| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, St. Bartholomew's Hospital, West Smithfield, London
EC1A 7BE, United Kingdom.
2
Present address: Department of Radiology, University of Michigan Hospitals,
MRI B2B311, 1500 E. Medical Center Dr., Ann Arbor, MI 48109-0030.
3
Department of Pediatric Oncology, St. Bartholomew's Hospital, West Smithfield,
London, EC1A 7BE, United Kingdom.
4
Department of Pathology, St. Bartholomew's Hospital, West Smithfield, London,
EC1A 7BE, United Kingdom.
Received February 25, 2000;
accepted after revision June 8, 2000.
Address correspondence to H. K. Hussain.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Thirty-four biopsies (biopsies of the abdomen, 32; of the chest, 2) were performed on 34 children at presentation under CT or sonographic guidance using 14-, 18-, or both 14- and 18-gauge needles. A minimum of two tissue cores was obtained. Most biopsies were performed under general anesthesia, permitting other procedures to be performed. The biopsy results were confirmed by subsequent surgical pathology, bone marrow biopsy, biochemical or clinical features, and follow-up examination.
RESULTS. The needle biopsy diagnoses were nephroblastoma (n = 11), neuroblastoma (n = 7), renal cell carcinoma (n = 2), synovial sarcoma (n = 1), non-Hodgkin's lymphoma (n = 2), clear cell sarcoma (n = 1), rhabdoid tumor (n = 1), pulmonary blastoma (n = 2), embryonal rhabdomyosarcoma (n 1), germ cell tumor (n = 1), adrenal carcinoma (n = 1), inflammatory tissue (n = 2), desmoplastic tumor of the mesentery (n = 1), and primitive neuroectodermal tumor (n = 1).
In 28 patients, the results were confirmed as correct (22 with surgery and 6 with follow-up examination). Four patients required additional biopsy. In two of these patients, the core biopsy showed inflammatory tissue only, and an open biopsy of a different site was performed; the other two patients did not respond to therapy on the basis of needle biopsy results, and an open biopsy altered the diagnosis. Two patients with widespread disease were excluded because they did not respond to treatment and were too ill to undergo an open biopsy. Only one significant complication was recorded.
CONCLUSION. Imaging-guided core biopsy is a safe and reliable means of obtaining sufficient tissue to make a confident histologic diagnosis of malignant pediatric tumors in a high percentage of patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our institution, core biopsies have replaced open surgical biopsies in children with suspected malignant lesions for the diagnosis of tumors that will be treated initially by chemotherapy or radiotherapy, but not for tumors considered resectable at the time of presentation. We present our experience with pediatric percutaneous biopsies in a tertiary referral center for pediatric oncology patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tumor Characteristics
Thirty-two biopsies were obtained from abdominal masses (retroperitoneal,
14; renal, 15, hepatic, 2; portal, 1) and two from thoracic masses (anterior
mediastinal, 1; chest wall, 1) (Fig.
1A,1B).
The sizes of the lesions were 2 cm (2 lesions), 2-4 cm (4 lesions), 4-8 cm (7
lesions), 8-12 cm (13 lesions), 12-16 cm (5 lesions), and 16-21 cm (3
lesions).
|
|
Technique
CT was used to guide the biopsy in 19 patients; sonography was used in 15
patients. Thirty biopsies were performed under general anesthesia, and four
under sedation and local anesthesia when the patients were older and more
cooperative.
Under the same general anesthetic episode, 12 patients had a bone marrow aspirate and biopsy in addition to the diagnostic tumor biopsy in the diagnostic radiology suite. The ages of the children who underwent biopsies under general anesthesia were less than 2 years (n = 10), 2-4 years (n = 9), and 4-6 years (n = 11).
The biopsies were performed using a Trucut needle assembled in an automated firing gun (Biopty; Bard, Crawley, West Sussex, United Kingdom). The needle size varied according to the site and size of the lesions. A 14-guage Trucut needle was used in 12 biopsies, 18-gauge in 17, and both 14- and 18-gauge needles in one. The choice of the needle size was determined by the size of the lesion, its location, and its proximity to vital structures. A large needle was used whenever it was believed safe to obtain large specimens; this is particularly important for the accurate histolgic classification of lymphoma and small round cell tumors.
