CONTENTS
13.1 Introduction 185
13.2 Indications and Contraindications 185 13.3 Alternative Diagnostic Tests 186 13.4 Prebiopsy Procedure 187 13.5 Biopsy Prucedure 187 13.5.1 Choice of Image Guidance 187 13.5.2 Patient Positioning 188 13.5.3 Sedation 189
13.5.4 Breathing Instructions 189
13.5.5 Preparation for Needle Insertion 190 13.5.6 Choice of Biopsy Needle 191 13.5.7 Choice of Syringe 191 13.5.8 Targeting the Nodule 191 13.5.9 CT Scan Parameters 194
13.5.10 Documentation of Needle Tip Location 194 13.5.11 Angled Approach 195
13.6 Complications 196 13.7 Post Biopsy Routine 198 13.8 Results 199
13.9 Cytology 200 13.10 Conclusion 202
References 202
13.1 Introduction
Transthoracic needle bi op sy (TNB) has become a widely used technique to evaluate pulmonary nod- ules. In part, the increased use of this technique has come about through the emergence of improved image guidance with CT scanning and most recently with CT fl uoroscopy. Performance of this procedure
13 Transthoracic Needle Bi op sy of Lung Nodules
D. F. Yankelevitz, D. Shaham, M. Vazquez, C. I. Henschke
D. F. Yankelevitz, MD, Professor of Radiology;
M. Vazquez, MD, Associate Professor of Clinical Pathology;
C. I. Henschke, PhD, MD, Professor of Radiology;
Weill Medical College of Cornell Uni ver si ty, 525 East 68th Street, New York, New York 10021, USA
D. Shaham, MD
Lecturer in Radiology, Hadassah University Hospital, Kiryat Hadassah, Jerusalem 91120, Israel
requires the skills of the radiologist to obtain the sample and the skills of a pathologist to interpret the sample provided. The performance of the procedure starts with careful consideration of the indications and contraindications, followed by choice of image guidance, technical considerations in performance of the procedure, risk reduction techniques and fi nal ly, interpretation of the specimen along with con sid er - ations for appropriate follow up. In this chapter we will discuss each of the above considerations in the chain of events that are involved in the per for mance of this procedure. As this chapter focuses on biopsy of pulmonary nodules we will not discuss the related topics of mediastinal or hilar biopsy.
13.2 Indications and Contraindications
The indication for performance of this procedure is
the presence of a pulmonary nodule. This can either
be the presence of a solitary nodule that is indeter-
minate or the presence of multiple nodules where
metastatic disease is con sid ered. The procedure can
be used to confi rm either malignant or benign condi-
tions. In terms of nodule size and location there are
no absolute con train di ca tions to the performance of
the procedure. It is clear however that both of these
factors can infl uence the technical diffi culty in the
performance of the pro ce dure. These factors must
be weighed by the ra di ol o gist in terms of his/her
own skill level in deciding whether to perform the
procedure. Con train di ca tions to the performance
of TNB are for the most part only relative. The
only absolute con train di ca tion is an uncooperative
patient (Lalli et al. 1978). In cases where a patient
is unable to remain still or follow instructions, the
procedure cannot be per formed. In special situa-
tions, where a diagnosis is considered imperative, the
procedure could be per formed while the patient is
sedated. The other relative contraindications include
the following:
1. Bleeding diathesis with international normalized ratio (INR) >1.3 or platelet count <50,000: When necessary, these can be corrected with trans fu sion even on an emergent basis (Klein and Zarka 2000).
2. Severe pulmonary dysfunction including em phy - se ma or bullous disease: In these cases, careful selection of biopsy path to minimize crossing of severely damaged lung is helpful. Ultimately, these underlying conditions increase the risk of pneumothorax; therefore it is necessary to apply appropriate risk reduction techniques and be fa mil iar with treatment of complications.
3. Contralateral pneumonectomy: The risk in this situation is the development of pnuemothorax in the remaining lung. Again careful attention to risk reduction and treatment when necessary are the main concerns.
4. Suspicion of hydatid cyst: Rupture of a hydatid cyst can cause widespread dissemination within the lung and pleural space (Stampfel 1982).
5. Diffi culty in positioning: On occasion patients may have diffi culty maintaining a position that allows the easiest access to the nodule. In general the shortest path to the nodule is chosen for the procedure, however on occasion the position may be uncomfortable and it may not be possible to maintain for the duration of the procedure. Care- ful attention to patient comfort prior to the start of the procedure, and once positioned per form ing the procedure as quickly as possible helps to alle- viate this problem. Occasionally patients must be positioned so that they are more comfortable even though this necessitates an alternative path for the needle to travel that may be longer and more dif- fi cult than the original.
6. Medications with anticoagulant effects: These in clude coumadin, aspirin and non steroidal anti-infl ammatory agents. These should generally be discontinued fi ve days prior to the procedure, in order to avoid excessive bleeding into the lung (Hirsh et al. 1989).
13.3 Alternative Diagnostic Tests
The choice of workup for pulmonary nodules can be quite challenging. It requires an understanding of the complex relationship of several probabili- ties. In the case of solitary nodules, this requires an un der stand ing of what the probability of malignancy
is for a given nodule based on the initial evaluation of radiologic and clinical fi ndings, as well as an un der - stand ing of how this initial assessment can be altered based on the results of additional diagnostic tests.
Additional considerations in regard to the workup include the availability of equipment and personnel, costs for various procedures, skill of the operator and risk of complications. These factors are quite com plex and it seems likely that there is no single par a digm for the work up of a nodule that is appropriate in all situations. Workup will vary from case to case and even different approaches for an identical nod ule may be necessary at different institutions.
Several diagnostic tests are often considered in the workup of pulmonary nodules. They range from totally non-invasive to minimally invasive to in va sive.
Under the category of non-invasive, the com mon ly considered diagnostic alternatives include con trast enhanced CT scanning and PET scanning. There is a large body of literature describing these tests.
However, the literature can be confusing with wide ranges reported for their respective sen si tiv i ties and specifi cities. Some general conclusions about these two approaches can be made. They both have dimin- ished accuracy for sub-centimeter nod ules. Both are prone to be false positive with active infl ammatory processes, and both tend to be false negative with low-grade cancers. Nevertheless, both of these pro- cedures continue to be improved and are active areas of research and development. Newer high-resolution multi-detector CT scanners should impact both of these tests and allow for more ac cu rate diagnosis of smaller nodules. In the case of nod ule enhance- ment, it will allow for higher resolution images of the nodule to be obtained quickly, and for PET scanning, mapping the PET images to high res o lu tion images obtained with CT may allow for ad di tion al attenua- tion correction.
Under the category of minimally invasive, bron- cho s co py is often considered as an alternative to TNB. However, the published reports of the accuracy of bronchoscopy are nearly uniformly lower than that of TNB. Bronchoscopy is generally limited to those cases where there is a positive bronchus sign.
