University of Arizona Cancer Center - Department of Internal Medicine Section of Hematology and Oncology

RADIONUCLIDES IN O N C O L O G Y * STEPHEN E. JONES

Radionuclides have a rapidly expanding role in clinical oncology. As 'tumourscanning' agents, they are of established importance in: 1) assessing the initial extent of spread (stage) of tumours (and thus influencing decisions concerning therapy), and 2) providing a means of evaluating response to therapy and detecting metastasis or recurrence (in some instances, before it becomes clinically overt). With the continuing improvements in equipment and imaging techniques and the development of more useful radiopharmaceuticals, turnout scanning in certain settings can provide the clinical oncologist with a highly sensitive, easy, and safe non-invasive supplement to or, where necessary, substitute for other diagnostic procedures. Radionuclides also have potential as therapeutic agents and as predictors of response to therapy. The usefulness of radionuclides in these roles depends on a number of factors. For example, with regard to radionuclides as scanning agents, certain limitations must be considered: 1) value in relation to other means of assessment (in terms of accuracy, tumour specificity, cost, and necessity for confirmation by other means); 2) propensity of a specific turnout for occurrence in or spread to organs or sites for which specific scans are available or tumour specificity is high; and 3) problems with background (normal tissue) uptake. In this review I have attempted to present the current uses and limitations of the radionuclides by turnout type from the perspective of the practicing medical oncologist rather than from the viewpoint of the specialist in nuclear medicine.

RADIONUCLIDES AS TUMOUR-SCANNING AGENTS The mechanisms of tumour localization with radiopharmaceuticals are considered in detail elsewhere; however, it is pertinent to my discussion to briefly review the Key-words: Cancer; Oncology; Radionudides; Radiopharmaceuticals; Tumour scanning. * Reproduced in part from Seminars in Nuclear Medicine 6, 331, 1976. With permission of Grune and Stratton, Inc., New York. Accepted for publication on March 24, 1977. La Ricerca Clin. Lab. 7, 143, 1977.

143

RADIONUCLIDES IN ONCOLOGY

major types of scans available for use in the diagnosis and management of cancer and which, directly or indirectly, can be considered 'tumour scans'. As proposed by LILIEN 2s, radionuclides may be classified as belonging to one of three general categories of tumour-scanning agents, depending on whether there is specific uptake by tumour and whether use is limited to one organ. Most of the commonly used radiopharmaceuticals are in the first category, 'lzonspecific, limited-use' agents. With these agents, abnormalities on scan represent alteration or displacement of normal tissue. Hence, space-occupying lesions may be due to non-malignant processes (e.g., abscess or cyst) or tumour, and other measures (such as liver biopsy, sonography, or arteriography) may be necessary to accurately define the etiology of the abnormality detected by scan. Because the radiopharmaceuticals in this category generally can each be used for only one specific organ (e.g., liver-spleen, brain, bone), additional scans with other radiopharmaceuticals are often necessary to evaluate potential involvement of other organs. The second category, 'specific, limited-use' radiopharmaceuticals, includes agents that are also relatively specific for a particular organ or type of tissue, but which because of some special property of tumour allow distinction between tumour and normal tissue with external imaging. Examples of this type of radiopharmaceutical would include 13'I for some thyroid cancers 38, lsli.19.iodocholesterol for functioning adrenal tumours 6, and radiolabeled antibodies to tumour-associated antigens 14,18,37 It is obvious that the use of radiopharmaceuticals in this category is still quite limited; however, their potential use, particularly employing radiolabeled antibodies, is great. Extensive effort has been directed toward the development of the third category of radiopharmaceuticals, 'specific, general-use' agents, which ideally would be taken up specifically by tumour and which could be used for a broad spectrum of tumour types. It is hoped that better imaging equipment and new tumour-seeking agents (e.g., ones with characteristics which approach those of the 'ideal' agent (tab. 2) will improve the reliability of these agents in this clinical setting. Although such an agent is not yet available, two close approximations are currently in clinical use: 67Ga-citrate 1 and rain-labeled bleomycin 3o. Both of these agents can be used to visualize a wide variety of tumours; however, neither is truly tumourspecific and the occurrence of 'false-positive' and 'false-negative' uptake is a significant limitation.

THE ROLE OF RADIONUCLIDES IN DEFINING THE EXTENT OF TUMOUR It has become increasingly important to the oncologist in choosing appropriate therapy for a particular patient to have accurate knowledge of the extent or degree of spread (stage) of tumour. 'Staging' is critical in treatment planning both for the radiation therapist and for the medical oncologist. Inadequate staging can result in less than optimal treatment or excessive local treatment when the disease is more disseminated than is apparent from physical examination or routine laboratory and X-ray studies. The need to define the extent of tumour prior to treatment has led to many new technical advances (e.g., improved endoscopic equipment, computerized tomography), laboratory tests (e.g., assays for cardnoembryonic antigen or other tumour markers), and radiographic procedures (e.g., whole-lung tomography, lymphangiography). In certain settings, the need for accurate information is so great that 144

S. E. JONES

'tumour-scanning' agents - - for staging -

-

for assessing response

therapeutic agents -- labeled antibodies labeled drugs -- components of combined modality treatment -

-

predictors oJ drug e~cacy Tab. 1 - C u r r e n t and p o t e n t i a l uses of radionu c l i de s in oncology.