Informed consent for anesthesia and biopsy was obtained from all parents. A full clotting screen, hemoglobin level, platelet count, and serum for cross matching were available before every biopsy. For biopsies carried out under CT guidance, a diagnostic CT scan was obtained after oral and IV contrast material was administered to delineate the bowel and vascular structures. Then the most suitable slice was chosen and the site of entry selected using the light beam on the CT gantry. The entry site was checked again using a small local anesthetic needle placed in the subcutaneous tissue. A 2-mm skin incision was made at the chosen site of entry. The biopsy needle was then introduced to a predetermined depth as judged on the scan, allowing for the 23-mm throw of the needle. The area was rescanned to check the position of the needle; if inadequate, the needle placement was readjusted, and the patient was scanned again until satisfactory needle positioning was achieved.
Under sonographic guidance, the most suitable site of entry was chosen and confirmed in two planes. The skin was then cleaned, and a 2-mm skin incision was made at the chosen entry site. The biopsy needle was introduced under real-time imaging to a predetermined depth, allowing for the throw of the needle. The position of the needle was then checked in two planes, and the gun was fired.
At least two tissue cores (2 tissue cores in 15 patients, 3 in 14 patients, and 4 in 1 patient) were obtained in all except one patient in whom it was believed unsafe to perform more than one pass because of bleeding tendency. For the second core, the same entry site was used with the needle positioned at a different angle; occasionally, a different entry site was chosen. A coaxial biopsy system was not used. The operator determined the number of and approach to the second and subsequent cores at the time of the biopsy; this determination depended on the size and quality of the specimens, technical feasibility, and safety of the procedure. A pathologist was not present on site, and the tissue samples were taken fresh to the histopathology department. The children were returned to the ward with clear observation instructions, which varied according to the site and technical difficulty of the biopsy. Complications were recorded.
All specimens were received fresh in the histopathology department. Imprints of tissue were prepared on glass slides and stained with toluidene blue. Most of each specimen was fixed in 10% formal saline, processed routinely, and embedded in paraffin wax. A small portion of each specimen was fixed in osmium tetroxide for electron microscopy. An even smaller portion was frozen for molecular biology analysis. Two-micrometer and 4-µm sections were cut from the paraffin was—embedded material and stained with H and E. The findings on imprint preparations and on H and E dictated which diagnostic tests would be performed. Immunohistochemical stains were performed in almost all patients. Antibodies used included protein gene product 9.5 (neural differentiation), desmin (muscle differentiation), lymphoid markers, and placental alkaline phosphatase (germ cell differentiation). Electron microscopy was performed in four patients. Molecular biologic analysis was not believed necessary for the diagnosis in any case. Diagnoses were made after examination of appropriately stained slides according to well-described histopathologic criteria. In many patients, more than one histopatholgist reviewed the slides.
Confirmation of Findings
Verification of the percutaneous biopsy results was made with surgical
pathology findings, bone marrow biopsy, biochemical features, the patient's
response to appropriate therapy, and clinical follow-up examinations for at
least 1 year. Nonverified biopsies were excluded from the analysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The analysis of our results is based on 32 diagnostic and nondiagnostic
biopsies with histopathologic verification of the results. The two diagnostic
but nonverified biopsies were excluded. On the basis of the final confirmation
(surgical pathology findings, 22; follow-up examination, 6), the percutaneous
biopsy diagnosis was correct in 28 (88%) of the 32 biopsies. These patients
were treated with primary chemotherapy or radiotherapy on the basis of the
results of the percutaneous biopsy alone. Twenty-two of 28 patients underwent
surgery to remove the residual tumor. The biopsy diagnoses in these patients
were nephroblastoma (n = 10), neuroblastoma (n = 6), renal
cell carcinoma (n = 2), clear cell renal sarcoma (n = 1),
rhabdoid renal tumor (n = 1), retroperitoneal synovial sarcoma
(n = 1), and pulmonary blastoma (n = 1). In all these
patients, the biopsy diagnoses were confirmed after examination of the
excision specimen. In the six patients who did not have surgical confirmation
(non—Hodgkin's lymphoma [n = 2], neuroblastoma [n =
1], embryonal rhabdomyosarcoma [n = 1], germ cell tumor [n =
1], adrenocortical carcinoma [n = 1]), the biopsy diagnoses were
concordant with the clinical, biochemical, and radiologic features of the
tumor; the predicted response to treatment; or clinical behavior of the tumor.
For example, the child with the germ cell tumor had markedly elevated
-fetoprotein levels, which supported the biopsy diagnosis, and
biochemical abnormalities associated with such tumors, including virilization
and precocious puberty. The other four patients responded satisfactorily to
treatment on the basis of needle biopsy diagnosis and underwent follow-up
examinations for at least 1 year.
