Intraoperative Gamma Probe-Directed Biopsy of Asymptomatic Suspected Bone Metastases

Transcription

1 Intraoperative Gamma Probe-Directed Biopsy of Asymptomatic Suspected Bone Metastases Lary A. Robinson, MD, Dianne Preksto, PA-C, Carlos Muro-Cacho, MD, and David S. Hubbell, MD Thoracic Oncology Program, Division of Cardiovascular and Thoracic Surgery, and Department of Pathology, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida College of Medicine, Tampa, Florida Background. Radioisotope bone scanning is frequently used in staging malignancies. However, false-positive results are common, and biopsy is usually required. In the absence of plain radiographic abnormalities or local symptoms, localization of the area of abnormal tracer activity at the time of open rib or sternum biopsy may be difficult. It often requires resection of a large portion of one or more ribs or the sternum to assure that biopsy of the target area was performed. In this setting, a small gamma probe underwent evaluation as an aid to precise intraoperative localization of increased tracer activity in the target bone. Methods. Ten patients with asymptomatic suspected osseous chest metastases by radioisotope bone scanning but with normal plain radiographs underwent open biopsy of 13 ribs and 1 sternum. Six to 12 hours before operation, each received an intravenous injection of 28 mci of technetium-99m oxidronate. The hand-held, pencil-sized gamma probe in a sterile sleeve was used to localize the area of greatest activity in the target bone, once the bone was exposed through a small incision. Biopsy of a 3-cm length of rib or portion of sternum was performed. In the first two rib biopsies, an intraoperative radiograph with a radiopaque marker on the rib confirmed that the correct rib was selected for biopsy. Intraoperative radiographs were not done on later cases. Results. The mean ratio of hot spot activity on the targeted rib to background counts on adjacent ribs was (range, 1.35 to 2.05), and the difference was easily discernible intraoperatively. The ratio of hot spot activity on the sternum was somewhat lower (1.22), but the target area was still easy to detect. An abnormal diagnosis to account for the increased tracer activity was found in each of the 13 ribs and 1 sternal biopsy in all 10 patients: metastatic squamous cell carcinoma (1 rib), metastatic prostatic adenocarcinoma (1 rib), lymphoma (2 ribs), localized hypercellular marrow (1 rib), medullary fibrosis/paget s disease of the bone (2 ribs), localized fibrosis/granulation tissue (1 rib), enchondroma (3 ribs), and chondroma (2 ribs, 1 sternum). The difference in background counts to hot spot activity was best with injection of the tracer 6 hours before operation. Conclusions. The intraoperative use of gamma counting is an easy, highly accurate aid (100% sensitivity) to localize areas of abnormal radioisotope uptake in suspected asymptomatic rib and sternal metastases. Use of this technique obviates the need to obtain intraoperative localizing radiographs to confirm accurate rib identification, thereby decreasing operative time. (Ann Thorac Surg 1998;65: ) 1998 by The Society of Thoracic Surgeons Almost any cancer may metastasize to bone, although carcinomas of the breast and prostate account for the majority of cases, with kidney, thyroid, and metastases often occurring, in that order of decreasing frequency [1]. Because accurate staging of these or any other malignancies is critically important to make rational therapeutic decisions, any suspected osseous metastases must be fully evaluated. Although bony metastases often cause localizing pain or even a pathologic fracture, occasionally metastases are occult and asymptomatic, and are suspected from abnormalities on a screening radioisotope bone scan. Accepted for publication Dec 9, Address reprint requests to Dr Robinson, Thoracic Oncology Program, H. Lee Moffitt Cancer Center and Research Institute, Magnolia Dr, Tampa, FL Favorite sites of osseous metastases are the ribs and occasionally