Minimally Invasive Sampling Techniques for Pancreatic and Pulmonary Nodules: A Review of the Evolution of Guidance and Sampling Methods

ABSTRACT

Tissue biopsy is frequently required in the workup of newly discovered nodules and masses detected by palpation, visual inspection, or imaging. Biopsy techniques include open surgical biopsy, core needle biopsy, fine‐needle biopsy (FNB), fine‐needle aspiration (FNA), and cryobiopsy. The last four techniques are utilized because of their minimally invasive nature and low patient morbidity. While FNA and FNB are often utilized at a number of body sites, cryobiopsy is becoming the dominant method for sampling of lung nodules. In this review, the authors investigate the utility, advantages, disadvantages, and diagnostic accuracy of minimally invasive biopsy techniques for lesions of the pancreas and lung.

1. Introduction

Newly discovered palpable nodules or masses and those detected on imaging studies often require tissue sampling. Traditionally, tissue diagnosis of lung and pancreatic lesions required open excisional or incisional biopsy, but the development of small biopsy needles and cryobiopsy devices, along with improved imaging and guidance techniques, have allowed more minimally invasive techniques to replace open biopsy. The optimal biopsy method depends on a host of factors including the site and orientation of the lesion requiring biopsy, the availability of radiologists with expertise in image guidance, trained endoscopists or bronchoscopists, availability of specialized needles or cryobiopsy devices, and cytopathologists or surgical pathologists trained in the evaluation of small biopsy specimens.

In this manuscript, the authors assess the available minimally invasive techniques and compare their advantages for sampling of pancreatic and pulmonary lesions. Significant advances in biopsy instruments (forceps, brushes, FNA, fine‐needle biopsy needles, and cryobiopsy probes) along with improved guidance methods (guide sheath, radial ultrasound probe, electromagnetic navigation, robotic‐assisted bronchoscopy) have significantly improved biopsy of lung and pancreatic lesions.

In the last two decades, a number of new small diameter needles with specialized cutting tips have been developed. Unlike the simple fine‐needle aspiration (FNA) needles which have beveled cutting tips and produce specimens composed of individual cells and small tissue fragments, these fine‐needle biopsy needles cut small cores which maintain histologic architecture. A variety of these cutting needles are commercially produced and designated by the names SharkCore (Medtronic, Minneapolis, MN), ProCore (Cook Medical, Bloomington, IN), and Acquire (Boston Scientific, Marlborough, MA). The precise design of these needles varies but all have complex cutting tips, some with an additional side slot design. The use of these needles varies with body site. This manuscript will discuss the use, advantages, and disadvantages of these fine‐needle biopsy needles in comparison to traditional fine‐needle aspiration needles and core biopsy needles. More recently, larger specimen samples can be obtained without an associated increase in diameter of sampling device by the use of cryobiopsy. Cryobiopsy samples provide a larger tissue fragment with the dual advantages of preservation of histologic architecture and an increased amount of tissue for molecular testing. Cryobiopsy is replacing cytologic methods at many institutions for pulmonary lesions. Guidance techniques have also improved allowing more successful sampling of peripheral lung nodules. These newer guidance methods for lung nodules include radial probe guidance, guide sheath techniques, electromagnetic navigational bronchoscopy, and robotic‐assisted bronchoscopy.

2. Biopsy Techniques for Pancreatic Masses and Cysts

Diagnosis of pancreatic solid or cystic lesions can be achieved by core needle biopsy (CNB), fine needle aspiration (FNA), or fine needle biopsy (FNB). Guidance techniques have varied and included ultrasound (US), computed tomography (CT), intraoperative visual and palpation guided biopsy, and endoscopic ultrasound (EUS) guided approaches. Each method has its unique advantages and limitations, but EUS guided FNB has recently become the most widely utilized method. Review of these methods highlights the advantages, disadvantages, diagnostic accuracies, and hazards of each method. While fine‐needle aspiration and fine‐needle biopsies are widely practiced for the diagnosis of pancreatic lesions, some authors have suggested limiting their use to metastatic or unresectable masses because these methods are insufficiently accurate for diagnosis of potentially resectable pancreatic tumors [1].

2.1. Traditional Core Biopsy

Core needles used for biopsy of pancreatic abnormalities vary in size from 11 to 20 gauge [2]. Guidance techniques have included CT, US, and palpation in the case of intraoperative biopsies [3, 4, 5, 6, 7, 8, 9, 10]. The earliest experience with CNB was associated with intraoperative biopsies [5, 6]. Percutaneous CT and US guided CNBs were also early methods for the investigation of pancreatic lesions [7, 8, 9, 10]. CNB uses the largest diameter needles (14G–19G). While large gauge needles have been used [2], biopsy needles as small as 18G are increasingly utilized [2, 8, 11, 12, 13, 14].

CNB of pancreatic masses has shown good diagnostic accuracy with most reported sensitivities to be between 90% and 100% [1, 4, 7, 8, 9, 10, 11, 12, 13]. Reported specificities range from 94.7% to 100% (Table 1) [1, 4, 7, 8, 9, 10, 11, 12, 13]. CNB appears useful in ruling out the presence of a malignancy with negative predictive values ranging from 0.42 to 1.0 [1, 4, 7, 8, 9, 10, 13]. Based on this data a positive result is highly predictive for the presence of a malignancy, but a benign result does not definitively exclude the presence of a malignancy [1, 4, 7, 10]. Core Needle biopsy has been associated with major complication rates as high as 2% [14, 15].

TABLE 1.

Accuracy data for core needle biopsy of pancreatic lesions.

Author	Number of cases	Overall accuracy	Sensitivity	Specificity	PPV	NPV
Huang [4]	—	95.8% (94%–100%)	94.4% (90%–100%)	97.9% (94%–100%)	—	76.2%
Bhatti [7]	153	91%	90%	95%	—	42%
Kahlriman [8]	250	98.4%	99%	94.7%	—	94.7%
Mitchell [9]	33	100%	100%	100%	—	100%
Yang [10]	88	93.3%	92.6%	100%	—	60%
Karlson [13]	89	—	91%	100%	100%	83%
Zech [14]	63	81%	78.1%	100%	100%	—

Abbreviations: NPV, negative predictive value; PPV, positive predictive value.

2.2. Endoscopic Ultrasound Guided Fine Needle Aspiration

The technique of guidance is similar for FNA and FNB. Thorough ultrasound examination is performed to determine the optimal approach path and the absence of vascular structures in the needle pathway. The ultrasound transducer head is placed against the gastric or abdominal wall, the catheter with enclosed needle is placed into the biopsy channel. The fine needle is then advanced through the gastric or duodenal wall and into the lesion. Following removal of the needle stylet, tissue acquisition is performed with multiple strokes of the needle under negative pressure. Material can be smeared for Giesma or Papanicolaou staining or formalin fixed for cell block preparation. Endoscopic ultrasound guided FNA (EUS‐FNA) was the first line procedure for definitive diagnosis of pancreatic malignancies until the development of new needle types used for fine needle biopsy (FNB). FNB has served as the dominant method for diagnosis of solid and cystic pancreatic lesions over the past two decades due to its high diagnostic accuracy, safety and low complication rate [16, 17, 18, 19, 20, 21, 22, 23, 24]. In many practices, EUS‐FNA remains the predominant method for ascertaining whether or not a pancreatic lesion is malignant. The technique has high diagnostic accuracy and a low rate of complications [16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. Reported accuracy for EUS‐FNA is approximately 90% for lesions over 10 mm [25].