Four biopsy results were considered incorrect; these results were pulmonary blastoma (n = 1), nephroblastoma (n = 1), and inflammatory tissue (n = 2). In two of these patients, the pathologic diagnoses made with the percutaneous biopsy were inaccurate and resulted in the initiation of inappropriate therapy. Both patients failed to respond satisfactorily to treatment and had surgical excision or an open biopsy of the mass.
In the first patient, surgical pathology of a mediastinal mass diagnosed on core biopsy as a pulmonary blastoma showed it to be a synovial sarcoma. The percutaneous biopsy in this patient was performed under sonographic guidance, and three tissue cores were obtained with an 18-gauge needle.
In the second patient, an open biopsy of an abdominal mass diagnosed on the core biopsy as a nephroblastoma altered the diagnosis to a neuroblastoma infiltrating the kidney. In retrospect, the error was caused by much of the biopsy specimen being composed of nonneoplastic kidney rather than tumor. This biopsy was performed under CT guidance using an 18-gauge needle to obtain two tissue cores.
The other two biopsies were considered incorrect because they were nondiagnostic and showed only inflammatory tissue when malignancy was strongly suspected; therefore, both patients required further diagnostic tests. An open biopsy in the first patient and a bone marrow biopsy in the second revealed the diagnoses of Hodgkin's disease and acute lymphoblastic leukemia, respectively. The percutaneous biopsy in the patient with Hodgkin's disease was obtained under CT guidance using both 18- (2 passes) and 14-guage (1 pass) needles; in the patient with acute lymphoblastic leukemia, only one tissue core was obtained under sonographic guidance using an 18-guage needle. The cause of biopsy failure in both cases was targeting and sampling error. In the patient with Hodgkin's disease, the subsequent open biopsy showed the abnormality being primarily in the bone and not in the surrounding soft tissue, which was percutaneously biopsied. In the patient with acute lymphoblastic leukemia, attempts were made to biopsy small liver lesions. We assume that intervening inflammatory tissue was incorrectly sampled.
The results of two biopsies, although diagnostic for malignancy, remain unverified and were excluded from the analysis. A 17-year-old patient presented with a large abdominal mass, having been treated successfully at age 5 years for a ganglioneuroblastoma. The needle biopsy showed a primitive neuroectodermal tumor that was different from the tumor occurring 12 years previously. The other patient also presented with a large abdominal mass and had a needle biopsy diagnosis of a desmoplastic small cell tumor of the mesentery. Both patients had widespread disease at the time of presentation. Neither responded to treatment and both were considered too ill to undergo an open biopsy.
One significant and two minor complications were recorded. These were local bleeding and infection (n = 1), transient hematuria (n = 1), and chest pain (n = 1). Both cases of bleeding occurred after renal biopsies. The first patient developed abdominal pain and distention shortly after the biopsy was performed under sonographic guidance with a 14-guage needle (2 passes). The hemoglobin level fell by 3 gm/L within 24 hr after the biopsy, followed 2 days later by fever and rigor. This reaction was presumed to represent intratumoral bleeding from the biopsy with resulting infection. The second patient developed mild macroscopic hematuria soon after the biopsy was performed under CT guidance with an 18-guage needle (2 passes); this patient's symptoms settled spontaneously after 48 hr. The third patient complained of chest pain after a mediastinal biopsy was performed under sonographic guidance with an 18-guage needle (3 passes); these symptoms were readily controlled with simple analgesia.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As with all retrospective studies, a uniform method of confirmation could not be used, a problem also encountered in other studies [3, 5,6,7]. Nevertheless, in 25 (78%) of our patients, tissue was subsequently obtained through surgical resection or biopsy. Somers et al. [3] reviewed 27 biopsies performed on 25 children with suspected malignancy, and, of those biopsies, 12 were correct. Hugosson et al. [5] recently reported the results of 90 procedures in 61 children, including 53 fine-needle aspirations and 37 needle core biopsies. The biopsies were performed on 27 malignant and 10 benign lesions using sonographic guidance, a 1.2-mm (19-guage) needle, and one or two passes. Proof of diagnosis was obtained at surgery in 14 children, repeated biopsy in eight, and clinical assessment in 39. Hoffer et al. [6] performed 15 imaging-guided biopsies of benign and malignant thoracic lesions. Thirteen were core biopsies performed with 16- to 21-guage needles and one or two passes. Two biopsies required surgical biopsy for more definitive tissue typing, and the remaining 11 were diagnostic. Of those, only five had histologic verification of the biopsy results and were correct. Klose et al. [7] reviewed the results of 44 biopsies performed on 41 benign and malignant lesions in 39 patients. Only 16 patients had malignant lesions and those underwent 17 biopsies. The biopsies in this series had proof of histologic diagnosis by surgical pathology (n = 11), previous or subsequent evaluation of the primary tumor in the case of metastatic disease (n = 5), bacteriologic evaluation (n = 6), or clinical course over 12 months.