the sternum, which may be readily evaluated and sampled. Radioisotope bone scanning is much more sensitive than plain radiographic studies [2]. Because false-positive bone scan abnormalities are common, histologic confirmation of a suspected bony metastasis is generally necessary. However, in the absence of plain radiographic abnormalities or local signs or symptoms, biopsy of a suspected osseous chest metastasis becomes a challenging problem for the thoracic surgeon. Localization of the exact area of abnormal tracer activity at the time of open biopsy may be difficult, often requiring resection of a large portion of one or more ribs or a large area of sternum to assure that biopsy of the target area was performed. In this setting, a small gamma probe [3] (Fig 1) underwent evaluation as an aid to precise intraoperative localization of increased tracer activity in the target bone by The Society of Thoracic Surgeons /98/$19.00 Published by Elsevier Science Inc PII S (98)

2 Ann Thorac Surg ROBINSON ET AL 1998;65: GAMMA PROBE AND BONE METASTASES 1427 Statistical Evaluation The count ratio is calculated as the peak count (in counts per second) measured over the target abnormality in the bone divided by the background counts from adjacent ribs. All numerical patient data are expressed as the mean the standard error of the mean. Comparisons between any two patient groups are made using analysis of variance (unpaired Student s t test) using a computerdriven statistical program [4]. Results are considered statistically significant when the p value is 0.05 or less. Fig 1. Neoprobe 1000 hand-held gamma counter and the smallest (pediatric) probe. Material and Methods Between April 1996 and May 1997, 10 patients with a known or suspected malignancy were found to have asymptomatic areas of increased tracer activity in ribs or the sternum on bone scan but the plain radiographs of the same areas were normal. Figure 2 illustrates an example of these radiographic findings. Technique The 10 patients underwent open biopsy of one or more ribs or the sternum (total 13 ribs and 1 sternum biopsied) under general anesthesia as an outpatient in almost all cases (8 of 10 patients). Six to 12 hours before operation, each patient received an intravenous injection of 28 mci of technetium-99m oxidronate (TechniScan HDP; Mallinckrodt Medical, St. Louis, MO), the standard dosage for a radioisotope bone scan. After induction of general anesthesia, prepping and draping of the patient, the hand-held gamma probe (Neoprobe 1000; Neoprobe Corporation, Dublin, OH) in a sterile plastic sleeve was used to localize the area of greatest tracer activity (measured in counts per second) on the skin of the chest wall. A 3- to 4-cm incision was made over this area and the targeted rib or sternum was exposed. The probe in the sterile sleeve was then used to localize precisely in the surgical wound the area of increased tracer activity in comparison with background counts on adjacent ribs as well as more distantly on the target rib (Fig 3). In the first 2 patients, an intraoperative cross-table lateral radiograph was obtained with a radiopaque marker on the rib in the wound to verify that biopsy of the correct rib was being performed. As our experience with the gamma probe grew, we eliminated these time-consuming, costly radiographs. A 3-cm portion of the targeted rib or the outer table of the sternum was removed subperiosteally. After a chest radiograph in the recovery room verifying the absence of a pneumothorax and that biopsy of the proper rib was performed, most patients were discharged home. Two patients were kept overnight because of a small pneumothorax requiring chest tube placement in each. Fig 2. (A) Radioisotope bone scan showing localized increased uptake of tracer in the left sixth rib anteriorly (left anterior oblique position in patient 6). (B) Rib detail films (left anterior oblique position) of the same patient showing no abnormalities in the left sixth rib (arrow) in the same area of the bone scan abnormality.

