Host immunity influences the impact of radiotherapy (RT) in cancer, but mechanistic connections remain obscure. In this study, we investigated the relationship of indoleamine 2,3-dioxygenase (IDO) systemic activity on clinical outcomes in RT-treated non–small cell lung cancer (NSCLC). IDO-mediated production of kynurenine and the kynurenine:tryptophan ratio in patient blood serum were determined for stage III NSCLC patients at times before, during, and after RT administration and then correlated to overall survival (OS), progression-free survival, and disease progression rate in patients. We found the impact of RT on these serum IDO markers to be heterogeneous in patients. On average, kynurenine:tryptophan ratios were reduced during RT but restored after RT. Notably, both baseline levels of kynurenine:tryptophan and changes in the levels of kynurenine after RT were significantly associated with OS. When combined, favorable change and favorable baseline corresponded with very long-term OS (median OS was not reached after 57 months of median follow-up). Favorable change combined with unfavorable baseline still corresponded with a lack of distant metastases. Our results suggest that RT alters IDO-mediated immune status in NSCLC patients and that changes in this serum biomarker may be useful to predict outcomes and perhaps personalize RT dosage to improve survival.

Significance: Radiotherapy appears to influence systemic IDO activity and to exert a significant impact on metastatic risk and overall survival, with possible implications for defining a biomarker to optimize radiation dose in patients to improve outcomes. Cancer Res; 78(3); 809–16. ©2017 AACR.

Non–small cell lung cancer (NSCLC) accounts for 80% to 85% of lung cancer, which is the leading cause of cancer death in the United States (1). Approximately 64% of patients with NSCLC require radiotherapy (RT) at least once during their course of disease (2). Recently, it has been reported that radiation-induced tumor killing can activate the immune system by generating tumor-specific antigens and converting the tumor into an individualized in situ vaccine (3). Animal studies have demonstrated that a combination of hypofractionated RT with immunotherapy generated synergetic effects on local tumor control and even abscopal effects for tumor killing outside the treatment fields (3–5). The RT abscopal effect has been reported in a few clinical cases, including one melanoma case in the New England Journal of Medicine (4).

Indoleamine 2, 3-dioxygenase (IDO) is an intercellular enzyme that converts the essential amino acid tryptophan into kynurenine through the IDO/kynurenine pathway (6). IDO depletes tryptophan in the tumor microenvironment and activates immune checkpoint via amino acid sensor general control nonderepressible 2 (GCN2; refs. 7, 8). Kynurenine, the direct metabolite of this pathway, and its further downstream metabolites, promotes potent immune suppression by enhancing Foxp3-regulatory T-cell functions and attenuating responses of effector T cells and natural killer cells (9, 10). Previous studies have validated IDO as a potent immune checkpoint in cancers and other chronic inflammatory diseases (11–13). Because tryptophan and kynurenine concentrations can be measured from patients' serum, IDO activity can be monitored by computing kynurenine to tryptophan (K:T) ratio. Clinical studies have reported that IDO activity correlated with the number of tumor-infiltrating lymphocytes in esophageal and colorectal cancers (14, 15), and elevated IDO activity correlated with poor clinical outcomes in several types of cancers (16–19), including lung cancer (20–22). These clinical data further support that IDO is an immune checkpoint that mediates antitumor immune activity. However, it is not clear how this IDO-mediated immune activity changes during or after RT in cancer patients, and whether these changes have any impact on tumor control or survival.

In this study, we hypothesized that RT can alter the IDO-mediated antitumor immune activity, which will affect tumor progression or survival in patients with NSCLC. To test this hypothesis, we assessed IDO-mediated immune activity by quantifying the key molecules associated with the IDO checkpoint, including kynurenine and the K:T ratio (a commonly used surrogate for IDO activity), at pre-RT, 2-week (during-RT), 4-week (during-RT), and post-RT (3 months after RT completion). We then performed survival analysis to correlate these parameters and their changes with progression-free survival (PFS), overall survival (OS), and distant tumor progression.

Patients and treatment

Patients with stage III inoperable/unresectable NSCLC who enrolled in Institutional Review Board–approved prospective protocols were eligible. The protocols were conducted according to requirements of Belmont Report and U.S. Common Rules. All subjects signed informed written consent before enrollment. All patients received conventionally fractionated daily RT, which was given using 3D conformal radiotherapy (3DCRT) as previously described (23, 24). Details of these prospective trials were summarized in Supplementary Table S1. Equivalent doses at 2 Gy fraction (EQD2) were computed for those received other than 2 Gy daily doses, using alpha/beta of 10.

Patients' follow-up and samples/clinical data collection

Patients were examined weekly during the course of RT, followed up approximately every 3 months during the first year, every 6 months during the second year, and then annually thereafter. Clinical and follow-up data, including age, gender, smoking history, histology, clinical stage, tumor volume, Karnofsky performance score (KPS), EQD2, and chemotherapy, were recorded prospectively up to 2 years after completion of RT. Red top blood collection tubes without anticoagulants were used for serum. Blood samples of patients were collected at up to 4 time points of pre-RT, 2- and 4-week during-RT, and post-RT (generally first follow-up at 3 months after RT completion).

