with a diameter of 20 gauge or smaller bore, by minimizing needle manipulation within the lesion, and by minimizing suction pressure applied by the syringe.

The most cost-effective medical device would be an instrument that is modest in cost and where the result can be obtained during the same patient visit to the physician's office, blood pressure measurement being the foremost example. The need for referral to specialized facilities with limited-throughput (eg, endoscopy) imaging, although often necessary, increases costs dramatically, and often results in diagnostic delay.

The major sources of sample inadequacy related to device design and biopsy procedure are shown in Table 3.

Needle biopsy is evolving from a simple to a complex procedure that involves not only cell/tissue retrieval but also immediate specialized fixation and preanalytical processing for multiple laboratory analyses. The clinical team needs to be fully aware of the diagnostic questions to be addressed in the context of the patient's clinical condition and the technical requirements for sample handling and for various analytical tests. Considerable manual skill is required for successful needle biopsies, especially FNAB. Sample yield has been shown to increase with operator experience^6,25,78,79  with imaging and with ROSE facilitating collaboration between radiologists and pathologists. Nonetheless, intrinsic operator manual skills commonly remain as a source of variable yield. Device improvement that reduces the operator skill to obtain an adequate sample is highly desirable.

By far, the most commonly used technologies are CNB⁸⁰  and FNAB,⁸¹  both of which rely on mechanical shear and cutting. In current practice, CNB needles derive from the tru-cut biopsy needle,⁸²  whereas FNAB was advanced into common use by Zajicek⁸³  in 1965. Additionally, through-the-needle microforceps^84,85  and variant devices such as SharkCore biopsy^86–88  have been used for specialized endoscopic ultrasound-guided (EUS) needle biopsy applications. As well, curved⁸⁹  or helical tipped⁹⁰  needles have been designed to reach less accessible sites. Typically, CNBs require an 18-gauge or larger-bore needle. With this size bore, risk of bleeding is present, particularly in vascularized lesions of thyroid or vascular tissues such as liver. For CNBs of breast lesions, there were no differences in diagnostic adequacy among 14-, 16-, and 18-gauge needles.⁹¹  Percutaneous lung lesion biopsy technique is controversial. In one report, a 14-gauge CNB provided more diagnostic information than a 22-gauge FNAB, albeit with a higher pneumothorax rate.⁹²  Because of pneumothorax and bleeding risks, needle size smaller than 14 gauge is recommended for lung needle biopsies.

Conversely, with EUS-guided biopsies of solid masses, FNAB required fewer passes and was equivalent in adequacy to CNB.⁹³  At some tissue sites, FNAB has been shown to have increased accuracy compared with CNB with lower complication rates.⁹⁴  In practice, for liver biopsies, a thinner-core 21-gauge needle has been shown to have equivalent yield to a 17-gauge needle.⁹⁵  For FNAB, with needle gauges 21 and higher, risk of bleeding is diminished. However, at gauge 25 and higher, risk of tissue shear with cell lysis or partial lysis as cells are drawn up the needle is increased. Compared with 27-gauge needles, 21- or 22-gauge needles have been shown to provide superior yield for thyroid⁹⁶  as well as head and neck lesions,²³  whereas 25-gauge needles were superior to 22-gauge needles for endoscopically biopsied pancreatic tumors and other masses.⁹⁷  Smaller-bore needles, for example 25-gauge, do not confer advantages⁹⁸  related to number of passes or reduction in complications. For FNAB in particular, design to lessen trauma to adjacent cells by minimizing cutting force during rotation is important.⁹⁹  Needle design is important, with conventional end-cut hypodermic needles demonstrating superior yield compared with side-cut devices.^75,100  For end-cut needles, reduction of insertion force by increasing cutting edge bevel angle and other modifications can result in longer cores with higher yield.¹⁰¹  Similarly, needles can be designed to minimize insertion force and to customize bevel length for particular tissues.^102–104  Needle design can also be customized to control deflection,¹⁰⁵  facilitating better positioning of the needle within the lesion. Biopsy needles can be localized within lesions by ultrasound.^106,107  Needle tip visibility increases with insonation angle, the angle of incidence of ultrasound waves relative to the detector.¹⁰⁸  To facilitate localization using ultrasound, needles, particularly those used for endoscopy, may have surface modifications such as polymer coating or etching to enhance echogenicity.¹⁰⁹  Recently, ultrasound vibration of the needle has been used experimentally to enhance needle visibility.¹¹⁰ 

Form factor is the term used to specify the size, configuration, physical arrangement, and specifications of device hardware. Biopsy guns used in CNB or syringes used in FNAB have bulk, and their form limits the available anatomical approaches for superficial lesions. As an example, to minimize the size of the instrument, an FNAB technique used for thyroid lesions particularly is to insert a needle alone directly guided by the operator. The needle is attached to a tube ≈30 cm in length, which is attached to a syringe held and operated by an assistant. Needle biopsies of deep lesions have lower adequacy rates compared with those of superficial lesions, related in part to needle deflection and in part to retention of material within the needle.¹¹¹  Core needle biopsy guns require maintenance, which, if deficient, is a cause of “misfires” and inadequate biopsies.