This implies that a bronchus is seen leading into the nodule. However, even with this favorable sign present, the procedure is still less accurate than TNB (Yankelevitz et al. 2000c).
Invasive procedures include thoracoscopic biopsy
and thoracotomy. The obvious advantage of these
procedures is that they obtain a larger amount of tis-
sue. However, as a diagnostic procedure, tho ra c ot o my
should generally be avoided. This is a major sur gi cal
procedure with attendant risks and morbidity associ- ated with it (Yankelevitz et al. 2000c). When neces- sary, thoracoscopy generally suffi ces to yield enough tissue and can be converted into a tho ra c ot o my if necessary. While claims of one hundred per cent sensi- tivity and specifi city have been attributed to thoracos- copy, the increasing use of this procedure to diagnose small nodules that are diffi cult for the surgeon to pal- pate have led to revision of this belief (Fig. 13.1).
One of the newer approaches being used to eval u ate small nodules relies on estimating their growth rates using serial CT scans (Yankelevitz et al. 2000c). This technique leverages the unique ability of CT scan ning to make accurate measurements. Once the volume can be accurately measured and the time be tween scans is known, it is relatively straight forward to estimate dou- bling times (Fig. 13.2). This technique is still relatively new and more work needs to be done to fully integrate its role into the di ag nos tic pro cess.
13.4 Prebiopsy Procedure
The majority of TNB’s are performed on an out pa tient basis. The ideal situation is to have the fi lms reviewed by the radiologist prior to scheduling the case. Once it is decided that the procedure is war rant ed, the patient can be placed on the schedule. The patient should be informed about what to expect prior to coming for the procedure. This includes a discussion of the risks and benefi ts, eating in struc tions, medications to either be continued or tem po rari ly be discontinued, and to
be prepared for the possibility of staying overnight in the event of com pli ca tions. It is also helpful to have written material that can be sent to the patient that contains in for ma tion regarding the procedure so that they can have any questions or concerns answered for them before they come. Laboratory tests are gener- ally required prior to the procedure to confi rm that there are no serious problems related to bleeding. The pathology department is made aware of the biopsy schedule in advance so that they are ready to evalu- ate the spec i men. In addition, many institutions are now re quir ing documentation of a recent history and physical examination. This can be obtained from the patients referring physician and must be recent. The ra di ol o gist generally obtains informed consent imme- diately before the procedure begins. The elements of in formed consent include a discussion of the risks and benefi ts, including details about the frequency of pneumothorax, hemmorahage and infection. Al ter - na tives are also discussed with the patient.
13.5 Biopsy Prucedure
13.5.1 Choice of Image Guidance
The majority of TNB are now performed under CT guidance. Flouroscopic guidance is also available, but is used less frequently at present. Occasionally, ultrasound can be used for guidance, but this is gen- er al ly limited to nodules that abut the chest wall or for
Fig. 13.1. a High resolution CT image of a nodule prior to attempted thoracoscopic resection. b High resolution CT image post attempted thoracoscopic resection shows the nod ule still present adjacent to the suture line
a b
mediastinal lesions (Gupta et al. 1999). In regard to the choice of fl uoroscopy or CT, there are ad van tag es and disadvantages when compared to each oth er. It is generally faster to perform the procedure under fl uo- roscopic guidance. With bi-plane fl u o ros co py the depth of the needle in relation to the nodule can be visualized and there is the benefi t of real time performance. How- ever, CT can identify lesions that are not well identifi ed with fl uoroscopy and doc u men ta tion of the needle tip within the nodule can be done with more confi dence using CT. Ultimately, weighing the relative merits of these approaches is often done on a case-to-case basis and rests in part with the radiologists experience and preference. CT fl uoroscopy has recently been intro- duced. This tech nol o gy allows for some of the advan- tages previously only found with routine fl uoroscopy, notably real time evaluation, to now be available in the
per for mance of CT guided procedures (White et al.
2000). CT fl uoroscopy is particularly useful in those cases where the nodule is small and poorly visualized by fl uoroscopy and for nodules that are located near the diaphragm where there is considerable motion and routine real time evaluation is helpful.
13.5.2 Patient Positioning
Patient positioning is generally done to allow for the most direct route to the nodule. There are however numerous considerations that can alter this. The choice of positions is either supine, prone or de cu - bi tus. Prone position is advantageous for several rea- sons, the primary one being that there is less chest
Fig. 13.2. (a) High resolution CT image of a small nodule. (b) High resolution CT image of the nodule obtained at a later time.
(c, d) Comparison of the 3-D renderings of the nodules respectively. Based on the time between scans and the change in volume, an estimate of the nodule volume doubling time can be easily performed
a b
c d
wall motion in this position. The ribs are attached to vertebral bodies in their posterior aspects and thus the posterior aspect does not move in the vertical plane when the patient breathes. Instead, it rotates in plane. This has been likened to the motion of the handles of a bucket. When the patient is supine, the anterior portions of the ribs move in the verti- cal plane and this causes motion of the needle with each breath. The decubitus position has the largest amount of motion associated with it. In this case, the dependent lung is relatively motionless and the ma jor i ty of motion occurs in the non dependent lung, which is generally the lung being biopsied. Another reason to prefer the prone position is that following the procedure the patient is instructed to lie with the biopsy site in the dependent position. This is useful in reducing the risk of developing a pneumothorax.
It is generally much easier for patients to lie on their backs for several hours post procedure than to lie prone. A fi nal reason for preferring the prone po si tion is that the patients do not have to visual- ize the needle actually entering them. While many patients simply close their eyes to avoid seeing this, for some it can be quite anxiety provoking.
Other considerations in terms of positioning re late to gaining the best access to the nodule. On oc ca sion there may be structures blocking the path of the needle. This includes bony structures such as
the scap u la or ribs as well as vascular structures such as the great vessels. Proper positioning is a major com po nent in planning the procedure and time spent on optimizing this is well worth it. The scapula can be rotated out of the way by placing the patients arm at their side and internally rotating the shoulder (Yankelev itz et al. 2000b). This generally moves the scapula laterally (Fig. 13.3). On occasion, it is help ful to place a pillow or folded sheet under the chest so as to allow the shoulder to rotate even fur ther and move the scapula laterally. This technique is also some- times helpful in spreading the ribs to allow for direct perpendicular access to the nodule with out having to advance the needle on an angle to the scan ning plane. Regarding the great vessels and clav i cle, it is often not possible to rotate these structures out of the biopsy plane. In cases where a nodule is lo cat ed near these structures, a prone approach is nec es sary, even though it may necessitate a longer path (Fig. 13.4).
13.5.3 Sedation
In general, conscious sedation is not necessary in the performance of TNB. The majority of patients are calm enough to allow for the performance of the procedure without any additional medication.
In addition, it is often desirable to have the patients cooperate with specifi c breathing instruction during the procedure. However, on occasion, after dis cus sion with the patient, it may be worthwhile to con sid er giving sedation. In those cases it is necessary to con- form to all relevant hospital policies in regard to this administration.