3 Table 1. Results of Patients Undergoing Gamma Probe-Directed Bone Biopsy No. Age (y) Sex Clinical Diagnosis Symptoms Bone Scan Uptake Plain Bone Radiograph AP (normal values) PSA (normal values) Bone Sampled Injection to Operation (h) Count Ratio Rib Pathology 1 65 M Carcinoma of prostate None L 7th rib & skull Normal 46 (37 107) 14.8 (0 4.0) L 7th rib Hypercellular marrow 2 44 F Lymphoma Vague chest pain (1 y) L ribs 6, 7 & 8 Normal 294 (80 258)... L 6th rib L 8th rib Lymphoma Lymphoma 3 62 M Sq cell ca of None R 9th rib, L 6th rib Normal 208 (80 258)... R 9th rib Metastatic sq cell ca 4 74 M Sq cell ca of 5 73 M Sq cell ca of 6 70 M Sq cell ca of 7 37 F Carcinoma of breast 8 66 M Carcinoma of prostate 9 54 F Sq cell ca of Pain in low back Spine, mult ribs bilat Fracture T11 spine 81 (37 147)... L 7th rib L 8th rib Paget s disease Paget s disease None R 10th & 11th Normal R 10th rib Enchondroma ribs R 11th rib Enchondroma Vague chest L 6th rib Normal L 6th rib Enchondroma wall pain None Body of the Normal Body of the Chondroma sternum sternum None L 6th rib Normal L 6th rib Metastatic prostate (25 150) ( ) carcinoma Vague chest L 6th & 9th Normal L 6th rib wall pain ribs (20 125) L 9th rib 1.63 Chondroma Chondroma M Melanoma None L 10th rib Normal L 10th rib Localized fibrosis, granulation tissue AP alkaline phosphatase (U/L); bilat bilaterally; L left; mult multiple; PSA prostate-specific antigen (ng/ml); R right; Sq cell ca squamous cell carcinoma ROBINSON ET AL Ann Thorac Surg GAMMA PROBE AND BONE METASTASES 1998;65:

4 Ann Thorac Surg ROBINSON ET AL 1998;65: GAMMA PROBE AND BONE METASTASES 1429 Fig 3. Intraoperative view of the gamma probe in a sterile sleeve being used to measure counts directly on a rib. Results Patient Population There were 3 women and 7 men in the group (Table 1). The mean age of the patients was years (median 63.5 years; range, 37 to 74 years). Six of the 10 patients had known active cancer in the primary site (carcinoma of the in 5, carcinoma of the prostate in 1) and were undergoing staging before treatment. The other 4 patients had their primary tumor in remission. In the 10 patients, the primary diagnosis in 5 is carcinoma of the, 2 have carcinoma of the prostate, and 1 each has carcinoma of the breast, lymphoma, and melanoma. Fig 4. Photomicrograph of a decalcified rib from patient 3 demonstrating metastatic squamous cell carcinoma. (Hematoxylin and eosin stain; 400 before 50% reduction.) Detection of Bone Abnormality Intraoperatively, the rib or sternum appeared grossly normal to the surgeon, even after resection of the abnormal area. When counted directly on the rib in the wound, the mean ratio of hot spot activity in the targeted rib to background counts (count ratio) on adjacent ribs was (range, 1.35 to 2.05; median, 1.59), and the difference was easily discernible intraoperatively (Table 1). The count ratio for the sternal biopsy was lower at 1.22, but the difference was still easily discernible. This ratio tended to be higher and easier to distinguish when patients received the injection of the radioactive tracer 6 hours before operation compared with 12 or more hours, although the difference was not statistically significant ( versus , respectively, p 0.157). In the first 2 patients, an intraoperative radiograph with a radiopaque marker attached to the target bone (a needle was superficially embedded in the bone by the surgeon through the open wound) was obtained to verify that biopsy of the correct rib was performed. Multiple radiographs were needed in both patients to see enough bony landmarks to count the ribs definitively. After these first 2 patients, confidence was gained in the gamma counting method and intraoperative radiographs were omitted. After the experience of using this technique in only a few patients, the total operative time for these cases decreased to 20 to 40 minutes, as intraoperative radiographs were no longer needed and the radioisotope counting was very precise and rapid. A postoperative chest radiograph was obtained and confirmed that biopsy of the correct rib was performed in all patients. Pathology An abnormal diagnosis to account for the increased tracer activity was found in each of the 13 rib and 1 sternal biopsy in 10 patients. However, only 30% of the patients (3 of 10 patients, 4 ribs) were found to have a bone metastasis. Figure 4 is a photomicrograph that illustrates one of the rib biopsy specimens containing bone marrow replaced by numerous clusters of malignant epithelial cells, inducing a desmoplastic reaction in the adjacent stroma, diagnostic of metastatic squamous cell carcinoma (patient 3). Figure 5 illustrates islands of mature cartilage with focal mineralization surrounded by fibroblastic proliferation diagnostic of an enchondroma from patient 6. This is one of the most common benign abnormalities found in this setting. Benign cartilaginous Fig 5. Photomicrograph of a decalcified rib from patient 6 showing benign cartilage growing in the bony trabecular spaces characteristic of an enchondroma. (Hematoxylin and eosin stain; 100 before 50% reduction.)

