Virus 
Hours Till Death 
Mean 
Patient Isolate( 1397/78) 
Osterrieth 
Tb virulent 
Hd 
Tb avirulent 
64,73,90,116 
95,116,139,163 
116,116,186,211 
211 ,-,-,- 
86 hrs. 
128 hrs. 
157 hrs. 
one death only 
no deaths 
Thus the virus isolated from the patient (1397/78), when compared to the 
relatively virulent starting strain (Osterrieth) of SFV, appears to have acquired 
considerable virulence for mice (mean hours to death decreased from 128 to 86), 
probably reflecting its multiple passages in animals. 
Also noteworthy, the relatively avirulent strain Hd used in these experiments is 
identical to the avirulent substrain of L10 that Liljestrom and Garoff utilized to 
isolate SFV4, the progenitor of the SFV/Helper 2 expression system (A. 
Helenius, Pers. Commun.) 
Because only four mice were injected with each SFV strain, however, the 
evidence suggesting increased virulence of the infecting SFV strain is preliminary. 
And although SFV strains with increased virulence in mice frequently have 
demonstrated increased virulence in guinea pigs and rabbits (Bradish et al., 1971; 
Smithbum and Haddow, 1944), published studies comparing SFV virulence in 
mice and humans do not exist. Thus an increased SFV virulence in mice 
suggests, but does not prove, increased virulence for humans. 
That the SFV strain involved in the fatal incident had been deliberately passaged 
to enhance its virulence apparently was known to SALS prior to circulating its 
final recommendations in January 1979, three months before the Willems et al. 
report was published. But the unpublished results from Willems and Kaluza, on 
the virulence in mice of the clinical isolate of SFV, evidently were unknown to 
SALS prior to the formulation of its recommendations. Phone interviews 
conducted (by G.F.T.) with two former SALS members involved with the 
reclassification of SFV indicated that SALS was unaware of these results. 
These results might well have altered the committee’s classification of SFV. For 
the SALS report that issued soon thereafter (SALS, 1980) stated that in 
developing recommended levels of practice and containment the committee 
members had "...recognized that different strains of the ‘same’ virus can vary 
markedly in pathogenicity. Differences in virulence occur naturally among strains 
and can also result from laboratory manipulations.... Assignment of a virus to a 
level was generally made on the basis of characteristics of the viral prototype or 
[464] 
Recombinant DNA Research, Volume 18 
