Free radical metabolites of toxic chemicals.

S.c. injection of tumour cells or small pieces of tumour irradiated to a dose of 22,000 rad evoked resistance to live challenge with the same tumour (a CBA strain fibrosarcoma induced with methylcholanthrene) 14 days later. This resistance was, however, overridden if the challenging inoculum was sufficiently large, and did not develop if the cells were irradiated to 100,000 rad. The resistance evoked by injection of 106 irradiated tumour cells was impaired by i.p. injection of 1 4 mg C. parvum 5 days before, and virtually abolished by a similar injection 11 days after, the irradiated cells. The effect of s.c. injection of a mixture of 106 irradiated cells and C. parvum 14 days before live challenge depended on the dose of C. parvum. With 0*7 mg the development of resistance was largely but not completely abrogated; 0-35 mg resulted in a lesser degree of abrogation, and 0-09 mg or 0*02 mg had little or no effect. IT was reported by Woodruff and Dunbar (1973) that simultaneous s.c. injection of irradiated tumour cells and i.p. injection of C. parvum to mice with small but actively growing fibrosarcoma isotransplants caused a more prolonged remission than either treatment alone. On the other hand Smith and Scott (1972) found that administration of C. parvum 7 days before immunization with irradiated tumour cells diminished the protective effect of the immunization, as judged by the effect of subsequent challenge with viable tumour cells. Taken together, these findings have important implications both for our understanding of the mode of action of C. parvum as an immunopotentiating agent and in relation to the possible use of C. parvum in combination with active specific immunization for the treatment of patients with cancer. The present experiments were therefore undertaken to see whether the observations of Smith and Scott could be repeated with our tumour system. As a preliminary we have studied the effect of variations in the preparation of the irradiated tumour material, the route of injection, and the quantity injected, on immunization in the absence of C. parvum. MATERIALS AND METHODS Mice.-7-9 week-old female CBA mice were used throughout. Tumour.-The tumour was a fibrosarcoma induced in a CBA female mouse with methyl-cholanthrene, and was in its 15th-17th transplant generation. In most of the experiments, we have used tumour cell suspensions prepared with pronase as described previously (Woodruff, Inchley and Dunbar, 1972), but for immunization we have also used …

IT was reported by Woodruff and Dunbar (1973) that simultaneous s.c. injection of irradiated tumour cells and i.p. injection of C. parvum to mice with small but actively growing fibrosarcoma isotransplants caused a more prolonged remission than either treatment alone. On the other hand Smith and Scott (1972) found that administration of C. parvum 7 days before immunization with irradiated tumour cells diminished the protective effect of the immunization, as judged by the effect of subsequent challenge with viable tumour cells.
Taken together, these findings have important implications both for our understanding of the mode of action of C. parvum as an immunopotentiating agent and in relation to the possible use of C. parvum in combination with active specific immunization for the treatment of patients with cancer. The present experiments were therefore undertaken to see whether the observations of Smith and Scott could be repeated with our tumour system. As a preliminary we have studied the effect of variations in the preparation of the irradiated tumour material, the route of injection, and the quantity injected, on immunization in the absence of C. parvum.

MATERIALS AND METHODS
Mice.-7-9 week-old female CBA mice were used throughout.
Tumour.-The tumour was a fibrosarcoma induced in a CBA female mouse with methylcholanthrene, and was in its 15th-17th transplant generation. In most of the experiments, we have used tumour cell suspensions prepared with pronase as described previously (Woodruff, Inchley and Dunbar, 1972), but for immunization we have also used small pieces of recently excised tumour. Tumour cells and pieces of tumour used for immunization were irradiated with a Westinghouse x-ray machine operating at 220 kV and 15 ma with HVL of 1-2 mm Cu under conditions of maximum back scatter, at a dose rate of 274 rad/min. Assessment of results.-The results were assessed by comparing the incidence of tumours and relative growth rates in the various treatment groups.
Differences in the incidence of tumours were often so clear-cut as to make statistical analysis unnecessary, but when this was not the case the probability (P) of the observed difference being due to random sampling, from published values applicable (unless otherwise stated) to a single-tailed test for fourfold tables (Diem and Lentner, 1970), is shown in the section on results. When more than half the mice in a group developed tumours the group, mean relative growth rate, together with its 95% confidence limits (i.e. t X standard error), was calculated. Only those mice in the group which actually developed tumours were included, and the relative growth rate for each individual treated mouse was taken as the ratio of the tumour diameter in that mouse on a particular day to the mean tumour diameter on the same day in untreated mice challenged with the same number of viable cells. The day was chosen so that the mean tumour diameter in the control mice was between 15 and 18 mm; it ranged from Day 14 to Day 24 depending on the size of the challenging inoculum.
Corynebacterium parvum.-A formalinkilled suspension of C. parvum strain CN 6134 (Batch WE2174) was kindly provided by Dr. A. Weinberg of the Wellcome Research Laboratories. In some experiments 0'2 ml of this suspension (containing 1-4 mg dry wt. organisms) was injected i.p.: in others C. parvum suspension mixed with irradiated tumour cells was injected s.c.