Lesions differ widely in their texture and hardness related to surrounding tissue. Control of needle force, deflection, direction, and tip position can be achieved exquisitely by tactile perception and manual operation. This fine tuning of control can be compromised or lost when automated devices such as CNB guns are used. With CNBs, linear translational motion into the lesion can proceed with some velocity, particularly if a device gun is used to thrust the needle into the lesion. The insertion velocity from a CNB biopsy gun can be more than 500 times that of an FNAB.⁴⁹  Insertion-force trauma is proportional to the square of the needle diameter and the energy required for insertion velocity. With automated guns, this ballistic impact energy is dissipated over a much larger tissue volume than the diameter of the needle. Hence, CNBs using guns can be much more traumatic to adjacent tissues than gentle manual translation, particularly in vascularized tissues such as lung and liver.

With FNAB, the bevel of the needle is used to cut tissue. Accordingly, the needle is propelled into the lesion with rapid serial clockwise and counterclockwise rotation as well as tilting. Further, to yield more tissue and obtain a more representative sample, the needle is withdrawn and then advanced internally through the lesion along paths in different directions than the first. This sampling technique is referred to as fanning, which also increases traumatic risk proportional to the number and vigor of the internal motions. Low-frequency vibration associated with this manual technique has been shown to increase insertion force, but on a much lesser scale than the ballistic impact of gun-driven CNB.¹¹² 

Fine-needle biopsy can be performed with suction (fine-needle aspiration) or less commonly, without suction (fine-needle capillary cytology).¹¹³  In both techniques, motion is required to obtain an adequate sample.^114–116  At least for thyroid needle biopsy, nonsuction techniques appear equivalent to suction techniques,^115,117  except in hypovascular lesions or where microcalcifications are present.¹¹⁸  Suction can be applied in various ways, including using syringes of various sizes, using a vacutainer, or, to achieve low suction pressure, slowly withdrawing a stylet from inside the needle, the slow-pull technique.^119,120 

Suction to obtain the sample is controversial,¹²¹  with some users obtaining adequate samples without suction.¹¹⁷  Too much suction is recognized as a factor that can increase bleeding risk, thereby degrading the sample.^119,122 

During needle biopsy procedures, pain is generated by the insertion force of the needle puncture through skin. This force can be reduced by using smaller-gauge needles and by modifying bevel design.^101,123  Less controllable is pain related to pressure from high-velocity linear translation (CNB with mechanical gun) and extensive needle translation within a lesion (FNAB). Pain arising from high-velocity linear translation or from multiple internal or external passes can limit the extent of the procedure and hence limit the yield.¹²⁴  Most needle biopsies into superficial tissues such as thyroid or breast are performed without local anesthesia. Nonetheless, pain arising from these procedures can be reduced, both by anxiety reduction strategies such as clear guidance for the patient about the procedure and by local anesthesia. For CNBs and deeper lesions, local anesthesia is often advantageous.

Tissue heterogeneity can limit sample yield. For both CNB and FNAB, tissue heterogeneity leads to sample fragmentation. For FNAB, tissue with a fibrous component reduces cell yield. Also, lesions with a cystic component may require more sampling to achieve sufficient yield.¹²⁵  Further, molecular heterogeneity, such as intratumor mutational heterogeneity, may be present even when cytologic appearances with light microscopy are relatively homogenous.

Blood in the sample reflects procedural trauma and can be decreased by using a smaller-bore needle and, with FNAB, by using gentler suction or omitting suction. Blood in the sample can obscure cellularity in FNABs. Perhaps more important, blood in the sample is a contaminant that can interfere with RNA extraction and amplification.

When the needle biopsy yields an insufficient sample, usually a further sample is required, with corresponding diagnostic delay in patient assessment. Although there often is a choice to perform an open biopsy, this results in further delay and incremental costs. In some situations, particularly for deep biopsies, patients may be too ill to tolerate a surgical procedure, and repeating the needle biopsy is the only positive pathway for diagnosis. When the sample is not representative, the patient is exposed to the potential hazards of a false-negative diagnosis. For patients, a minimally invasive needle biopsy procedure that reliably provides a sample adequate for definitive diagnostic results and facilitates timeliness in diagnosis has immense value in terms of anxiety reduction and confidence in their caregivers.