Measurements of serum tryptophan and kynurenine

Serum tryptophan and kynurenine concentrations were measured using a high-performance liquid chromatography system (Shimadzu LC20) as described before with minor modification (25, 26). A total of 25 μL of serum samples were diluted with equal volumes of 30 mmol/L NaAc (pH 4.0) and deproteinated with perchloric acid. Kynurenine was detected on an UV channel at 360 nm, and tryptophan was detected on a fluorescence channel at 285 nm excitation and 365 nm emission. Samples were analyzed using Lab Solution software (Shimadzu). Samples were quantified with external standards, and at least one quality control sample was randomly inserted into every plate for reference. For quality control purposes, double-blinding duplicate/triplicate testing verified the assay reproducibility to be over 95%.

Statistical analysis

OS and PFS were the primary endpoints. OS was computed from RT start to the date of death of any cause. PFS was defined from start of RT to the date of any progression or death. Patients were censored at last follow-up if progression or death had not occurred. The Kaplan–Meier log-rank test was applied to compare the survival difference between groups. Clinical variables with log-rank P < 0.05 under univariate analysis were selected as covariants for adjustment in multivariate analysis. Multivariate Cox proportional hazards model was used to estimate HRs with 95% confidence interval (95% CI). The IDO checkpoint–associated molecules from the 4 time points were compared using repeated ANOVA (equivalent to paired t test for two groups). Considering 2 markers × 2 time points in survival analysis, Bonferroni correction for multiple testing was performed for statistical significance at a P < 0.0125 (0.05/4) level. All statistical analyses were performed using IBM SPSS Statistics 22.0 (IBM, Inc.). Scatter plot figures of IDO dynamic changes were generated using GraphPad Prism (version 5.01).

Study population and overall clinical outcomes

Of 182 patients enrolled, 110 patients staged III with quality samples available for IDO testing formed the primary study population. There were 103 Caucasian (94%) and 29 (26%) females. The median age was 66 years. All patients (100%) received definitive dose (≥60 Gy) of 3DCRT, and 103 (94%) of whom underwent concurrent platinum-based chemotherapy. Table 1 details the patient demographic and clinical features. The median follow-up time was 56 months (95% CI, 49–63 months). The median OS and PFS were 23 (95% CI, 17–29) and 11 (95% CI, 7–15) months, respectively. Univariate analysis showed that age (P = 0.016), gender (P = 0.040), tumor volume (P = 0.010), KPS (P < 0.001), and EQD2 (P = 0.014) were significant for OS. Younger, female, smaller tumor volume, higher KPS, and higher EQD2 had better OS. However, only tumor volume (P = 0.041), KPS (P = 0.010), and EQD2 (P = 0.004) were significant for PFS (Table 1). These clinical factors were thus selected as clinical covariants for further multivariate analysis of IDO checkpoint–associated parameters.