13.5.4 Breathing Instructions
With fl uoroscopic guidance, the degree of in spi ra tion can be monitored in real time. This allows for the relationship between the needle and the nodule to be observed and instructions are given to stop breathing when the alignment between them ap pears correct.
Once this occurs, the needle is ad vanced towards the nodule with continuous ob ser va tion. With CT guidance, except in the case of CT fl uoroscopy, real time observation is not possible, and the needle is advanced without its orientation being directly observed. Each adjustment of the nee dle is per- formed based on review of the last set of CT images.
The needle is adjusted and a new set of im ag es is
Fig. 13.3. CT image of patient in the prone position. The left arm is located at the patient’s side and is rotated internally.
This pulls the scapula laterally allowing much greater access to the lateral portion of the lung. Note the difference between the two sides. The lateral positioning of the needle allows it to be placed directly into the left upper lobe. Had the entry site been more medial, the major fi ssure (arrows) would have had to be crossed and that would have led to an increased chance for pneumothorax
then obtained. In terms of specifi c breathing instruc- tions, different instructions are given de pend ing on the situation. When nodules are located in the upper lobes and the patient can be placed in the prone posi- tion, the amount of motion in the lungs is quite small and patients are simply re quest ed to breathe gently throughout the procedure. The relationship between the advancing needle and the nodule is suffi ciently stable to allow for this ap proach. In patients where the nodule is located near the diaphragm or when the biopsy is performed with the patient in the supine position, it is often nec es sary to give specifi c breath- ing instructions. The pur pose of these instructions is to have the nodule in the identical location each time the needle is being ad vanced. In the case of CT fl uoroscopy, the advancing needle can be visualized in relation to the nodule. However, there are some important differences in comparison to routine fl uo- roscopy in that only a single axial plane is visualized and there is a slight, approximately 0.5 second, offset for the images to update. Nevertheless, this technique is quite useful in those cases where there is motion due to breathing (White et al. 2000).
13.5.5 Preparation for Needle Insertion
Once the patient has been correctly positioned and instructed, the next step is to mark the skin entry site. With fl ouroscopy, this can be done by placing a forceps into the imaging plane and observing its location relative to the nodule. Once the tip of the
forceps is in the correct location, its location at the skin surface is marked. With CT, a set of images are obtained for initial localization of the nodule. A plane is chosen for the skin entry site. This is not always the image plane that includes the nodule, as there are occasions where planes above or below must be chosen because of a structure blocking a directly perpendicular approach. Next, a set of mark ers is placed on the patient in the chosen entry plane.
These markers can be as simple as a set of blood drawing needles taped together and the purpose is to allow for localization within the scanning plane.
A new set of images is now obtained in the scanning plane, with the markers in place. The marker in the optimal location is now chosen. Once the marker is chosen, the laser light on the scanner is turned on to show the scanning plane on the skin surface and now the combination of the laser light and the chosen marker defi nes the precise entry site. The skin is marked, generally with a felt tip marker.
Once the skin entry site is chosen, the skin is cleansed with antiseptic solution and the sur- round ing area is covered with a sterile drape. Local an es the sia is now given. It is important to give suffi - cient anesthesia to make the procedure virtually pain free. This involves numbing by instillation of anes- thesia from the skin surface all the way to the parietal pleu ral surface. There are sensory nerve fi bers on the pa ri etal pleural surface, and while it is not necessary to directly insert the needle into it, infusion into ad ja cent tissue allows for anesthesia to diffuse into the pleura. Documentation of the anesthetic needle ad ja cent to the parietal pleural surface is useful, and
Fig. 13.4. a The nodule located in the apex of the right lung is closer to the anterior surface of the lung than to the pos te ri or surface.
However, both the fi rst rib and the subclavian vein blocks direct access. b The biopsy is performed in the prone position. There is nothing obstructing a direct path toward the nodule, however, the needle must travel a greater distance to reach its target
a b
the creation of a small amount of local swelling at the pleural surface caused by the anesthetic agent is helpful in determining that suffi cient anesthesia has been given. With experience, operators can actually feel when the needle has reached the parietal pleural surface while they are instilling the anesthetic.
13.5.6 Choice of Biopsy Needle
There are numerous types of biopsy needles. They come in various lengths and gauges. Broadly, how- ev er, there are two basic types of needle designs that are used. This includes the co-axial design and the single shaft (non co-axial). Choice between these two is generally a matter of preference. However, each type has certain advantages compared to the other. Using the co-axial technique, multiple samples can be obtained with a single pleural puncture. In this way, if the initial sample is insuffi cient, additional material can be obtained without the additional risk of puncturing the pleura and causing pneu motho rax.
The advantage of the single needle is that it is gener- ally of thinner gauge, as it does not require an outer cannula, and is more fl exible. The thinner gauge may be helpful in reducing risk of pneu motho rax and bleeding, while the increased fl exibility is found by some operators to be useful in guiding the needle to the correct location. Additionally, in cases where the initial sample is inadequate it may be use ful to move the needle to a very different portion of the nodule where more viable, diagnostic tissue can be obtained.
This would offset the benefi t of the co-axial approach, where the additional samples are obtained in close proximity to the original sample. However, each additional puncture of the pleural surface increases the risk of pneumothorax.
Another distinction between the types of needles is whether they are of the aspiration type or cutting type. Aspiration needles only obtain material suit able for cytologic interpretation, while cutting nee dles obtain material that allows for histologic eval u a tion.
Needles suitable for obtaining histologic ma te ri al have either a circumferential cutting tip, a side slot that acts as a receptacle, or a spring-loaded cutting edge over a side slot. The spring-loaded nee dles generally have a throw distance of one to two centi- meters. In some of the designs, this is ad just able. In general, obtaining cytologic material is suf fi cient to diagnose malignancy. However, to diagnose benign conditions, it is sometimes helpful to have histologic material. The cytologic criteria necessary to establish
benign diagnoses are less well defi ned, nevertheless, these criteria continue to be refi ned.
Yet another distinction among the needles relates to the shape of the needle tip. There are two basic types, symmetrical and beveled. Both are suffi ciently sharp so as to pass through the skin without fi rst making a small incision. The main difference relates to guidance. Symmetrical tips tend to travel in a straight line whereas beveled tips tend to travel in the direction opposite the bevel. Some have found this later effect useful in guiding or steering the nee dle.
13.5.7 Choice of Syringe
Once the needle is inserted into the nodule the next step is to aspirate material for cytologic inspec- tion. Attaching a syringe to the needle and creating neg a tive pressure accomplish this. The amount of neg a tive pressure that can be created is the same re gard less of the size of the syringe (Yankelevitz et al. 1995). Once a vacuum is created, it creates the full amount of negative pressure possible regardless of whether a 50 cc syringe or a 5 cc syringe is used.
The only dif fer ence is that it requires more effort to create this vacuum with the larger syringe. For practi- cal pur pos es, it is easy to use the smaller size syringes.