It is important to recognize that an inadequate needle biopsy has costs that extend far beyond those of the needle operator, needles, and the biopsy room. Impact to the value-chain costs include costs of specimen transport and preparation, cytotechnologist screening costs, special diagnostics such as genomics and immunocytochemistry, and pathologists' professional assessment and reporting fees. Further, costs for any attributable misdiagnosis and for inappropriate therapy and opportunity cost of delayed therapy should be included.

The total financial costs of FNAB and CNB are far from trivial. As calculated for a hospital setting in the United States in 2015, FNAB for a palpable breast lump costs $522, rising to $3800 for ultrasound-guided FNAB. The corresponding cost for image-guided CNB was estimated at $5946.¹²⁶  For bone tumors, relative costs have been estimated for image-guided FNAB at $3500, CNB at $4000, and open surgical biopsy at $5700.^127,128  These estimates are costs per procedure and do not reflect total diagnostic costs per case that result from repeat procedures related to inadequate first biopsy.

Clinically, the concept of adequacy must always relate to the clinical application and the reliability and ease by which the problem can be assessed by needle biopsy. Needle biopsies have many different applications, ranging from diagnostic assessment of a mass by cytology or grading of disease by determining specific tissue architectural features to specific assessment of biomarkers, microorganisms, or genomic changes to retrieval of tumor cells for culture or subsequent xenografts. An important emerging application is the retrieval of stem cells for culture and reimplantation to assist tissue regeneration. As well as technical advances to accommodate new analytical needs, specialty groups have advocated for quality standards and education to enhance needle biopsy adequacy. Although needle biopsies can be classified as superficial biopsies, using a direct needle device, and as deep biopsies, often incorporating the biopsy needle in an endoscopic device, biopsy quality is an important concern for all tissue sites. Specific issues related to needle biopsy diagnostic adequacy at selected tissue sites are discussed below.

The clinical problem is usually whether a thyroid nodule is malignant and, if so, what is its histologic (or molecular) phenotype. Malignant nodules are usually treated with subtotal thyroidectomy, which uncommonly has surgical hazards, but commonly has the need for the patient to receive thyroid hormone for life. Therefore, the treatment of choice for benign nodules is usually hemithyroidectomy, a more conservative operation. Because the thyroid gland is quite vascular, FNAB is preferred as the needle biopsy mode.^129,130  The problems of adequacy are illustrated by Gharib et al¹³¹  in an analysis of almost 11 000 thyroid FNABs. Of these biopsies, 21% were deemed nondiagnostic. A subset of 148 patients received surgery, and of these, 26 (18%) had malignancy. A further 1192 (11%) had an FNAB diagnosis of suspicious for malignancy. Of these, only 29% were found to be malignant at surgery, indicating that the total of inadequate cases approached 30%.

This is similar to the reviewed literature.¹³²  Among the 7071 cases (64%) diagnosed as benign, 18 cases (2%) were later found to be malignant. Review of these cases indicated that 10 had inadequate samples and 8 had misinterpretation. However, it could be argued that misinterpretations would be fewer if the samples had more tumor cells. Similarly, the incidence of malignancy confirmed histologically in nondiagnostic thyroid FNABs was 5%, about the same incidence rate as for the total FNAB thyroid population.¹³³  Currently, molecular-signature tests using FNAB have been devised to assess thyroid nodules.^19,134  These tests appear to have adequacy problems similar to those of conventional FNAB. Recently, immunohistochemical markers such as PD-L1 have been shown to be important in distinguishing benign from more aggressive thyroid follicular lesions.¹³⁵  Similarly, a molecular signature has been devised for papillary thyroid cancer that can predict the cases most at risk for metastatic spread.¹³⁶  Therefore, it would be desirable to have increased cell yield from FNAB so that cytology, immunocytochemical markers, and, if deemed necessary, molecular markers could be performed on the same FNAB. A confounding variable is the perception that thyroid cancer is overdiagnosed.¹³⁷  This suggests that cytologic criteria for suspicious for malignancy are not stringent enough. Recently, thyroid FNAB has been used to further classify suspicious lesions by the more detailed Bethesda classification system.^138,139  Another adequacy problem particularly relating to thyroid lesions with high vascularity is blood contamination. This problem can be addressed by reducing procedural trauma, namely taking care to insert the needle gently and to limit the suction vacuum. Further, thyroid nodules are often cystic, at least in part. With FNAB, increased sampling may be required in these lesions to obtain representative cells.¹²⁵ 

