Egbert Oosterwijk
University Medical Center Nijmegen, Nijmegen, The Netherlands

Monoclonal Antibody G250 in Renal Cell Carcinoma: Issues of Specificity, Heterogeneity and Efficacy

Anecdotal serological evidence has suggested the existence of Renal Cell Carcinoma (RCC)-associated antigens absent in normal kidney tissue. Evidently these would be ideal mAb targets, provided that expression in the normal tissues is limited. To specifically target RCC we developed monoclonal antibody (mAb) G250. Immunohistochemical fine specificity analysis demonstrated homogeneous antigen expression in clear cell RCC, the most prevalent RCC subtype, and no antigen expression in the majority of normal tissues, including normal kidney. In clinical trials radiolabeled mAbG250 has been shown to target metastatic RCC, and biopsy based studies have revealed that the absolute and relative amounts accumulated mAbG250 are among the highest ever reported (up to 0.5 %ID/g). In contrast to the experience with most other mAbs, saturation of accessible G250 epitopes was observed. Thus, the optimal protein dose level for imaging and radioimmunotherapy was rather low (5-10 mg mAb). While this might pose a problem for high dose treatment, loss of immunoreactivity has not been observed, even at high specific activity levels.

Despite homogeneous antigen expression and the occurrence of saturability, mAbG250 uptake is almost invariably heterogeneous. Within tumor specimens regional differences in 131-iodine-mAbG250 uptake were as high as two orders of magnitude. Our initial analysis showed that factors other than antigen expression are important for adequate mAb targeting. Obviously, non-homogeneous mAb distribution will hamper any therapeutic efficacy, and therefore we have been interested in defining parameters governing (intratumoral) mAb distribution. In a dual labeling study where patients received two injections of iodine radiolabeled mAbG250 separated by 4 days, distribution analysis showed that particular tumor regions demonstrated high mAb uptake, whereas other tumor regions remained cold. Analysis of antigen expression, (neo)vasculature, and necrosis did not reveal any relation with mAb uptake other than that high antigen expression was a prerequisite for high mAbG250 uptake. We suggested that tumor parameters governing mAbG250 uptake do not alter significantly within the time period studied (4 days) and that multiple radiolabeled antibody injections, administered within short time periods, will target the same areas within a tumor, and thus will not solve the problem of heterogeneous tumor uptake of antibody.

This phenomenon may play a more prominent role in larger tumors that are most likely built up by different subpopulations of cells with intrinsic differences. One factor that might influence the intratumoral biodistribution is mAb metabolism. In vitro as well as in vivo data did not give evidence that this factor played a prominent role: in RIT trials 131-I-G250 imaging was possible 3-4 weeks post administration, and in animal models mAbG250 tumor retention was invariably high, suggesting minimal mAb metabolisation. Recently we have obtained evidence in animal models as well as in metastatic RCC patients that part of the heterogeneous 131-iodine labeled mAbG250 uptake may be explained by intrinsic internalization differences. Thus, (part of) the observed heterogeneous mAbG250 accumulation was actually the net result of mAb internalizing vs. non-mAb internalizing tumor cells. For example, the use of radiometals may lead to a more homogeneous radiation absorbed dose, possibly leading to satisfactory anti-tumor effects.

Although dosimetric analyses indicated that radiation absorbed doses close to tumor-sterilizing levels seemed achievable, anti-tumor activity in metastasized RCC patients has been anecdotal: single high dose therapy at MTD as well as rapid fractionated dose therapy has not lead to satisfactory responses. Translation of the exquisite targeting abilities of mAbG250 into mAbG250 guided therapy still remains a major challenge. Currently, mAbG250 trials alone or in combination with cytokine treatment are in progress.

Increased insight in molecular mechanisms leading to G250 expression have revealed that G250 expression is not solely restricted to clear cell RCC, but is also expressed in hypoxic regions of a variety of tumors. Since hypoxic tumor regions are notoriously radiotherapy resistant, mAbG250-based clinical trials seem warranted in non-RCC tumor types.