A 10 cc syringe represents a good compromise in size, as it is still relatively easy to create a vacuum, and at the same time, if a small amount of air enters the system during the aspiration, the syringe is usually large enough to still maintain suffi cient negative pres sure to effectively complete the aspiration.
13.5.8 Targeting the Nodule
Depending on the size of the nodule, it is often quite
diffi cult, if not impossible to directly align the needle
in the soft tissue so that it will be on course to di rect ly
enter the nodule. As an example, lets assume a situa-
tion involving placing a needle into a one cen ti me ter
nodule that is ten centimeters deep to the pleural sur-
face. After giving local anesthetic, the bi op sy needle
is inserted into the soft tissues. If this biopsy needle
is off target by as little as two degrees, then by the
time it has advanced the full ten cen ti me ters, it will
miss the target. For smaller nodules, the degree of
accuracy needs to be even higher. Proper alignment
of the needle with this degree of accuracy is quite
challenging by itself, however when small degrees of
patient motion are added, it becomes ap par ent that directly targeting the nodule from the soft tissues in a straight line is generally not possible (Fig. 13.5).
Therefore some mechanism of steering the needle is necessary. A technique that is useful for this purpose is often referred to as ‘bevel steering’ (Yankelevitz et al. 1996a) (Fig. 13.6). This tech nique allows for the needle tip to be repositioned without completely withdrawing it, thus avoiding re-punc tur ing the pleura. After identifying that the tip of the needle will not enter the nodule, the technique in volves the following steps
1. The needle is partially withdrawn
2. The bevel is turned so that it faces in the direction opposite to the direction that it must move 3. The needle has pressure placed on it to direct it
towards the nodule
4. The skin and soft tissues (to the extent possible) are pulled with the operators free hand to ex ag - ger ate the angle of the needle shaft towards the nodule.
Fig. 13.5. Illustration of a one-cm. nodule 10 cm. beneath the skin surface. If the needle is angled by as little as two degrees away from the line perpendicular to the center of the nodule, the needle will miss the nodule completely. This implies that for nodules that are small and deep, there needs to be a tech- nique to compensate for minor misalignments
Fig. 13.6. Bevel steering is a technique used to reposition the needle tip without completely removing the needle. It in volves several steps. a Shows needle has missed the nodule. b The needle is partially withdrawn and rotated so that the bevel faces in the direction opposite to its intended course. c Torque is applied to the needle and the soft tissues are pulled so as to redirect the needle towards the nodule. d The needle is now advanced and the amount that it has been re-directed is assessed. This process may need to be repeated several times
a
c
b
d
This approach can cause the needle tip to sig nifi - cant ly change its location (Fig. 13.7). The degree of effectiveness is infl uenced by a number of factors.
This includes the depth of the lesion and the soft tis sue thickness. When these are large, the degree to which the needle tip can be re-positioned may not be suffi cient. In addition this approach often requires
multiple attempts. The degree to which the needle is withdrawn and the amount of force placed on it can vary each time. Once expertise in using this tech- nique is mastered, it is possible to make fi ne ad just - ments to the needle tip location and this becomes indispensable when performing biopsies of small le sions (Fig. 13.8).
Fig. 13.8. a The needle tip is minimally off course for this small nodule. b The technique of bevel steering is applied and the tip has moved several mm. towards the center of the nodule
a b
Fig. 13.7. a The needle tip is just lateral to the nodule. b Using the technique of bevel steering the needle is re-po si tioned, however, the degree of adjust- ment moves the needle to the other side of the nodule. Thus with a single re-ad just ment and only partially withdrawing the needle, the tip has moved by over one cm. c The process is now repeated, ex cept that steering is now in the opposite direction and to a lesser extent. The tip is now in the center of the nodule. It is often necessary to repeat this process several times to po si tion the needle tip in the desired location
a
c b
13.5.9 CT Scan Parameters
This primarily relates to choice of mA and slice thick ness. As a general rule, the lowest dose that allows for evaluation of the needle in relation to the nodule is required. With modern scanners this in volves an mA setting of 40 or lower. Radiation dose reduction is important because it is often necessary to perform multiple images through the same tissue volume during the course of the procedure.
The slice thickness is generally chosen in relation to the size of the nodule. In order to be certain that a single CT image contains only nodule and is not a combination of lung and nodule, the slice thickness should be less than half the diameter of the nodule.
In this way contiguous slices will include at least one image that contains no partial volume effects.
A sim ple, general rule of thumb for choosing slice thickness is as follows:
1. For nodules greater than three centimeter in di am e ter a CT slice thickness of one centimeter or fi ve millimeters.
2. For nodules between one centimeter and three centimeters a slice thickness of fi ve millimeters 3. For nodules between fi ve millimeters and one
cen ti me ter a slice thickness of three millimeters 4. For nodules less than fi ve millimeters a slice
thick ness of three millimeters for guidance and one millimeter for fi nal localization.
When performing the procedure, CT images are usually obtained in sets of three contiguous images, with the central image located at the point where the needle tip is expected to be located, upon its being advanced. In this way there is an image above and below the needle tip location.
13.5.10 Documentation of Needle Tip Location
This is a critical step in proper performance of TNB, and one that is frequently left out. Perhaps the single most important cause of an inconclusive biopsy re sult is not having the needle tip located within the nodule.
There are a variety of reasons that this can occur. This includes partial volume effects where the tip actually appears to be located in the nodule when in fact it is not (Fig. 13.9). It can also occur when the tip passes through the nodule and may actually be located in a plane beyond the nodule (Fig. 13.10). There are three ways to be certain that the needle tip is actu- ally visualized (Yankelevitz and Henschke 1993) (Fig. 13.11).
1. Identifi cation of a distinctive feature of the tip such as a notch.
2. An intense shadowing artifact emanates from the tip when it is perpendicular to the scan ning plan.
3. Images above and below the tip document that no additional portion of the needle is distal to the tip.
Fig. 13.9. a Picture of a needle placed outside of a synthetic nodule (along the z-axis). b CT scan of the synthetic nodule and needle gives the appear- ance of the needle actually being within the nodule. This occurs due to partial volume effect. The CT slice thickness may be wide enough to include both a portion of the nodule as well as the needle
a
b
Localization of the tip and documenting that it is inside the nodule by obtaining three images that prove there is no partial volume effect is an essential part of the procedure (Yankelevitz et al. 1993).
This is particularly important when indeterminate ma te ri al is obtained and there is consideration of fol- low ing a non-specifi c result. Without confi dence that a genuine sample from within the nodule has been ob tained, this cannot even be a consideration.
13.5.11 Angled Approach
When possible, a direct approach with the needle directly above the nodule and traveling within the scanning plane is preferable. However, this is not always possible due to overlying bony structures or other impediments, such as large vessels, that make an angled approach necessary in order to avoid these structures. With bi-plane fl uoroscopy the relation between the nodule and needle can be observed in real time. Adjustments can be made as the needle is being advanced and its spatial orienta- tion with the nodule is continuously being assessed.