-- inexpensive (can be repeated as often as needed) safe (for personnel and patients) -- highly specific (taken up exclusively by tumour and not normal tissues) -- widely useful (taken up by many or most types of tumour) readily and easily imaged highly sensitive (can detect small metastatic /oci) reliable (negligible false-positive and false-negative rates) -

-

-

-

-

-

-

-

Tab. 2 - Characteristics of an ideal tumour-imaging agent.

additional surgical procedures are often used: bone marrow core and liver biopsy in patients with certain non-Hodgkin's lymphomas 2,20, exploratory laparotomy with splenectomy in patients with Hodgkin's disease 24,2s and laparoscopy in patients with ovarian cancer 39. The information obtained from these procedures has clearly influenced the choice, and consequently the results, of therapy and has led to improved survival for many patients 20.2s. However, many of these procedures have inherent limitations including expense and associated morbidity, which would preclude use in certain patients or frequent repetition during follow-up. Mthough less specific, nuclear medicine diagnostic procedures are, in contrast, easily performed, associated with much lower morbidity and high sensitivity, and they can be repeated safely as often as necessary. The usefulness of currently available scans in the staging of specific turnout types is considered in the following sections (tab. 3). MALIGNANT LYMPHOMA S

Although some general diagnostic yield could be expected from performing each of the four types of scans shown in tab. 3 [bone, liver-spleen, brain, general-use ('tumour-specific')] in each patient with newly diagnosed malignant lymphoma, the diagnostic yield can be greatly improved if the following points are considered. Malignant lymphomas usually arise in the lymphatic system, appear to spread via contiguous lymphatic channels to adjacent lymph node regions, and eventually disseminate in a relatively characteristic manner according to histopathologic type, natural 145

RADIONUCLIDES IN 0NCOLOGY

history, etc. 20,24,2s.With Hodgkin's disease, the likelihood of finding involvement of certain sites (with either clinical or pathologic staging procedures) at the time of initial presentation is generally low: bone marrow (about 5%), skeletal system ( < 5 % ) , liver ( < 5 % ) , brain (<1%)24. With the non-Hodgkin's lymphomas, the likelihood of finding involvement of the bone marrow, skeletal system, or liver at the time of initial presentation is considerably higher 20. For example, involvement of the bone marrow and liver is encountered in about one-half of patients with nodular lymphoma at presentation, whereas localized extra-lymphatic ('E') spread, including involvement of the skeletal system, is observed more often in patients with diffuse lymphomas. Bone marrow scanning There is little published information on bone marrow scanning during the initial staging of patients with lymphomas and further study of the potential uses of these types of scans is required. It is possible to image erythroid bone marrow with agents such as rain-chloride 29 and this procedure may prove to have some value in finding lymphoma in the bone marrow or assessing bone marrow 'reserves'. However, this site is readily evaluated with needle or open bone marrow core biopsy, which indicates reliably and easily the presence of bone marrow involvement 20,24 Bone scanning Lymphomatous involvement of the skeleton is uncommon at the time of diagnosis except in patients who have specific symptoms of bone pain or otherwise unexplained elevations of serum calcium or alkaline phosphatase 41. When bone involvement is found, it may be focal, readily included in conventional radiation therapy ports, and have generally favourable prognostic implications 24,as, or it may be widespread. We have found a very close association between abnormalities on bone scan and constitu-

type of tumour scan type of tumour

lymphomas muhiple myeloma breast melanoma lung colon other gastrointestinal tumours gynecologic(pelvic) turnouts sarcomas

prostate

bone

liver-spleen

brain

general-use

2+ i+ 4+ 2+ 3+ 0 0 0 2+ 4+

i+ 0 i+ 2+ 3+ 3+ 2+ i+ 2+ I+

-4-

2+

0 0 2+ 3+ 0 0 0 0 0

1+ 3+ 1+ 3+ 3+ 3+ 3+

-4--

-[-

Tab. 3 - Relative usefulness of radionuclides for initial staging of neoplasms (judged on the basis of a number of factors: likelihood of turnout to spread to a given site, specificity of available agents, frequency of false-positive or false-negative studies, etc.). 146

S. E. JONES

no. of patients with feature abnormal bone scan* presence of bone pain absence of bone pain abnormal radiographic studies* normal radiographic studies elevated alkaline phosphatase normal alkaline phosphatase elevated calcium normal calcium Hodgkin's disease non-Hodgkin's lymphoma, nodular non-Hodgkin's lymphoma, diffuse pathologic stage (Ann Arbor) I, IE, II, IIe III, III~ IV bone marrow involvement no bone marrow involvement

normal bone scan*

11 6 11 6 5 7 1 10 8 1 8 2 q¢~

i 14 ii

5°

* Abnormal: interpreted as being consistent with lymphomatous involvement of the skeletal system. Normal: included several examples of 'benign' abnormalities (hyperostosis frontalis, scoliosis, osteo-

arthritis, etc.). ** One false-positive scan and one scan indicating localized bone involvement (stage IIe). ° Of five patients with normal bone marrow biopsy specimens, three had only focal involvement of bone by lymphoma and one had lung cancer at autopsy metastatic to bone. Tab. 4 - Correlation between results of first bone scan and clinical, laboratory and pathologic features at time of scanning in 26 patients with lymphoma. (After SCHECHTERet al. 41, modified).

tional symptoms, widespread lymphoma, abnormal skeletal radiographs, and bone marrow involvement 41. In several cases we have found bone scanning to be especially useful in detecting skeletal involvement in patients free of pain, with normal laboratory tests, or with normal or nonspecific radiographs (tab. 4). Abnormal bone scans indicative of bone involvement by iymphoma are found more often in patients with diffuse non-Hodgkin's lymphomas rather than nodular lymphoma (tab. 5). Others have suggested that routine bone scanning during staging for Hodgkin's disease might detect a much higher frequency of localized involvement [localized extra-lymphatic ('E') spread] because it is more sensitive than plain X-rays ,i, but a larger series of consecutively studied patients with Hodgkin's disease and other lymphomas will be necessary to define the actual value of routine bone scanning during staging in these patients. Our own approach has been to utilize bone scanning in the initial staging of patients with lymphoma if clinically indicated (e.g., bone pain, unexplained elevation of serum calcium, or alkaline phosphatase) or if the finding of bone disease would influence our choice of therapy. 147