A large number of individual studies and meta‐analyses have been published documenting the method's diagnostic accuracy (Table 2). The majority of studies have reported very high specificity ranging from approximately 90% to 100% [16, 17, 18, 19, 20, 21, 22, 23, 24]. Associated positive predictive values (PPV) are similarly high ranging from 73.8% to 100% [20, 22, 23, 25]. Sensitivity data are more variable and range from 65% to 97% [16, 17, 18, 19, 20, 21, 22, 23, 24]. Likewise, negative predictive values (NPV) range from a low of 32.6% [20] to a high of 90.9%. The relatively poor sensitivity and low NPV suggests that a benign diagnosis does not exclude a malignancy. This has led some authors to question if EUS‐FNA is useful in patients with resectable pancreatic masses [1, 25].

TABLE 2.

Accuracy data for endoscopic ultrasound‐guided fine‐needle aspiration of pancreatic lesions.

Author	Number of cases	Overall accuracy	Sensitivity	Specificity	PPV	NPV
Chen [16]	1860	—	92%	96%	—	—
Raut [17]	233	92%	91%	100%	—	—
Sagami [18]	271	85.8%	77.4%	89.2%	73.8%	90.9%
Tian [19]
22 g	1824	—	89%	100%	—	—
25 g	1824	—	90%	99%	—	—
Zhang [20]
ROSE	97	94.8%	94.4%	100%	100%	58.3%
No ROSE	97	70.1%	65.1%	100%	100%	32.6%
Wu [21]	18	80.3%	83.3%	80.0%	90.9%	71.4%
DelVecchio‐Blanco [22]	308	69%	63%	100%	—	—
Yoshinaga [23]	246	96.2%	97.2%	88%	100%	81.4%
Chen [24]	4840	—	89%	96%	—	—

Rapid onsite evaluation (ROSE) has been proffered as a method for improving adequacy rates and diagnostic sensitivitiy [20, 26, 27]. These studies have consistently demonstrated improved diagnostic sensitivity when ROSE is used. Sensitivity improved from 63.7% to 91.7% in one study and from 65.1% to 94.4% in another with the use of ROSE [20]. Other factors have been shown to impact diagnostic accuracy including needle size, lesion size, operator skill, and repeat EUS‐FNA sampling [22]. Needle size has been shown to impact diagnostic sensitivity with 25G needles outperforming 22G needles in some studies [19, 28, 29, 30]. 25G needles have a pooled sensitivity of 90% vs. 87% for 22G needles in a meta‐analysis [19]. A more recent study confirmed their findings with sensitivity improving from 0.85 (22G needle) to 0.93 (25G needle) [29]. The diagnostic accuracy of EUS‐FNA is also impacted by nodule size [18, 30, 31]. Pancreatic nodules under 10 mm were associated with a lower diagnostic accuracy than larger nodules [18, 30, 31]. The presence of chronic pancreatitis has also been shown to lower the diagnostic accuracy of EUS‐FNA [32].

A number of adverse events or complications are associated with EUS‐FNA [33, 34]. Reported complications occur in 0%–10.5% of cases [33]. Pancreatitis and abdominal pain are the most frequently reported complications, but fever and hemorrhage have also been reported. The more serious complication of needle track seeding has been reported in a few studies [33]. Yane et al. [33] reported a 5‐year cumulative needle tract seeding rate of 3.8% and Yoon et al. [34] reported peritoneal seeding associated with EUS‐FNA of intraductal papillary mucinous neoplasms in approximately 2.3% of cases.

2.3. Endoscopic Ultrasound‐Guided Fine‐Needle Biopsy

In an attempt to improve specimen adequacy and diagnostic accuracy of ultrasound guided needle sampling of pancreatic lesions a variety of small caliber cutting needles have been developed [35, 36]. These differ in the shape and placement of the cutting edges, with two basic types of fine needle biopsy needles currently available for clinical practice: end cutting or side cutting (reverse cutting) needles.

The first iteration was the Menghini‐type coring device proposed in 1998 [37] which was unsuccessful due to its poor penetration and lower diagnostic yield compared to FNA. The first‐generation needle biopsy device was the Quick‐core biopsy device (Cook‐Medical, Winston‐Salem, NC) which was quickly abandoned as it was stiff and technically difficult to use [38]. The second‐generation biopsy needles were the EchoTip ProCore (Cook Medical, Bloomington, IN, USA). These needles have an opening on the side, with a reverse bevel architecure [39]. ProCore FNB required a lower number of passes but had a similar diagnostic yield to EUS‐FNA [40, 41, 42]. Third‐generation EUS FNB needles are known as crown cut needles. They are either Franseen Type with three symmetrically distributed cutting edges or the fork‐tip type needle. Acquire (Boston Scientific, Marlborough, MA, USA), SonoTip TopGain needle (Mediglobe, Rohrdorf, Germany), and the Echotip Acucore (Cook Medical, Winston‐Salem, NC, USA) are examples of the Franseen needle type and SharkCore (Medtronic, MN, USA) is an example of the fork‐tip needle [42]. Figure 1 shows an example of the cutting end of the Acquire needle.

FIGURE 1.

Photograph of an Acquire needle demonstrating the multiple points with bevels at the distal end of the needle. This arrangement of cutting bevels improves the cutting nature of the needle to produce intact true cores. [Color figure can be viewed at wileyonlinelibrary.com]

A number of studies have compared these needles with each other and with EUS‐FNA. EUS‐FNB outperforms EUS‐FNA in terms of higher diagnostic yield, retained tissue architecture, and total tissue area [43, 44]. EUS‐FNB has a higher diagnostic performance [45]. EUS‐FNB has a higher diagnostic accuracy, sample adequacy, and requires fewer needle passes compared to EUS‐FNA and has minimal adverse events [46]. EUS‐FNB is preferred for sampling solid pancreatic masses over EUS‐FNA based on a predominance of scientific literature [45, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57]. EUS‐FNB has been consistently shown to have high diagnostic accuracy, often higher than EUS‐FNA (Table 3) [43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59].

TABLE 3.

Accuracy data for endoscopic ultrasound‐guided fine‐needle biopsy of pancreatic lesions.

Author	Number of cases	Overall accuracy		Sensitivity		Specificity
		FNA	FNB	FNA	FNB	FNA	FNB
Paduani [43]	45	80%	90%	80%	91%	—	—
Oppong [52]	108	75%	84%	71%	82%	—	—
Yousri [47]	50	98%	98%	98%	98%	100%	100%
Fujita [59]	105	85.3%	88.2%	—	—	—	—
VanRiet [57]	685 (meta‐analysis)	80%	87%	—	—	—	—

Few studies have directly compared end tip to side tip needles. In a meta‐analysis performed by Gkolfakis et al. [48], end tip (Acquire, Boston Scientific) outperformed side tip or reverse‐bevel needles (Echotip Procore, Cook Medical) for adequacy and accuracy. Similar findings were reported for the fork‐tip (SharkCore, MedTronic) needle. Gkolfakis et al. [48] showed similar findings when end‐cutting needles were compared to FNA needles with end cutting FNB needles outperforming FNA needles. Other studies have confirmed the improved accuracy of FNB over FNA [49, 50, 51, 52, 53, 54, 55] Studies have confirmed the superiority of the Franseen style needle over reverse tip and side tip designs [56, 57, 58, 59, 60].