In the case of CT (unless CT fl ouroscopy is used), it is more chal leng ing and an understanding of basic geometric considerations is necessary in order to avoid missing the nodule (Yankelevitz et al.
1995). The most com mon situation where an angled approach is nec es sary is when a small nodule is being biopsied and it is located directly beneath a rib. Nodules larger than the width of the rib allow for direct access through an intercostal space. In those that are small er, the needle must travel on an angle through the intercostal space adjacent to the nodule, but the CT section that in cludes the intercostal space may not include the nod ule. Thus, the length of the needle from skin surface to the nodule is never seen on a single image. The needle passes through succes- sive CT sections until it reaches the nodule. When it be comes necessary to use this approach, it is gener- ally best to start above the rib and angle downwards.
In this way, the in ter cos tal vessels that run in a groove on the un der sur face of the ribs can be avoided. The
Fig. 13.11. CT image of a side slotted needle aligned with the CT scanning plane and inside a water bottle. The needle tip can be identifi ed with certainty when a distinctive feature, such as the notch (arrow) is visualized or when a shadowing artifact is seen extending from the tip. In this example, both of these features are present
Fig. 13.10. a The needle tip appears to be piercing the nod ule on this single image. b However, on the adjacent image, it is appar- ent that the tip extends beyond the nodule. Had a specimen been obtained at this time, it would have been in ad e quate. This demonstrates the importance of obtaining images above and below the estimated needle tip location
a b
correct angle that the needle needs to enter the skin can be determined by counting the number of con- tig u ous sections that the needle will pass through as it advances from the skin entry point until reaching the nodule. This in for ma tion, along with knowledge of the depth of the nodule relative to the skin entry point allows for estimation of the correct angle (Yan- kelev itz et al. 1995) (Fig. 13.12). When the nod ule is directly be neath the skin entry point, then a single image shows the entire length of the needle along with the nodule. When the nodule is located on a slice next to the skin entry point, and when the needle is properly angled and advanced into the nod ule the CT images would show half of the needle length in the image con tain ing the entry point and the remaining half of the needle would be visualized in the image containing the nodule. If the skin entry point were two slices over, then each image would contain one
third of the length of the needle. The higher the degree of an gu la tion, the smaller the nee dle length will appear on contiguous slices. The length of needle visualized on each successive CT image will always appear the same, and will begin on each successive image at the same level that it ended on the previous one (Yankelev itz et al. 1995) (Fig. 13.13).
Angling the CT gantry is an alternative approach to advancing the needle on an angle relative to the scanning plane. It allows for visualizing the needle and the nodule in a single plane while avoiding any overlying structures (Stern et al. 1993). In this way, the CT image plane can be made to align with the intercostal space and the nodule. The needle is then advanced directly along this plane. This approach avoids the need to focus on the relative lengths of the needle seen on successive images, although fi nding the proper angle so that the needle is parallel to the angled gantry plane during its entire course can also be problematic.
13.6 Complications
The most frequent complication related to TNB is pneumothorax (PTX). The reported rate that this occurs varies greatly, ranging from 15–60% (Klein and Zarka 2000). There are several reasons to ac count for this wide variation. With the increasing use of CT guidance for TNB, it is possible to detect very small PTX that may not even be visible on chest x-ray. Therefore, depending on the threshold of amount of PTX that is reported, the number of cases reported will vary. In addition, patient populations vary and it can be expected that patients that have a larger degree of emphysema will have a higher rate of PTX (Miller et al. 1988). The choice of needle will also affect the rate of PTX. Larger gauge needles will cause more PTX than small gauge needles (Klein and Zarka 2000). While the relationship between either overall amount of emphysema or gauge of the needle and the development of PTX is not exactly known, and there have even been reports suggesting a very limited relationship (Cox et al. 1999), it is obvi- ous that some relationship must exist and that both of these factors relate to the proportion of pa tients that develop PTX. Perhaps more important than the fre quen cy of PTX is the proportion of PTX that need to be treated. There has also been wide varia- tion in the range that PTX requiring treatment has been re port ed as well, ranging from 5%–25% with
Fig. 13.12. In this illustration, the nodule is located beneath the rib and does not allow for direct access within the scan- ning planes that contain the nodule (S-1 and S-0). In order to reach the nodule, the needle must be inserted in a different scanning plane (S+1). Needle N1 is properly angled to avoid the rib and reach the nodule. Needle N2, while entering the skin in the same CT scanning plane, is not properly angled and would miss the nodule. The length of N1 in each CT image is represented by the vertical line (I). This length is constant for each scanning plane and is equal to the distance (H) that the nodule is located beneath the rib. For N2 the length seen on each CT image is represented by (r) and thus it can be predicted that when this needle is advanced, it will miss the nodule. On image (S+1) it has length (I) only be cause it has not yet passed through the entire imaging plane
an average of 7% (Klein and Zarka 2000). Treat- ment of large PTX is generally required once the PTX reaches about 30% or the patient is de vel op ing symptoms (Moore 1997). It is generally pre ferred to treat the PTX with a one-piece self-con tained chest tube that can be inserted by the ra di ol o gist. These one-piece chest tubes are easily inserted and allow
the patients to remain am bu la to ry once they are in place. Patients are generally ad mit ted to the hospital for one night once these tubes have been inserted.
Hemorrhage is the next most frequently oc cur ring complication. It results in hemoptysis in only 5%
of patients, and is generally very limited. Patients should be made aware of the possibility of this
e
Fig. 13.13. This set of fi ve (a–e) contiguous 3 mm images shows the appearance of the needle as it progresses towards the nodule. On each successive image as the needle advances toward the nodule, the length of the needle is seen to be con- stant, except for the last image where the needle does not pass through the entire scanning plane. In each successive image, the needle starts at the same height where it ended in the previous image
a
c
b
d
oc cur ring before the procedure starts since it can be quite alarming if they are not alerted to its pos- si bil i ty and that it will stop quickly. At least to some extent every patient undergoing TNB develops some degree of hemorrhage. It may be too small an amount to be visualized with fl uoroscopy and in some cases may even be diffi cult to identify with CT. Hemoptysis is usually preceded by a cough. It is generally most se vere when large vessels, particularly arteries, are punc tured. Although quite rare, particularly with the use of small gauge needles, severe hemorrhage is the most frequent cause of death following TNB (Pro to pa pas et al. 1996).
One of the least frequent, but most severe com pli - ca tions is air embolism. This occurs when air enters the pulmonary venous system and can lead to sys tem ic air embolism. Air embolism can cause my o car dial infarction, arrhythmia, stroke and death. Once air embolism is suspected the patient should be placed in the left lateral decubitus position or in Trendelenberg position to prevent residual air in the left atrium from entering the systemic circulation. The patient should be placed on 100% oxygen and general symptomatic support should be provided. Patients may then need to be transferred to hy per bar ic oxygen units for further treatment (Klein and Zarka 2000).