RADIONUCLIDES IN ONCOLOGY

Liver scanning Although the overall incidence of lymphomatous involvement of the liver at the time of presentation is relatively low, in certain clinico-pathologic settings, liver involvement is much more likely: mixed cellularity or lymphocyte depletion type of Hodgkin's disease or older patients with constitutional symptoms or abdominal/ inguinal node presentations. Estimation of liver involvement on the basis of hepatomegaly or abnormal liver function test results is unreliable 24. Liver scanning as an initial staging procedure also has serious limitations. Although large focal filling defects in the liver or spleen are frequently proven to be due to lymphoma 31,42, organomegaly or non-uniform uptake of radionuclide are more common findings and may or may not be due to lymphoma. For example, involvement of the liver is documented histologically in nearly one-half of patients with nodular lymphocytic lymphoma undergoing liver biopsy 2; however, such involvement frequently consists of little more than small focal subcapsular infiltrates of cytologically abnormal cells. Liver scans in these patients are often normal or reveal only nonspecific abnormalities (e.g., slight hepatomegaly or patchy uptake of isotope). Confirmation of involvement in these cases requires careful histologic examination of liver or spleen. Thus, radionuclide scanning usually does not help the clinician in determining whether either or both of these organs are involved by lymphoma. Brain scanning Lymphomatous involvement of the central nervous system (CNS) is rare at the time of initial presentation, and there is little indication for brain scanning unless

patients with bone involvement histologic type frequency

%

3/15 7/123 4/47 4/52 0/I5 18/252

20 6 9 8 0 7

disuse

undifferentiated histiocytic mixed lymphocytic, poorly differentiated lymphocytic, well differentiated

total nodular histiocytic mixed lymphocytic, poorly differentiated lymphocytic, well differentiated

total

1/30 1/81 1/79 0/9 3/199

Tab. 5 - Frequency of bone involvement in 451 patients with non-Hodgkin's lymphomas (all stages, therapy). (After SCHECHTERet al. **, modified).

no prior

148

S. E. JONES

scans

disease present

positive negative

51 (TP)

total

54

3 (FN)

no evidence of disease 13" (FP) 8 (TN) 21

* Includes five patients with false-positivediffuse lung labeling. Tab. 6 - Overall results of scanning versus disease activity in panents with lymphoma. TP = truepositives; FP = false-positives;FN = false-negatives; TN = true-negatives.

specific symptoms or signs of neurologic disease are present. However, in childhood lymphoma or in diffuse lymphocytic, convoluted cell lymphoma (probably of T-lymphocyte origin) the likelihood of CNS involvement is much greater 34, and an initial brain scan as well as a cytologic examination of the cerebrospinal fluid are indicated. Moreover, during follow-up of patients with non-Hodgkin's lymphomas of the diffuse type, the appearance of focal neurologic findings - even those suggesting a spinal cord compression syndrome - is an indication for brain scanning as part of the evaluation for diffuse leptomeningeal lymphoma 16

'Turnout-specific', general-use scanning 'Turnout-specific', general-use scans have some value in staging patients with lymphoma, but the data generated by these techniques are 'soft', that is, confirmation of lymphomatous involvement by other techniques is usually required. Two radiopharmaceuticals have had considerable evaluation in patients with lymphomas: ntln-bleomycin and 67Ga. In a study of mln-bleomycin turnout scanning in patients with lymphoma, we found that abnormalities on scan accurately reflected sites of lymphomatous involvement in 88% of the 41 patients with known active lymphoma (tab. 6). Published results with 67Ga are similar: LILIEN et al. 3o summarized the experience in 469 patients with active lymphoma and found that lymphoma was successfully imaged in 75% of cases. With both agents, however, a sizeable fraction of patients with active lymphoma (12-25%) do not accumulate sufficient radioactivity in sites of lymphomatous involvement to permit differentiation between lymphoma and normal tissue uptake ~5.2,26,43. These 'false-negative' scans are a significant problem with the currently available agents and may be due to technical factors (equipment, imaging energy), size of the lesion (nodes or masses less than 2 cm in diameter seldom can be imaged), or failure of the turnout to take up the radiopharmaceutical. With regard to the latter possible factor, a recent report has suggested that there might be variation in uptake of 67Ga by different histologic types of non-Hodgkin's Iymphomas, with histiocytic tumours being imaged most successfully 27. However, all histologic types of lymphomas were imaged equally well in our series of patients receiving lnln-bleomycin 22. Enthusiasm for the high overall true-positive rates obtained with current 'turnoutspecific' agents in patients with lymphoma must be tempered by several clinically important factors. First, when there are multiple sites of involvement (as is frequently the case with lymphomas), the scan may detect most but not necessarily all sites of 149