5 1430 ROBINSON ET AL Ann Thorac Surg GAMMA PROBE AND BONE METASTASES 1998;65: tumors (enchondroma or chondroma) accounted for six (43%) of the bone scan abnormalities in this patient population. Of particular note, no rib was sampled within 4 cm of the costal chondral junction. In all patients, only rib (not costal cartilage) or bony sternum was sampled. Therefore, the presence of cartilage in the form of a chondroma or enchondroma seen in six biopsy specimens is not an artifact related to the location of the biopsy. Two of the patients with masses (patients 5 and 6), who had a demonstrated benign cause of their bone scan abnormality, subsequently underwent curative resections of their primary cancer. The bone biopsy result played a pivotal role in determining the future treatment or lack of treatment in all patients. Enzyme Studies Serum alkaline phosphatase is generally considered to be one of the clinical indicators of metastatic disease, especially to bone [5]. In this series, only 1 patient (no. 2) of the 10 had a slightly elevated alkaline phosphatase, and that patient was found to have lymphoma in the rib biopsy specimen. The other 2 patients with rib biopsy results positive for metastatic disease had normal serum alkaline phosphatase levels. Prostate-specific antigen is a highly sensitive serum marker for the detection of prostate cancer, although various conditions, such as benign prostatic hypertrophy, may cause false positives [6]. The 2 patients with prostate cancer (nos. 1 and 8) both had elevated prostate-specific antigen levels but only 1 patient (no. 8) had metastatic prostate cancer in the rib biopsy specimen. Morbidity and Mortality Inadvertent entry into the pleural cavity occurred in 2 patients during rib resection requiring overnight hospitalization after insertion of a small chest catheter. There were no other complications and no mortality. There was no complication from the radioisotope or the use of the small gamma camera. Comment Accurate staging of cancers is essential in determining rational treatment strategies based on clinical trials. The presence of metastatic disease to bone is considered to be distant metastatic spread in almost all cancers and is classified as stage IV [7]. Treatment for this stage then becomes nonoperative and generally chemotherapy is preferred, with radiotherapy added for localized, symptomatic lesions. Some cancers have a predilection for osseous spread, and the highest incidence is with carcinoma of the prostate and of the breast followed by the kidney, thyroid, and [1]. However, because treatment and prognosis are based on stage, which is dependent on the presence or absence of bone metastases, it is critically important to prove that suspected lesions are truly metastatic cancer or otherwise patients may be denied potentially curative therapy. Bony metastases may be readily evident if they cause localized pain, swelling, or even pathologic fractures. The serum alkaline phosphatase level may be elevated in these patients as a nonspecific clinical indicator of osseous metastases, although nonmalignant causes such as osteomalacia, hyperparathyroidism, osteitis deformans, healing fractures, and various types of hepatobiliary disease may also raise levels of this enzyme [5]. Occasionally, metastases are occult and essentially asymptomatic and are suspected from a screening radioisotope bone scan. With some cancers, bone scanning is routine in the staging workup, and asymptomatic abnormalities may be found. In other cancers, such as non-small cell carcinoma of the, bone scanning is recommended only if there is a clinical indicator of bony involvement such as bone pain or elevated levels of serum alkaline phosphatase or calcium. Rarely are bony metastases found in patients with non small cancer who are completely free of any clinical indicators [8, 9]. The most common sites of bony metastases in decreasing frequency are the vertebrae, pelvis and sacrum, femur, and ribs [1]. The typical plain radiographic finding is that of decreased density or osteolytic activity at the site of metastatic disease. Much less common is the finding of increased bone density (osteosclerosis or osteoblastosis), and it is generally associated with prostate cancer or the result of hormone treatment of breast cancer [1]. However, plain bone radiographs are relatively insensitive indicators of metastatic disease as more than 50% of the bony trabecula must be destroyed, such as in a vertebral body, before it is radiographically apparent [2]. Radioisotope bone scanning is a far more sensitive technique