Other infrequent complications include ma lig nant seeding of the biopsy track, which has a re port ed occurrence rate of .012% (Ayar et al. 1998), va s ova gal reactions (Moore 1998), and lung torsion fol low ing large pneumothorax (Fogarty and Dudek 1995).
Inadvertent puncture of the pericardium can lead to hemopericardium (Man et al. 1998).
13.7 Post Biopsy Routine
Following TNB, a routine protocol is useful for reduc- ing the risk of complications. The most common of these complications is PTX. There are several tech- niques that have been used to try and diminish its likelihood of occurrence. The most useful of these is to place the patient with the biopsy site in the depen- dent po si tion. This is presumed to be effective for the fol low ing reasons. The weight of the lung is now pressing down on the biopsy site and acts to prevent leakage of air (analogous to compressing a bleeding site). In addition, the alveolar air spaces contract in the de pen dent portion of the lung making leakage of air more diffi cult, and there is also less movement of the dependent lung during routine breathing. Gen-
erally it is suggested that the patient remain in this de pen dent position for about two hours. During this time the patient is encouraged to breathe normally and remain relatively motionless. One diffi culty with this approach is that it is often diffi cult for patients to remain prone for long periods of time. In those cases, patients are initially placed prone and allowed to turn over into a supine position once they feel un com - fort able. This is one of the reasons that prone position is preferable during the performance of the biopsy, so that the patients can lie supine afterwards.
During the immediate post biopsy period, some patients experience pleuritic type chest pain with- out actually having a PTX. In these patients the pain typ i cal ly lasts for less than one hour. It is frequently re lieved with minor analgesics.
Post biopsy radiographs are ordered to check for the development of PTX two to four hours post pro ce dure. If symptoms develop prior to this, ra dio - graphs are obtained immediately. If patients have no PTX at two hours and are feeling well, they are dis charged after being given explicit instructions re gard ing development of symptoms and what to do and who to contact. They are advised to avoid any activity that will make them breath heavily for at least 24 hours and to avoid strenuous activity for at least three days. In patients that do have a PTX noted on the two-hour radiograph, an additional ra dio graph can be obtained one hour later to assess for stability, provided the patient is asymptomatic. While there is no absolute cut off in size of a PTX that mandates treatment, it is generally initiated when the PTX reaches approximately 30% or when the patient is symptomatic, as mentioned above.
Several options exist for treatment of PTX
al though all of these involve the removal of air from
the pleural space. In patients that develop PTX while
still in the biopsy suite, aspiration of the PTX can be
performed with a small removable temporary cath-
e ter such as a two inch 18 gauge intravenous catheter
(Yankelevitz et al. 1996b). The catheter is attached
to a one-way valve and air is aspirated from the pleu-
ral space. Patients are placed on nasal oxygen during
this time and it is kept on for at least one-hour post
procedure. This procedure is nearly always suc cess ful
in initially removing the PTX and in about half of the
patients the PTX either does not recur or recurs to
a lesser extent (Fig. 13.14). In those patients where
the PTX enlarges or in those that develop it during
the recovery period and are becoming symptomatic,
chest tube insertion is necessary. While there are
many varieties of chest tubes available, the small
bore tubes attached to one-way valves are nearly
al ways suffi cient for treatment. Larger tubes, inserted by the chest surgeons are rarely needed.
Aside from instructions about limiting activity, it is advisable to contact the patient the next day to con fi rm that they are asymptomatic. Generally, the results from the pathology, once they are available, are given to the referring physician so that (s)he can discuss issues related to management and possible follow-up directly with the patient.
13.8 Results
There are three basic categories of results obtained with TNB. This includes diagnosis of malignancy, specifi c benign diagnosis and non-specifi c benign
diagnosis. For nodules above a threshold size (1–
1.5 cm, depending on experience) there is general agreement about the accuracy of TNB in diagnosing malignancy. It is generally considered to be over 90%
sensitive, with less than 1% false positive rate (Taft et al. 1980). In one series, among 218 cases for which both cytology as well as tissue diagnosis obtained by biopsy, resection or autospy were available, 11 of the cytology diagnoses were initially thought to be false positives. However, upon re-review of the surgi- cal biopsies, it was felt that they failed to sample the lesion and therefore the aspiration biopsy diagnoses were accurate with a specifi city of 100% and (Cagle et al. 1993). With smaller size, the accuracy declines.
However, this should not be viewed as a limitation of TNB in mak ing a diagnosis, rather it should be viewed as a tech ni cal challenge in placing the tip of the needle in the nodule. Assuming that this can be
Fig. 13.14. a A small right upper lobe nodule with patient in the prone position. It was felt that even though the needle would pass through the fi ssure, thus increasing the pos si bil i ty of pneumothorax (PTX), that the likelihood of success would be improved using this approach. b Needle tip is seen within the nodule. c Several minutes post procedure, a post biopsy CT image shows a moderate size PTX (arrow). d Following aspiration of the PTX using a small catheter (a portion of the catheter is seen in the soft tissue) the PTX is almost completely resolved. On follow-up CXR several hours later, only a minimal PTX could be identi- fi ed, and the patient was subsequently discharged
a b
c d
accomplished, there is no reason that the accuracy for diagnosing malignancy in small nodules should not be equiv a lent to that of large ones. There is even reason to believe that biopsy of small lesions may be more accurate once the technical factors of per- forming the biopsy are removed. In small nodules there is gen er al ly less necrosis and surrounding infl ammation and therefore the nodule is primarily composed of actual tumor cells, thus yielding more reliable specimens.
The diagnosis of specifi c benign disease is gen- er al ly more diffi cult than that of malignancy. In gen er al, this requires a larger amount of tissue for the di ag no sis. There are several categories of specifi c be nign diagnosis that can be made. This includes be nign tumors such as hamartomas, infectious nod ules such as tuberculomas, and non infectious gran u lo mas such as rheumatoid nodules. The litera- ture supporting the accuracy of making these diag- noses with TNB can be quite confusing, with results vary ing from 16% to 68% (Klein and Zarka 2000).
These widely varying results can at least in part be ex plained by three factors: sampling error, amount of specimen and pathologic interpretation. As be nign nodules are generally smaller than malignant ones, it is more likely to miss the nodule and not ob tain any representative specimen. Missing the lesion is prob a bly the single most important factor in not obtaining a diagnosis both in benign and malignant nodules (Westcott 1995). The amount of tissue that can be obtained with cutting needles compared to simple aspiration needles can be substantially in creased, and in some cases cutting needles provide suffi cient additional material to allow for specifi c benign di ag no sis. The extent to which this occurs is not precisely known and is in part related to the skill of the in ter pret ing pathologist. The skill set nec- es sary for mak ing specifi c benign diagnosis based on cytology is quite high and not always available, thus in many sit u a tions larger amounts of tissue may be necessary to allow for diagnosis.
The category of non-specifi c benign diagnosis is quite controversial. There have been numerous re ports that suggest that in the absence of a specifi c benign diagnosis, malignancy cannot be excluded and therefore a non-specifi c benign diagnosis can- not be accepted (Liptay 1999). This argument is gen- er al ly made in the context of ruling out lung cancer, with the implicit recognition that not diagnosing lung cancer early on is quite serious. The coun ter - point to this view is that non-specifi c diagnosis can be accepted as benign with the following conditions (Yankelevitz et al. 1997):
1. That there has been careful documentation of the needle tip within the nodule
2. That when possible, more than a single sample has been obtained from different portions of the nodule
3. That the specimen contain material other than simply blood, such as fi brous tissue or non-spe- cifi c infl ammatory tissue
4. That a careful plan exists for continued follow-up of the nodule to assess for change
When these conditions are met, the proportion of nodules with non-specifi c TNB results that are ul ti mate ly found to be malignant is quite low, on the or der of 5–10%.
13.9 Cytology
The interdisciplinary collaboration between the cy tol o gist and the radiologist is imperative to achieve a high diagnostic accuracy. The necessity for proper specimen preparation and immediate microscopic assessment of procured specimens cannot be over- em pha sized. It is preferable to schedule cases se quen - tial ly on the dame day, at a rate of approximately one case per hour. In this way, the same cytologist can be easily scheduled to work with the radiologist for the entire day. An assistant to the cytologist is assigned to place fresh stains and all supplies that the cytolo- gist may require in a designated area in the CT suite by the CT table on a counter adjacent to a sink. The supplies include Diff-Quik® (Dade AG, Dudin gen, Switzerland) stains and a jar of fresh water, gloves, slides, slide labeling pens, Coplan jars with alcohol, Cytolyt®, formalin jars, sterile cell medium such as RPMI® (for fl ow cytometry), needles, a watch glass, cardboard slide holders, and specimen bags. The Diff-Quik® Stain is a water soluble modifi cation of the Wright Stain which is rapidly performed on air-dried smears. The methylene blue component of the Diff-Quik® Stain accounts for a basophilic stain ing of nucleoli and cytoplasm and the meta- ch ro mat ic (red to purple) staining of the nuclei is due to the azure component. The Eosin component of the stain accounts for the yellow-red staining of the cy to plasm.
Ideally, a multiheaded microscope is utilized on-
site for the evaluation of Diff-Quik® stained air-dried
smears so that the cytologist and radiologist can view
the sample together and decide, based on the evalua-
tion of adequacy, whether an additional biopsy needs to be performed. This decision is not only based on the ability to make a diagnosis but, if benign, whether there is suffi cient material for cul ture and sensitivity studies or, if malignant, for im mu nochem i cal stud- ies to assist in determining the primary site of the malignancy. The initial concern when evaluating an aspirate on-site is the iden ti fi ca tion of cancer and if possible, its classifi cation as small cell or non small cell carcinoma. Small cell car ci no ma is composed of small metachromatic cells showing nuclear molding and streaking, granular chromatin, inconspicuous nucleoli, and scant cy to plasm leading to a mark- edly increased nuclear to cy to plas mic (N/C) ratio (Fig. 13.15). Non-small cell car ci no ma is composed of larger cells with enlarged nu clei, conspicuous nucle- oli and eccentric cy to plasm which is frequently dense or vacuolated in ad e no car ci no ma or keratinized in squamous cell car ci no ma (Fig. 13.16) (Kamiya et al.
1995). If mucin is iden ti fi ed in the cytoplasm of the cells or is found extracellularly in the background of the smear, a di ag no sis of mucinous adenocarcinoma can be con fi dent ly made (Roger et al. 1976).
The subtle nuclear morphology seen in low grade carcinomas is best seen on Papanicolaou Stain. Ad di - tion al ly, the Papanicolaou Stain easily identifi es the keratinized cells in squamous cell carcinoma by their orangeophilic cytoplasm. Squamous cell car ci no ma may show extensive necrosis and the aspirated mate- rial shows liquifaction. In these instances, the fl uid can be expelled from the syringe along the pe rim e ter of a watch glass so that any cellular particles will adhere to the watch glass. The particles can then be smeared onto a slide and the air-dried smear stained with Diff-Quik® to render an immediate in ter pre ta tion which is usually diagnostic. The re main ing material is best preserved in Cytoloyt® for ThinPrep slides.
If necessary, special stains for or gan isms or immu- nostains can be performed on ad di tion al ThinPrep monolayered cellular prep a ra tions from the same sample. The tissue con cen tra tion technique utilizing the watch glass can be ap plied to bloody aspirates as well (Giri and Vazquez 2000). The bloody material is expelled along the pe rim e ter of the watch glass, the particles are col lect ed and smeared, and then the clotted blood is sub mit ted in formalin for a cell block (paraffi n em bed ded blood clot) from which standard H&E stained cut sections are made. In general, if the im me di ate in ter pre ta tion of the bloody aspirate is ma lig nant, the cell block will have an abundance of cells for evaluation by immunohistochemistry.
Occasionally, smooth round tumors are aspirated that prove to be carcinoid tumors or hamartomas.
Fig. 13.15. Diff-Quik Stain of small cell cell carcinoma (high power). The aspirate shows a population of small cells with metachromatic nuclei, inconspicuous nucleoli and scant cy to - plasm. The nuclei are fragile and show molding and streak ing.
Note the golden red blood cells in the smear as well
Fig. 13.16. Diff-Quik Stain of non-small cell carcinoma (high power). The aspirate shows clusters of cohesive large cells with pleomorphic metachromatic nuclei, prominent ba so philic nucleoli, and abundant basophilic cytoplasm
The former is characterized by a monotonous pop- u la tion of round to spindled cells with granular nu cle ar chromatin and scant cytoplasm (Fig. 13.17) that can show a neuroendocrine pattern of growth.
Ves sels can be a prominent feature on the smears and their identifi cation helps to distinguish carcinoid tu mors from high grade neuroendocrine carcinoma.
Since carcinoid tumors tend to be bloody, a cell block preparation should be made for staining with chro- mogranin and synaptophysin in order to con fi rm the cytologic diagnosis. Aspirates of hama r to mas show fi brillary myxoid ground substance, vari able amounts of fi broadipose tissue and cartilage and clusters of bland ciliated and reactive bronchial cells.
When the tumor is composed entirely of the fi bril lary
myxoid ground substance, its ho mo ge neous appear- ance on CT scan can arouse suspicion for malignancy.
Also, the ground substance can mim ic mucin and the proliferation of benign bron chi oloal ve o lar cells, adenocarcinoma (Fig. 13.18). This pitfall must be recognized in order to avoid a false positive diagnosis of mucinous ad e no car ci no ma.
