Modeling of Non-Small Cell Lung Cancer Volume Changes during CT-Based Image Guided Radiotherapy: Patterns Observed and Clinical Implications

Table 3

Individual patient, tumor, treatment, and tumor response characteristics. Patient IDs starting with an “L” had a long RT course while those starting with an “S” had a short one. Best fit values of model parameters with likelihood-based confidence ranges in square brackets.

Pt.; ID	Day 0 volume; (cc)	Model	Response; type	Td; (days)	GF	𝑇pot (days)	𝛾	𝛾rec	𝐾; (%)	𝐾𝑝; (%)	𝐾𝑞; (%)	𝐷	Rini	Rind	Mean Δ%
L1	8.5	M Res	B	200		5.6			15 [14–16]			0.5 [0.46–0.95]	0.20 [0.19–0.21]	3.4 [3.1–3.7]	8.0
		St Rec		200	0.15	13.9	0.01	0.10*  [0.08–0.11]		29 [28–30]	1 [0.5–2]	0.56 [0.47–0.71]			7.0
L2	13.2	M	C	100		9.3			1.8 [1-2]			0.2 [0.2–1]			3.5
L3	4.6	St Rec	B	400	0.05	27.8	0	0.37^‡  [0.19–0.81]		49 [47–50]	1 [0.5–2]	0.6 [0.48–1]			10.2
L4	161.1	M Res	A	1000		5.5			10 [9–12]			0.3 [0.25–0.4]	0.31 [0.28–0.33]	3.2 [2.7-3.7]	4.0
L5	1.8	M Res	A	150		5.6			16 [14–33]			0.35 [0.2–0.7]	0.36 [0.32–0.39]	6.2 [5.8-6.8]	5.1
		St Rec		150	0.25	5.6	0	0.02^†  [0.02–0.02]		7 [6-7]	1 [0.5–2]	0.36 [0.25–0.42]			5.4
S1	3.2	St	A	25	0.06	22.8	0			90 [90–100]	5 [5–9]	0.2 [0.2–0.28]			8.4
S2	12.1	St Rec	B	25	0.10	14.3	0	0.26^§  [0.05–0.54]		90 [86–99.9]	1 [0–6]	1 [0.85–0.9999]			6.8
S3	2.0	M	D	250		5.6			8 [5–11]			0.999 [0.25–1]			6.6
S4	2.0	M	C	ND		ND			25 [5–44]			ND			4.0
S5	23.5	M	C	150		8.4			9 [5–12]			0.2 [0.2–0.4]			3.7
		St		150		8.4				90 [38–100]	2 [0.01–6]	0.2 [0.1–0.62]			5.1
S6	8.1	M	D	45		5.6			13 [11–21]			1 [0.57–1]			6.6
S7	4.7	St	A	30	0.06	23.6	0			99 [98–100]	23 [23–25]	0.2 [0.16–0.29]			9.2
S8	1.2	St Rec	B	1000	0.05	27.7	0	0.48^§  [0.27– 0.68]		65 [59–85]	1 [0–7]	0.9 [0.44–1]			7.2
S10	3.2	St	A	15		9.7	0.01			99 [85–100]	23 [20–25]	0.2 [0.15–0.31]			13.1
S11	1.7	M	D	15		5.8			22 [20–23]			0.99 [0.95–1]			23.9
S12	0.7	St Rec	B	75	0.05	28.0	0	0.74^§  [0.23–0.79]		76 [74–95]	0.01 [0.01–4]	0.75 [0.1–0.97]			7.2
S13	4.1	St	A	1000		5.5				90 [57–99]	53 [53–99]	0.2 [0.2–0.25]			14.6

Abbreviations:
Pt.: patient; Mean Δ%: average percent difference between data and fit; minimal (M) and standard (St) models are defined in the text, with the respective growth parameters (Td, GF, 𝑇_pot, and 𝛾) and killing parameters (𝐾,𝐾𝑝, 𝐾𝑞, and 𝐷). For the resistance models, two alternative resistance models (Res) were considered: (i) initial resistance, with parameter Rini (fraction of resistant cells at 𝑡=0); (ii) induced resistance, with parameter Rind (probability of induced resistance per fraction delivered). The recruitment model (Rec) is characterized by the parameter 𝛾rec (fraction of quiescent cells recruited into proliferation per day).
Response types: A: “high” kill and “slow” dying rate; B: “high” kill and “fast” dying rate; C: “low” kill and “slow” dying rate; D: “low” kill and “fast” dying rate.
*Recruitment from 4th week after treatment.
†Recruitment from 3rd week after treatment.
^‡Recruitment on days 1–13 after treatment.
§Recruitment on days 1–3 after treatment.