Open access Original research J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by Clinical and molecular parameters associated to pneumonitis development in non-­small-­cell lung cancer patients receiving chemoimmunotherapy from NADIM trial Belén Sierra-­Rodero ‍ ‍ ,1 Alberto Cruz-­Bermúdez ‍ ‍ ,1 Ernest Nadal ‍ ‍ ,2 Yago Garitaonaindía,1 Amelia Insa,3 Joaquín Mosquera,4 Joaquín Casal-­Rubio,5 Manuel Dómine,6 Margarita Majem ‍ ‍ ,7 Delvys Rodriguez-­Abreu,8 Alex Martinez-­Marti ‍ ‍ ,9 Javier De Castro Carpeño,10 Manuel Cobo,11 Guillermo López Vivanco,12 Edel Del Barco,13 Reyes Bernabé Caro,14 Nuria Viñolas,15 Isidoro Barneto Aranda,16 Santiago Viteri,17 Bartomeu Massuti,18 Raquel Laza-­Briviesca,1 Marta Casarrubios,1 Aránzazu García-­Grande,19 To cite: Sierra-­Rodero B, Atocha Romero,1 Fernando Franco,1 Mariano Provencio ‍ ‍ 1 Cruz-­Bermúdez A, Nadal E, et al. Clinical and molecular parameters associated to pneumonitis development in non-­small-­cell lung ABSTRACT increased clonality of intratumoral T cells; and elevated Background Pneumonitis (Pn) is one of the main plasma levels of several growth factors (EGF, HGF, VEGF, cancer patients receiving chemoimmunotherapy from immune-­related adverse effects, having a special ANG-1, PDGF, NGF, and NT4) and inflammatory cytokines NADIM trial. Journal for importance in lung cancer, since they share affected (MIF, CCL16, neutrophil gelatinase-­associated lipocalin, copyright. ImmunoTherapy of Cancer tissue. Despite its clinical relevance, Pn development BMP-4, and u-­PAR). 2021;9:e002804. doi:10.1136/ remains an unpredictable treatment adverse effect, whose Conclusions Although statistically underpowered, our jitc-2021-002804 mechanisms are mainly unknown, being even more results shed light on the possible mechanisms behind Pn obscure when it is associated to chemoimmunotherapy. development, involving innate and adaptative immunity, ►► Additional supplemental Methods In order to identify parameters associated to and open the possibility to predict patients at high risk. If material is published online only. treatment related Pn, we analyzed clinical variables and confirmed, this may allow the personalization of both, the To view, please visit the journal surveillance strategy and the therapeutic approaches to molecular parameters from 46 patients with potentially online (http://​dx.d​ oi.​org/​10.​ manage Pn in patients receiving chemoimmunotherapy. 1136/j​ itc-​2021-0​ 02804). resectable stage IIIA non-­small-­cell lung cancer treated with neoadjuvant chemoimmunotherapy included in the NADIM clinical trial (NCT03081689). Pn was defined as BS-­R and AC-­B are joint first clinical or radiographic evidence of lung inflammation INTRODUCTION authors. without alternative diagnoses, from treatment initiation to The programmed death-1 (PD1)/PD ligand AC-­B and MP are joint senior 180 days. 1 (PD-­L1) pathway is one of the several mech- authors. Results Among 46 patients, 12 developed Pn (26.1%). anisms that tumor cells use to repress natural Sex, age, smoking status, packs-­year, histological antitumor immune activity.1 Development of Accepted 29 July 2021 subtype, clinical or pathological response, progression-­ antibodies that specifically target these mole- free survival, overall survival and number of nivolumab cules has permitted weaponizing the host cycles, were not associated to Pn development. Regarding immune system against cancer, boosting the © Author(s) (or their molecular parameters at diagnosis, Pn development employer(s)) 2021. Re-­use immune response against the tumor.2 was not associated to programmed death ligand 1, Nevertheless, these molecules are also permitted under CC BY-­NC. No commercial re-­use. See rights TPS, T cell receptor repertoire parameters, or tumor involved in the maintenance of immunologic and permissions. Published by mutational burden. However, patients who developed Pn homeostasis and self-­ molecules immunotol- BMJ. had statistically significant lower blood median levels of erance, preventing excessive autoimmunity For numbered affiliations see platelet to monocyte ratio (p=0.012) and teratocarcinoma-­ derived growth factor 1 (p=0.013; area under the curve throughout life. Thus, its inhibition may as end of article. (AUC) 0.801), but higher median percentages of natural well lead to autoimmune-­like adverse events, Correspondence to killers (NKs) (p=0.019; AUC 0.786), monocytes (p=0.017; by disturbing the normal immunoregulation Dr Alberto Cruz-­Bermúdez; AUC 0.791), MSP (p=0.006; AUC 0.838), PARN (p=0.017; of the body. These immune-­relates adverse ​alberto.​cruz.b​ ermudez@​gmail.