Both FNAB and CNB have been used to assess bone lesions. For lytic neoplastic lesions, FNAB and CNB have comparable efficacy, with about 85% adequacy.¹⁴⁰  Fine-needle aspiration biopsy has the advantage of less trauma and the opportunity to assess adequacy during the procedure by the ROSE technique. Using FNAB to identify the lesion followed by CNB to obtain additional tissue is an alternate technique.¹⁴¹  For identifying active microorganisms by culture from vertebral lesions, a literature review demonstrated that 63% of needle biopsies were inadequate.¹⁴²  Computed tomography–guided CNB was shown to be effective in 81% of radiolucent lesions.¹⁴³  Inadequate CNBs are more likely to occur with either very lytic or very sclerotic bone lesions.^144,145  Of note, 40% of nondiagnostic CNBs became diagnostic on repeat CNB, indicating an opportunity for technical improvement.¹⁴⁶  Experience with sacroiliac joint needle biopsies in ankylosing spondylitis patients reflects the difficulties of obtaining adequate samples from an anatomically difficult site for which immunohistologic data are desirable.¹⁴⁷  Using a 17-gauge CNB, tissue could be obtained in 69% of cases, but only 33% of cases were suitable for immunohistochemistry.

Bone marrow aspiration biopsy is widely used to assess hematologic disease by cytology, immunohistochemistry, flow cytometry, and genomic techniques. Normal hematopoietic marrow is soft and easily aspirated.¹⁴⁸  Bone marrow has heterogeneity, limiting the adequacy of aspiration from a single site.¹⁴⁹  When fibrosis is present or when marrow architecture is required, a CNB is indicated. Adequacy is dependent on needle design,¹⁵⁰  strategy for needle insertion into the iliac crest,¹⁵¹  and operator manual skill.¹⁵²  If both marrow cytology and marrow architecture are needed, separate aspiration and core biopsies are recommended.^153,154  Of note, in an extensive survey from 2001 to 2011 of 6374 bone marrow samples in 32 academic centers, adequacy remained a major concern, with adequacy criteria for CNB centered on core biopsy length and capacity for morphologic evaluation.¹⁵⁴  Altogether, 11% of CNBs were noncontributory. In part this can be attributed to the observation that on average only 74% of the biopsies contained evaluable bone marrow. Bone marrow cells contain bone-specific alkaline phosphatase. Accordingly, a high ratio of bone-specific alkaline phosphatase to total alkaline phosphatase indicates that the aspirate is derived primarily from marrow.¹⁵⁵ 

Needle biopsies of muscle are performed to assess muscle fiber type in high-performance athletes,^156,157  to assess mitochondria function related to genetic or metabolic abnormalities,^158,159  and to diagnose neuromuscular disorders.¹⁶⁰  For muscle fiber typing, cross-sectional profiles of about 250 fibers are required.¹⁵⁶  Typically, 60 mg of muscle is required for fiber typing, whereas 100 to 200 mg (which usually requires multiple cuts) is required for mitochondrial function assessment. Modification of needle technique to apply suction may increase yield.¹⁶¹  Although open biopsy is usually preferred for diagnosis of neuromuscular disease, CNB has the advantages that multiple areas can be sampled, that tissue can be obtained immediately for cell culture or genetic studies even though multiple passes may be required, and that with a 14-gauge needle only about 15 mg of tissue per pass is acquired.¹⁶⁰ 

Synovial needle biopsy is used to diagnose and assess various types of inflammatory arthritis and can be applied to both large and small joints.^162–165  The yield and diagnostic information provided by ultrasound-guided needle biopsy of synovium using 14-gauge needles is comparable to that of forceps biopsy visualized by arthroscopy.^166,167  However, for initial diagnosis of arthritis type, adequacy problems remain, as 18% of biopsies showed inadequate tissue and only 16% of biopsies resulted in a specific diagnosis.¹⁶⁸  For detection of tophaceous gout or calcium pyrophosphate dihydrate crystals in tissues, FNAB followed by direct examination of needle content with polarized light microscopy is minimally invasive and effective.^169,170  In addition, synovial biopsy can be effective in diagnosing periprosthetic infections prior to revision arthroplasty,¹⁷¹  but is not superior to synovial fluid aspiration for microbial culture.¹⁷² 

Both FNAB¹⁷³  and CNB⁶⁸  have been used for soft tissue tumors, and both modalities have limitations. Computed tomography–guided FNAB was totally inadequate for diagnostic assessment in 20% of cases and provided very limited information in a further 10%. However, when adequate lesional cells are present, FNAB proved to be accurate in defining tumor phenotype in 97% of cases.¹⁷³  Core needle biopsy has been shown to be effective for diagnosis of smooth muscle neoplasms, but in 50% of cases, samples underrepresent the grade compared with excision.⁶⁸  In more than 90% of cases, 2 or more cores were taken during the biopsy procedure. It is important to recognize that FNAB, CNB, and open biopsy each have similar accuracy rates as does prior magnetic resonance imaging.¹²⁸  Therefore, choosing the most appropriate biopsy device for the lesion is an important medical decision. For example, for lesions less than 1 cm in diameter, image-guided CNB is unlikely to be effective.¹⁷⁴ 

