Attempts to employ virus-induced tumors in therapeutic-type tests date back to the early days of the Rous sarcoma (2S) and the exploratory studies of Rous (24) on the effects of irradiation. However, because of the unpredictability of virus source materials, during earlier years, and the wide variations obtained in the biological results of different experiments (see 4), serious consideration was not given, until recently, to the use of this tumor in actual screening and therapeutic-testing programs. During the last few years much has been learned concerning the biology of both the virus and the tumors which it induces, and systematic studies have been begun on the use of virus-induced Rous sarcomas in chemotherapeutic research.

Among the first to investigate this tumor with the newer bioexperi-mental techniques and standardized preparations of virus were Groupe and his associates (17), who studied the tumor-therapeutic and virus-prophylactic effects of the biological product, xerosin (25), which had been found to modify the reactions of test animals to certain viruses of infectious diseases (26). They found that daily injections of xerosin significantly delayed the appearance of Rous sarcomas in chickens, whether the injections were begun before (prophylactic tests) or after (therapeutic tests) infection with the virus. Tumors induced with strong doses of the virus were not further altered, nor was life prolonged by continued xerosin treatment. On the other hand, tumors induced with weaker doses of the virus were converted to a slowly growing, "atypical" type in more than 50 percent of the test animals. The same result was obtained even when initiation of the treatment of each chicken was delayed unti its tumor had become grossly visible. Although a few of the atypical tumors retrogressed, the great majority of them continued to grow slowly and eventually killed their hosts in those experiments in which the animals were maintained for sufficient periods.

The present interest in the studies of Groupe and associates is not so much in xerosin itself as a potential antitumor agent, but in the testing techniques that were developed by these investigators during the study of this and other factors which modify response to the Rous sarcoma virus (27, 28). The latent-period response to this virus had been shown to be a relatively efficient indicator reaction for bioassays, and one that could be subjected to statistical control and interpretation by relatively simple graphic methods (2, 9). Using the information that microscopic tumor foci may be present within 3 days after virus inoculation (29), Groupe devised a therapeutic-test procedure in which candidate compounds were administered beginning on the 3d day after virus inoculation. It was anticipated that any antitumor effect on the microscopic foci would be manifested by a delay in the appearance of grossly visible tumors, and that simple daily inspections for the presence of tumors would provide the basis for a highly sensitive and statistically analyzable criterion of antitumor activity. This has proved to be the case. The graphic statistical methods used in analyses of the data were the same as those previously illustrated by Group6 et aZ. (27, 28)-see (9)-for a biometric discussion.

Table 1 gives examples of types of results, having different degrees of statistical significance, taken from actual preliminary screening experiments carried out in Groupe's laboratory. Although statistically significant delays in the appearance of experimental tumors as compared with controls, such as that illustrated in the bottom row of the table, provide a simple and highly sensitive test for antitumor activity, it should be emphasized that such accumulative percentage values alone do not tell the whole story. The response of a given chicken is scored as positive on the day the tumor appears, and it contributes full weight to the computation of the final accumulative percentage value, regardless of the fact that it may grow very slowly and may be only a small fraction of the size of the control tumors at the termination of the preliminary test.

Table 1. Types Of Results In Preliminary Screening Tests Based Upon Latent-Period Delay Of Rous Sarcoma (Groupe Et At.)

Candidate microbial preparation

Number of chicks

Accumulative percent with tumors on day:







(None) Ex-100 Ex-98 Ex-89

30 15 15 15

3 7 0 0

40 40 13 7

93 80 47 53

97 93 87 93

>0.05 0. 05>P>0. 01 <0. 01


No significant difference Borderline significance Significant

Similar differences in size, after a period of 10 days, were employed by Johnson and associates (30, 31) in another method of measuring antitumor activity with Rous sarcomas induced with standardized doses of the virus. These investigators studied primarily the prophylactic effects of their compounds but were able to demonstrate a high degree of reproducibility of quantitative inhibition, as measured by the weight ratios of experimental to control tumors.

In table 2 are shown the summarized results of prophylactic effects on the Rous sarcoma virus in tests carried out in the laboratories of Johnson and Groupe with compounds that had previously been found to produce antitumor effects in some other, nonviral test system. Although some of the compounds of the separate studies were the same, the series was not identical in the two laboratories. As may be seen, approximately half of those tested in each laboratory were detected by the Rous sarcoma prophylactic tests.

On the other hand, only 3 out of a total of 38 compounds effective in other systems were detected by Group6 and associates in therapeutic tests initiated 3 days after Rous sarcoma virus infection. They are listed in table 3. It is of interest that the 3 compounds shown are among those which have been found to be most effective in the actual treatment of neoplasms in human subjects.

Table 2. Results With Rous Sarcoma Of Compounds Effective In Other Systems

Prophylactic (antiviral) effects*

Number of compounds giving result:


Test criterion

Virus inhibition (+)

No significant inhibition (-)

Total No.

Johnson et al. Groupe et al.

Tumor weight Latent-period delay

11 11

13 10

24 21

*Daily injections begun 1 or 2 days before virus Infection.

Table 3. Results With Rous Sarcoma Of Compounds Effective In Other Systems

(Group6 et al.)

Therapeutic (antitumor) effects*

Chemical compound

Latent-period delay test


P<0. 01


P<0. 01

Triethylenemelamine (TEM)

.05>P>0. 01

35 other compounds

No significant delay

*Dally injections begun 3 days after virus infection.

In a pilot screening program in search of effective microbial products in cultures of Actinomycetes and gram-negative bacteria, 8 culture fractions (supernatants, nitrates, or cell extracts) were found to produce latent-period prolongations of borderline significance (0.05>P>0.01), and 6 were found to produce highly significant prolongations (P<0.01) out of a total of 245 examined by the preliminary therapeutic test. More critical studies on these 6 most effective products are now under way.