ROSE in EUS‐FNB does not provide a significant benefit [61, 62]. EUS‐FNB with and without ROSE were compared in the largest multicenter, randomized control trial performed to date. ProCore, SharkCore, and Acquire needles were included. ROSE was performed using the touch imprint cytology technique [62]. Both groups showed comparable diagnostic accuracy [62]. A meta‐analysis confirmed that ROSE did not improve diagnostic accuracy in FNB performed with end‐cutting needles but could still have a role when reverse bevel needles are used [61]. ASGE guidelines recommend against the routine use of ROSE for EUS‐FNB. ROSE should be considered when available in the following circumstances: repeat FNB due to initial nondiagnostic sample, the lesion is not clearly seen or is obscured by artifact, an FNA needle will be used, and a preliminary diagnosis is desired to guide immediate treatment decisions [46].

Macroscopic on‐site evaluation (MOSE) of EUS‐guided FNA specimens of pancreatic masses has been proposed as a simple, feasible, and readily available alternative to ROSE [63, 64, 65]. MOSE was introduced by Iwashita et al. in 2015 [63]. The material obtained is expelled onto a glass slide and examined grossly for the presence of cores. These cores are defined as whitish or yellowish pieces of tissue with an apparent bulk (not including paste‐like or liquid‐like specimens) and are termed macroscopically visible cores (MVC). When these MVCs measured ≥ 4 mm, the sample showed a sensitivity of 93.1%. A randomized control trial conducted by Chong et al. [64] showed no advantage in terms of improving diagnostic yield but identified a reduction of the number of needle passes when MOSE was used [63, 64, 65].

Fine needle biopsy specimens are fixed in 10% neutral buffered formalin for paraffin embedding and histologic review.

2.4. Summary of Needle Biopsy Methods for Solid Pancreatic Masses

• Because of the relatively low negative predictive value of all three needle biopsy techniques, some authors have suggested that needle biopsies, including CNB, FNA, and FNB, be performed only in patients with unresectable or metastatic pancreatic masses.

• Needle biopsy has a small but not negligible needle tract seeding rate.

• FNB appears to have a slightly superior accuracy over FNA for diagnosis of pancreatic neoplasms.

• End cutting needles appear to have higher accuracy than side cutting (reverse bevel needles).

• FNB needles appear to yield more tissue for ancillary studies than FNA needles.

• EUS‐FNB appears to be the optimal biopsy technique due to its combination of high accuracy, high tissue volume acquisition, and low risk of complications

2.5. Needle Biopsy of Pancreatic Cystic Masses

EUS‐guided through the needle biopsy (TTNB) enables the retrieval of histological specimens which help to differentiate among the various cyst types. Two intracystic devices currently are in use: Moray microforceps (Steris, Mentor, OH, USA) and the MicroBite (MTW, Endoskopie Manafakture, Wesel, Germany). The indications for EUS‐TTNB are not well defined. Currently, three criteria need to be fulfilled for their use: biopsy can impact clinical management by reducing unnecessary surgical intervention, changing the clinical diagnosis, or modifying the follow‐up. One agreed upon indication is a morphologically indeterminate cyst [66]. Identification of high‐grade dysplasia may be hindered by the fact that the cyst lining is heterogenous, and this technique only samples the cyst wall at the trajectory of the TTNB device. The high rate of serious adverse events, up to 9.9%, has limited its adoption by endoscopists [67]. EUS‐FNB and EUS‐FNA have an advantage over EUS‐TTNB because of the fanning technique the endoscopists can use to sample more regions of the cyst wall [68]. EUS‐FNB has been used to sample the cyst wall with good results [68].

3. Guidance Techniques for Pulmonary Lesions

The identification of a new pulmonary nodule on imaging studies may require analysis of a tissue sample to determine the nature of the pathologic process present. Biopsy of a pulmonary nodule may be associated with or preceded by lymph node sampling for staging [69, 70]. Biopsy of a pulmonary nodule can be performed by wedge biopsy, video‐assisted thoracoscopic surgery (VATS), core needle biopsy, forceps biopsy, bronchial brushing, fine‐needle aspiration, fine needle biopsy, or cryobiopsy. Biopsy guidance techniques include computed tomography (CT), fluoroscopic guided transbronchial biopsy, flexible bronchoscopy, endobronchial ultrasound‐guided transbronchial biopsy (EBUS‐TB), electromagnetic navigational bronchoscopy (ENB), or robotic‐assisted bronchoscopy. Choice of guidance technique, biopsy method, and needle type or cryobiopsy depends on the location of the pulmonary nodule and the need to sample mediastinal and hilar lymph nodes (Table 4). Peripheral nodules can be sampled by transthoracic‐percutaneous core needle biopsy or FNA (PCNB or PCNFA), endobronchial ultrasound‐guided transbronchial biopsy using a radial probe with or without guide sheath, electromagnetic navigational biopsy (ENB), robotic‐assisted bronchoscopy (RAB), or guide sheath transbronchial FNA. Central pulmonary nodules and hilar/mediastinal nodules are sampled by forcep biopsy, flexible bronchoscopy with transbronchial fine‐needle aspiration (TBFNA) (Wang needle technique), flexible bronchoscopy with fluoroscopy‐guided FNA, endobronchial ultrasound‐guided transbronchial fine‐needle aspiration (EBUS‐TBFNA), endoscopic ultrasound‐guided transbronchial fine‐needle biopsy (EBUS‐TBNB), endoscopic ultrasound‐guided transesophageal FNA (EUS‐FNA), or fine needle biopsy (EUS‐FNB). Cryobiopsy has recently become the dominant method for sampling of pulmonary lesions. Each guidance technique and biopsy device is associated with its own limitations, complications, and diagnostic accuracy (PPV and NPV) [71]. Various guidance techniques (flexible bronchoscopy, ultrasound‐guided transbronchial, radial probe guide sheath, electromagnetic navigation, and robotic‐assisted bronchoscopy) can be paired with different tissue acquisition devices (FNA, FNB, forceps, brush, or cryobiopsy).

TABLE 4.

Biopsy techniques for central and peripheral pulmonary nodules.

Central lesions

Transthoracic/percutaneous core needle biopsy

Transthoracic/percutaneous FNA

EBUS‐TBFNA

EBUS‐TBFNB

EUS‐transesophageal FNA

Peripheral lesions

Transthoracic/percutaneous core needle biopsy

Transthoracic/percutaneous FNA

EBUS‐TBNA with radial probe

EBUS—with guide sheath

Electromagnetic navigation bronchoscopy (ENB) with FNA or FNB

Robotic‐assisted bronchoscopy (RAB)

3.1. Thoracotomy With Wedge Resection

Pulmonary nodules and interstitial lung disease can be investigated by thoracotomy with wedge resection. This procedure may be both diagnostic and therapeutic [72]. Wedge resection is associated with requirements for hospitalization, chest tube placement, and significant cost. Diagnostic accuracy for wedge resection is assumed to be high, but little if any data is published specifically addressing its diagnostic accuracy. Because of its potential for significant patient morbidity and high cost, more minimal diagnostic techniques are now favored for initial diagnosis.