Needle biopsy is the diagnostic modality of choice for the assessment of palpable lumps suspicious for breast cancer. With an expert operator, FNAB is comparable to CNB, with more than 85% adequacy, but as FNAB requires more manual skill, CNB is the most common modality used.^175–179  An option is to perform FNAB and, if necessary, subsequent CNB biopsy.¹⁸⁰  For nonpalpable breast lesions, ultrasound-guided FNAB is a useful diagnostic modality.^181,182  For primary breast cancers, it is also desirable to assess estrogen receptor, progesterone receptor, and HER2 status, and, currently, biomarkers for specific drugs such as PI3K. Again, although CNB is the current standard, either FNAB or CNB techniques can be used.^176,183  Further, during neoadjuvant chemotherapy, it is useful to sample the tumor by needle biopsy to assess ribosomal RNA disruption, an indicator of long-term chemotherapy efficacy.^184,185  Also, to assess regional spread to axillary lymph nodes, ultrasound-guided FNAB can be used and can often obviate the need for a sentinel node biopsy.^186,187  If adequacy assessment is limited to the general probabilistic categories positive, suspicious, atypical, epithelial proliferative, unremarkable, and nondiagnostic, then adequacy can be very high.²⁸  However, for lesions classified as suspicious only, clusters have been used to assess breast FNAB adequacy.^188,189  As a diagnosis of invasive ductal carcinoma may be made on very few cells, or, as in some circumstances, abundant cells may be associated with benign disease, there is a consensus not to weigh highly cell counts or cell clusters by themselves, but rather to provide cell adequacy assessment in the context of clinical and imaging findings.¹⁹⁰  Of concern for late deleterious consequence of biopsy methodology, there is one report that distant metastasis at more than 5 years was higher in patients diagnosed with CNB compared with FNAB.¹⁹¹ 

Needle biopsies of lung can be characterized by indication and location as (1) diffuse lesions (infection, neoplasia, pneumoconiosis, idiopathic); (2) mass, lung parenchyma; or (3) mass, associated with bronchus.

For diffuse lesions and parenchymal masses biopsied percutaneously, the same considerations for FNAB versus CNB apply as for other tissue sites.¹⁹²  However, larger needle gauges, usually required for nonneoplastic lung diseases,¹⁹³  present greater risks of bleeding and pneumothorax.¹⁹⁴  Bleeding and pneumothorax complications are adversely affected by pleura-lesion distance more than 21 mm^194,195  and by oblique insertion angle more than 51°.¹⁹⁶  For mediastinal lymph node assessment and masses biopsied endoscopically via the bronchus, special instrumentation for endobronchial ultrasound-guided biopsy is required.^197,198  Further, as a small portion of non–small cell lung cancers respond to therapy directed against driver mutations such as epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase (ALK), PD-L1, and proto-oncogene tyrosine-protein kinase (ROS1), a biopsy for molecular assessment of driver mutations is necessary.¹⁹⁹  Presently, these biopsies require ROSE to achieve 90% adequacy.⁴ 

Accessing pancreatic tumors via endoscopy of the gastrointestinal tract is a technically challenging procedure and usually requires ultrasound guidance (EUS). Both FNAB and CNB techniques are used and instrument designs continue to evolve.⁸⁰  With a comparable number of passes there appears to be no differences in adequacy rates between CNB and FNAB for diagnosis of malignancy.²⁰⁰  Adequacy is high only if multiple passes are performed and, for FNAB, if the samples evaluated immediately by ROSE.^201–203  However, EUS needle biopsy has advantages over open biopsy to sample chemotherapy-naïve tumors and to provide material for genomic analysis, for example KRAS mutations²⁰⁴  and immunophenotype analysis.²⁰⁵  Against the significant hazards of open biopsy, inadequacy rates of EUS needle biopsy can be tolerated. Endoscopic ultrasound-guided needle biopsy is also extremely useful for accessing and selecting metastatic tumor present in the peritoneum.²⁰⁶  The use of a stylet during this procedure is controversial, but there is an emerging consensus that the stylet presents few advantages for specimen acquisition.^207,208  The wet suction technique, whereby the endoscopic catheter is filled with saline prior to insertion of a needle into the lesion and application of suction, may enhance yield.²⁰⁹  Yield may also be enhanced by bevel design.²¹⁰  However, adequacy is a substantial issue even with FNAB of pancreatic cystic lesions.²¹¹ 