RADIONUCLIDES IN 0NCOLOGY

disease for the reasons previously enumerated. Second, nonspecific, non-turnout related accumulations of radioactivity ('false-positives') can occur. In some areas, such as the mediastinum (about 5% of scans with 67Ga or tllIn-bleomycin) or the lungs (10% of studies with mIn-bleomycin), these accumulations can be readily discerned from tumour uptake with plain chest X-rays 22.26,43. Because the predominant route of excretion for 67Ga is through the gastrointestinal tract, false-positive abdominal uptake has been a frequent problem with that agent 1,15,26, 27.43, but it is an uncommon problem (3% of scans) when mIn-bleomycin is used 22.3o.Nonetheless, the distinction between involvement of retroperitoneal lymph node sites v e r s u s nonspecific gastrointestinal tract uptake can be quite troublesome at times. With mIn-bleomycin, some sites such as the liver and bone marrow are much more difficult to assess because of background uptake by normal tissue. Although we did find that bone marrow involvement was frequently associated with a patchy or non-uniform uptake of mIn-bleomycin on scan 22, this is a rather unreliable method for identifying bone marrow involvement. Finally, in almost all of the cases of lymphoma studied with nlln-bleomycin scanning at our institution, the sites of involvement by lymphoma could be readily and more accurately detected by other means, such as careful physical examination, conventional X-ray procedures, or readily obtained biopsies. Thus, scanning with this agent provided little cAinically useful information that we did not already have. This fact served as a basis for a careful site-by-site comparison of the results of mlnbleomycin scanning with the presence or absence of lymphoma. Among 731 sites of potential lymphomatous involvement, the true-positive rate was 90%; the falsenegative rate, 10%; the false-positive rate, 4%; and the true-negative rate, 96% (tab. 7). Similar results have been obtained with 67Ga in the evaluation of patients with Hodgkin's disease 26,43 Thus, turnout scanning with current general-use agents plays only a supplementary role to other more standard procedures in staging patients with lymphomas and clinical decision-making based solely on the findings from these scans is not yet reliable. MULTIPLE MYELOMA

Radionuclide scanning has a much smaller role in the initial staging and management of multiple myeloma. Since liver, spleen, and brain involvement are rare at the time of diagnosis, routine performance of these scans has little clinical value. However, bone scanning and general tumour scanning are of somewhat greater clinical use.

scan result

individual sites involved by lymphoma

positive negative

129 (TP) 15 (FN)

total

144

sites uninvolved by lymphoma 24 (FP) 563 (TN) 587

Tab. 7 - Accuracyof scanning 731 valuable sites. Abbreviations as in tab. 6. 150

S, E. JONES

Bone scanning

At least two-thirds of patients with multiple myeloma have radiologically detectable bone involvement. In general, these patients have a high turnout cell mass (in excess of 1 x 10 '2 myeloma cells) in the body, most of which is located in the bone marrow and adjacent bone 9. In contrast to other neoplasms that metastasize to bone and are associated with both osteolytic and osteoblastic processes, myeloma is almost always entirely osteolytic with very little new bone formation. Osteolysis in myeloma appears to result from secretion of a peptide by myeloma cells, which activates osteoclasts and leads directly to bone resorption 3s and, hence, radionuclides requiring new bone formation for their incorporation will obviously be less useful in myeloma than in other forms of cancer. Although bone scanning is occasionally useful in documenting some sites of involvement in myeloma, reliance on scans to assess the extent of skeIetal disease may grossly underestimate the degree of involvement. The benefit of greater sensitivity in detecting changes in bone disease compared with conventional radiography that exists for diseases like lymphoma and breast cancer is not realized in myeloma. The role of bone scanning in the assessment of healing of osteolytic lesions in patients responding to chemotherapy (e.g., to determine whether new bone formation is occurring at a rate greater than that prior to therapy) must still be determined. General-use tumour scanning These agents have limited useft, lness in multiple myeloma. In our own studies of labeled bleomycin, for instance, we found that the typical scan in a patient with myeloma revealed only a general and rather uniform reduction in bone marrow radioactivity, although extra-osseous sites occasionally accumulated radioactivity z2.30. In this sense, the bulk of turnout residing in the bone marrow was not imaged, probably as a result of displacement of normal erythroid elements, which appear to be responsible for the uptake of rain 29. In one case, however, we observed a striking uptake of radioactivity in amyloid deposits about the hip and shoulder girdles. Labeled bleomycin compounds may be of some value in defining systemic amyloidosis in patients with myeloma, but study of additional cases will be necessary.

B R E A S T CANCER

Bone scanning Bone scanning has become the most important use of radionuclides in the management of breast cancer. Such scans are a sensitive initial screen for bony metastases and frequently will detect involvement of the skeletal system in 30-50o70 of patients with normal radiographic studies of the same region 36. False-negative scans are uncommon. However, abnormalities on scan may be nonspecific (not due to cancer) in 3-10O7o of studies s3,36, and specific X-ray views of the abnormal regions on scan must be obtained to exclude benign causes for the abnormalities observed. We now use bone scanning for staging in breast cancer in two major clinical settings. Early disease - Bone scans should be performed preoperatively in patients with clinical or mammographic findings suspicious for breast cancer. When bone scanning

151

R A D I O N U C L I D E S IN O N C O L O G Y

initial screen: results of bone scan

confirmatory study: results of radiographs

treatment based on final assessment of skeletal involvement

normal

not done if bone scan normal

adjuvant treatment: * eight courses of adriamycin + cydophosphamide ± local irradiation; after treatment, annual bone scans

equivocal or abnormal

normal or 'benign' abnormality

adjuvant treatment: same as above except follow-up bone scans obtained every 6 months

abnormal

abnormal indicating metastatic disease

ineligible for adjuvant program; treatment for advanced cancer

* Treatment is part of an investigational adjuvant breast cancer program ~ofor patients at high or intermediate risk for recurrence. Tab. 8 - Role of bone scanning in influencing the immediate postoperative treatment of patients with early (stage II) breast cancer at University of Arizona Cancer Center.