3.2. Video‐Assisted Thoracic Surgery

Video assisted thoracic surgery (VATS) is a safe and efficient alternative to open lobectomy for patients with early‐stage non‐small‐cell carcinoma [73]. The procedure involves making a small incision in the chest wall through which a thoracoscope is inserted to allow visualizing of the lung and resection of detected nodules up to and including lobectomy. While formal surgery, it is associated with less operative trauma and shorter recovery times than wedge resection.

Because open wedge resections and VATS are not true minimally invasive techniques, they will not be further discussed here.

3.3. Percutaneous/Transthoracic Needle Biopsy

Percutaneous/transthoracic needle biopsy can be performed with either core needles or FNA.

3.3.1. Percutaneous/Transthoracic Core Needle Biopsy

Percutaneous‐transthoracic needle biopsy can use either core needles or fine needles under computed tomographic guidance (CT) and is effective for both central and peripheral nodules [74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84]. Ultrasound guidance can be used with similar accuracy [85, 86]. The increased need for ancillary testing has resulted in increased use of core needle biopsy (CNB) and FNB needles. Percutaneous transthoracic biopsy is traditionally performed under CT guidance. The lungs are imaged and the nodule is localized. Its position is marked on the chest wall with a “BeBe” and confirmed with repeat CT. The distance between the skin surface and the nodule is measured and the depth marked on the biopsy needle with a “steristrip.” Usually, a coaxial method is used where a larger guide sheath needle is inserted to the proper marked depth just peripheral to the nodule. The location of the coaxial needle is confirmed. After proper placement of the coaxial needle is confirmed, multiple FNAs or CNBs may be performed. The coaxial needle is withdrawn at the end of the sampling procedure.

Computed tomographic guided core needle biopsy (CT‐CNB) was one of the earliest minimally invasive methods for the investigation of pulmonary nodules and remains a standard technique for sampling peripheral lung nodules [74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84]. The increased need for sufficient material for molecular testing has resulted in its increased usage for both central and peripheral lung nodules despite its association with a significant incidence of biopsy‐related complications [77, 80, 84]. The accuracy of CT‐CNB is reported in Table 5. Sensitivity and specificity of percutaneous CT‐CNB depend on lesion size, with smaller lesions having poorer diagnostic accuracy [77, 85]. A number of factors have been shown to impact diagnostic accuracy, including the solid versus sub‐solid nature of the nodule, location of the nodule, number of passes, and needle type [76, 78].

TABLE 5.

Diagnostic accuracy of CT‐guided transthoracic core‐needle biopsy of lung nodules.

Author	Accuracy	Sensitivity	Specificity
Huang [77]	83.7% (< 15 mm) 96.8% (> 15 mm)	—	—
Liu [75]	95.1%	92.8%	100%
Basiri [76]	83%	75.4%	91.4%
Teixeira [78]	93.0% (< 10 mm)	NG	NG
Greif J [79]	81.7%	—	—
Lopez‐Hänninen [80]	95%	96%	100%
Tipaldi [81]	81% < 20 mm 92% > 20 mm	NG	NG
Besa [82]	91.5%	90.6%	NG
Tian [83]	94.6%	92.0%	98.6%

Abbreviation: NG, not given.

Core needle biopsy has a reportedly higher diagnostic accuracy than FNA for the specific diagnosis of benign pulmonary lesions (81.7% vs. 16.7%) [79]. For both benign and malignant lesions CNB has a high diagnostic accuracy of 95% [80] with full core needles having inferior sensitivity to tru‐cut needles (77% vs. 91%) [81]. In reported series and systematic reviews, the sensitivity of CNB has ranged from 90.6% to 94% and the specificity of CNB has ranged from 98% to 100% [74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85]. Nodule size impacts diagnostic accuracy of percutaneous CNB but the technique remains sufficiently accurate for nodules < 2 cm [76, 77, 78, 82, 87, 88]. Ultrasound can be used for transthoracic CNB guidance with similar accuracy to that of CT guidance (sensitivity 88% and specificity 100%) [85, 86]. A number of complications are associated with percutaneous core needle biopsy with pneumothorax and hemorrhage being most common [77, 89, 90, 91]. The complication rate is variable with the rate of pneumothorax being approximately 24% and the rate of significant hemorrhage being between 7% and 36% [77, 91]. Risk factors associated with pneumothorax include: lesion size, length of needle path, lesion location, presence of obstructive lung disease and operator experience [77, 91]. Most pneumothoraces do not require a chest tube placement. Risk factors for hemorrhage include long needle path, small lesion size, non‐ plural contact lesions and restrictive lung disease [77, 88].

Despite the not insignificant risks for complications, CNB remains a frequent choice for biopsy of pulmonary nodules due to its ability to obtain adequate material for molecular testing and other ancillary testing.

3.3.2. Percutaneous/Transthoracic Fine Needle Aspiration

Due to concerns regarding CNB associated complications and a desire to improve diagnostic accuracy, FNA was combined with the transthoracic/percutaneous approach previously described for core needle biopsy. The procedure is similar but multiple FNAs are taken and their adequacy confirmed by ROSE. Diagnostic accuracy of CT guided percutaneous/transthoracic FNA (CTPTFNA) is reportedly good [92, 93, 94, 95, 96, 97, 98]. Reported diagnostic sensitivity ranged from 76% to 90.85% [92, 93, 94, 95, 96, 97, 98]. Reported specificity is as high as 100% [92, 93, 94, 95, 96, 97, 98]. Diagnostic accuracy is higher for large nodules (> 0.8 cm) than it is for small nodules (≤ 0.7 cm) [85, 86, 97, 98]. Core needle biopsy appears slightly more accurate than FNA in some studies but is similar in others [99, 100, 101]. Use of a coaxial needle biopsy system appears to improve diagnostic yield [89]. The combination of CNB with FNA improved diagnostic accuracy for percutaneous needle biopsy [101]. In one study, diagnostic sensitivity and specificity were 89.7% and 83.4% for FNA and 75.6% and 84.7% for CNB but improved to 95.5% and 94.5% when FNA and CNB were combined [101].

Percutaneous/transthoracic needle biopsy of lung nodules using either CNB or FNA is associated with a number of complications including hemorrhage, pneumothorax or air embolism. The reported risk for pulmonary hemorrhage after percutaneous CT guided biopsy has varied from 7% to 36% [90, 91, 95, 97, 99] with the incidence varying with severity. The use of semiautomatic cutting needles and the coaxial technique appears to reduce the risk of serious hemorrhage [90]. Post‐biopsy pneumothorax varies in frequency with most cases not requiring chest tube placement. Risk of pneumothorax is related to needle size with larger needles having a greater risk for pneumothorax. Pneumothorax occurs in approximately 38% of cases when 18G needles are used for biopsy [91]. The risk of complication is higher when the biopsy needle transverses aerated rather than non‐aerated lung. A potentially serious complication associated with CT‐guided percutaneous lung biopsy is air embolism [102, 103, 104]. Risk factors for air embolism are coughing during procedure, hemoptysis during procedure, distance between the mass and pulmonary vein, cavitary lesion, lesion location, number of passes and abnormalities of coagulation [102, 103, 104]. Risk of an air embolism varies but is usually very low (0.2%) but air embolism can rarely be fatal [102, 103, 104].