Liver lesions illustrate most of the applications and challenges of needle biopsies. There are 4 common technical approaches to liver needle biopsy: blind percutaneous biopsy, ultrasound-guided percutaneous biopsy, EUS biopsy, and transvenous (transjugular) biopsy.²¹²  The choice of approach is determined by clinical needs. For diffuse disease, blind biopsy is possible but ultrasound guidance desirable^213–215 ; for an accessible solitary or predominant mass, ultrasound-guided percutaneous biopsy is used. Frequently, diseased liver has multiple nodules and/or nodules that are not easily accessible. In these cases, EUS biopsies are preferable.²¹⁶  Transvenous biopsies from the internal jugular vein through the vena cava are reserved for lesions adjacent to the venous system or where there is significant bleeding risk or ascites.²¹⁷  Similarly, FNAB, CNB, and occasionally microforceps²¹⁸  modes are used depending on the application. There is considerable controversy over which modality is most effective.²¹⁹  Core needle biopsies can provide tissue architecture, but at increased risk of bleeding complications. However, for medical diseases, CNBs still have a considerable inadequacy rate, in part because adequate assessment is dependent on architectural presence of sufficient portal tracts.^69,220  Although more portal tracts are retrievable with larger-bore needles, this may not make a practical difference in diagnostic results.⁹⁵  Increasing the CNB bore from 18 to 16 gauge has been shown to increase yield.²²¹  With the aim of increasing yield and reducing complications, a CNB device has been modified to facilitate multiple passes from a single percutaneous puncture.²²²  Fine-needle aspiration biopsy is less traumatic but does not provide tissue architecture information. Further, needle biopsy applications range from assessing tissue architecture (eg, hepatitis or transplant reaction), diagnosing a mass (eg, cancer²²³ ), and assessing antiviral drug load²²⁴  to assessing presence of specific biomarkers or microorganisms.²²⁵  Needle biopsies continue to be useful in the assessment of chronic hepatitis and its immune reactions.²²⁶ 

Needle biopsies of kidney are used for 2 different applications: assessment of medical disease and diagnosis of renal masses. For medical diseases, percutaneous needle biopsy is an essential diagnostic tool to assess glomerular, vascular, and tubulointerstitial disorders²²⁷  and allograft reaction.²²⁸  Typically, CNB is the mode of choice for glomerular diseases, where a yield of more than 20 glomeruli is deemed adequate for assessment.²²⁹  Bleeding, hematuria, and pain are the major complications occurring in more than 15% of patients^230,231  and are more common with increased needle size.²³²  For glomerular disease, the biopsy objective is to obtain enough glomeruli for assessment without bleeding or hematuria. With CNB, strategies to obtain this objective involve using larger needles,^232–236  biopsy guns,^232,237  coaxial needles,²³⁸  and ultrasound guidance to position needle insertion where glomeruli have dense distribution through the entire needle path.^239,240  Comparing 18-gauge needles, increased core fragmentation was found using disposable needles.²⁴¹  Another strategy to increase adequacy and safety is to use a needle guide with its path monitored by ultrasound.²⁴²  Similar techniques can be used to enhance adequacy for renal mass biopsies (kidney tumors)²⁴³  but a significant inadequacy rate remains.²⁴⁴  To enhance diagnostic adequacy and decrease clinical complications, specialized imaging, including optical coherence tomography^245,246  and Raman spectroscopy²⁴⁷  performed through an image-guided needle, is an emerging alternative approach to renal mass biopsy. Small kidney masses less than 4 cm in diameter present special diagnostic and clinical management challenges, addressable by needle biopsy.^248,249  DNA methylation biomarkers obtained by needle biopsy are being explored to characterize specific tumor type and aggressiveness.²⁵⁰  To reduce complications, FNAB is an increasingly favored modality for assessing renal masses.²⁵¹  Inadequacy rates of either FNAB or CNB can be reduced by combining the 2 techniques.²⁵²  For FNAB containing spindle cells, immunocytochemistry may increase diagnostic specificity,²⁵³  and for lymphoid lesions, flow cytometry.²⁵⁴ 

Prostate needle biopsies are very widely used for primary diagnosis²⁵⁵  and are increasingly used for posttreatment surveillance²⁵⁶  and, using biomarkers, for prognosis.²⁵⁷  The extensive histologic heterogeneity as measured by Gleason score differences within an individual cancer mandates multiple sampling of tumor at the same procedure,²⁵⁸  at times using a saturation technique, extending to 12 or more needle biopsies.^258,259  To enhance adequacy, much effort has been made to improve the approach route for needle biopsy^260,261  and, in needle design, to increase core length.²⁵⁵  Accuracy of needle targeting is affected by mode of imaging guidance²⁶²  and by tip deflection, which in turn is affected by prostate size and tissue texture.²⁶³ 

