It is generally agreed that the clinical staging of prostate cancer is highly inaccurate- be it the local, regional or systemic extent of disease. Focusing on the regional and systemic extent, the traditional work-up has consisted of a bone scan and computerized tomography (CT) or magnetic resonance imaging (MRI) to evaluate the regional nodes. Bone scans clearly have increased sensitivity over the metastatic bone survey, but they still miss a proportion of cases. In addition, bone scans lack specificity.
Attempts to evaluate the pelvic nodes with CT scan or MRI has not proven particularly effective. These imaging modalities define metastatic nodes strictly by size--i.e., nodes greater than 1 cm are defined as abnormal. By definition, therefore, nodal involvement less than 1 cm is interpreted as normal and, conversely, inflammatory or hyperplastic nodes might be erroneously read as neoplastic. That is, CT and MRI lack both sensitivity and specificity. One study (5), which primarily examined the ability of sonography and MRI to evaluate the local extent of disease, also provided data on the ability of MRI to detect nodal disease. In this study 189 patients had both MRI and pelvic lymph node dissection. 23 of the 189 patients were found to have nodal involvement; the MRI was able to detect only 1 of the 23 (4%). Similarly low sensitivity has been noted in CT studies (6). Because of these poor results, CT and MRI have been all but abandoned for pelvic lymph node staging.
Systemically administered mAbs, by being able to circulate and bind to a specific antigenic target, offer promise to increase both the sensitivity and specificity of current imaging techniques. In the case of prostate cancer, the target antigen would not have to be tumor-specific, organ-specificity would suffice. Any uptake outside the prostate itself would represent metatstatic disease.
Initial attempts to image prostate cancer began with mAbs to the established prostate-related antigens PSA and prostatic acid phosphatase (PAP) which had or approached requisite specificity. Attempts to image with mAbs to PSA met with little success (7). mAbs to PAP appear somewhat better. Babian et al (8) used systemically administered mAb PAY 276 to PAP in 25 patients with stage D2 disease. They studied total doses ranging from 5 to 80 mg of mAb: at each dose level, 1 mg of mAb was labelled with 111Indium (111In), the remainder consisting of unlabelled mAb. The antibody scan visualized disease in 19 of the 25 patients (76%). At dose levels of 15 mg and higher, all patients had at least one site of disease visualized. When the mAb scans were compared to the conventional bone scan and plain bone films-the conventional studies being considered definitive-the mAb scans were able to detect 7 of 32 lesions (22%) at the 20 mg dose, 31 of 58 lesions (53%) at the 40 mg dose and 101 of 134 lesions (75%) at the 80 mg dose. At the 80 mg dose level, which had the best result, the false negative rate was 25% while the false positive rate was 2.3%. The proportion of "false negatives" actually reflecting lack of specificity of the bone scan cannot be determined. 2/3 of the patients in this study had elevated serum PAP levels which was not thought to interfere with imaging. 8 of 16 patients tested developed human anti-mouse antibody, 6 of these at doses of 40 mg or higher. Only one patient in the series had documented nodal disease. His nodes were not visualized on the mAb scan. In another study of a different mAb to PAP, Leroy et al have used 123I-mAb 227A via peri-prostatic injection to image regional nodes prior to surgery (9). At doses of 100-400 ug of the F(ab')2 fragment in 32 patients, they could successfully image regional nodes with a sensitivity of 93% and a specificity of 83%.
Following trials of mAbs to PSA and PAP came another murine mAb originally designated 7E11-C5.3 and later renamed CYT-356. This mAb was initially produced by Horoszewicz et al (10) by immunizing mice with the human prostate cancer cell line LNCaP. The antigen detected by this mAb has recently been cloned and sequenced by Israeli et al (11) and characterized as a 100 kd glycoprotein which bears some homology to the transferrin receptor. The antigen includes an apparent transmembrane region and, combined with its apparent prostate organ-specificity, have led to the designation, prostate-specific membrane (PSM) antigen.
A phase I clinical trial of 111In-labelled CYT 356 was carried out in 80 patients utilizing doses from 0.1 to 10.0 mg (12, 13). 48 of the patients had stage D2 disease; the remaining 32 patients were pre-surgical patients scheduled for a staging lymph node dissection. 43 of the 48 stage D2 patients had bone metastasis on conventional bone scan. 25 of these 43 patients (58%) had positive CYT-356 antibody scans. Hormonal therapy did not compromise this sensitivity as 21 of 33 patients (64%) on hormonal therapy also had positive antibody scans. 4 of these 33 patients had disease detected by mAb scan but not by bone scan. In 2 of these 4 patients, the detected lesions could be confirmed. 8 patients in the study had biopsy-confirmed soft tissue metatstasis. CYT-356 detected disease in 5 of these 8 (63%). The detected sites included lung (n=2), retroperitoneal (n=1), mediastinal (n=1) and supraclavicular lymph nodes (n=1).
Thirty of the 32 pre-surgical patients actually underwent surgery. mAb CYT-356 imaged the prostate in 17 of the 30 (57%). Pelvic nodal disease was pathologically confirmed in 4 patients. The antibody scan detected nodes in 2 of these 4. CT or MRI was positive in 1 of these 4 patients.
A phase II study was subsequently done in 76 patients who were at high risk for nodal involvement and who were scheduled for pelvic lymph node dissection. These patients received 0.5 mg 111In-CYT-356. 32 of these 76 patients were found to have pathologically involved nodes. 21 of these 32 patients had both CT and mAb scans. The mAb scans were positive in 11 of the 21 patients (sensitivity=52%); the CT detected disease in 2 of the 21 (sensitivity=10%). Both of the patients with positive CT had positive mAb scans. Of 44 patients with pathologically negative nodes, the mAb scan was negative in 42 (specificity=95%).
An additional study with CYT-356 nearing completion is being performed in patients who have undergone radical prostatectomy or definitive radiotherapy but who have biochemical evidence of relapse (PSA > 0.8 ng/mL after surgery or 2 consecutive rising PSA determinations post RT) but no detectable disease on bone scan, CT and/or MRI (transaxial or endo-rectal). While this study is not yet complete, one center has reported on 20 such patients (14). 14 of 20 patients (70%) had positive mAb scans. All 14 were positive in the prostatic bed and 12 of the 14 were positive in lymph nodes. 9 of the 14 patients underwent biopsy of the prostatic bed. 7 of these 9 (78%) with a positive scan were biopsy-positive. 2 of 3 patients with a negative scan were biopsy-negative. Repeat injections and scans were performed in 3 patients after androgen ablation. 2 of the 3 demonstrated decreased activity in previously noted sites of nodal disease (Fig 4). It is worthy of emphasis that the sites of disease detected by mAb scan were not detectable by conventional imaging studies.
Although results are still relatively preliminary, the indications are that mAbs can indeed be utilized to image prostate cancer metastasis with potentially greater sensitivity and specificity than conventional studies. This would represent a significant advance in allowing improved selection of appropriate therapy, both for newly diagnosed patients as well as patients who relapse after initial therapy.