Table 1

Comparison of OS and PFS based on clinical characteristics of patients

OSPFS
Clinical factorsPatients (n)Death n (%)MST (months)PaHR (95% CI)Progression n (%)MPT (months)PaHR (95% CI)a
Age (years) 
 ≤66 59 38 (64) 31 0.016 1.00 (ref.) 44 (75) 16 0.235 1.00 (ref.) 
 >66 51 41 (80) 14  1.03 (1.01–1.05) 42 (82)  1.01 (0.99–1.04) 
Gender 
 Male 81 63 (78) 22 0.040 1.00 (ref.) 66 (81) 12 0.583 1.00 (ref.) 
 Female 29 16 (55) 41  0.56 (0.32–0.97) 20 (69) 10  0.87 (0.53–1.44) 
Smoking 
 No 4 (57) 43 0.402 1.00 (ref.) 5 (71) 19 0.599 1.00 (ref.) 
 Yes 103 75 (73) 23  1.54 (0.56–4.21) 81 (79) 11  1.27 (0.52–3.15) 
Histology 
 Adenocarcinoma 33 18 (54) 31 0.061 1.00 (ref.) 23 (70) 11 0.918 1.00 (ref.) 
 Squamous cell 38 28 (74) 16  1.83 (1.01–3.31) 29 (76) 10  1.12 (0.65–1.93) 
 NOS 39 33 (85) 19  1.93 (1.08–3.42) 34 (87) 14  1.09 (0.64–1.86) 
Tumor volume (10 cc) 
 ≤16 51 35 (69) 25 0.010 1.00 (ref.) 38 (74) 14 0.041 1.00 (ref.) 
 >16 51 39 (76) 22  1.02 (1.00–1.03) 41 (80)  1.01 (1.00–1.02) 
KPS 
 ≤80 43 36 (84) 13 <0.001 1.00 (ref.) 38 (88) 0.010 1.00 (ref.) 
 >80 67 43 (64) 32  0.95 (0.93–0.98) 48 (72) 14  0.96 (0.94–0.99) 
EQD2 (Gy) 
 ≤70 63 50 (79) 18 0.014 1.00 (ref.) 53 (84) 0.004 1.00 (ref.) 
 >70 47 29 (62) 32  0.97 (0.95–0.99) 33 (70) 19  0.97 (0.95–0.99) 
Chemotherapy 
 No 5 (71) 0.243 1.00 (ref.) 5 (71) 0.476 1.00 (ref.) 
 Yes 103 74 (72) 23  0.58 (0.23–1.44) 81 (79) 12  0.72 (0.29–1.78) 
OSPFS
Clinical factorsPatients (n)Death n (%)MST (months)PaHR (95% CI)Progression n (%)MPT (months)PaHR (95% CI)a
Age (years) 
 ≤66 59 38 (64) 31 0.016 1.00 (ref.) 44 (75) 16 0.235 1.00 (ref.) 
 >66 51 41 (80) 14  1.03 (1.01–1.05) 42 (82)  1.01 (0.99–1.04) 
Gender 
 Male 81 63 (78) 22 0.040 1.00 (ref.) 66 (81) 12 0.583 1.00 (ref.) 
 Female 29 16 (55) 41  0.56 (0.32–0.97) 20 (69) 10  0.87 (0.53–1.44) 
Smoking 
 No 4 (57) 43 0.402 1.00 (ref.) 5 (71) 19 0.599 1.00 (ref.) 
 Yes 103 75 (73) 23  1.54 (0.56–4.21) 81 (79) 11  1.27 (0.52–3.15) 
Histology 
 Adenocarcinoma 33 18 (54) 31 0.061 1.00 (ref.) 23 (70) 11 0.918 1.00 (ref.) 
 Squamous cell 38 28 (74) 16  1.83 (1.01–3.31) 29 (76) 10  1.12 (0.65–1.93) 
 NOS 39 33 (85) 19  1.93 (1.08–3.42) 34 (87) 14  1.09 (0.64–1.86) 
Tumor volume (10 cc) 
 ≤16 51 35 (69) 25 0.010 1.00 (ref.) 38 (74) 14 0.041 1.00 (ref.) 
 >16 51 39 (76) 22  1.02 (1.00–1.03) 41 (80)  1.01 (1.00–1.02) 
KPS 
 ≤80 43 36 (84) 13 <0.001 1.00 (ref.) 38 (88) 0.010 1.00 (ref.) 
 >80 67 43 (64) 32  0.95 (0.93–0.98) 48 (72) 14  0.96 (0.94–0.99) 
EQD2 (Gy) 
 ≤70 63 50 (79) 18 0.014 1.00 (ref.) 53 (84) 0.004 1.00 (ref.) 
 >70 47 29 (62) 32  0.97 (0.95–0.99) 33 (70) 19  0.97 (0.95–0.99) 
Chemotherapy 
 No 5 (71) 0.243 1.00 (ref.) 5 (71) 0.476 1.00 (ref.) 
 Yes 103 74 (72) 23  0.58 (0.23–1.44) 81 (79) 12  0.72 (0.29–1.78) 

Abbreviations: EQD2, the 2 Gy-per-fraction equivalent dose; MPT, median PFS; MST, median OS; NOS, non-otherwise specified.

aBy univariate analysis. Age, tumor volume, KPS, and EQD2 were analyzed as continuous variables.

Dynamics of kynurenine and K:T ratio at different time points

Serum kynurenine and K:T ratio changed heterogeneously through 4 time points from pre-RT, to 2- and 4-week during RT, and post-RT (Fig. 1A and B). Mean kynurenine concentrations did not change significantly at 2 weeks of RT, but increased at 4 weeks during RT, and increased further post-RT. Kynurenine concentrations post-RT were significantly higher than that of other time points. The mean K:T ratio decreased from pre to 2 weeks and to 4 weeks during RT, and increased sharply post-RT. The K:T ratio at the post-RT time point was significantly higher than that of pre- and during-RT time points. K:T ratio at 2 weeks was significantly lower than that of pre-RT under the paired t test analysis (P = 0.007).

Figure 1.

Dynamic changes of IDO-associated molecular activity during and post RT. The top plot shows individual and mean activity levels, and the bottom plot shows spaghetti plots for kynurenine (A) and K:T ratio (B) at four time points. P values shown are from paired t tests. Error bars at figures show 95% CI.

Figure 1.

Dynamic changes of IDO-associated molecular activity during and post RT. The top plot shows individual and mean activity levels, and the bottom plot shows spaghetti plots for kynurenine (A) and K:T ratio (B) at four time points. P values shown are from paired t tests. Error bars at figures show 95% CI.

Close modal

Baseline IDO biomarkers and outcome

Kynurenine at pre-RT was not significantly correlated with OS (HR = 1.03; 95% CI, 0.73–1.45; P = 0.887) or PFS (HR = 0.99; 95% CI, 0.71–1.37; P = 0.927; Table 2). K:T ratio at pre-RT correlated significantly with OS (HR = 2.36; 95% CI, 1.21–4.59; P = 0.011) and PFS (HR = 1.77; 95% CI, 1.02–3.05; P = 0.042; Fig. 2A and B; Table 2) after adjusting for the clinical significant factors. However, only the P value for OS remained significant after Bonferroni correction. High K:T ratio correlated significantly with poorer survival.