Lymph node needle biopsy presents special challenges as the suspected lesion may be firm or soft, may be present in more than one node, may occupy only part of an enlarged node, and may represent an extrinsic (metastatic carcinoma, infection) or intrinsic (reactive lymphadenopathy, lymphoma) process. Accordingly, image guidance is imperative in both choosing the target node and guiding the needle to the lesion.²⁶⁴  The choice of needle biopsy technique is dependent on lesion size, anticipated texture, location, and whether tissue architecture is required.¹⁸⁷  In general, FNAB is preferred if the lesion is small (<1 cm), soft, or close to vital structures; where tissue architecture is not required; or in veterinary applications where CNB is not practical.²⁶⁵  Core needle biopsy is reserved for lesions where tissue architecture is needed or where distinction between lymph node reaction and neoplasia is difficult.²⁶⁶  Nonetheless, both FNAB and CNB have significant inadequacy rates and require multiple passes.^111,264  Diagnostic assessment of lymphoid lesions and lymphoma with needle biopsy samples can be quite complex, requiring several different analytical modalities, including flow cytometry, and mandating that samples be sufficient and representative.²⁶⁷  This is facilitated by placing the samples in a cell-preservation medium and, after retrieving the tissue fragments for formalin fixation, using the fluid medium for flow cytometry.²⁶⁷  Lymph node FNAB for cytology can be used as an independent prognostic indicator for metastatic breast cancer,²⁶⁸  as can the presence of CYFRA 21 keratin biomarker.²⁶⁹  Fine-needle aspiration biopsy of lymph node can also be used in allergy studies to assess effectively the complex immune reaction following intradermal antigen challenge.²⁷⁰  Underuse of FNAB in this application relates in part to capacity to analyze lymphocyte subsets and in part to difficulty in accessing sufficient quantity of cells.

Needle biopsy is used to obtain samples for cytology,^271,272  for microbial stains and culture,^273,274  for genomic probes from localized sites of infection,^275–277  or for flow cytometry of viral antigens.²⁷⁸  Each type of sample has particular requirements related to representative cells and handling, and each is fraught with inadequacy related to sample representativeness. Cytologic analysis is used to detect reactions typical of active infection, such as human immunodeficiency virus lymphoid reaction,²⁷⁹  granulomatous inflammation particularly related to tuberculosis,^225,280  or viral cytopathic effects.²⁷¹  At some sites, such as mediastinal lymph nodes, endoscopic transbronchial FNAB has been ineffective to identify respiratory infections.²⁸¹  However, percutaneous needle biopsy, either FNAB or CNB, can be effective to diagnose pulmonary infection.²⁸²  Concentration of FNAB by centrifugation has been shown to improve detection rate of tubercle bacilli by microscopy,²⁸³  culture,²⁸⁴  and molecular diagnostics.²⁸⁵  For vertebral osteomyelitis, the microbial yield is often inadequate.²⁸⁶  Choice of sample site can be important, with better diagnostic yield obtained from adjacent soft tissue rather than bone in cases of pyogenic vertebral abscess.²⁸⁷ 

Needle biopsies are widely used for many diagnostic applications because, whatever the limitations of adequacy, these modalities offer the safest and most effective means of obtaining cells and cell components for diagnostic purposes. Some opportunities to improve needle biopsy adequacy are shown in Table 4 and discussed further below.

The conventional definition of needle biopsy adequacy, presence of recognizable cells indicative or diagnostic for the clinical lesion, is insufficient for current expectations of quality in medicine and for precision medicine envisaged for the future. Diagnostic adequacy for needle biopsies needs to be defined in quantitative terms related to precise fit for purpose (clinical intended use). This definition may include provision of sufficient cells not only to identify tumor but also tumor phenotype, and, when relevant to therapy, to assess cells by immunocytochemistry and genomic profiling for companion diagnostic markers. A practical issue is that the menu of appropriate tests may be known only upon immediate retrieval of the sample. Just as for intraoperative consultation, these situations require the presence of a pathologist. Cost-effective diagnostic adequacy is a further goal that involves reducing trauma (small-gauge needle), using more effective equipment, FNAB rather than CNB, and reducing the number of passes to obtain a sample. Ideally, sampling would be so effective that ROSE techniques would not be required. Further, efforts to achieve consensus of biopsy adequacy definitions will foster development of more quantitative definitions of adequacy and serve to stimulate development of better needle biopsy procedures.

In recent years, advances in needle biopsy technology have been directed toward complex devices that can obtain samples from deep sites such as bronchus or pancreas. The need for multiple passes and frequent failure to obtain sufficient samples for molecular analysis appears to be clinically acceptable at present but is far from ideal. Greater recognition of needle biopsy inadequacy and more rigorous audit of biopsy adequacy is a first step toward addressing the technical problems. Needle biopsy design needs to be focused on ergonomics, reducing dependence on operator manual skills, and minimizing the duty cycle, setup, and sampling time. Biopsy needle design considerations should incorporate the functionality to retrieve the lesional sample irrespective of its intrinsic hardness, texture, or heterogeneity. Retention of operator tactile perception, “feel,” as feedback for lesion texture and needle position should be considered as a positive device-design feature. In addition, there is room for improvement in biopsy localization techniques. Further, adequacy may be enhanced in the future by reducing the mass of material required by substitution of cytologic diagnostic criteria for histologic architectural criteria and substitution of diagnostic biomarkers for cell morphology criteria.²⁸⁸ 