3.4. Flexible Fiberoptic Bronchoscopy

Flexible fiberoptic bronchoscopy is an early minimally invasive technique for the biopsy of visible mucosal and immediately submucosal bronchial lesions [105, 106, 107, 108, 109]. Successful utilization of this method requires the presence of a visible mucosal or submucosal lesion. Flexible bronchoscopy can utilize forceps biopsy, bronchial brushing, or washings in addition to fine needle aspiration or fine needle biopsy [107, 108, 109]. The technique is most accurate for central lesions but can sample nodules not directly in contact with a bronchus when used with fluoroscopic or CT guidance [105, 106, 107, 109]. Diagnostic sensitivity for the technique varies with lesion location and biopsy method, with forceps biopsy having higher sensitivity than bronchial brushings and washings [109, 110, 111, 112, 113, 114].

The technique uses a flexible bronchoscope with a biopsy channel. The bronchoscope is introduced into the trachea and visualizes the mucosa of the carina and bronchi. The lesion of interest is visualized and forceps biopsy, bronchial brushings, or FNA is performed.

3.4.1. Flexible Bronchoscopy With Fine‐Needle Aspiration

Development of the flexible bronchoscope allowed direct visualization of mucosal and immediately submucosal nodules and masses and their sampling by FNA. For FNA, the lesion is visualized and a sheath with enclosed FNA needle is brought to the area of interest with the sheath and enclosed needle pressed against the lesion. The needle is then extended and multiple strokes of the FNA needle are performed. The needle is then retracted into the sheath. Sheath and needle are withdrawn from the bronchoscope. Material is expelled from the needle onto glass slides and smears made. When used without imaging guidance, fiber optic flexible bronchoscopy (FOFB) tissue sampling is limited to abnormalities no more peripheral than the segmental bronchi. Imaging using CT or fluoroscopy can extend the range of sampling further peripherally.

Flexible bronchoscopic biopsy of visible and submucosal lesions has good accuracy including sensitivity and specificity [105, 106, 107, 108]. The reported sensitivity has ranged from 36% to 92% for FNA [106, 107, 108, 109, 110, 111]. Lower sensitivities exist for brushing (39.0% to 64.8%) and forceps biopsy (56.6%) [109, 112, 113, 114]. When these methods are used in combination, the sensitivity is 92.7% [106]. Other studies have reported significantly higher sensitivity for forceps biopsy (82.7%) [108].

Transbronchial needle aspiration can be used to investigate peripheral lung lesions [71, 115, 116, 117, 118]. Fluoroscopic or CT guidance can be used, with fluoroscopic guidance reported sensitivity being 62%, which is higher than bronchial washings (24%) or brushing (27%) [115]. Complications occur in 2.2% of patients, with pneumothorax occurring in 1.7% of patients [118].

3.5. Endobronchial Ultrasound Guided Transbronchial Techniques

Because of the limitations of flexible bronchoscopy and the development of ultrasound probes, endobronchial ultrasound‐guided transbronchial FNA and FNB were developed. Endobronchial ultrasound‐guided transbronchial FNA and FNB can be used for both cancer diagnosis and staging. EBUS‐transbronchial biopsy can sample central lung cancers, mediastinal and hilar lymph nodes as well as peripheral cancers. The ultrasound probe used depends on the location of the target lesion. Two types of ultrasound probes are available. The convex probe is used for sampling of central lung nodules and mediastinal or hilar lymph nodes. Radial probes are used to sample peripheral lung lesions. The ultrasound probes have different characteristics. The convex probe has an angle of view of 90° while the radial probe has an angle view of 360°. The convex probe has a depth of penetration of greater than 5 cm while the radial probe has a more limited depth of penetration of between 4 and 5 cm. Resolution is superior for the convex probe. Finally, in most cases the radial probe cannot sample in real time while the convex probe can [119]. Table 6 lists the characteristics of the convex and radial probes.

TABLE 6.

Characteristics of ultrasound probes used for biopsy of pulmonary lesions.

	Convex	Radial
Angle of view	90°	360°
Field of view	110°	360°
Depth of penetration	> 5 cm	4–5 cm
Resolution	Good	Fair
Real‐time Sampling Availability	Yes	No

Note: Modified from Reference Avasavala et al. [119].

The tissue sampling of peripheral pulmonary nodules may be challenging when traditional EBUS‐transbronchial sampling cannot reliably reach the lesion. A number of methods have been used to extend the peripheral reach of traditional EBUS using a convex probe. Among these are radial probe EBUS (rEBUS), guide sheath rEBUS, electromagnetic navigation, and robotic‐assisted bronchoscopy.

3.5.1. Endobronchial Ultrasound‐Guided Transbronchial Fine‐Needle Aspiration and Fine‐ Needle Biopsy

Endobronchial ultrasound‐guided transbronchial sampling can use one of three biopsy instruments: FNA, fine‐needle biopsy (FNB), or cryobiopsy probe [120, 121]. FNA and FNB have been the most frequently used methods, but cryobiopsy is gaining popularity. Diagnostic accuracy of EBUS TBNA using a convex probe for lung nodules has been reported as having a sensitivity of 66.7% and a specificity of 82% [121]. Other authors have shown that EBUS‐TBNA obtains a tissue diagnosis in a high percentage of cases [122]. The impact of needle gauge on diagnostic accuracy has been variable. Jeyabulan et al. [123] found that 21 g needles outperformed 22 g needles. Nakajima et al. [124] and Yarmus et al. [125] found no difference in diagnostic yield. EBUS‐TBNA can be used for both central and peripheral lung nodules using different ultrasound probes or guide sheaths [126]. EBUS‐TBNA has a reported diagnostic rate of 79% for peripheral lesions when appropriate ultrasound probes or guide sheaths are used [127]. EBUS‐TBNA is associated with a number of complications which impact approximately 1.36% of patients [128]. Most common of these is hemorrhage, which occurs in 0.2%–0.68% of patients. Significant pneumothorax occurs in 0.53% of patients, and infection and 0.22% of patients [128].

Patients with high stage non‐small cell carcinomas may benefit from adjuvant chemotherapy [129]. Accurate nodal staging is important. Traditionally, positron emission tomography (PET) has been used to evaluate mediastinal and hilar lymph nodes [130]. EBUS‐TBNA and mediastinoscopy can be used to investigate mediastinal lymph nodes with EBUS‐TBNA being associated with lower risks and costs than mediastinoscopy [131]. EBUS‐TBNA shows similar accuracy to mediastinoscopy and can replace mediastinoscopy in evaluation of mediastinal lymph nodes [132]. EBUS‐TBNA can accurately separate clinical N0 from clinical N1 patients [133]. The overall false‐negative rate for EBUS‐TBNA evaluation of mediastinal lymph nodes was 5.5% with a sensitivity of 60% and a specificity of 100% [133]. The associated negative predictive value was 93.4%. EBUS‐TBNA has shown a sensitivity of 90% and the specificity of 100% [134]. EBUS‐TBNA identified four of six cases with false negative PET/CT scan findings [134]. EBUS‐TBNA appears accurate for identification of positive mediastinal lymph with an accuracy of 93.5% and a negative predictive value of 92.6% [135]. From a study of over 2000 patients, it appears that PET‐CT positivity alone was nonspecific for malignancy and insufficient for guiding management [136]. EBUS‐TBNA has been shown to improve diagnostic accuracy for evaluating mediastinal lymph nodes and clinical staging [137]. EBUS‐TBNA identifies additional lymph nodes as positive in 7.4% of PET negative nodes [137]. In patients having non‐small‐cell lung cancer with a radiologically normal mediastinum, EBUS‐TBNA will disclose positive lymph nodes in 12.8% of cases [138]. Despite high diagnostic accuracy of EBUS‐TBNA for evaluation of lymph node, some authors have recommended the performance of mediastinoscopy to confirm benign results by EBUS‐TBNA [139].

