Human experimentation: a rule gone awry?

The U.S. EPA’s new Protections for Subjects in Human Research rule, which came into force on 7 April 2006, was born of a need to tighten the ethical guidelines controlling nonmedical human experimentation. The rule was ostensibly designed to offer people greater protection in pesticide toxicity experiments. But just two weeks after its coming into force, a coalition of labor and environmental interest groups filed suit against the EPA, challenging the rule’s legality and ethics. Against a backdrop of claims of industry influence, financial interests, and bipartisan rhetoric, the Second Circuit Court of Appeals in New York City must now determine whether this rule safeguards Americans against unethical experimentation or sells them out to big business.

Non-small cell lung carcinomas (NSCLC), surprisingly, could be visualised scintigraphically with McAbs not intended initiallly to react with them. Anti-CEA McAb (Riva et al., 1986), anti-osteosarcoma (Perkins et al., 1986), an antibody directed against c-myc oncogene product (Chan et al., 1986) and Fab'2 fragments directed against hepatitis virus (Kalofonos et al., 1988) have been used successfully to this end. In the present work, the first purpose was to determine whether Po66, a McAb prepared by immunisation against lung squamous cell carcinoma (LSqCC) (Dazord et al., 1987), was more effective than an unrelated immunoglobulin for NSCLC imaging. The second objective was to evaluate the value of immunoscintigraphy (IS) among other investigative means in the diagnosis of NSCLC.

Material, patients and methods
Monoclonal antibody and radiolabelling procedure Po66, an IgG I McAb, immunoprecipitated a 47 kDa molecular weight antigen and was prepared as reported earlier (Dazord et al., 1987). Briefly, Balb/c mice were immunised with enzymatically dissociated cells from a patient's LSqCC. Mouse immune cells were fused with SP2/0 plasmocytoma and McAb Po66 was selected from the hybrids obtained. Po66 consistently reacted with LSqCC, with half the adenocarcinomas tested and not with small cell lung carcinoma. It did not recognise normal tissues except distal renal tubules and gastric or bronchial serous glands (Dazord et al., 1987). The Po66 batch designed for human use was purified from ascites obtained from i.p. grafted Balb/c mice. The ascitic fluid was precipitated in 40% saturation ammonium sulphate, dialysed against 10 mM, pH 8, phosphate buffer and eluted through a DEAE ion-exchange column with a 10-150 mM, pH 8, phosphate buffer gradient. The antibody was subjected to the controls recommended by the European Commission (Commission of the European Community, Drafting Group Biotech/Pharmacy, 1986). A mouse IgGI monoclonal immunoglobulin, Py, without known specificity, was taken as control and processed like Po66. Samples of the antibodies Po66 and Py were respectively radio-iodinated with iodine-131 or iodine-125 by the iodogen method (Fraker & Spek, 1978) and purified from free iodine by elution through a Dowex anionic exchanger column equilibrated with PBS containing 0.3% human serum albumin. The protein bound radioactive fraction averaged 90%. Each new preparation of iodine-131 labelled Po66 was tested against lung cancer cells using a competitive radioimmunoassay with unlabelled antibody and the ability to accumulate into tumours was checked in xenografted nude mice.
Selection and monitoring ofpatients After the patients had given informed consent, 29 males and one female, ranging from 42 to 77 years, were selected for the trial. The diagnosis of malignancy was histologically established by endoscopic bronchus biopsy. All patients had a conventional plain chest X-ray and a thoracic CT scan when required to determine eligibility for surgery. Immunoscintigraphy (IS) with Po66 was performed in 27 patients with primary bronchial carcinoma (26 LSqCCs and one adenocarcinoma) and in six patients with recurrent tumour. Moreover, three distant metastatic localisations were investigated. Fifteen of the 27 patients explored for primary tumour were operated upon and 10 of them had a CT scan. In the patients operated upon, the reactivity of Po66 with the tumour was evaluated by peroxidase staining on deep frozen surgical specimen as described previously. Radiolabelled Po66 was administered to the patients after blocking thyroid and stomach radioiodine uptake by ingestion of 1 g potassium iodide and 0.5 g potassium perchlorate, 1 h before injection of the antibody and daily throughout the 12 days of scintigraphic exploration. There was no premedication against a possible allergic effect. The radiolabelled antibody was diluted in Po66 AND LUNG-SQUAMOUS CELL CARCINOMA 150 ml of isotonic solution and infused intravenously over a period of 45 mn. Each patient received I mg of intact antibody labelled with 70 MBq iodine-1 31.