Because needle biopsy involves less time, trauma, and cost, it is the preferred first choice over open biopsy when representative material can be obtained. However, needle biopsy is still challenged by sample adequacy. This can be addressed by improvement in needle biopsy technology and by improved assessment criteria, including cytology and deployment of biomarkers that are more amenable to adequacy via a needle than criteria such as tissue architecture originally developed for open surgical specimens. Currently, there is considerable effort to develop liquid biopsy,^14,15  sampling of circulating tumor cells, nucleic acids, from peripheral blood. For most applications, direct needle biopsy of the lesion would be preferred for specificity if specimen adequacy, trauma, and costs are comparable or superior.

Often, the needle biopsy operator is different from the physician making therapeutic decisions on behalf of the patient. This makes it even more essential that adequate clinical information be provided to the operator so that careful attention is made to obtain an adequate biopsy. Regular feedback on success rate through recording of proficiency for each procedure and its result is an essential quality assurance tool to analyze deficiencies and improve biopsy performance.

Education and competence assessment for needle biopsy applies to both knowledge and capability to understand the intended use, as well as the capacity to perform the biopsy itself. The professional personnel involved, who include the referring physician, the biopsy operator, the pathologist, and the laboratory technologists, must be informed about the purpose of the biopsy, the diagnostic possibilities, and the appropriate priorities and techniques for specimen handling and preanalytical processing, as well as the practical limitations of these procedures. These requirements are far from trivial. In the coming era of precision medicine, needle biopsy clinical indications and methods for analysis are evolving rapidly, on a continuing basis. In the past, the art of performing needle biopsies was taught and learned largely from medical folklore: “See one, do one, teach one.” This resulted in a wide range of operator capability. Recognizing this deficiency and the observation that competency in needle biopsy technique tends to increase with experience,²⁸⁹  professional societies, through their guidelines and standard-setting activities, have stimulated the development of objective competence-based training^{230,290–296}  programs to rectify this problem. The need remains to develop teaching tools, including simulation-based training^297–300  and continuing education, that describes well the principles of needle biopsy technique and the characteristics of an adequate biopsy.³⁰¹ 

Because FNAB devices have a much lower intrinsic cost and facility requirement, this is the needle biopsy mode of choice in developing countries.^302,303  This provides a strong incentive to refine FNAB to enhance adequacy of yield and ease of use, a goal that will enhance cost-effective diagnostics in developed nations as well. Performance of the needle biopsy is just one step in the diagnostic assessment process. For superficial biopsies in particular, fee schedules are modest, inhibiting development and deployment of novel devices that can improve sampling. This results in increased downstream costs of processing and time in attempts to interpret inadequate samples. A costing solution that will facilitate new technology development is to bundle the technical diagnostic costs of the entire procedure, including needle biopsy and cell and tissue processing and analysis, as well as the opportunity costs for repeating inadequate biopsies.

Precision medicine will require more frequent, more detailed, and more quantitative assessment on each tissue sample, often involving advanced immunochemical and genomic techniques. At the same time, sample access at deep sites, patient safety, and facility costs are mandating the need for smaller samples than can be obtained by open surgical biopsy. For these reasons, needle biopsy has become a key preanalytical process for diagnosis and follow-up assessment of many diseases, especially tumors. For successful precision medicine, a needle biopsy involves much more than a skilled operator, a skilled imager, and a skilled pathologist. The needle biopsy process starts with the clinical question regarding the intended purpose of the biopsy and continues through the biopsy itself, followed by elaborate specimen handling and prioritizing the scant tissue for various analyses. The analysis itself may involve several members of the laboratory team with production of a unified report that impacts directly on the personalized therapy of the patient.

Presently, needle biopsy adequacy can be improved throughout the entire process by educating professionals on process steps, by communicating clearly with all the professionals involved in the intended diagnostic use, by planning for technical challenges of specimen handling on scant tissue, and by providing high-quality analyses and reporting, as well as by tracking all aspects of biopsy performance through quality assurance programs. In the future, needle biopsy device advances, improvements in needle biopsy specimen handling, and new developments in analytical test modalities will be seen as key enablers for diagnostics in precision medicine.

We thank Daniele Generali, MD, PhD; Anne Hsieh, BASc, MASc, PhD; Ed Korb, MB BCh, FRCPC, DABR; Lisa Boreanaz, BASc, MASc; and John Soloninka, BSc Eng Physics, P Eng, MBA for their stimulating discussions about needle biopsy adequacy and their helpful suggestions on preparation of this manuscript.