is used in this manner, metastases to the skeleton are detected in about 10% of cases s; some of these patients could therefore be spared from aggressive surgical procedures. I n general, more surgeons are now performing scans as part of their preoperative staging evaluation, and this trend will continue with further recognition of the value of bone scanning prior to mastectomy. I f a scan has not been performed preoperatively it should be done as soon after mastectomy as is feasible. Unfortunately, more nonspecific chest wall abnormalities are seen in the area of the mastectomy when scans are done postoperatively. W e are currently evaluating the results of bone scanning as part of an adjuvant breast cancer treatment program for patients with stage I I disease 4o. Involvement of axillary lymph nodes or a large primary lesion are adverse prognostic factors and indicate a high likelihood of occult, distant micrometastases 13. A remarkable decrease in the recurrence rate following mastectomy is now being observed with the use of adjuvant chemotherapy in these high-risk patients 3.12. Axillary lymph node involvement, in addition to its relationship to micrometastases, is also associated with a greater likelihood of overt metastatic disease to the skeleton s, and we use bone scanning in our own adjuvant chemotherapy program to help us distinguish patients with overt metastatic disease from those who are only at high risk for occult micrometastases. Tab. 8 summarizes how we use the results of a screening bone scan in our adjuvant breast cancer research program. A d v a n c e d disease - W e frequently obtain bone scans for patients we see in consultation who have a history of breast cancer or who are experiencing their first recurrence of disease. I n many instances, bone disease is found to be more extensive than is suspected clinically 36; in other instances, the scan serves as a useful base-line examination for later follow-up observations. Liver scanning

Although liver scanning may indicate gross hepatic involvement by breast cancer, it has the same drawbacks for initial staging of patients with breast cancer as it does 152

S. E. JONES

for patients with lymphoma. We generalIy perform liver scans only if we have a high index of suspicion of metastatic disease to that organ (e.g., hepatomegaly, abnormal liver function tests, particularly an elevated serum alkaline phosphatase). As with most other solid turnouts, large focal filling defects on liver scan generally correlate with metastatic disease to the liver, but small, miliary lesions or diffuse infiltration may be missed completely on scan, or the scan may reveal only nonspecific abnormalities. If treatment decisions will be influenced by whether or not there is metastatic breast cancer in the liver, we usually require more direct confirmation (e.g., percutaneous needle biopsy, laparoscopy with biopsy, laparotomy with direct biopsy).

Brain scanning Brain scanning in patients with early breast cancer has little value unless the patient has definite neurologic findings or symptoms.

General-use tumour scanning In most cases, primary breast lesions or gross metastatic disease can be imaged with 67Ga or mIn-bleomycin, but these scans rarely provide the clinician with information that he does not already have from careful physical examination or conventional radiography. One potential use of such radiopharmaceuticals is for the preoperative non-invasive prediction of axillary lymph node metastases and this is under active investigation at the present t~me. LUNG CANCER

Radionuclides play an important role in the initial evaluation of patients with bronchogenic carcinoma, particularly patients with squamous cell or oat cell carcinoma, to screen for hepatic or bony metastases. Although brain metastases are detected in only about 5% of all patients with lung cancer undergoing initial evaluation, there is a higher incidence in patients with oat cell carcinoma ,7. Brain scans should be done routinely for patients with this diagnosis. The general-use tumour scans are of considerably less clinical value in patients with lung cancer because, although 75-95% of these turnouts accumulate radioactivity and can be imaged after injection of 6TGa or mln-bleomycin *'30, there is generally no difliculty in detecting the tumour with conventional chest X-rays. MELANOMA

There is little published information on the use of tumour scanningin the staging of patients with malignant melanoma. We have found that nearly 90% of metastatic deposits can be visualized after injection of mIn-bleomycin and about 70% of these tumours can be imaged with 67Ga 3o. Hence, we use a mIn-bleomycin tumour scan in the evaluation of patients who have regional node involvement, invasive primary lesions (e.g., Clark's levels 3-5), or their first recurrence (e.g., subcutaneous nodule) to attempt to define more extensive disease. On occasion, this has been successful. Other scans (liver, bone, brain) are generally performed as clinically indicated. Patients with widespread melanoma or neurologic symptoms or signs should have

153

R A D I O N U C L I D E S IN O N C O L O G Y

a brain scan because this tumour frequently involves the CNS, and treatment of metastatic disease in this site requires whole-brain irradiation or, occasionally, surgery. OTHER SOLID TUMOURS OF ABDOMINAL ORIGIN

Cancers arising in the gastrointestinal tract or reproductive organs can certainly metastasize to lungs, bone, or brain, but they are more likely to spread within the abdomen. Hence, the general-use radiopharmaceuticals like mIn-hleomycin have been quite useful to us in the management of these turnouts 30. Often when we are asked to see patients postoperatively, the bulk of the tumour has been removed but there is residual disease in the abdomen. Physical examination may be difficult or the findings uncertain and we are confronted with having to plan treatment without any easy means of assessing the extent of residual turnout. 70-90% of these turnouts (e.g., pancreas, colon, ovary, cervix) take up mIn-bleomycin and this agent appears to be superior to 6VGain this group of neoplasms 30 (probably because there is much less difficulty with gastrointestinal tract accumulation with mIn-bleomycin). Preoperative liver scans are particularly useful in patients suspected of having a malignancy arising in the gastrointestinal tract 7,10 Early recognition of probable metastatic disease to the liver may spare some patients more aggressive cancer operations (e.g., Whipple procedure, colostomy) and thus influence treatment decisions. The surgeon who is aware of the results of liver scanning may elect to perform a relatively simple end-to-end anastomosis for an obstructing lesion in the colon and at the same time he should attempt to confirm the suspicion of metastatic disease to the liver by appropriate biopsy. Biopsy of the liver is especially important in pelvic and gastrointestinal cancers, since not all nodules observed on the surface of the liver at the time of operation nor all of the filling defects noted on scan are due to cancer. Confirmatory biopsies should be obtained in all cases.