Imaging
Thirty-three IS were performed using a large field of view gamma-camera (Acti CGR) fitted with a high energy parallel hole collimator and 128 x 128 word-mode computer storage (MDS A2). Planar chest images of 15min each were performed on days 3, 6, 9 and 12 after administration of McAbs and additional images were done when necessary. 99Tcmlabelled macroaggregates were delivered siinultaneously to outline the lungs. Image subtraction procedures were used only to improve the contrast and in all cases diagnosis was made on the basis of presubtraction images.

Biodistribution
To study the distribution of antibody Po66, seven patients received 1 mg of non-specific antibody Py labelled with 7 MBq of iodine-125 together with '31I-labelled Po66.
Radioactivity counting was made on three different samples of blood and of various tissues: tumour, normal lung, muscle and bone. A double isotope counting for '3'I and 1251 was achieved. 131I radioactivity was counted directly from a predetermined window while, for 125, the contribution of '3'I in the 1251 window had to be subtracted. The results were expressed in MBq per gram of tissue. A specific localisation index (SLI) was calculated as follows: The injection was well tolerated by all patients and no immediate or delayed adverse reaction occurred except in one case of moderate and transient blood pressure fall without loss of consciousness.
In vivo distribution To assess the immunological specificity of the tumoral accumulation of Po66, seven patients received 31 I-labelled Po66 together with '251I-labelled Py, a non-related IgG I immunoglobulin. Surgical specimens were sampled 6 days after i.v. injection of both antibodies. The specific localisation indices (SLI) are shown in Figure I and Table I. SLI in tumour was always above 1 (1.40-5.67, mean 3.17) and was significantly higher than the SLI found in healthy lung (0.60-1.89, mean 1.11) and in other tissues (0.32-2.00, mean 1.07). The ratio of '"'I uptake in tumour as related to healthy lung varied from 1.49 to 5.84 (mean 3.05). The comparison of ISs performed in the same patients 3, 6, 9 and 12 days after Po66 injection showed that vascular radioactivity declined more rapidly than tumour bound radioactivity (Figure 2b). The cardiac image, indeed, was constantly apparent on day 3 and decreased in the following days until complete disappearance on day 12. Thus, tumours, especially those located in the mediastinum area, were more easily evidenced at later than at early intervals after administration of McAb. Kidney and liver were never detectable scintigraphically after 6 days.
Comparison of immunoscintigraphy data with other investigations Tables II and III show the results of the 33 immunoscintigraphic investigations performed and other related data. Table It refers to patients with primary chest tumour and  Table III to patients in recurrence. All patients had histologically proven carcinoma as shown by endoscopic biopsy and 15 of them were surgically confirmed. As can be seen in Table II, 20 out of 27 primary cancers were visualised by     Figure 3 and all six patients with recurrent tumour had a positive IS (Table III). Thus, the overall sensitivity of IS, as defined by the percentage of known tumours detectable by this method of investigation, was 78%.
In seven instances, IS did not allow any visualization of radiologically present tumours. In two cases (patients 1 and 2) the Po66-associated antigen could not be detected by the immunoperoxidase technique performed on a surgical sample. In two cases (patients 3 and 4), the surgery report indicated tumour size under 2cm in diameter. In the last three cases (patients 5, 6 and 7), paradoxical results were obtained: the tumour seen radiologically was not visualised by IS but an intense McAb uptake occurred in the contralateral field (Figure 4). An important defect of vascularisation of the tumour, as judged from scintigraphy with 9"Tcm labelled macroaggregates, was noticed in these patients, suggesting that poor tumour vascularisation constitutes a limitation to IS. Patient 6 was operated upon and a significant mediastinal and contralateral involvement was found. The other two patients had no anatomical control but they died from severe respiratory failure within the 3 months following surgery.
In patients 8, 9, 16 and 19, IS showed lesions of greater extension than expected from radiological data ( Figure Sa  and b). In patient 16 the involvement of left upper lobe detected by IS but not by radiological means had been noticed in the endoscopic examination. In patients 9 and 19, the extension was confirmed by CT scan 3 months after IS. In patient 8, surgery showed that the contralateral and mediastinal involvement corresponded to widespread small tumoral granulations.
In the patients of Table III, IS was done as a systematic investigation 4-6 months after initial surgery. Six local relapses were discovered by IS while conventional chest Xray in four cases and endoscopic examination in three cases did not allow any tumour detection (Figure 2a and b). The tumour localisations evidenced by IS were confirmed later by CT scan in all patients.