RESULTS
The results are summarized in Tables  I and II. Pretreatment with irradiated cells alone Mice injected s.c. with 106 irradiated cells (22,000 rad) were almost completely resistant to challenge with 104 or 105 viable tumour cells 14 days later, but were susceptible to challenge with 106 or 107 cells (Table I and group 2, Table  II). There is a suggestion ( Table I) that larger (107 or 108) doses of irradiated cells were marginally less effective than 106 cells (comparison of the pooled number of takes after immunization with 107 or 108 irradiated cells with the number after immunization with 106 irradiated cells in a two-tailed test gives P = 0.02). J.v., and possibly also i.p., injection of 106 irradiated cells was less effective than s.c. injection. Irradiated pieces of tumour tissue were as effective as irradiated cell suspensions. Cells irradiated to 100,000 rad did not render mice resistant to subsequent live challenge (Table I).

Pretreatment with C. parvum alone
In conformity with many previous experiments with this tumour, i.p. injection of 1-4 mg C. parvum 3 days before s.c. inoculation of 105 viable tumour cells resulted in slower development of the tumour though it did not significantly influence the number of takes. A similar injection 19 days before tumour inoculation, and s.c. injection in various doses   Table II after challenge with 105 viable cells: P --0 001), but did not differ significantly from the incidence in untreated mice; and the mean growth rate, though slower than in untreated mice, was similar to that in mice which received C. parvum alone (Table II, group 4). In mice which received 1-4 mg C. parvum i.p. on Day -19 and 106 irradiated cells s.c. on Day -14, the incidence of tumours was greater than in mice which received irradiated cells alone (comparing groups 7 and 2 of Table II after challenge with 104 and 105 viable cells, the difference in each case is just significant at the P = 0*05 level), and very much less than in untreated mice (comparing groups 7 and 1 of Table II after challenge with 104viable cells: P < 0001).
It thus appears that, under the conditions of the experiment, the resistance normally evoked by pretreatment with irradiated cells was completely abrogated by a large i.p. injection of C. parvum 11 days after the cells and partly abrogated by a similar injection 5 days before the cells.
All mice which received 106 irradiated cells mixed with 0 7 mg C. parvum developed tumours after challenge with 105 or 107 viable cells, and the mean relative growth rate was the same after 107 cells, and only marginally less after 105 cells, than in the corresponding untreated controls. After challenge with 104 viable cells, however, the incidence of tumours in the treated mice was significantly less than in the controls (comparing groups 9 and 1 of Table II after challenge with 104 viable cells: P = 0-001). It thus appears that at this dosage C. parvum abrogated the development of resistance to a great extent but not completely. Smaller doses had little effect. A dose of 0 35 mg (Table II, group 10) appeared to cause slight abrogation as judged by the response to 105 cells, though not by the response to a larger challenge. With a dose of 0 09 mg, the immunizing effect of the irradiated cells was, if anything, actually increased (compare groups 2 and 11 of Table II). Neither the difference in tumour incidence nor the difference in tumour growth rate in mice which developed tumour are themselves significant, but if the group mean tumour diameter on Day 15 of all animals in the group are compared by a t test (scoring animals without tumours as having tumours of diameter 0) the difference based on a single-tailed test is just significant at the P = 0.05 level (t = 1*96 for 10 degrees of freedom). DISCUSSION It is apparent that, with the tumour used in these experiments, the resistance to live challenge which develops after injection of an optimal dose of irradiated tumour cells is not absolute but can be over-ridden if the challenging inoculum is sufficiently large. It is also clear that the immunogenicity of irradiated tumour cells is lost if the dose of irradiation is too large. It seems likely that these conclusions would hold good also for other immunogenic tumours.
The observations on the abrogation of the resistance resulting from injection of irradiated tumour cells by injection of C. parvum confirm and extend those of Smith and Scott (1972), and highlight the need for caution in clinical trials of combined procedures of this kind. The degree of abrogation depends inter alia on the dose of C. parvum, the time at which it is given in relation to the irradiated cells, and the route of injection.
When C. parvum is mixed with irradiated cells large doses are more prone to cause abrogation than small doses, and there is a suggestion that with a very small dose (in these experiments 0 09 mg), instead of abrogation, immunization may actually be enhanced. When the dose is still further reduced there is no effect in either direction.
Our attempts to elucidate the mechanism underlying this phenomenon are as yet inconclusive. As we have already briefly reported (Woodruff, Ghaffar and Dunbar, 1975), spleen cells from mice pretreated with C. parvum and irradiated tumour cells may be cytotoxic for the tumour in vitro even though the mice show no increased resistance to challenge with viable tumour cells. In the light of this finding we suggested that the abrogation of resistance in vivo might be caused by a blocking factor in the serum, and preliminary experiments appeared to lend some support to this conjecture. We do not know, however, to what extent, if any, the cell-mediated cytotoxicity in vitro is specific for the tumour, and further investigation has failed to confirm the existence of a blocking factor. Another possibility, which is currently being investigated, is that administration of C. parvum under the conditions of the experiment results in the development of suppressor T-cells. For the time being, however, the question of mechanism must be regarded as 8ub judice.
This study was supported by grants from the Cancer Research Campaign.