The most common benign aspirate yields only non-specifi c fi ndings including alveolar mac ro- ph ag es, infl ammatory cells, ciliated bronchial cells and re ac tive alveolar pneumocytes. Fragments of me sothe li um may be aspirated when the nodule is pleural based and these mesothelial sheets must not be false ly interpreted as adenocarcinoma. Some- times, acid fast orgasnisms and fungal forms can be iden ti fi ed with special stains on smears of necrotiz- ing gran u lo mas. In general however, a nonspecifi c be nign diagnosis on transthoracic aspiration biopsy is not considered defi nitive for the work-up of a pul mo nary nodule. As mentioned above, follow-up is based on presence or absence of growth on sub se - quent CT scans and other clinical factors, as well as certainty regarding needle tip positioning.
13.10 Conclusion
TNB is a highly accurate procedure for the diagnosis of pulmonary nodules. With careful attention to tech- ni cal factors, nodules of any size in any location may be biopsied. The close collaboration with a skilled
cytologist is essential and it is also quite help ful to develop close working relations with the cli ni cians so that they have confi dence in the results. This is particularly helpful in those cases where non spe cifi c results have been obtained. Following the pa tients with repeat CT scans to assess the nodules for interval change or scheduling them for repeat biopsy may be used to establish the benign nature of a nod ule and avoid unnecessary surgery.
References
Ayar D, Golla B, Lee JY et al (1998) Needle-track metastasis after transthoracic needle biopsy. J Thorac Imaging 13:2–6 Cagle PT, Kovach M, Ramzy I (1993) Causes of false results
in transthoracic fi ne needle lung aspirates. Acta Cytol 37:
16–20
Cox JE, Chiles C, McManus CM et al (1999) Transthoracic needle aspiration biopsy: variables that affect risk of pneu motho rax. Radiology 212:165–168
Fogarty JP, Dudek G (1995) An unusual case of lung torsion.
Chest 108:575–578
Giri D, Vazquez MF (2000) “Pick and Smear” tissue con cen - tra tion technique for image guided FNA specimens. Acta Cytol 45:889–890
Gupta S, Takhtani D, Gulati M et al (1999) Sonographically guided fi ne-needle aspiration biopsy of lytic lesions of the spine: technique and indications. J Clin Ultrasound 27:123–129
Hirsh J, Salzman EW, Harker L et al (1989) Aspirin and other platelet active drugs: relationship among dose, ef fec - tive ness and side effects. Chest 95:12S–18S
Kamiya M, Uei Y, Shimosato Y (1995) Cytologic features of peripheral squamous cell carcinoma of the lung. Acta Cytol 39:61–68
Fig. 13.17. Papanicolaou Stain of a carcinoid tumor (medium power). The aspirate shows small uniform oval to spindled cells with fi nely granular nuclear chromatin, inconspicuous nucleoli and scant cytoplasm
Fig. 13.18. Diff-Quik Stain of a hamartoma (low power). The aspirate shows myxoid stroma with bland mesenchymal cells and delicate vessels. There is an accompanying proliferation of benign bronchioloalveolar cells with uniform nuclei, in tra - nu cle ar inclusions and smooth round nuclear membranes
Klein JS, Zarka MA (2000) Transthoracic needle biopsy. Radiol Clin North Am 38:235–266
Lalli AF, McCormack LJ, Zelch M et al (1978) Aspiration bi op sies of chest lesions. Radiology 127:35–40
Liptay MJ (1999) Solitary pulmonary nodule: treatment op tions. Chest 116 [Suppl 6]:517S–518S
Man A, Schwarz Y, Greif J (1998) Case report: Cardiac tam- pon ade following fi ne needle aspiration (FNA) of a me di - as ti nal mass. Clin Radiol 53:151–152
Miller KS, Fish GB, Stanley JH et al (1988) Prediction of pneu- motho rax rate in percutaneous needle aspiration of the lung. Chest 93:742–745
Moore EH (1997) Needle-aspiration lung biopsy: a com pre - hen sive approach to complication reduction. J Thorac Imaging 12:259–271
Moore EH (1998) Technical aspects of needle aspiration lung biopsy: a personal perspective. Radiology 208:303–318 Protopapas Z, White CS, Miller BH et al (1996) Transthoracic
needle biopsy: results of a nationwide survey (abstract).
Radiology 201:270–271
Roger V, Nasiell M, Linden M et al (1976) Cytologic dif fer en tial diagnosis of bronchiolo-alveolar carcinoma and bron cho- gen ic carcinoma. Acta Cytol 20:303–307
Saleh HA, Haapaniemi J, Khatib G et al (1998) Bronchioloal- veolar carcinoma: diagnostic pitfalls and im mu no cy to- chem i cal contribution. Diagn Cytopathol 18:301–306 Stampfel G (1982) Anaphylactoid reaction – a rare com pli -
ca tion after fi ne needle aspiration of the lung. Radiology 22:329–330
Stern EJ, Webb WR, Gamsu G (1993) CT gantry tilt: utility in transthoracic fi ne-needle aspiration biopsy. Work in prog- ress. Radiology. 187:873–874
Taft PD, Szyfelbein WM, Greene R (1980) A study of vari abil i ty in cytologic diagnoses based on pulmonary aspiration specimens. Am J Clin Pathol 73:36–40
Westcott JL (1995) Needle biopsy of chest lesions. In: Taveras JM, Ferrucci JT (eds) Radiology, vol 1, chap 45. Lippincott, Philadelphia
White CS, Meyer CA, Templeton PA (2000) CT fl uoroscopy for thoracic interventional procedures. Radiol Clin North Am 38:303–322
Yankelevitz DF, Henschke CI (1993) Needle-tip localization for CT-guided biopsies. J Thorac Imaging 8:241–243
Yankelevitz DF, Henschke CI, Davis SD (1993) Percutaneous CT biopsy of chest lesions: an in vitro analysis of the effect of partial volume averaging on needle positioning. AJR Am J Roentgenol 161:273–278
Yankelevitz DF, Hayt D, Henschke CI (1995) Transthoracic needle biopsy. What size syringe? Clin Imaging 19:
208–209
Yankelevitz DF, Davis SD, Chiarella D et al (1996a) Needle-tip repositioning during computed-tomography-guided tran s- tho rac ic needle aspiration biopsy of small deep pul mo nary lesions: minor adjustments make a big difference. J Thorac Imaging 11:279–282
Yankelevitz DF, Davis SD, Henschke CI (1996b) Aspiration of a large pneumothorax resulting from transthoracic nee dle biopsy. Radiology 200:695–697
Yankelevitz DF, Henschke CI, Koizumi JH et al (1997) CT- guided transthoracic needle biopsy of small solitary pul- mo nary nodules. Clin Imaging 21:107–110
Yankelevitz DF, Reeves AP, Kostis WJ et al (2000a) Small pul- mo nary nodules: volumetrically determined growth rates based on CT evaluation. Radiology 217:251–256
Yankelevitz DF, Vazquez M, Henschke CI (2000b) Special tech niques in transthoracic needle biopsy of pulmonary nodules. Radiol Clin North Am 38:267–279
Yankelevitz DF, Wisnivesky JP, Henschke CI (2000c) Com- par i son of biopsy techniques in assessment of solitary pul- monary nodules. Semin Ultrasound CT MR 21:139–148