THE ROLE OF RADIONUCLIDES IN THE SUBSEQUENT FOLLOW-UP OF PATIENTS WITH CANCER These agents have established roles in the follow-up and management of patients with cancer. The various scans can be used for serial screening for recurrent cancer if therapy has been given with curative intent. They also permit some evaluation of response to treatment (radiation therapy or chemotherapy)1.21.-.30,n,36.41. Finally, scans may be used to detect relapse before it is otherwise clinically apparent 22. The relative usefulness to the practicing oncologist of these procedures during follow-up of the patient with cancer is indicated in tab. 9. Bone scanning has become the easiest and probably most reliable means of assessing response to chemotherapy in patients with known skeletal metastases for two reasons. First, although relief of bone pain with chemotherapy occurs frequently (particularly in patients with breast cancer), this has been criticized as a subjective rather than an objective criterion of response. Second, conventional radiographs may be extremely slow to show bone healing, although new bone formation, detectable by scanning, may have been occurring for a considerable period of time. Thus, serial bone scans appear to permit more rapid objective assessment of response to treatment with chemotherapy 21.n. 154

S.E.

JONES

type of tumour scan yl~c of

gmphomas aultiple myeloma ,tease

aelanoma mg olon pther gastrointestinal tumours cnecologic (pelvic) tumours arcomas ,rostate

bone

liver-spleen

brain

general-use

3+ 2+ 4+ 2+ 3+ 0 0 0 1+ 4+

-i nt -

2+ 0 3+ 3+ 3+ 0 0 0 1+ 0

2+

0 t+ 2+ 2+ 3+ 2 1+ 1+ 1+

-4-

I+ 3+ 3+ 3+ 3+ 3+ 3+ --I-

Tab. 9 - Relative usefulness of radionudides for subsequent management of neoplasms. (Based on several considerations such as the frequency with which a tumour is likely to spread to a particular site with time, the accuracy of the scan in identifying involvement of that site, and the clinical setting).

Serial liver scanning is useful in following selected patients (such as those with known hepatic metastases) for response to treatment. Brain scans are indicated in patients who have developed neurologic symptoms or signs and probably in patients with tumours with a known propensity for spread to the CNS. For example, brain metastases develop in as many as 40% of patients with oat cell carcinoma of the lung t7. Indeed, 'prophylactic' whole-brain irradiation is included in many current programs for the treatment of this tumour. Patients with histiocytic lymphoma, leukemia, melanoma, and breast cancer also may first relapse in the CNS, which is a 'drug sanctuary', since most cancer chemotherapeutic agents do not cross the blood-brain barrier. 10 % of our patients with breast cancer who are responding in peripheraX sites (e.g., lungs, skin, nodes, bone) to combination chemotherapy have CNS manifestations as the first sign of relapse 10. Early detection of brain metastases or leptomeningeal tumour is important so that neurologic deficits may be minimized with prompt and appropriate treatment. Computerized axial tomography, a recent major advance in the early detection of CNS lesions, may eventually supplant some of the uses of brain scanning. The role one attributes to the general-use turnout scans in the follow-up of patients with cancer depends on whether there are better means of follow-up or assessment of response. In patients with abdominal cancers, we often do not have adequate methods of follow-up and have relied heavily on the use of these scans coupled with abdominal sonography. In patients with lymphomas, we have relied more on conventional means of assessing response, although mIn-bleomycin scan results do generally reflect the course of the disease 22. The greatest limitation of these agents at this time is their nonspecific uptake in normal tissues, particularly in patients who are thought to be free of disease by clinical criteria. For example, in 21 patients with lymphoma in remission, only eight scans with lnln-bleomycin were completely normal; 13 were judged to be falsely positive ". Similar findings have 155

RADIONUCLIDES IN ONCOLOGY

been observed with 6VGa. However, it is in this group of patients who are apparently in remission that we most want accurate information about occult, active disease, and it is in this group that these general-use tumour scans are currently the least reliable. It is hoped that better imaging equipment and new tumour-seeking agents will improve the usefulness of these procedures in the follow-up of patients with lymphoma and other cancers.

RADIONUCLIDES AS THERAPEUTIC AGENTS

It is appropriate in this paper to consider the therapeutic use of radionuclides as one component of combined modality therapy for cancer. For example, at the University of Arizona, a combined therapeutic modality approach has been developed for patients with thyroid cancer who are at high risk for local or distant recurrence after conventional treatment. As first reported by DURIE et al. 8 15 of 65 patients with thyroid cancer to date were considered eligible for this program, which consists of total thyroidectomy with neck dissection (if there are positive lymph nodes) followed by 131I and repetitive courses of combination chemotherapy employing the drugs known to have activity in thyroid cancer (adriamycin, bleomycin, vincristine, and melphalan). The combined use of 131I and chemotherapy is based on the assumption that different mechanisms of tumour cell kill might be exploited. Well-differentiated residual thyroid cancer cells are more likely to take up a31I whereas more undifferentiated or anaplastic cells might be eliminated with chemotherapy. Although the results of this pilot program are quite preliminary, the data suggest that this combined approach will be of value for patients with high-risk thyroid cancer. The radiolabeting of drugs (particularly tumour-seeking drugs like bteomycin) or antibodies (such as antibodies against tumour-associated antigens) is also a potential means of delivering lethal irradiation directly to the cancer cell and may become a practical form of therapy in the future.