Table 2.

Association between IDO activity and changes with OS and PFS of NSCLC

OSPFS
Time pointsIDO activitiesPatients (n)Death n (%)MST (months)PaHR (95% CI)aProgression n (%)MPT (months)PaHR (95% CI)a
Pre-RT (n = 85) Kynurenine 
  ≤Median 43 30 (70) 22 0.887 1.00 (ref.) 33 (77) 0.927 1.00 (ref.) 
  >Median 42 30 (71) 18  1.03 (0.73–1.45) 32 (76) 12  0.99 (0.71–1.37) 
 K:T ratio 
  ≤Median 43 25 (58) 36 0.011 1.00 (ref.) 30 (70) 14 0.042 1.00 (ref.) 
  >Median 42 35 (83) 12  2.36 (1.21–4.59) 35 (83)  1.77 (1.02–3.05) 
Post-RT (n = 76) Kynurenine 
  ≤Median 38 22 (58) 36 0.384 1.00 (ref.) 26 (68) 15 0.235 1.00 (ref.) 
  >Median 38 27 (71) 25  1.07 (0.92–1.24) 29 (76) 16  1.08 (0.95–1.23) 
  K:T ratio          
  ≤Median 38 21 (55) 41 0.277 1.00 (ref.) 26 (68) 25 0.739 1.00 (ref.) 
  >Median 38 28 (74) 24  1.44 (0.75–2.78) 29 (76) 14  1.10 (0.62–1.98) 
Post/Pre (n = 57) Kynurenine 
  ≤Median 28 13 (46) 41 0.006 1.00 (ref.) 16 (57) 15 0.025 1.00 (ref.) 
  >Median 29 22 (76) 28  1.35 (1.09–1.68) 23 (79) 14  1.31 (1.04–1.66) 
 K:T ratio 
  ≤Median 29 18 (62) 32 0.278 1.00 (ref.) 19 (65) 17 0.268 1.00 (ref.) 
  >Median 28 17 (61) 28  1.62 (0.68–3.87) 20 (71) 11  1.57 (0.71–3.46) 
OSPFS
Time pointsIDO activitiesPatients (n)Death n (%)MST (months)PaHR (95% CI)aProgression n (%)MPT (months)PaHR (95% CI)a
Pre-RT (n = 85) Kynurenine 
  ≤Median 43 30 (70) 22 0.887 1.00 (ref.) 33 (77) 0.927 1.00 (ref.) 
  >Median 42 30 (71) 18  1.03 (0.73–1.45) 32 (76) 12  0.99 (0.71–1.37) 
 K:T ratio 
  ≤Median 43 25 (58) 36 0.011 1.00 (ref.) 30 (70) 14 0.042 1.00 (ref.) 
  >Median 42 35 (83) 12  2.36 (1.21–4.59) 35 (83)  1.77 (1.02–3.05) 
Post-RT (n = 76) Kynurenine 
  ≤Median 38 22 (58) 36 0.384 1.00 (ref.) 26 (68) 15 0.235 1.00 (ref.) 
  >Median 38 27 (71) 25  1.07 (0.92–1.24) 29 (76) 16  1.08 (0.95–1.23) 
  K:T ratio          
  ≤Median 38 21 (55) 41 0.277 1.00 (ref.) 26 (68) 25 0.739 1.00 (ref.) 
  >Median 38 28 (74) 24  1.44 (0.75–2.78) 29 (76) 14  1.10 (0.62–1.98) 
Post/Pre (n = 57) Kynurenine 
  ≤Median 28 13 (46) 41 0.006 1.00 (ref.) 16 (57) 15 0.025 1.00 (ref.) 
  >Median 29 22 (76) 28  1.35 (1.09–1.68) 23 (79) 14  1.31 (1.04–1.66) 
 K:T ratio 
  ≤Median 29 18 (62) 32 0.278 1.00 (ref.) 19 (65) 17 0.268 1.00 (ref.) 
  >Median 28 17 (61) 28  1.62 (0.68–3.87) 20 (71) 11  1.57 (0.71–3.46) 

Abbreviations: MPT, median PFS; MST, median OS.

aFrom multivariate Cox proportional hazards regression models by adjusting for tumor volume, KPS, EQD2. IDO parameters were analyzed as continuous variables. P < 0.0125 was considered statistical significance.

Figure 2.

Baseline IDO-mediated immune activity and treatment outcomes. Pre-RT K:T ratio for OS (A); pre-RT K:T ratio for progression-free survival (B).

Figure 2.

Baseline IDO-mediated immune activity and treatment outcomes. Pre-RT K:T ratio for OS (A); pre-RT K:T ratio for progression-free survival (B).

Close modal

Post-RT IDO biomarkers and outcome

None of the post-RT levels in kynurenine or K:T ratio was significant for OS or PFS. Among various RT-induced changes of these IDO checkpoint molecules, only change in kynurenine was significant. Using median as the cut-off, greater levels of post/pre kynurenine correlated significantly with worse OS (HR = 1.35; 95% CI, 1.09–1.68; P = 0.006) and poorer PFS (HR = 1.31; 95% CI, 1.04–1.66; P = 0.025; Fig. 3A and B and Table 2). However, only the P value for OS remained significant after Bonferroni correction.

Figure 3.

Changes of IDO-mediated immune activity and treatment outcomes. Changes of kynurenine after RT (post/pre kynurenine) associated with OS (A) and progression-free survival (B).

Figure 3.

Changes of IDO-mediated immune activity and treatment outcomes. Changes of kynurenine after RT (post/pre kynurenine) associated with OS (A) and progression-free survival (B).

Close modal

Combined effects of baseline and change of IDO biomarker on OS, PFS, and metastasis

To study the combined effects of baseline and change of IDO activities, patients were stratified into the following four groups according to the medians of pre-RT K:T ratio and post/pre kynurenine: (1) low pre-RT K:T ratio and low post/pre kynurenine (favorable baseline and favorable change); (2) high pre-RT K:T ratio and low post/pre kynurenine (unfavorable baseline and favorable change); (3) low pre-RT K:T ratio and high post/pre kynurenine (favorable baseline and unfavorable change); (4) high pre-RT K:T ratio and high post/pre kynurenine (unfavorable baseline and unfavorable change). There were significant differences in OS (P = 0.002; Fig. 4A) and PFS (P = 0.013; Fig. 4B) for these four groups. Patients in Group 1 achieved extremely long-term OS (the median OS has not been reached after up to 57 months of median follow-up), but a relatively low PFS of 14 (10–17) months. Patients in Group 2 had median OS and PFS of 25 (14–36) and 17 (0–45) months, respectively. The median OS and PFS (and their 95% CIs) were 36 (22–50) and 18 (0–45) months for Group 3, and 12 (8–17) and 6 (2–9) months for Group 4, respectively.

Figure 4.

Combined effects of baseline IDO biomarkers and post-RT change with survival and tumor progression. Patients were stratified into four groups according to the medians of pre-RT K:T ratio (baseline immune activity) and post/pre-kynurenine (change of immune activity). Group 1, favorable baseline and favorable change; Group 2, unfavorable baseline and favorable change; Group 3, favorable baseline and unfavorable change; Group 4, unfavorable baseline and change. A, Comparison of OS in four groups. Group 1 showed extremely long-term OS. B, Comparison of PFS in four groups. C, Comparison of tumor local/regional and distant progressions in four groups. The favorable baseline (Groups 1 and 3) corresponded to less local/regional progression, whereas Group 2 had 0% distant progression. D, Comparison of OS for patients developed distant metastasis for Groups 1, 3, and 4.

Figure 4.

Combined effects of baseline IDO biomarkers and post-RT change with survival and tumor progression. Patients were stratified into four groups according to the medians of pre-RT K:T ratio (baseline immune activity) and post/pre-kynurenine (change of immune activity). Group 1, favorable baseline and favorable change; Group 2, unfavorable baseline and favorable change; Group 3, favorable baseline and unfavorable change; Group 4, unfavorable baseline and change. A, Comparison of OS in four groups. Group 1 showed extremely long-term OS. B, Comparison of PFS in four groups. C, Comparison of tumor local/regional and distant progressions in four groups. The favorable baseline (Groups 1 and 3) corresponded to less local/regional progression, whereas Group 2 had 0% distant progression. D, Comparison of OS for patients developed distant metastasis for Groups 1, 3, and 4.

Close modal

Pattern of failure analysis was performed for these four groups to study how baseline and change of the IDO differently affect OS and PFS. Distant and local/regional failure were 6/13 (46%) and 1/13 (8%), 0/15 (0%) and 7/15 (47%), 8/19 (42%) and 2/19 (11%), and 4/10 (40%) and 4/10 (40%) for Groups 1, 2, 3, and 4, respectively (P = 0.019; Fig. 4C). Furthermore, those patients who developed distant progression in Group 1 had a median OS of 41 months, which was longer than that in group 3 and group 4 (23 and 9 months; P = 0.057; Fig. 4D).

This study demonstrated that (1) RT caused significant reductions in IDO activity (represented by K:T ratio) during-RT, but IDO increased significantly post-RT (after more radiation); (2) IDO at baseline and post-RT correlated significantly with OS and PFS; (3) RT-induced change in kynurenine post-RT was significant for either OS or distant progression in an unusual manner (median OS did not reach, corresponded to 88 months mean OS with favorable baseline, and 0% distant progression under unfavorable baseline). These results support our hypotheses that RT can alter the IDO-mediated antitumor immune activity.

This study demonstrates that IDO activity at baseline is an important prognostic marker for patients with NSCLC treated with fractionated RT. This is in agreement with previous studies that baseline IDO immune status plays a prognostic role after surgery or chemotherapy (16, 17, 20–22). Our data demonstrated low activity of IDO, i.e., low level of kynurenine or low K:T ratio (favorable baseline), was significantly associated with better survival. The prognostic significance of baseline IDO biomarkers in our patients treated with RT may serve as a strong validation of above studies and clinically confirms the immune-suppressive role of IDO in NSCLC (27). Although studies with larger sample sizes are needed for RT, this has clinical significance as knowing the prognosis of an individual patient may help triage the care plan of that patient.

More importantly, this study is the first to demonstrate that fractionated RT in NSCLC induces significant change of IDO-mediated immune activity level, and such change correlates with OS, PFS, and unusual treatment outcomes in combination with baseline IDO activity. Studies in animal models have shown that hypofractionated RT can generate synergetic immune responses (3, 5), and there are speculations and single-case reports about activation of antitumor immunity by hypofractionated RT (3, 4). However, there are no reports in human studies showing that RT can generate antitumor immunity in a significant number of patients. Recent results in KYNOTE-001 phase I trial suggested that previous treatment with RT in patients with advanced NSCLC results in longer PFS and OS with pembrolizumab treatment than that seen in patients who did not have previous RT (28). This finding supported that RT affects host immunity in NSCLC patients. Our study showed that patients with favorable RT-induced change in combination with favorable baselines (Group 1) corresponded to extremely long-term OS (>60% OS up to 57 months of median follow-up), suggesting RT-induced immune activity change may generate memory T cells for long-term tumor control. In addition, patients with favorable change and unfavorable baseline (Group 2) had 0% distant progression, suggesting that this RT-induced antitumor immunity may control occult tumor cells outside the RT fields, a phenomenon similar to abscopal effects. However, it is unclear why patients in Group 1 did not have similar “abscopal” effect as in Group 2. The underlying biological mechanism for this phenomenon will require further research. We speculate that relatively small amount of antitumor immune cells were activated in Group 1 patients due to their favorable baseline status. They were not sufficient to control the occult diseases outside the RT fields in a short time, and local salvage treatment was required and able to control the distant metastasis developed from these occult diseases. Our data showed that the metastasis patients still had 41-month median OS, supporting our speculation.

Most interestingly, our data showed that RT can lead favorable or unfavorable changes of IDO immune activity varied with time point (dose level) or patient level. It is unclear what patient-specific factors influence these heterogeneous RT responses. We hypothesize that both underdose and overdose of RT will lead to unfavorable changes of immune activity. Underdose will not be sufficient to generate the favorable change, whereas overdose will damage the immune system and revert into the unfavorable status. Therefore, there is an optimal RT dose for maximal immune activation. Because patients and tumors have heterogeneous radiosensitivities, the optimal RT dose could be different for different patients. We found that median K:T ratio and median kynurenine concentration reduced in the initial phase of RT (at 2-week, lower doses). This statistically significant K:T ratio reduction suggests that IDO activity may be suppressed by antitumor immunity in some patients as early as 2 weeks after starting RT. Though median kynurenine levels tended to increase, median K:T ratio remained at low levels in the middle phase of RT (at 4-week), whereas both increased at later stages (post-RT status, high doses). Therefore, changes of these IDO-associated molecules during-RT may serve as potential biomarkers to determine the optimal individualized RT dose for each patient by determining when to stop excessive RT. However, a slight increase of IDO or kynurenine concentration during RT may not be an indication of unfavorable change, because immune response to radiation damage of other normal tissues may also partially contribute to the increase. Further study is required to understand better the underlying mechanisms.

In addition, the findings from this study may be relevant to blood biomarker to guide multimodality clinical trials. Anti-IDO agents are available commercially, and clinical trials using various IDO inhibitors are in progress (29). Animal studies have revealed synergies between RT and anti-IDO agents (30). A pilot study using an IDO inhibitor after chemotherapy in stage III–IV NSCLC patients has generated promising results (31). A key question about combining RT with IDO inhibitor therapy and other immunotherapy drugs is the optimal RT dose and optimal time points to give IDO inhibitors. RT doses that activate antitumor immunity but do not cause excessive immune depletion are desirable.

This study has limitations. One of them is serum sample availability. The numbers of available samples at different time points were diverse, which may affect study results, especially for intergroup comparisons. Also, the study was not powered for a subgroup analysis on chemotherapy–IDO activity association. Previous studies have shown synergy between chemotherapy and IDO inhibitors in promoting clinical responses. It is possible that potential synergistic effects of chemotherapy may have partially biased our results. Rather than a sole effect of RT, the observed phenomenon could be the combined effects of chemotherapy and RT.

A.L. Mellor is a consultant/advisory board member for NewLink Genetics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: W. Wang, L. Huang, J.-Y. Jin, Y. Zang, F.-M. (Spring) Kong

Development of methodology: W. Wang, L. Huang, J.-Y. Jin, A.L. Mellor, F.-M. (Spring) Kong

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): W. Wang, L. Huang, A.L. Mellor, F.-M. (Spring) Kong

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): W. Wang, L. Huang, J.-Y. Jin, H. Wu, W. Pi, L. Li, A.L. Mellor, F.-M. (Spring) Kong

Writing, review, and/or revision of the manuscript: W. Wang, L. Huang, J.-Y. Jin, S. Jolly, H. Wu, W. Pi, A.L. Mellor, F.-M. (Spring) Kong

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): W. Wang, J.-Y. Jin, L. Yan, F.-M. (Spring) Kong