3.5.2. Endobronchial Ultrasound‐Guided Transbronchial Fine‐Needle Biopsy

Following the successful utilization of FNB for diagnosis of pancreatic lesions, FNB needles were applied to the investigation of pulmonary lesions and mediastinal lymph nodes. Diagnostic accuracy and safety have been good using 22G and 25G needles [140]. Two types of needles are available for FNB represented by “end‐cut” and “side‐cut” needles. The Franseen‐Acquire needle (Figure 1) along with the SharkCore needle are the two most commonly used “end‐cut” needles. The ProCore and the Temno (Merit Medical, South Jordan, UT, USA) are the most commonly used “side‐cut” needles. Use of the Franseen/Acquire needle demonstrated a diagnostic rate of 92.1% and provided sufficient material for ancillary testing [141]. In a study by Kramer et al. [142], the Acquire 22G TBNB provided improved quality tissue specimens over those obtained by traditional FNA needles. Lanfranchi et al. [143] showed that 25G FNB needles had a higher diagnostic accuracy and yield superior material to 22G needles, but the difference did not achieve statistical significance. “Side‐cut” needles are characterized by a slant tip and a cutting slot at the side. Few studies have directly compared the “end‐cut” and “side‐cut” designs. Yip SWY et al. [144] compared classic FNA needles, Franseen needles and Temno side‐cut needles. Technical success rates were 88%, 95% and 89% for the needles respectively. Diagnostic accuracy was lower for FNA needles (75%) than with FNB needles (Franseen (100%) and Temno (96%)). Based on their study, Yip SWY et al. [144] concluded that the “end‐cut” style needle was superior to the other needle types. Gangadharan et al. [145] confirmed the data reported by Yip SWY et al. [144] These authors stated that the Franseen (end‐cut) needle appeared superior for the identification of malignancy in lymph nodes in comparison to the other types of needles [145].

3.5.3. Radial Ultrasound Probe Guided Tissue Acquisition

The radial probe (rEBUS‐D‐TB lung biopsy) allows more peripheral sampling by FNA, FNB, bronchial brushing, or forceps biopsy. The radial probe can be extended more distally than that possible using the conventional convex probe [146]. The diagnostic yield of rEBUS‐D‐TBLB was 67% but depends on lesion size and the bronchopulmonary segment involved [146, 147, 148, 149, 150, 151]. The procedure is performed by inserting a small ultrasound probe through the working channel of a flexible bronchoscope with or without a guide sheath. Imaging allows accurate determination of the lesion's exact position. After the position of the nodule is documented, the radial probe is withdrawn from the guide sheath and sampling is performed by FNA, FNB, brushing, or forceps biopsy [147]. The combining of radial probe EBUS with a guide sheath may improve accuracy and adequacy of sampling [148, 149, 150, 151]. The use of a guide sheath appears to improve the accuracy of sampling when combined with radial probe guidance [148, 149, 150, 151].

3.5.4. Virtual Bronchoscopic Guide Sheath Navigation With Fluoroscopic Guidance

In some cases, peripheral pulmonary lesions detected on CT cannot be located by EBUS‐ guide sheath techniques alone. In these cases, accurate sampling can be achieved by using CT to plan the bronchial route by review of CT images before initiation of bronchoscopy. Virtual bronchoscopic navigation systems exist to navigate a guide sheath through the bronchial system [151]. Flexible bronchoscopy is utilized to insert the guide sheath into the appropriate bronchus and the sheath and biopsy instrument are moved to the lesion under fluoroscopic guidance [151]. This method improves specimen adequacy and diagnostic accuracy when rEBUS‐GS could not localize the peripheral pulmonary lesion [151].

3.5.5. Endoscopic Ultrasound‐Guided FNA and FNB

In certain cases, endoscopic ultrasound‐guided transesophageal FNA (EUS‐TEFNA) can be used for the investigation of mediastinal lymph nodes and central lung nodules [152, 153, 154]. The EUS‐TEFNA method offers some advantages over the EBUS‐TBFNA approach in that the needle will cross less lung tissue when sampling mediastinal lymph nodes and some central lung nodules. This can be important in patients with poor lung function [152]. In a study comparing EUSFNB and EUSFNA, sensitivity and diagnostic accuracy were better for EUSFNB versus EUSFNA with sensitivities being 100% versus 78.7% and diagnostic accuracy being 100% versus 78.6% for the diagnosis of peri‐esophageal lung lesions [152]. Others have reported similar success for both lung lesions and mediastinal lymph nodes [155, 156, 157]. Nasir et al. [155] demonstrated that the accuracy and sensitivity of EUS‐FNA were both 94.5% [155]. Giri et al. [157] demonstrated that EUS FNA using a transesophageal route was associated with a 95.4% sample adequacy and diagnostic accuracy of 93.4%. Complications were observed in 0.7% of procedures.

Endoscopic ultrasound with bronchoscope guided FNA (EUS‐B‐FNA) can be used in patients with poor respiratory function and appears to improve diagnostic yield for lung cancer to 100% without adverse outcomes [158]. Combining EBUS‐TBNA and EUS‐B‐FNA may further improve diagnostic accuracy and specimen adequacy [159].

3.5.6. Electromagnetic Navigational Bronchoscopy

Peripheral pulmonary nodules are difficult to sample by the usual EBUS bronchoscopy techniques and alternate guidance methods have been developed to sample these lesions. Among these techniques is electromagnetic navigation bronchoscopy (ENB) which can accurately sample peripheral pulmonary nodules [160]. The technique combines navigational assistance with sheath‐needle guidance to localize the lesion and sample it by forceps biopsy, FNA, or FNB. The system is complex and requires virtual bronchoscopy planning software, a “location board” which emits low frequency electromagnetic pulses, a guide sheath, a steerable catheter, and a “locatable guide” which allows tracking of the guide and sampling instrument [160]. Initial steps begin with mapping of the pathway through the bronchi to the targeted nodule. This initial pathway planning uses CT scans (using prespecified slice thickness and image overlap) to plan a pathway along the bronchi. The combined CT views allow a three‐dimensional image to be constructed facilitating the bronchoscopist to navigate the sheath and needle along the mapped path. Electromagnetic guidance begins using an electromagnetic sensor incorporated into the tip of the steerable navigation catheter [160]. The steerable navigation catheter is inserted into the guide catheter and both are inserted into the bronchoscope's working channel. When the bronchoscope can proceed no further distally, the steerable navigation and guide catheters are advanced and steered by manipulation of the steerable catheter's handle. The catheter's position and direction in relation to the target are accessed by comparing the relative position of the sensor to the CT road map. At each bronchial bifurcation, the steerable catheter and associated guide catheter are directed by manipulation of the steerable catheter's handle to flex the catheter tip into the appropriate bronchus toward the target. Once the target is reached, the guide catheter is fixed in place and the steerable catheter withdrawn. The target is then sampled using forceps, FNA, or FNB needles. The sampling instrument is withdrawn and material is processed. Because the guide catheter is locked in place, multiple samples can be taken.

