Table 2.

Key performance indices for detection of alterations in classic oncogenic drivers in NSCLC

Patient group	Genomic alterations	Sensitivity (%)		PPV (%)		Specificity (%)
		SPU	PLA	SPU	PLA	SPU	PLA
Entire cohort (n = 71)	Drivers	50.0	63.5	96.3	97.1	99.8	99.8
	EGFR	25.0	70.8	100	100	100	100
	KRAS	83.3	75.0	100	100	100	100
	ALK	75.0	37.5	85.7	100	98.5	100
	ROS1	66.7	66.7	100	100.0	100	100
	BRAF	100.0	100.0	100	50.0	100	99.0
	MET	0.0	0.0	na	na	100	100
	RET	100	100	100	100	100	100
Centrally located lung tumors (n = 39)	Drivers	63.0	59.3	94.4	100	99.5	100
	EGFR	25.0	75.0	100	100	100	100
	KRAS	85.7	57.1	100	100	100	100
	ALK	71.4	28.6	83.3	100	97.1	100
	ROS1	66.7	66.7	100	100	100	100
	BRAF	100	100	100	100	100	100
	MET	nd	nd	nd	nd	nd	nd
	RET	100	100	100	100	100	100
Smokers (n = 33)	Drivers	68.8	68.8	100	100	100	100
	EGFR	33.3	66.7	100	100	100	100
	KRAS	75.0	75.0	100	100	100	100
	ALK	100	0.0	100	na	100	100
	ROS1	100	100	100	100	100	100
	BRAF	100	100	100	100	100	100
	MET	0.0	0.0	na	na	100	100
	RET	nd	nd	nd	nd	nd	nd
LUSC (n = 15)	Drivers	100	100	100	100	100	100
	EGFR	100	100	100	100	100	100
	KRAS	nd	nd	nd	nd	nd	nd
	ALK	100	100	100	100	100	100
	ROS1	nd	nd	nd	nd	nd	nd
	BRAF	nd	nd	nd	nd	nd	nd
	MET	nd	nd	nd	nd	nd	nd
	RET	nd	nd	nd	nd	nd	nd

PLA plasma, PPV positive predictive value, SPU sputum supernatant. All sensitivity, PPV, and specificity values are expressed in percent. na, not applicable. nd, no alteration in indicated gene detected in reference (i.e. tumor biopsy) samples