Study supervision: L. Huang, J.-Y. Jin, F.-M. (Spring) Kong

We are grateful to Theodore S. Lawrence, MD and PhD, for his valuable review comments of the article. This work was supported in part by NIH grant R01CA142840 [Principal Investigator: F.-M. (Spring) Kong] and a start-up award [Principal Investigator: F.-M. (Spring) Kong] from Indiana University School of Medicine.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Miller
KD
,
Siegel
RL
,
Lin
CC
,
Mariotto
AB
,
Kramer
JL
,
Rowland
JH
, et al
Cancer treatment and survivorship statistics, 2016
.
CA: Cancer J Clin
2016
;
66
:
271
89
.
2.
Tyldesley
S
,
Boyd
C
,
Schulze
K
,
Walker
H
,
Mackillop
WJ
. 
Estimating the need for radiotherapy for lung cancer: an evidence-based, epidemiologic approach
.
Int J Radiat Oncol Biol Phys
2001
;
49
:
973
85
.
3.
Formenti
SC
. 
Silvia Formenti on the promise of combining radiotherapy and immunotherapy to treat cancer
.
Oncology (Williston Park)
2016
;
30
:
289
92
.
4.
Postow
MA
,
Callahan
MK
,
Barker
CA
,
Yamada
Y
,
Yuan
J
,
Kitano
S
, et al
Immunologic correlates of the abscopal effect in a patient with melanoma
.
N Engl J Med
2012
;
366
:
925
31
.
5.
Lee
Y
,
Auh
SL
,
Wang
Y
,
Burnette
B
,
Wang
Y
,
Meng
Y
, et al
Therapeutic effects of ablative radiation on local tumor require CD8+ T cells: changing strategies for cancer treatment
.
Blood
2009
;
114
:
589
95
.
6.
Platten
M
,
Wick
W
,
Van den Eynde
BJ
. 
Tryptophan catabolism in cancer: beyond IDO and tryptophan depletion
.
Cancer Res
2012
;
72
:
5435
40
.
7.
Munn
DH
,
Sharma
MD
,
Baban
B
,
Harding
HP
,
Zhang
Y
,
Ron
D
, et al
GCN2 kinase in T cells mediates proliferative arrest and anergy induction in response to indoleamine 2,3-dioxygenase
.
Immunity
2005
;
22
:
633
42
.
8.
Sharma
MD
,
Baban
B
,
Chandler
P
,
Hou
DY
,
Singh
N
,
Yagita
H
, et al
Plasmacytoid dendritic cells from mouse tumor-draining lymph nodes directly activate mature Tregs via indoleamine 2,3-dioxygenase
.
J Clin Invest
2007
;
117
:
2570
82
.
9.
Schafer
CC
,
Wang
Y
,
Hough
KP
,
Sawant
A
,
Grant
SC
,
Thannickal
VJ
, et al
Indoleamine 2,3-dioxygenase regulates anti-tumor immunity in lung cancer by metabolic reprogramming of immune cells in the tumor microenvironment
.
Oncotarget
2016
;
7
:
75407
24
.
10.
Pallotta
MT
,
Orabona
C
,
Volpi
C
,
Vacca
C
,
Belladonna
ML
,
Bianchi
R
, et al
Indoleamine 2,3-dioxygenase is a signaling protein in long-term tolerance by dendritic cells
.
Nat Immunol
2011
;
12
:
870
8
.
11.
Munn
DH
,
Mellor
AL
. 
IDO in the tumor microenvironment: inflammation, counter-regulation, and tolerance
.
Trends Immunol
2016
;
37
:
193
207
.
12.
Mbongue
JC
,
Nicholas
DA
,
Torrez
TW
,
Kim
NS
,
Firek
AF
,
Langridge
WH
. 
The role of indoleamine 2, 3-dioxygenase in immune suppression and autoimmunity
.
Vaccines (Basel)
2015
;
3
:
703
29
.
13.
Liu
H
,
Liu
L
,
Liu
K
,
Bizargity
P
,
Hancock
WW
,
Visner
GA
. 
Reduced cytotoxic function of effector CD8+ T cells is responsible for indoleamine 2,3-dioxygenase-dependent immune suppression
.
J Immunol
2009
;
183
:
1022
31
.
14.
Zhang
G
,
Liu
WL
,
Zhang
L
,
Wang
JY
,
Kuang
MH
,
Liu
P
, et al
Involvement of indoleamine 2,3-dioxygenase in impairing tumor-infiltrating CD8 T-cell functions in esophageal squamous cell carcinoma
.
Clin Dev Immunol
2011
;
2011
:
384726
.
15.
Brandacher
G
,
Perathoner
A
,
Ladurner
R
,
Schneeberger
S
,
Obrist
P
,
Winkler
C
, et al
Prognostic value of indoleamine 2,3-dioxygenase expression in colorectal cancer: effect on tumor-infiltrating T cells
.
Clin Cancer Res
2006
;
12
:
1144
51
.
16.
Ino
K
. 
Indoleamine 2,3-dioxygenase and immune tolerance in ovarian cancer
.
