TABLE 3.

Effect of baseline polymorphisms on occurrence of D30N substitution^a

Amino acid position	No. of patients with D30N substitution (%)		P valueb
	Polymorphism absent	Polymorphism present
10	22/47 (46.8)	3/8 (37.5)	0.715
12	20/43 (46.5)	5/12 (41.7)	1.000
13	19/42 (45.2)	6/13 (46.2)	1.000
15	19/42 (45.2)	6/13 (46.2)	1.000
35	17/43 (39.5)	8/12 (66.7)	0.114
36	19/40 (47.5)	6/15 (40.0)	0.764
37	19/40 (47.5)	6/15 (40.0)	0.764
41	23/45 (51.1)	2/10 (20.0)	0.092
62	20/41 (48.8)	5/14 (35.7)	0.537
63	7/14 (50.0)	18/41 (43.9)	0.762
64	17/43 (39.5)	8/12 (66.7)	0.114
72	22/47 (46.8)	3/8 (37.5)	0.715
77	19/40 (47.5)	6/15 (40.0)	0.763
93	20/44 (45.5)	5/11 (45.5)	1.000

^aSequence analysis was performed on HIV protease genes obtained from plasma samples from 55 patients at baseline prior to nelfinavir therapy. Polymorphisms which occurred in >10% patients (≥6 changes/55 baseline sequences) were identified based on comparison with the consensus sequence as described in Materials and Methods. The proportion of patients that acquired the D30N substitution following nelfinavir therapy was then calculated for each of the two groups (polymorphism absent or present). 

^bA test of equality of the two proportions was conducted by Fisher’s exact test.