Some studies [4, 8,9,10] lack histologic proof of the percutaneous biopsy results; for example, in the study of Sabbah et al. [4], 15 children with malignant abdominal tumors underwent 15 biopsies, with the results of only four confirmed by electron microscopy (n = 3) and surgery (n = 1). Baran et al. [8] reviewed 20 biopsies performed on 20 children with suspected or known malignant (16 biopsies) or inflammatory lesions (4 biopsies), and none had histologic proof of the biopsy results. Similarly, van Sonnenberg et al. [9] reviewed the results of diagnostic and therapeutic interventional radiologic procedures in 100 children, including 30 fine-needle aspirations and biopsies; of those, 18 were malignant lesions. Surgical pathology was warranted on two occasions only when the biopsy results were nondiagnostic. The number of malignant biopsies in the study by Towbin and Strife [10], which also looks into interventional techniques in children, is too small (n = 5) to draw a significant conclusion.
In our series, surgical confirmation was not always possible or necessary. In one of our patients subsequently shown to have acute lymphoblastic leukemia, the final diagnosis was made on the basis of bone marrow biopsy. In another group of six patients, surgical excision was not feasible or acceptable clinical practice. Two of these patients had lymphoma. The accuracy of needle core biopsy in non—Hodgkin's lymphoma has previously been documented [12,13,14]. Both patients were treated on the basis of this diagnosis and responded well to treatment. The child with neuroblastoma responded well to treatment, as did the child with embryonal rhabdomyosarcoma. Another two patients with germ cell tumor and adrenocortical carcinoma had associated clinical and biochemical abnormalities confirming the percutaneous biopsy diagnosis. Two biopsy results were unconfirmed and were excluded from the analysis even though they may have been correct.
In the past, pathologists have had difficulty in classifying childhood tumors on the basis of a needle core of tissue; however, this difficulty is less a problem today because of the widespread availability of immunohistochemical stains and access to gene amplification procedures [15, 16]. If an adequate tissue sample is ensured and examined by an experienced pathologist, core biopsies should be adequate for a complete histologic diagnosis in most patients. This is borne out by our results.
The major cause of an unsuccessful needle core biopsy in our series was targeting or sampling error, either because of failure to obtain the appropriate tissue or because of the unrepresentative tissue sampled. In two of our patients, this problem resulted in the acquisition of only inflammatory tissue. In another patient, the problem lead to an incorrect pathologic diagnosis whereby a synovial sarcoma was misdiagnosed as a pulmonary blastoma. This error may have been related in part to the small size of the sample. Furthermore, synovial sarcomas are difficult to diagnose histologically, and misdiagnosing such tumors is a recognized error [17]. In one patient, part of the kidney was sampled rather than the adjacent neuroblastoma, resulting in misdiagnosing the tumor as a nephroblastoma.
Both of these drawbacks can be minimized by obtaining large and multiple tissue samples from different sites of the tumor and avoiding areas of necrosis [16]. Crush and drying artifacts, which affect interpretation, can be reduced by using modern sheath Trucut needles in a biopsy gun and with the prompt placement of the tissue cores in appropriate preservative media [18, 19].
The size of the samples obtained by percutaneous biopsies should be adequate for both routine and research studies. We found that the size of the sample shows no consistent correlation with the diagnostic quality of the material, and bigger was not necessarily better, particularly if the tissue sampled was necrotic or if nondiagnostic reactive tissue within or around the tumor was biopsied. Furthermore, the size of an adequate specimen varies with the morphologic and diagnostic features of the tumor, and there is no predetermined size of biopsy sample that is adequate for all circumstances. For example, size is not crucial in uniform tumors. By contrast, morphologically heterogeneous tumors would require more than one tissue sample from multiple tumor sites [19]. Other tumors such as lymphoma in which microscopic architectural organization plays a significant role may require relatively large samples for a complete diagnosis and classification [6, 12,13,14, 19].