RADIONUCLIDES IN THE PREDICTION OF RESPONSE TO ANTICANCER DRUGS Another potential and special role for certain types of radiopharmaceuticals is in predicting tumour responsiveness to a particular drug. 13'I is currently the only radiopharmaceutical in use that does predict response to therapeutic doses of the agent in certain metastatic thyroid cancers 38 During our initial evaluation of labeled bleomycins as tumour-imaging agents, we became interested in learning whether patients whose tumours accumulated radioactivity after the administration of mln-bleomycin would respond to therapeutic doses of bleomycin. Only 2 of 13 patients (15%) who received treatment with bteomydn had objective regressions of tumour, despite the obvious fact that radioactivity had appeared in known tumour sites in all cases shortly after the injection of labeled bleomycin 23. We concluded that n'In-bleomycin turnout scanning does not predict responsiveness to subsequent treatment with bleomycin, a fact that now appears obvious since 70-80% of all solid tumours, most of which do not respond to treatment with bleomycin, can be imaged after injection of H'In-bleomycin 30 The lack of predictive value of mIn-bleomycin in this regard points out our current lack of knowledge about both the pharmacokinetics of this radiopharmaceutical and 156

S. E. JONES

the mechanisms and sites of drug a n d / o r radionuclide uptake after i.v. administration. I n mice, t u m o u r uptake of a tightly labeled tritiated bleomycin does correlate w i t h w h e t h e r or n o t the turnout will r e s p o n d to treatment w i t h bleomycin tg. I n man, a m o r e realistic evaluation of bleomycin for this purpose will be possible w h e n a tightly labeled bleomycin w i t h fully p r e s e r v e d biologic activity is synthesized and the mechanisms of t u m o u r u p t a k e are b e t t e r understood. STCo-bleomycin already appears to b e a b e t t e r and more specific tumour-imaging agent (although there are serious p r o b l e m s w i t h disposal of radioactive urine after its administration). T h e p o t e n t i a l uses of other labeled drugs, particularly those that are highly active against a w i d e variety of tumours, are obviously great, and further advances in this area can be anticipated.

SUMMARY Radionuclides already have a major rote in the daily practice of oncology and will, undoubtedly, be of even greater importance in the future. The variety of current and potential applications is shown in tab. 1. Their major use at this time is, in the broadest sense, for 'tumour scanning', which includes the evaluation of specific organs for the presence of tumour (usually with different radiopharmaceuticals for each organ) and the entire body (generalized tnmour searches with radiopharmaceuticals like 6VGa-citrate or rain-labeled bleomycin). The clinician uses these agents in the initial evaluation of extent of tumour (staging) and in the subsequent management of the patient with cancer to assess response to treatment, to reveal early relapse, and to assist in making decisions concerning treatment. The other potential roles for radionuclides are also briefly considered in this review.

REFERENCES I ) ANDREWSG.A., E~wa}ms C.L.: Tumour Scanning with Gallium 67 - J. Amer. reed. Ass. 233, 1100, 1975. 2) BAGLE¥C. M., Jr., THOMAS L.B., JOHNSON R.E. et al.: Diagnosis of Liver Involvement by Lymphoma: Results in 96 Consecutive Peritoneoscopies - Cancer 3I, 840, 1973. 3) BONADONNAG., BR~SSAMmNOE., VALAGUSSAP. et al.: Adjuvant Chemotherapy with Cyclophosphamide, Methotrexate, and 5-Fluorouracil for Early Breast Cancer - New Engl. J. Med. 294, 405, 1976. 4) CELLERINOA., FILIPPz P.G., CHIANTARETTOA. et al.: Operative and Pathologic Survey of 50 Cases of Peripheral Lung Tumours Scanned with 67 Gallium - Chest 64, 700, 1973. 5) C~Al~K~SN.D., MALMUDL. S., CASWET.LT. et al.: Preoperative Bone Scans. Use in Women with Early Breast Cancer - J. Amer. reed. Ass. 233, 516, 1975. 6) CONN J.W., MORtTA R., COHEN E.L. et al.: Primary Aldosteronism: Photoscanning of Tumours after Administration of u'I-19-Iodocholesterol - Arch. intern. Med. 129, 417, 1972. 7) DaviEs R. J'., V~RNON M., CROFT D.N.: Liver Snaps and the Detection of Clinically Unsuspected Liver Metastases - Lancet i, 279, 1974. 8) DU~IE B. G. M., HELLMAND., O'MARA R. E. et aL: Multimodality Treatment for High Risk Thyroid Carcinoma - Proc. Amer. Ass. Cancer Res. 17, 277, 1976. 9) DUl~I~B. G. M., SALMONS.E.: A Clinical Staging System for Multiple Myeloma: Correlation of Measured Myeloma Cell Mass with Presenting Clinical Features, Response to Treatment, and Survival - Cancer 36, 842, 1975. 10) FEE H.J., PlmKOl, E.K., CAMERONJ.L. et al.: Liver Scanning in Patients with Suspected Abdominal Turnout - J. Amer. reed. Ass. 230, 1675, 1974. 11 ) FERRANTA., RODHAINJ., MICHAUXJ. L. et al.: Detection of Skeletal Involvement in Hodgkin's Disease: A Comparison of Radiography, Bone Scanning, and Bone Marrow Biopsy in 38 Patients - Cancer 35, 1346, 1975. 12) FISHER B., CAVnmNE P., ECONOMOU S.G. et al.: L-Phenylalanine Mustard (L-PAM) in the Management of Primary Breast Cancer. A Report of Early Findings - New Engl. J. Med. 292, 117, 1975. 157