​ effects (irAEs) have a significant impact on AUC 0.790), and E-­Cadherin (p=0.022; AUC 0.788). In com addition, the immune scenario of Pn after neoadjuvant patients, leading in some cases to disrup- Dr Mariano Provencio; treatment involves: high levels of neutrophils and NK tion of treatments and to life-­ threatening ​mprovenciop@​gmail.c​ om cells, but low levels of B and T cells in peripheral blood; situations.3 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 1 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by The combination of immune checkpoint inhibitors Common Terminology Criteria for Adverse Events V.4.0. with chemotherapy, has not only increased survival, but Investigators determined whether adverse events were toxicity as well, making it an arising issue as chemoimmu- treatment related according to the study protocol and notherapy is being positioned in the frontline treatment standard regulatory requirements. of some tumors, and in particular, in locally or metastatic A central committee confirmed all Pn cases and consid- non-­small-­cell lung cancer (NSCLC).4 5 ered adverse events related with neoadjuvant treatment One of the main irAEs in terms of incidence and fatality those developed within 180 days after the administration rate is pneumonitis (Pn).6 Patients with lung cancer have of first dose of neoadjuvant nivolumab. Patients with Pn the highest rates of Pn,7 specially addressed in advanced were considered to be those who had clinical and/or NSCLC, where its documented incidence using combina- radiological findings of Pn of any grade (such as dyspnea tion of anti PD-­(L)1 with chemotherapy oscillates from and pleural pain), once other clinical entities had been 1.3% to 6.5%, with a fatal rate between 0% and 0.4%. ruled out (namely infectious diseases, pulmonary throm- However, rates of reported dyspnea oscillate from 13.6% boembolism, or tumor progression, among others). to 23.4%, which may suggest that the real rate of Pn may Clinical variables as well as molecular parameters (at be underestimated.8–12 Pn incidence of chemoimmuno- diagnosis and after neoadjuvant treatment, along with therapy is still less known in the NSCLC perioperative their treatment variation) were compared between scenario, where few studies have reported efficacy results, patients who developed Pn as an adverse effect to treat- with no specific data reported on Pn incidence.13–16 ment and patients who did not, in order to search for In addition, although carboplatin or paclitaxel have a markers associated with this irAE. reasonably safe pulmonary toxicity profile, it has been described that chemotherapy alone can produce Pn PD-L1 TPS, tumor mutational burden and specific mutations and fibrosis, being the incidence higher when these Basal PD-­ L1 tumor proportion score (TPS), tumor treatments are combined.17 18 Thus, potentially modi- mutational burden (TMB) and specific mutations were fying the possible mechanisms behind anti-­PD(L)1 Pn retrieved from.13 Briefly, PD-­L1 immunohistochemistry and contributing to increased pulmonary toxicity of assay (22C3 pharmaDx, Code SK006; Dako, Glostrup, chemoimmunotherapy. Denmark) was used to assess diagnostic PD-­ L1 TPS Some risk factors have been proposed, such as prior following manufacturer’s instructions. lung disease, prior lung radiotherapy, combination with TMB Library generation and sequencing of samples copyright. anti-­CTLA-4,19 or the levels of some immune-­ related was performed on an Ion Chef System and S5 Sequencer. molecular parameters. However, the scarcity of data on Specifically, 20 ng of extracted DNA was treated with heat the biological mechanisms, has resulted in a lack of reli- labile uracil-­DNA glycosylase and used for library prepa- able biomarkers to identify patients at risk and limited ration using the Oncomine Tumor Mutation Load Assay. understanding on how to best treat them.20 21 Eight samples were loaded at 50pM and sequenced onto The aim of this study is to identify potential clinical and an Ion 540 chip. Reads were aligned to hg19 using Torrent molecular biomarkers associated to Pn development after Suite 5.12 and BAM files were transferred to Ion Reporter neoadjuvant chemoimmunotherapy in NSCLC patients. 5.12 for variant calling. TMB was computed using the TMB filter chain and the TMB algorithm 3.0. Germline variants were filtered out using a germline filter-­chain METHODS based on population databases. Additionally, specific Study population and design mutations using the Oncomine Variants 5.12 filter whose We analyzed treatment related adverse events and clin- variant allele frequency was greater than 5%, reached ≥60 ical variables of 46 patients with potentially resectable of coverage and a p≤0.05 were reported. stage IIIA NSCLC in the neoadjuvant setting with chemo- immunotherapy included in the NADIM clinical trial Blood counts (NCT03081689), an open-­label, multicenter, single-­arm Patients had laboratory blood tests before each 21-­day phase II trial done at 18 hospitals in Spain. treatment cycle to monitor complete blood cell counts Patients included in the study were treated with and