2.1 Dosimetric Analysis
Patient CT-scans with a slice thickness of 0.9 mm or 3 mm showing pneumonitis at the time of first occurrence were analyzed. The treatment planning software Eclipse versions 15.6 and 16.0 (Varian Medical Systems, Palo Alto, Santa Clara, CA, USA) was used to accurately delineate the volume encompassing the radiological extensions of the pneumonitis. The derived contours were validated and approved by experienced specialists in radiology and nuclear medicine. Contours were transformed to the RT planning CT using deformable image registration applying a demon’s algorithm [20]. In case of overlap with the gross tumor volume (GTV), the GTV was subtracted from the pneumonitis contours to ensure solid tumor mass not to contribute to the assessment of pneumonitis. Three dimensional voxel-wise dose data was converted to 2 Gy equivalent doses (EQD2) based on the Linear Quadratic Model (LQM) [21] using a Matlab (MATLAB R2019b, The MathWorks Inc., Natick, MA, USA) [22] script. An α/β ratio of 3 for normal lung tissue was assumed [23]. For the pneumonitis volume, relevant dosimetric data was extracted such as the volume fraction receiving at least 20 Gy (V20Gy), 40 Gy (V40Gy), mean dose and the volume receiving 20 Gy in cm3. Dose data were categorized into 3 dose levels: low dose (LD) comprising doses below 20 Gy, intermediate dose (ID) comprising doses ranging from 20 Gy to 40 Gy and high dose (HD) with a minimum dose of 40 Gy. DVHs were extracted and the area under the curve (AUC) was derived. For the total lung, the original mean lung dose (MLD) and the V20Gy were extracted.
2.2 Patient data
Our method was tested using patient data as depicted in Table 1. Ninety patients, who received thoracic R(C)T with (39) or without ICI (51) in a time interval of 110 days around R(C)T between 2010 and 2022 at our institute were collected. Data was conducted based on patient data files and imaging data. Patient follow-up after definitive treatment included clinical examination and chest CT scans 6 weeks after therapy and every 3 to 6 months for 3 years, every 6 months for 2 years followed by once yearly intervals. Follow-up schedules after palliative treatment were based on a patient individual basis. Eighty-six patients with primary lung cancer and 3 patients with lung metastases and one with pleural carcinomatosis were included. RT fractionation schemes varied with total doses from 30 to 66 Gy and single doses between 1.8 Gy and 3.0 Gy. Sixty-one patients received definitive (meaning curatively intended) R(C)T +/- ICI and 29 patients were treated in palliative intention as listed in Table 1. From the 39 patients, who received ICI therapy, all were treated with PD-L1 or PD-1 inhibitors. Out of the group receiving ICI therapy, the majority of 23 patients (59 %) received Durvalumab.
Table 1: Patient characteristics., 1R(C)T abbreviates radio(chemo)therapy, 2ICI stands for immune checkpoint inhibition 3CTx stands for chemotherapy, 4SD abbreviates standard deviation. 5MWU stands for Mann-Whitney-U test
Patient characteristics
|
|
|
|
|
|
|
|
R(C)T1+ICI2
|
R(C)T+ICI [%]
|
R(C)T
|
R(C)T [%]
|
p-value
|
Test
|
No. of patients
|
39
|
43
|
51
|
56
|
|
|
No. females
|
14
|
35.9
|
11
|
21.6
|
0.159
|
Chi-square
|
No. males
|
25
|
64.1
|
40
|
78.4
|
0.159
|
Chi-square
|
Median Age [a] (min;max)
|
69 (47;83)
|
|
62 (49;85)
|
|
0.058
|
MWU5
|
Pulmonary Co-morbidity
|
15
|
38.5
|
14
|
27.5
|
0.268
|
Chi-square
|
Active or former smokers
|
25
|
64.1
|
34
|
66.7
|
0.483
|
Chi-square
|
No. of patients with lung metastases
|
2
|
5.1
|
1
|
2.0
|
0.407
|
Chi-square
|
No. of patients with primary lung tumors
|
36
|
92.3
|
50
|
98.0
|
0.191
|
Chi-square
|
CTx3
|
35
|
89.7
|
37
|
72.5
|
0.043
|
Chi-square
|
concomitant CTx
|
16
|
41.0
|
16
|
31.4
|
0.343
|
Chi-square
|
Prior thoracic RT
|
2
|
5.1
|
1
|
2.0
|
0.191
|
Chi-square
|
Definitive R(C)T
|
25
|
64.1
|
36
|
70.6
|
0.516
|
Chi-square
|
Median time between ICI & RT (min;max) [d]
|
14 (0;76)
|
|
-
|
|
|
|
No. of pneumonitis
|
16
|
41.0
|
16
|
31.4
|
0.578
|
Chi-square
|
Mean onset time after RT (SD4) [d]
|
100.0 (49.73)
|
|
74.9 (59.97)
|
|
0.102
|
MWU
|
Median onset time after RT (min;max) [d]
|
87 (14;190)
|
|
54 (0;198)
|
|
0.102
|
MWU
|
The time interval between R(C)T and ICI therapy varied between 0 and 76 days.
Additional chemotherapy was administered in 35 (89.7 %) cases in the R(C)T+ICI group and in 37 cases (72.5 %) in the R(C)T only group. Three patients had a history of thoracic RT, 1 in the R(C)T only group with a time interval of more than 3 years and 2 in the R(C)T+ ICI group with minimum of 11 months prior to radioimmunotherapy.
In total, 59 patients (65 %) were former or active smokers, 25 (64.1 %) in the R(C)T+ICI group and 34 (66.7 %) in the R(C)T group. Twenty-nine patients suffered from pulmonary comorbidities, 15 (38.5 %) in the R(C)T+ICI and 14 (27.5 %) in the R(C)T only group.
2.2.1 Pneumonitis definition
Pneumonitis was diagnosed based on clinical and/or radiological findings and was graded according to the Common Terminology Criteria for Adverse Events (CTCAE) v5 [16]. All grades were included. Clinical symptoms covered coughing, dyspnea and thoracic pain. Radiological findings encompassed a variety of findings such as cryptogenic organizing pneumonia (COP), with ground-glass and consolidative opacities. Nonspecific interstitial pneumonia (NSIP), another form of interstitial lung disease, presents with ground-glass and reticular opacities, indicating thickening of the interstitial lung tissue [11,17–19].
2.3 Statistical Analysis
Exploratory statistical analysis was performed using IBM SPSS Statistics version 28.0.1.1 (14). Univariate analysis and analysis of significance was performed using chi-squared tests for categorial variables. For numeric data, we applied Mann-Whitney-U (MWU) tests. Statistical significance level was set at p < 0.05.
In a first step, statistical analysis was performed for the entire data set. In a second step, the data set was divided into two groups of patients to reduce the impact of biologically different dose schemes. One group contained patients, who received definitive R(C)T ± ICI, and the other group summarized patients who received palliative R(C)T ± ICI. Cases in these subgroups were matched pairwise according to their planning target volumes (PTVs) in order to reduce the interfering influence of non-matching irradiated volumes on the radiation dose-volume correlation.