RADIONUCLIDES IN ONCOLOGY

13) FISHERE. R., GREGoRIoR.M., FISHER B.: The Pathology of Invasive Breast Cancer. A Syllabus Derived from Findings of the National Surgical Adjuvant Breast Project (Protocol No. 4) Cancer 36, 1, 1975. 14) GHOSE T., GucLu A., TAI J. et al.: Antibody as Carrier of t31I in Cancer Diagnosis and Treatment - Cancer 36, 1646, 1975. 15) GREENLAWR.H., WEINSTEIN M.B., BRILL A.B. et ai.: 6'Ga-Citrate Imaging in Untreated Malignant Lymphoma. Preliminary Report of Cooperative Group - J. nucl. Med. 15, 404, 1974. 16) GRIFFINJ. W., THOMPSONR. W., MITCmNSONM.J. et al.: Lymphomatous Leptomeningitis Amer. J. Med. 51,200, 1971. 17) HANEENH.H.: Should Initial Treatment of Small Cell Carcinoma Include Systemic Chemotherapy and Brain Irradiation? - Cancer chemother. Rep. 4, 239, 1973. 18) HOFFERP.B., LA:rHROPK., REFETOFFS. et al.: Use of Radioiodine-Labeled Anti-CEA Antibody as a Tumor Scanning Agent - J. nucl. ivied. 14, 408, 1973. (Abstract). 19) ICHIKAWAT., UMEgAWAH., OHASHI S. et al.: Animal Experiments Confirming the Specific Effect of Bleomycin against Squamous Cell Carcinoma - In: Progress in Antimicrobial and Antineoplastic Chemotherapy. Proceedings of the 6th International Congress of Chemotherapy. Tokyo, 1969. University Park Press, Baltimore, 1970; vol. 2, p. 315. 20) JONES S.E.: Non-Hodgkin Lymphomas - J. Amer. med. Ass. 234, 633, 1975. 21) JoNEs S.E., DURIE B.G.M., SALMON S.E.: Combination Chemotherapy with Adriamycin and Cyclophosphamide for Advanced Breast Cancer - Cancer 36, 90, 1975. 22) JoNEs S.E., LILIEN D. L., O'MARA R.E. et al.: Indinm-lll Bleomycin Turnout Scanning in Lymphoma - IVied. pediat. Oncol. 1, 11, 1975. 23) JONES S. E., SALMONS. E., DUXlEB. G. M.: Failure of Indium-Ill-Labeled Bleomycin Tumour Scanning to Predict Response to Bleomycin (NSC-125066) Treatment - Cancer chemother. Rep. 58, 733, 1974. 24) KAPLANH.S.: Hodgkin's Disease - Harvard University Press, Cambridge, Mass., 1972. 25) KAPLANH.S., ROSENBERGS.A.: The Management of Hodg!dn's Disease - Cancer 36, 796, 1975. 26) KAY D. N., McCv,EADYV.R.: Clinical Isotope Scanning Using 67Ga-Citrate in the Management of Hodgkin's Disease - Brit. J. Radiol. 45, 437, 1972. 27) LEvi J.A., O'CoNNELL M.J., MURPHY W.L. et al.: Role of 67Gallium Citrate Scanning in the Management of Non-Hodgkin's Lymphoma - Cancer 36, 1690, 1975. 28) LILIEN D.L.: The Current Status of Tumour Imaging - J. Amer. med. Ass. 230, 735, 1974. 29) LILIEN D. L., BERGENH. G., ANDERSOND. P. et al.: rain-Chloride: A New Agent for Bone Marrow Imaging - J. nucl. Med. 14, 184, 1973. 30) LILIEN D. L., JONES S.E., O'MARA R.E. et al.: A Clinical Evaluation of Indium-Ill Bleomycin as a Tumour-Imaging Agent - Cancer 35, 1036, 1975. 31) LIPTON M.J., DENAI~o G.L., SILVEKMANS. et al.: Evaluation of the Liver and Spleen in Hodgidn's Disease. I. The Value of Hepatic Scindgraphy - Amer. J. Med. 52, 356, 1972. 32) LLOYDR.E., JONES S.E., SALMONS.E. et al.: Combination Chemotherapy with Adriamycin (NSC-123127) and Cyclophosphamide (NSC-26271) for Solid Tumours: A Phase II Trial Cancer Treat. Rep. 60, 77, 1976. 33) LOEPFLERR. K., DISIMONE R.N., HOWLANDW.J.: Limitadons of Bone Scanning in Clinical Oncology - J. Amer. med. Ass. 234, 1228, 1975. 34) LUKES R.J., COLLINS R.D.: New Approaches to the Classification of the Lymphomata Brit. J. Cancer 31 (Suppl. 2), 1, 1975. 35) MUNDYG.R., RAISZ L. G., COOPERR.A. et al.: Evidence for the Secretion of an Osteoclast Stimulating Factor in Myeloma - New Engl. J. Med. 29I, 1041, 1974. 36) O'MAmt R.E.: Bone Scanning in Osseous Metastatic Disease - J. Amer. reed. Ass. 229, 1915, 1974. 37) O~DERS. E., BLOOMERW. D., JONES A. G. et al.: Radionuclide Immtmoglobulin Lymphangiography: A Case Report - Cancer 35, 1487, 1975. 38) POCHINE.F.: Thyroid Adenocarcinoma: A Functioning Tumour - Lancet i, 94, 1969. 39) ROSENOFPS.H., YOUNGR. C., ANDERSONT. et al.: Peritoneoscopy: A Valuable Staging Tool in Ovarian Carcinoma - Ann. intern. Med. 83, 37, 1975.

158

S, E. JONES

40) SALMONS.E.: Progress in Adjuvant Chemotherapy of Early Breast Cancer - Ariz. Med. 32, 108, 1975. 41) SCHECHTERJ.P., jom~s S.E., WOOLFENDEN J.M. et al,: Bone Scanning in Lymphoma Cancer 38, 1142, 1976. 42) SILVE~MANS., DENAItDO G.L., GLATSTEINE. et al.: Evaluation of the Liver and Spleen in Hodgkin's Disease. II. The Value of Splenic Scintigraphy - Amer. J. Med. 52, 362, 1972. 43) TURNERD. A., PINSKY S.M., GOTTSCHALKA. et al.: The Use of ~7Ga Scanning in the Staging of Hodgkin's Disease - Radiology 104, 97, 1972.

Requests/or reprints should be addressed to: STEPHEN E. JONES Department o~ Internal Medicine Section o~ Hematology and Oncology University o~ Arizona Cancer Center Tucson, Ariz. 85724 - USA

159