to detect osseous metastases than plain radiographs. Only 5% to 15% destruction of bone is necessary before neoplastic deposits are detected by bone scanning, which is appreciated much earlier than with plain radiographs [2]. Only 3% of patients with radiographically documented evidence of metastases have no abnormality on radioisotope bone scanning [2]. For the bone scan to be positive with increased uptake, the 99m-technetium-labeled phosphorous compound will need to localize to areas of increased blood flow or diffuse intraosseous bone formation [10]. Most metastatic tumors will cause both osteoclastic bone resorption as well as new bone formation. The technetium diphosphonate complex appears to bind to the surface of the hydroxyapitite crystal during new bone formation and not to the tumor itself. However, a variety of normal or benign conditions as well as primary bone tumors may also result in a hot spot on bone scan leading to a false-positive result in a search for osseous metastases. Some of these conditions that may cause a false-positive result are as follows [10]: Normal structures Base of skull, facial bones, inferior tip of scapula Sternomanubrial and corpus-manubrial joints Alae of sacrum Kidneys and bladder Variant anatomy

6 Ann Thorac Surg ROBINSON ET AL 1998;65: GAMMA PROBE AND BONE METASTASES 1431 Soft tissue abnormalities Injection site Dental abscess Calcific tendinitis or myositis Postoperative scar Hydronephrosis or hydroureter Soft tissue osseous metaplasia Bony abnormalities with increased blood flow Osteomyelitis, osteitis Fracture (recent or healing) Increased blood flow and bone formation Eosinophilic granuloma Paget s disease Fibrous dysplasia Renal osteodystrophy Osteoid osteoma Aseptic necrosis, cysts Hyperostosis frontalis interna Osteitis pubis Sudeck s atrophy Benign bone tumors Chondroma Fibroma Primary malignant bone tumors Osteosarcoma Ewing s sarcoma Chondrosarcoma Metastatic tumors The most common abnormality causing a positive bone scan in the current study (6 of 14, 43% of all biopsies) is the benign cartilaginous tumor chondroma [11]. When a chondroma is centrally located in bone it is called an enchondroma, but it is termed a chondroma when it expands through the cortex. Chondromas are relatively common representing 13.4% of benign tumors in bone, although the real incidence is unknown as they are generally asymptomatic. Usually, chondromas occur sporadically, most commonly in the small bones of the hands and feet, although they may occur in thin bones such as the ribs or scapulae, and rarely in the sternum (patient 7). Multiple chondromas may occur associated with conditions such as Ollier s disease, Maffucci s syndrome, and other rare syndromes. A chondrosarcoma may develop late in patients with these unusual disorders. Unless they become very large, chondromas are asymptomatic and generally are not seen on plain radiographs, or at best all that is seen is a subtle, localized area of central rarefaction in the bone. Most chondromas are found incidentally as a hot spot on a radioisotopic bone scan performed during the metastatic workup of a patient with a known or suspected malignancy. A biopsy often follows and the diagnosis of chondroma is made. In the asymptomatic patient, the false-positive rate in routine bone scans is significant, ranging from 55% in cancer [9], up to 71% in various cancers (current study). Among radiologists, it is commonly believed that a positive bone scan abnormality in the face of normal plain radiographs in an asymptomatic patient with known cancer indicates a bony metastasis. The current study would suggest just the opposite, that is, the bone scan abnormality in this setting is more likely benign. Magnetic resonance imaging, especially of the spine, appears to be highly sensitive in confirming the presence or absence of metastases if the radiographic results are typical [2]. However, a positive bone scan in a patient with a suspected or known malignancy generally should have histologic confirmation of the metastasis before deciding on the final staging. The presence of a positive bone scan in an asymptomatic patient with normal plain radiographs presents a difficult clinical problem to the surgeon planning a bone biopsy. Ribs, when positive on bone scan, are probably the easiest bones to sample technically, but localization of the exact site of the bone scan hot spot generally is difficult. Depending on the quality and definition of the bone scan image, it may be challenging to define exactly which rib is hot. Likewise it is often difficult to