Electromagnetic navigational biopsy (ENB) appears to be the most expensive and requires specialized equipment and training but is an accurate technology for obtaining tissue from peripheral lung nodules [161, 162, 163, 164]. Qi et al. [162] compared R‐EBUS to electromagnetic navigation bronchoscopy for the diagnosis of peripheral pulmonary nodules. This study demonstrated similar diagnostic yields of 74.6% (R‐EBUS) and 78.2% (ENB) for the two methods. Gildea et al. [165] confirmed the findings of Qi et al. [162] Eberhardt et al. [166] compared EBUS, ENB, and the combination of the two and showed that the diagnostic yield of the combination (88%) was superior to either method alone: EBUS (69%) and ENB (59%). The ENB technique is especially helpful for the diagnosis of peripheral lung lesions in areas that cannot be sampled by conventional methods, with ENB obtaining a diagnosis in 77% of patients [167].

Recent large studies have shown excellent success with ENB. In a multi‐center study of 1157 cases, ENB was successful in 1092 (94.4%) cases in obtaining tissue [168]. A malignancy was identified in 44.3% and a negative diagnosis in 608 of 1092 cases (55.7%). One year follow‐up showed that 270 of the original 608 negative diagnoses were false negatives (36%). The 12‐month diagnostic yield was calculated as 72.9% (true positives + true negatives/total cases). Molecular analysis was attempted in 30.7% of cases with adenocarcinoma. Tissue was adequate for complete testing in 75 of 87 cases (86.2%) [168]. Similar findings were observed at a two‐year interval [169].

ENB was compared to CT guided percutaneous sampling for peripheral indeterminate pulmonary nodules. CT guided sampling was more often diagnostic than ENB 86.0% vs. 66.0% [170]. Yield for molecular testing was achieved by the two approaches 88.9% versus 82%. The two groups had similar rates of major complications (hemorrhage or pneumothorax) [170].

Other smaller studies have shown similar results. Oh et al. [171] in a study of 30 patients undergoing ENB demonstrated that 29 of 30 (96.6%) had adequate material for review. The diagnostic accuracy was 68.0% (17/25). Only one patient had a pneumothorax, and no clinically significant hemorrhages occurred. Importantly, of the 13 patients' material submitted for molecular analysis, 12 of 13 (92.3%) had successful molecular testing.

In a large multi‐center‐based proposal for guidelines formulated by a panel of experts two observations were made [172]. First, the guided bronchoscope approach is safer than a percutaneous approach. A second observation was that guided bronchoscopic sampling of a parenchymal lung lesion was comparable to percutaneous biopsy for adequacy of material for comprehensive biomarker testing. Studies have shown that ENB is a safe and accurate technique [173, 174].

3.5.7. Robotic‐Assisted Bronchoscopy

Robotic‐assisted bronchoscopy (RAB) uses multiple thin slice CT images to produce a “road map” to allow navigation of a guide sheath to the pulmonary nodule. The guide sheath contains a miniature camera which allows the bronchoscopist to navigate the sheath to the pulmonary nodule. After the guide has been firmly fixed against the lesion, the camera probe is withdrawn and the sampling device (brush, forceps, FNA, FNB, and in some cases the cryoprobe) is introduced into the sheath and sampling performed.

3.6. Biopsy Devices for Lung Nodules

Most of the guidance techniques described above can be paired with forceps biopsy, brushing techniques, FNA, and FNB. Cryobiopsy, which often provides a larger sample size, can be paired with EBUS ENB guide sheath methods and robotic‐assisted bronchoscopy.

3.6.1. Transbronchial Cryobiopsy

Because of the increasing need for adequate material for comprehensive molecular testing and immunocytochemical studies, alternate sampling techniques have been investigated. Techniques resulting in small tissue specimens with retained histologic architecture and increased cross‐sectional area were developed, including cryobiopsy. Cryo techniques have a long history of use, including cryoablation of lung tumors. Cryobiopsy takes a relatively large specimen allowing histologic evaluation of architecture and has been used for evaluation of non‐neoplastic pulmonary lesions [175].

Transbronchial cryobiopsy has been used for investigating diffuse parenchymal lung disease [176]. The method appears to be able to accurately diagnose specific types of diffuse parenchymal lung disease in 70%–80% of cases [176]. A major advantage of cryobiopsy is that it may provide a larger tissue specimen with better preservation of tissue architecture than other techniques such as bronchial washings, FNA, or FNB. While originally developed for diagnosis of diffuse parenchymal disease, it can also investigate discrete pulmonary nodules [177].

The technique uses a long flexible insulated catheter attached to a blunt metal tip [175]. The metal tip can be rapidly cooled to very low temperatures using the Joule‐Thompson effect which can reduce the metal tip to a temperature of −79°C using either carbon dioxide or nitrous oxide in a few seconds. This rapid cooling of the probe tip allows cryoadhesion of tissue to the probe tip. The adherent tissue is then mechanically avulsed, the catheter and probe are withdrawn through the bronchoscope and prepared for examination and processing. The cryobiopsy catheter and tip are usually guided to the peripheral lesion using CT generated “road maps” and electromagnetic navigational methods.

For optimal treatment of patients with lung cancer, the testing of multiple molecular markers is required including EGFR, KRAS, ALK, and immunohistochemical testing for PD‐L1. Thus, adequate tissue for testing all these markers is required. Tissue adequacy is in part determined by cross‐sectional area (see Figure 2 for comparative cross‐sectional areas). Several authors [178, 179, 180, 181, 182] have compared medium cross‐sectional areas of specimens from transbronchial forceps biopsies to cryobiopsies. Table 7 documents the median size of the biopsy sample for each technique. In the author's experience, cryobiopsy does yield larger specimens than forceps biopsy, but less than core needle biopsy, similar to that reported by Chung et al. [185], Ehab et al. [181] and Akthas et al. [182]. This data demonstrates that cryobiopsy produced biopsies of greater size than transbronchial forceps biopsies [178, 179, 180, 181, 182]. The increased size of the cryobiopsies yielded superior samples for PD‐L1 testing [180]. Cryobiopsy yielded samples with significantly more DNA and RNA than transbronchial forceps biopsy [179]. Cryobiopsy is associated with significantly fewer specimens inadequate for molecular testing than transbronchial forceps biopsy [181].

FIGURE 2.

Biopsy tissue samples shown for size comparison; all were photographed at 20× and H + E stained. (A) 20 g core biopsy specimen. (B) Forceps biopsy sample. (C) Acquire biopsy specimen. (D) Cryobiopsy specimen. [Color figure can be viewed at wileyonlinelibrary.com]

TABLE 7.

Characteristics of specimens obtained by common techniques for sampling pulmonary nodules.

Technique	Diagnostic yield	Diagnostic accuracy (Sen, Spec)	Mean sample area a	Adequacy rate for molecular testing
Percutaneous core biopsy	85%–96% [183, 184]	(92%, 100%) [183, 184]	4.5–18 mm2 b	89% [183, 184]
Percutaneous FNA c	63%–67% [183, 184]	(86%, 100%) [183, 184]	1.5–2.5 mm2 c	85% [185]
EBUS FNA	Not reported	(91.2%, 100%) [186]	1.5–2.5 mm2 c	64.0%–95.3% [186, 187, 188, 189, 190, 191]
Electromagnetic navigational biopsy	51.4% [187]–96.6% [193]	68% [192]–79% [193]	2.0 mm2	92% [191]
Forceps biopsy	44.4% [188]–72% [194]	47.5%–80% [178, 179, 180, 181, 182]	2.0 mm2	81.5% [178, 179, 180, 181, 182]
Cryobiopsy	77% [194]–84.4% [185, 193]	Sensitivity is 92%–96% [195]	3.0 mm2	99.1% [185]

^a Mean sample size based on measurements of specimens obtained at the University of Missouri.

^b Sample size depends on needle gauge. Numbers given are for 20 and 18 g needles.

^c Average aggregate size for cells in cell block excluding red blood cells.

While cryobiopsy has the advantages of larger sample size, better preservation of histologic architecture, and less crush artifact, it does have some limitations. It appears that pneumothorax is absent with cryobiopsy [181] but moderate bleeding has been reported in 36.6% of patients [182].

3.7. Summary of Guidance and Biopsy Methods for Pulmonary Nodules and Mediastinal Lymph Nodes

• Imaging guidance techniques differ for central and peripheral lesions.

• Core needle biopsy, fine‐needle biopsy, and fine‐needle aspiration have largely replaced wedge resection for investigation of lung lesions

• EBUS‐TBNA and EBUS‐TBNB are the primary biopsy methods for mediastinal lymph nodes and central lung nodules.

• EBUS‐TBNB has slightly higher diagnostic yield and diagnostic accuracy than EBUS TBNA.

• Sampling of hilar and mediastinal lymph nodes by EBUS TBNA or EBUS‐TBNB is an important component of the workup of a lung nodule

• EBUS TBNA and EBUS‐TBNB have diagnostic yields and accuracies similar to that of mediastinoscopy for sampling of mediastinal and hilar lymph nodes.

• Peripheral lung nodules historically have been investigated by transbronchial/percutaneous core needle biopsy or FNA, but concerns over rates of pneumothorax and air embolism have reduced the use of these methods.

• Currently, peripheral lung nodules are sampled by either electromagnetic navigation or radial ultrasound probe guidance with forceps biopsy, bronchial brushing, or guide sheath directed FNA.

• Electromagnetic navigation biopsy and radial probe biopsy had diagnostic sensitivities of about 70%.

• EUS‐transesophageal FNA and FNB have good diagnostic accuracy for central lesions and lymph nodes and appear safer than traditional EBUS TBFNA or EBUS‐TBNB in patients with poor respiratory function.

3.8. Summary and Conclusions for Sampling of Pulmonary Nodules

Selection of the optimal method for tissue sampling of a pulmonary nodule depends on analysis of a number of issues including patient safety, patient morbidity, accuracy of sampling method, availability of adequate material for ancillary testing including molecular testing, and cost of the method used. Balancing these factors requires careful consideration of the technique's strengths and weaknesses in light of the patient's clinical condition and the availability of experienced radiologists, pulmonologists, and pathologists.

Initial techniques were surgical and included thoracotomy with wedge resection and VATS. Both of these techniques yield large volume samples but are associated with variable degrees of morbidity and are not minimally invasive procedures. These surgical techniques are associated with longer post‐procedural recovery times and are more costly.

The development of the flexible bronchoscope and its later combination with ultrasound guidance largely replaced open wedge biopsy and VATS. Direct flexible bronchoscopy with visual guidance was largely limited to visible bronchial lesions. Biopsy methods included forceps biopsy, bronchial brushing and washings, as well as fine needle aspiration. The use of fluoroscopic or CT guidance could extend the sampling distance somewhat. Diagnostic yield and accuracy were good, and forceps biopsies yielded specimens of larger size than cell blocks obtained by FNA or brushing methods (Table 7). Unfortunately, direct fiber optic sampling could not reliably sample hilar or mediastinal lymph nodes or peripheral pulmonary nodules, leading to its limited use in modern bronchoscopy.

Percutaneous/transthoracic FNA and core needle biopsy techniques are highly accurate with sensitivities and specificities being 92% and 100% respectively for core needle biopsy and 86% and 100% for FNA [183]. Percutaneous core biopsy yields specimens adequate for molecular testing in 89% of cases (Table 7) [183, 184]. Percutaneous core biopsy has a diagnostic yield superior to that of cryobiopsy but inferior adequacy rates for molecular testing (89% [183, 184] vs. 99% [185]).

Major complication rates (pneumothorax requiring chest tube or symptomatic hemorrhage) were similar for both core needle biopsy and cryobiopsy (1.3% [186, 187] vs. 1.8% [188]) but minor pneumothorax is more common with percutaneous biopsies (19.1% [186]).

Ultrasound guided FNA is able to sample hilar and mediastinal lymph nodes as well as central lung nodules. EBUS transbronchial sampling when used with FNA, diagnostic yield and accuracy are good [189]. Specimen size is more limited than with other methods (forceps biopsy, core biopsy or cryobiopsy) but studies [190, 191, 192, 193] have shown that sufficient material for ancillary immunohistochemical studies (TTF‐1, P40, PD‐L1) and molecular testing is obtained in many cases (Table 7) [190, 191, 192, 193].

Electromagnetic navigational biopsy has good diagnostic yield ranging from 44.4% to 96.6% depending on biopsy instrument [187, 192, 194]. The method (using forceps) has a 92% adequacy rate for molecular testing [191]. This adequacy rate is inferior to that achievable for cryobiopsy [185] but similar to that seen with percutaneous core needle biopsy [183, 184]. The cost of electromagnetic navigational biopsy is high, ranging from $7000 to $15,000 based on University of Missouri website for patient information (Table 8).

TABLE 8.

Costs for pulmonary biopsy techniques based on University of Missouri website estimator.

Percutaneous core biopsy	$1850–$4800
EBUS FNA	$5000–$8000
Electromagnetic navigational biopsy	$7000–$15,000
Cryobiopsy	$4500–$9000

Cryobiopsy has good diagnostic yield ranging from 77% [188] to 84.4% [185] and is associated with a sensitivity for malignancy of up to 96% [194]. Cryobiopsy has an excellent rate of adequacy for successful molecular testing reported to be as high as 99.1% [185, 194]. The rate of clinically significant complications is low with reported rates of 1.8% [195].

The choice of optimal biopsy technique involves multiple factors including patient safety, diagnostic accuracy, and adequacy of material for molecular testing. Currently, based on diagnostic accuracy and the rate of specimen adequacy for molecular testing, three guidance and biopsy techniques currently dominate the sampling of pulmonary nodules. Three sampling methods are percutaneous core needle biopsy, electromagnetic navigation biopsy, and cryobiopsy. These techniques provide sufficient material for molecular testing in at least 90% of cases [183, 184, 185, 188, 193, 194, 195]. Clinically significant complications occur at similar rates of approximately 1.3%–1.8% [186, 192, 195] making these techniques clinically viable. The comparison of cost associated with each technique impacts selection of the method used. ENB is the most expensive of the methods, with cost estimates ranging between $7000 and $15,000 (Table 8), and percutaneous core biopsy the least expensive, with the cost estimate of between $1850 and $4800 (Table 8). Based on the combination of procedure cost, diagnostic accuracy, and adequacy for molecular testing, percutaneous core needle biopsy and cryobiopsy appear to be the most advantageous. Both these diagnostic techniques depend on histologic rather than cytologic specimen evaluation. Further comparison studies will be necessary to confirm the optimal method.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.