Discussion
The present clinical trial showed the ability of McAb Po66 to visualise non-small cell lung carcinoma (NSCLC) with a 78% sensitivity. These results confirm the work by Kalofonos et al. (1988) showing, in a series of 14 patients, that NSCLC could be easily detected using a F(ab')2 fragment directed against human milk fat globules and reacting histologically with the tumours. These authors, like others (see Introduction), showed that NSCLC could be non-specifically visualised by unrelated antibodies. The non-specific retention of circulating proteins or macromolecules in cancer has been documented and seems related to the particularities of vasculature of tumours (Jain, 1987; O'Connor & Bale 1984). In our own study, the SLI measurements performed showed  This observation confirms comparable investigations by Kalofonos et al. (1988) and shows very clearly that McAbs specially intended to react with tumour should be prefered to unrelated proteins. Interestingly, the mean SLI value with Po66 in humans was of 3.2 and did not differ significantly from the 3.4 result obtained with xenografted tumours in mouse (Dazord et al., 1987), although most studies show a decrease of SLI in humans when compared to animal studies.
The main problem concerning IS of lung cancer is its place among other tools of investigation. Lung cancers are partic-Figure 5 Patient no. 8. a, Chest X-ray. Lung carcinoma radiologically limited to the right upper lobe. b, IS, anterior view, showing an intense uptake of the monoclonal antibody in the mediastinum and a diffuse uptake in the two lung fields. Surgical control showed that the contralateral and mediastinal involvement corresponded to widespread small tumoral granulations. ularily accessible to radiological investigations and the interest of IS might appear relatively restricted. In fact, in our short series of 33 patients, IS gave information not given by other means of investigation in at least eight instances. In four patients at the preoperative stage (nos 8,9,16 and 19), IS showed lesions more extended than those detected by CTscan and there was some evidence that these scintigraphic abnormalities corresponded to real tumour involvements. The possibility that the lesions corresponded to disseminated small tumours was suggested in one case. Likewise, in the patient with recurrent tumour, IS detected lesions not evidenced by plain chest X-ray in four of the six patients investigated.
Although the present prospective study is still in progress, an attempt can be made to define the role of IS in the follow-up of LSqCC patients. It seems obvious that IS cannot be used for the initial diagnosis which is done by clinical examination, plain chest X-ray and especially by endoscopic biopsy. We infer, however, that IS may be helpful to evaluate the extension of the lesions discovered. Possible mediastinal and contralateral involvement, which contraindicate surgery, might be assesed by IS. Furthermore, IS seems valuable in those patients who have been operated upon and in whom recurrence is suspected.