localize precisely the anteroposterior location of the rib abnormality based on the scan alone when the plain bone radiographs are normal. As a result, the surgeon is often forced to remove a large grossly normal-appearing area of one or even two ribs just to be sure that the bone scan abnormality is contained in the specimen. In the muscular or obese patient, it may be difficult to determine and count at operation (within one or two ribs) which is the target rib for removal, especially in the lower ribs posteriorly. Consequently, most patients require many intraoperative radiographs with a radiopaque marker adherent to the rib in the wound to verify that the correct numbered rib will be biopsied. Because a frozen section analysis on bone with a thick cortex, such as rib, is not possible intraoperatively and the final histologic diagnosis awaits lengthy decalcification and processing, 7 to 10 days may pass before the surgeon finds out whether biopsy of the correct area was performed. If the final diagnosis is normal bone, then there is even more uncertainty about whether the correct area was sampled. To address this difficult and frustrating clinical problem for the surgeon, Little and associates [12] advocated preoperative injection of the target rib and the overlying skin with methylene blue in the nuclear medicine department guided by the gamma camera, followed by immediate excisional biopsy of the blue portion of the rib in the operating room. Although this tattooing technique appears to work in experienced hands, it prolongs and complicates the operative episode by requiring careful coordination of the nuclear medicine department schedule while keeping an operating room available to immediately accept the rib biopsy patient for operation so that the blue dye does not have time to spread to other ribs. The technique described in this study requires only an intravenous injection of radioisotope 4 to 12 hours preoperatively followed by a very short operative procedure precisely pinpointing the target hot spot. The extreme accuracy of this technique obviates the need for timeconsuming intraoperative localizing radiographs. Use of the gamma probe (Neoprobe) to guide the biopsy of ribs is more direct, far quicker, less cumbersome especially in obese patients, and is likely to allow far more cost-

7 1432 ROBINSON ET AL Ann Thorac Surg GAMMA PROBE AND BONE METASTASES 1998;65: effective bone biopsies. This method has no side effects and has proven in this series to be 100% sensitive in finding bony abnormalities. Because the hand-held, intraoperative gamma counter is gaining widespread application for multiple uses in surgical oncology such as for nodal mapping in breast cancer and melanoma [3, 13], this device is becoming a common fixture in most major hospitals and should be considered by the thoracic surgeon for use in guiding the biopsy of suspected asymptomatic osseous chest metastases. References 1. Clain A. Secondary malignant disease of bone. Brit J Cancer 1965;19: Rogers LF. Secondary malignancies of bone. In: Juhl JH, Crummy AB, eds. Paul and Juhl s essentials of radiologic imaging, 6th ed, Philadelphia: Lippincott, 1993: Albertini JJ, Cruse CW, Rapaport D, et al. Intraoperative radiolympho-scintigraphy improves sentinel lymph node identification for patients with melanoma. Ann Surg 1996; 223: Glantz S. Primer of biostatistics. Version 3.0. New York: McGraw-Hill, Henry JB, ed. Clinical diagnosis and management by laboratory methods, 19th ed. Philadelphia: Saunders, 1996: Stamey TA, Yang N, Hay AR, et al. Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med 1987;317: Beahrs OH, Henson DE, Hutter RVP, Kennedy BJ. Handbook for staging of cancer, 1st ed. Philadelphia: Lippincott, 1993: Michel F, Solèr M, Imhof E, Perruchoud AP. Initial staging of non small cell cancer: value of routine radioisotope bone scanning. Thorax 1991;46: Ichinose Y, Hara N, Ohta M, et al. Preoperative examination to detect distant metastasis is not advocated for asymptomatic patients with stages 1 and 2 non small cell cancer. Chest 1989;96: Wahner HW, Brown ML. Role of bone scanning. In: Sim FH, ed. Diagnosis and management of metastatic bone disease, 1st ed. New York: Raven, 1988: Unni KK. Chondroma. In: Unni KK, ed. Dahlin s bone tumors, 5th ed, Philadelphia: Lippincott-Raven, 1996: Little AG, DeMeester TR, Kirchner PT, et al. Guided biopsies of abnormalities on nuclear bone scans. J Thorac Cardiovasc Surg 1983;85: Albertini JJ, Lyman GH, Cox C, et al. Lymphatic mapping and sentinel node biopsy in the patient with breast cancer. JAMA 1996;276: