Intralesional radiolabelled monoclonal antibody therapy.

Despite twenty years of scientific research, systemically administered radiolabelled MAb's have failed to demonstrate any significant therapeutic benefit. The fundamental barrier to success, in brain tumours, is the low percentage of the injected antibody taken up per gram of tumour (< 0.001%). This study builds on these experiences but utilises modern advances in stereotactic neurosurgery and computer graphics. It investigates the role of intralesional administration in an attempt to increase local uptake at first in experimental tumours and then in patients whose symptoms and eventual outcome are dictated by their local tumour. In the laboratory, the percentage of the injected activity taken up per gram of tumour (%ID/g) was three times greater following intralesional (IL) rather than systemic administration (Student's t-test, p=0.012). Whereas, doses to the systemic organs were significantly lower following IL administration (Student's t-test, p=0.041). Antibody specificity was found to be an important advantage for two reasons. Firstly, there was a 14% greater %ID/g with specific (ANL/11/52) rather than non specific control (ICR2) MAb (Student's t-test, p=0.047). Secondly, analysis of the autoradiografts showed that the volume of radioactivity in tumour was significantly higher for the specific MAb (Student's t-test p=0.041). Furthermore a sharpe cut off of activity was demonstrated at the tumour/adjacent muscle junction. As a result, the radioactivity tended to conformed to the shape of the tumour at a submacroscopic level. In seven patients, the feasibility of intralesional 131I-labelled anti-tenascin (81C6) therapy has been demonstrated. It was well tolerated, mean doses to normal tissues were low (<0.4Gy), doses to tumour (46Gy) and the % of the injected activity retained per gram of tumour (0.46%ID/g) were 100-1000 times higher than previous studies using systemic administration. The precise distribution of radioactive uptake, in relation to the anatomical features, was evaluated by two novel approaches developed specifically for this project. Firstly, SPECT images were registered with the pre-treatment diagnostic scans, using surface landmarks. Secondly, at post mortem in two patients, 3-D scintillation counting of multiple sections was correlated with the histological features. Prominent tumour specificity was demonstrated with less than 1% of the radioactivity found in normal brain. With one or two catheters, however, the activity did not distribute throughout the whole tumour. The impact of higher infused doses and multiple catheters on improving the dose and distribution throughout the tumour now require investigation. We believe that with these improvements, this technique has a realistic chance of improving local tumour control in this group of patients.