Curr Opin Obstet Gynecol
2011
;
23
:
13
8
.
17.
Pelak
MJ
,
Snietura
M
,
Lange
D
,
Nikiel
B
,
Pecka
KM
. 
The prognostic significance of indoleamine-2,3-dioxygenase and the receptors for transforming growth factor beta and interferon gamma in metastatic lymph nodes in malignant melanoma
.
Polish J Pathol
2015
;
66
:
376
82
.
18.
Chamuleau
ME
,
van de Loosdrecht
AA
,
Hess
CJ
,
Janssen
JJ
,
Zevenbergen
A
,
Delwel
R
, et al
High INDO (indoleamine 2,3-dioxygenase) mRNA level in blasts of acute myeloid leukemic patients predicts poor clinical outcome
.
Haematologica
2008
;
93
:
1894
8
.
19.
Cavia-Saiz
M
,
Muniz
P
,
De Santiago
R
,
Herreros-Villanueva
M
,
Garcia-Giron
C
,
Lopez
AS
, et al
Changes in the levels of thioredoxin and indoleamine-2,3-dioxygenase activity in plasma of patients with colorectal cancer treated with chemotherapy
.
Biochem Cell Biol
2012
;
90
:
173
8
.
20.
Suzuki
K
,
Kachala
SS
,
Kadota
K
,
Shen
R
,
Mo
Q
,
Beer
DG
, et al
Prognostic immune markers in non-small cell lung cancer
.
Clin Cancer Res
2011
;
17
:
5247
56
.
21.
Creelan
BC
,
Antonia
S
,
Bepler
G
,
Garrett
TJ
,
Simon
GR
,
Soliman
HH
. 
Indoleamine 2,3-dioxygenase activity and clinical outcome following induction chemotherapy and concurrent chemoradiation in stage III non-small cell lung cancer
.
Oncoimmunology
2013
;
2
:
e23428
.
22.
Suzuki
Y
,
Suda
T
,
Furuhashi
K
,
Suzuki
M
,
Fujie
M
,
Hahimoto
D
, et al
Increased serum kynurenine/tryptophan ratio correlates with disease progression in lung cancer
.
Lung Cancer
2010
;
67
:
361
5
.
23.
Chapet
O
,
Kong
FM
,
Quint
LE
,
Chang
AC
,
Ten Haken
RK
,
Eisbruch
A
, et al
CT-based definition of thoracic lymph node stations: an atlas from the University of Michigan
.
Int J Radiat Oncol Biol Phys
2005
;
63
:
170
8
.
24.
Kong
FM
,
Hayman
JA
,
Griffith
KA
,
Kalemkerian
GP
,
Arenberg
D
,
Lyons
S
, et al
Final toxicity results of a radiation-dose escalation study in patients with non-small-cell lung cancer (NSCLC): predictors for radiation pneumonitis and fibrosis
.
Int J Radiat Oncol Biol Phys
2006
;
65
:
1075
86
.
25.
Huang
L
,
Lemos
HP
,
Li
L
,
Li
M
,
Chandler
PR
,
Baban
B
, et al
Engineering DNA nanoparticles as immunomodulatory reagents that activate regulatory T cells
.
J Immunol
2012
;
188
:
4913
20
.
26.
Laich
A
,
Neurauter
G
,
Widner
B
,
Fuchs
D
. 
More rapid method for simultaneous measurement of tryptophan and kynurenine by HPLC
.
Clin Chem
2002
;
48
:
579
81
.
27.
Soliman
H
,
Mediavilla-Varela
M
,
Antonia
S
. 
Indoleamine 2,3-dioxygenase: is it an immune suppressor?
Cancer J
2010
;
16
:
354
9
.
28.
Shaverdian
N
,
Lisberg
AE
,
Bornazyan
K
,
Veruttipong
D
,
Goldman
JW
,
Formenti
SC
, et al
Previous radiotherapy and the clinical activity and toxicity of pembrolizumab in the treatment of non-small-cell lung cancer: a secondary analysis of the KEYNOTE-001 phase 1 trial
.
Lancet Oncol
2017
;
18
:
895
903
.
29.
Vacchelli
E
,
Aranda
F
,
Eggermont
A
,
Sautes-Fridman
C
,
Tartour
E
,
Kennedy
EP
, et al
Trial watch: IDO inhibitors in cancer therapy
.
Oncoimmunology
2014
;
3
:
e957994
.
30.
Li
M
,
Bolduc
AR
,
Hoda
MN
,
Gamble
DN
,
Dolisca
SB
,
Bolduc
AK
, et al
The indoleamine 2,3-dioxygenase pathway controls complement-dependent enhancement of chemo-radiation therapy against murine glioblastoma
.
J Immunother Cancer
2014
;
2
:
21
.
31.
Iversen
TZ
,
Engell-Noerregaard
L
,
Ellebaek
E
,
Andersen
R
,
Larsen
SK
,
Bjoern
J
, et al
Long-lasting disease stabilization in the absence of toxicity in metastatic lung cancer patients vaccinated with an epitope derived from indoleamine 2,3 dioxygenase
.
Clin Cancer Res
2014
;
20
:
221
32
.

Supplementary data