Status of Radiolabeled Monoclonal Antibodies for Diagnosis and Therapy of Cancer

Almost exactly one decade ago, in an editorial published in theNew England Journal of Medicine [1], I noted that "progressis slow but sure" in the development of monoclonal antibodiesfor clinical use. At that time, only muromonab-CD3 (OrthocloneOKT3) was approved for human use to prevent rejection of kidneytransplants. In the ensuing 10 years, only one oncologic monoclonalantibody product, satumomoab pendetide (OncoScint CR/OV, Cytogen,Princeton, New Jersey) [2] has been approved by the FDA. Progresssurely has been slow.

A Break in the Log-Jam?

What an oncologist wants to know is, has the log-jam been broken?or, said another way, will any of these products be availablefor me to evaluate in my patients? The answer is a conservativemaybe. In the last year, the FDA has presented two oncology imagingproducts to advisory committees: Verluma (NR-LU-10, NeoRx, Seattle,Washington) [3] for imaging extensive small-cell carcinoma ofthe lung, and CEA-Scan (arcitumonab, Immunomedics, Morris Plains,New Jersey) [4] for imaging cancer of the colon or rectum. Bothof these antibodies have been recommended for approval. Divgi'sarticle discusses the clinical results with OncoScint and CEA-Scan,but doesn't describe the results of the Verluma studies.

Several factors noted by Divgi deserve emphasis. OncoScint (anindium-111-labeled intact anti-TAG72 immunoglobulin G [IgG]) wasapproved for one-time extrahepatic imaging of presurgical patientswith colorectal or ovarian cancer. It appears that these two restrictionshave limited the use of OncoScint in the clinic. Recently, basedon data presented by Cytogen, the FDA has decided to allow OncoScintto be used multiple times, but the restriction to extrahepaticimaging has been retained. By contrast, CEA-Scan (a technetium-99m-labeledanticarcinoembryonic antigen [CEA] Fab' fragment), appears tobe useful in the liver. Although CEA-Scan is superior to CT inextrahepatic areas, it is equivalent and complementary to CT inthe liver (ie, CT detects some liver metastases missed by CEA-Scanand CEA-Scan detects disease missed by CT). Furthermore, unlikeOncoScint, CEA-Scan appears to be initially nonimmunogenic (<1% induction of human anti-mouse antibodies [HAMAs], as measuredin an antifragment HAMA assay).

Although the published information is sparse, data presented byNeoRx at the FDA's Oncologic Drugs Advisory Committee meetingin December 1995 suggest that Verluma was sensitive for the detectionof extensive small-cell carcinoma of the lung in an 89-patienttrial (not controlled by surgical confirmation of the imagingresults). (By contrast, both OncoScint and CEA-Scan were usedin more than 200 patients in whom surgical confirmation was performed.)

After Verluma and CEA-Scan, what are likely to be the next radiolabeledantibodies available for diagnostic use in the clinic? Phase IIItrials with ProstaScint for prostate cancer and CEA-Scan for non-small-celllung cancer are underway. Phase III studies of the human antibody,88BV59H21-2, for colorectal cancer also are ongoing, as are studiesemploying intraoperative probes with many of the antibodies discussedin the article (iodine-125-labeled CC49, as noted by Divgi, aswell as several technetium-99m-labeled antibodies).

Even Slower Progress With Therapeutic Use

As slow as progress has been in diagnostics, progress with therapeuticuse of radiolabeled monoclonal antibodies in oncology has beeneven slower. As Divgi correctly points out, nonmurine antibodiesconjugated with cytotoxic isotopes other than iodine-131 are desirable.To date, the therapeutic use of murine monoclonal antibodies hasbeen limited by immunogenicity and bone marrow toxicity. Onlyin patients with hematologic malignancies have useful antitumoreffects been seen regularly. The ideal candidate agent for solidtumor therapy has yet to be convincingly defined. Theoretically,a humanized IgG or F(ab')2, labeled with a strong beta-emittingradioisotope, should offer the best chance for success.

In the realm of nonradiolabeled monoclonal antibodies, other strategiesneed to be evaluated more extensively. For example, a phase IIItrial of a humanized anti-HER2 antibody is underway in patientswith breast cancer. Also, promising studies are progressing withvarious strategies for pretargeting nontoxic antibodies to thetumor site to avoid systemic toxicity when the cytotoxic agentis administered.

Summary

In summary, it is clear that the use of radiolabeled murine monoclonalantibodies for radioimmunodetection may soon be regularly employedin oncologic practice. Radioimmunotherapy, on the other hand,is still far from established in clinical practice.

References:

1. Pinsky C: Monoclonal antibodies: Progress is slow but sure.N Engl J Med 315:704-705, 1986.

2. Collier B, Abdel-Nabi H, Doerr R, et al: Immunoscintigraphyperformed with In-111-labeled CYT-103 in the management of colorectalcancer: Comparison with CT. Radiology 185: 179-186, 1992.

3. Breitz H, Sullivan K, Help W: Imaging lung cancer with radiolabeledantibodies. Semin Nucl Med 23:127-132, 1993.

4. Moffat F, Pinsky C, Hammershaimb L, et al: Clinical utilityof external immunoscintography with the IMMU-4 99mTc-Fab antibodyfragment (CEA-Scan; arcitumoMoAb) in patients undergoing surgeryfor carcinoma of the colon and rectum. J Clin Oncol vol 14, 1996(in press).