Unfortunately, our series is too small to properly detect a difference in accuracy depending on the size of the Trucut needle used and number of tissue cores obtained; however, the incorrect biopsies were performed in a similar fashion to the correct biopsies using 14-, 18-, or both 14- and 18-gauge needles to obtain a minimum of two tissue cores, except one liver biopsy, in which only one tissue core was obtained with an 18-gauge needle. The error in this case was believed to be related to a combination of nonrepresentative tissue being sampled and small sample size. Other authors have used similar techniques with needle sizes ranging from 14- to 22-guage [3,4,5,6,7,8,9,10] and various numbers of tissue cores, including one [4, 6, 7] or a minimum of two [3, 5]. Some authors have not documented the number of tissue cores they have taken [8,9,10].
Despite the widespread use of diagnostic needle core biopsy in the practice of adult oncology [11,12,13,14], it is less commonly used in children in whom, traditionally, open surgical biopsies have been used for the initial diagnosis of malignancy. This practice may be related to several factors, including fear of accidental puncture of vital organs, increased morbidity from hemorrhage and infection in ill children, and other complications such as tumor seeding along the needle tract. Since 1981, several reports have been published confirming the accuracy, safety, and cost-effectiveness of this technique in children [3,4,5,6,7,8,9,10]. The main indication for needle core biopsy in our series was the primary diagnosis of malignancy when chemotherapy or radiotherapy was planned as the initial form of therapy, regardless of the tumor stage, which was similar to most other studies [3, 6,7,8, 10, 16, 20] except the study by Sabbah et al. [4], in which only inoperable tumors were biopsied. Percutaneous biopsies were not performed when surgical resection was likely to be chosen as the first line of treatment because it was believed that the biopsy findings would not alter the treatment.
Percutaneous biopsies have many advantages over open surgical biopsies. They are quick and easy to perform, cause less morbidity (an important consideration in sick children), and allow treatment to be started early without having to wait for postoperative recovery. Complications such as significant bleeding and performation of vital structures are uncommon with the use of modern, high-resolution imaging modalities. Large feeding vessels and areas of high vascularity within tumors can be equally well seen on CT and Doppler flow imaging, and thereby avoided. Other possible complications such as tumor seeding along the needle tract may occur if large needles are used to biopsy malignant lesions, particularly nephroblastomas [21,22,23]; however, such a complication is exceedingly rare with the use of modern, sheathed Trucut needles [19]. No tumor recurrence along the needle biopsy tract was apparent on imaging in any of our patients during follow-up examinations performed at least 1 year after the procedure.
In contrast to other series [3, 4, 6,7,8,9,10, 20] and somewhat similar to the series by Hugosson et al. [5] in which 39 of their biopsies were performed with the patient under general anesthesia, most of our biopsies were performed with the patient under general anesthesia, enabling other procedures such as bone marrow aspiration to be performed under the same anesthetic episode; this may also reflect the younger age of our study population. No increased morbidity was encountered as a result of general anesthesia. With care and adequate precautions, the procedure is safe; in our study and in accordance with other studies [3,4,5,6,7,8,9,10], only one significant complication was recorded in which the patient had intratumoral bleeding and infection after a renal biopsy, requiring blood transfusion and antibiotic therapy. The problem of hemorrhage after biopsies of nephroblastomas is well documented [22, 23]; it is related to the size and fragility of the tumor and is further facilitated by infection. The child who developed microscopic hematuria after a renal biopsy was treated conservatively and did not require blood transfusion.
No consistent relationship between the incidence of complications and the size of the needle or number of tissue cores obtained was noted, although our numbers are too small to make a significant correlation.
In agreement with other authors [3,4,5,6,7,8,9,10], 16, 20], we found that percutaneous core biopsies are accurate, safe, and cost-effective, and we believe that if radiologic and pathologic expertise is available and adequate tissue samples can be obtained, a complete histologic diagnosis can be made in most patients. The major advantage of a percutaneous biopsy is that the child is spared the morbidity of surgery and a possible delay in initiating therapy. We showed that imaging-guided core biopsies have a successful outcome in most children with malignancy. We recommend that this technique be used whenever possible as an alternative to open surgical biopsy in children with suspected malignancy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. J. Jackman, R. L. Birdwell, and D. M. Ikeda Atypical Ductal Hyperplasia: Can Some Lesions Be Defined as Probably Benign after Stereotactic 11-gauge Vacuum-assisted Biopsy, Eliminating the Recommendation for Surgical Excision? Radiology, August 1, 2002; 224(2): 548 - 554. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