biochemical parameters. Hemoglobin, leucocytes, nivolumab (360 mg), paclitaxel (200 mg/m2) and carbo- neutrophils, eosinophils, lymphocytes, monocytes, plate- platin (area under the curve (AUC) 6; 6 mg/mL per lets and LDH levels were retrieved from laboratory blood min), on day 1 of each 21-­day cycle, for three cycles before tests at two timepoints: prior treatment initiation and surgical resection.13 Main exclusion criteria were the pres- after neoadjuvant treatment but before surgery. Addi- ence of active autoimmune or infectious disease, ongoing tionally, Lung Immune Prognostic Index (LIPI), neutro- systemic corticosteroid or other immunosuppressive phil to lymphocyte ratio (NLR), monocyte to lymphocyte therapy, history of symptomatic grade 3 or 4 interstitial ratio (MLR), platelet to lymphocyte ratio (PLR), derived lung disease, clinically significant concurrent malignan- NLR (dNLR), and platelet to monocyte ratio (PMR) were cies and previous treatment with checkpoint inhibitors. calculated. dNLR was defined as neutrophils/(leucocytes-­ Adverse events and abnormal laboratory findings neutrophils). LIPI was estimated as the sum of two factors, were graded according to the National Cancer Institute dNLR greater than three and lactate dehydrogenase 2 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by (LDH) greater than 333 IU/L, defining three groups Statistical analysis (good, 0 factors; intermediate, 1 factor; poor, 2 factors).22 To test for clinical and molecular data categorical associa- tions with Pn development, we performed a Fisher’s exact Immunophenotyping and cytokine levels test, in variables such as sex, histology, smoking status, Cell and plasma isolation, immunophenotyping of pathological and clinical response, number of cycles, peripheral blood mononuclear cells (PBMCs) and immu- specific mutations and LIPI score. Non-­parametric Mann-­ noassays for detecting cytokines in plasma were obtained Whitney U test was performed for clinical variable analysis and processed as previously described 23 . Briefly, 29 (age and packs-­year) and quantitative molecular param- samples of peripheral blood mononuclear cells (PBMCs) eters, such as blood counts, cytokines, TCR repertoire at baseline or after neoadjuvant treatment, but presur- and immune populations. To analyze changes related to gery, were isolated by lymphoprep gradient centrifu- neoadjuvant treatment specific of Pn development, differ- gation and cryopreserved with freezing medium until ences between paired postsamples and presamples were use for flow cytometry analysis. Immunophenotyping of calculated for each patient group and Mann-­Whitney U general and specific subpopulations of circulating natural test was performed. Kaplan-­Meier survival analysis was killer (NK) cells (CD3-­CD56+), T cells (CD3+), cytotoxic performed using the log-­rank test in order to identify T cells (CD3 +CD8+), helper T cells (CD3 +CD4+), T-­NK difference between patients who developed pneumonitis like cells (CD3 +CD56+), B cells (CD3-­CD19+) and mono- and those who did not. The receiver operating character- cytes (CD14+) with different activation and checkpoint istic (ROC) curve analysis was calculated to predict the markers was determined by multicolor panels on a MACS potential value of the parameters as predictors of pneu- Quant 10 cytometer (Miltenyi Biotec) and analyzed with monitis development, and AUC values over 0.75 were the FLOWJO software, as previously described.23 When considered good enough predictors to be included in data from different antibody panels were available for this study. Since the research was designed as a discovery the same cell population median value among them was study, p values were not adjusted in order to maximize used. Information derived from anti-­PD-1 clone PD1.3.1.3 the finding of new biomarkers and the generation of new labeling in posttreatment samples was not included in the hypothesis. P<0.05 was considered statistically significant analysis due to nivolumab binding competition to PD-1 for all tests. Finally, the post hoc nature of the analyses, receptor. together with the large number of variables analyzed and Plasma fraction of 30 patients at baseline and 34 patients the small number of patients, requires confirmation of copyright. after neoadjuvant treatment was collected after gradient the biomarkers found in independent cohorts. centrifugation and stored diluted 1:2 with RPMI at −80°C. Levels of 200 cytokines related to cancer biomarkers were measured using G-­Series Human Cytokine Antibody Array RESULTS 4000 (RayBiotech) following manufacturer protocols. Clinical characteristics The cohort of patients included 46 patients with NSCLC T cell receptor repertoire stage IIIA, according to American Joint Committee Tissue and blood samples were used for the analysis of the on Cancer seventh edition criteria, recruited across 18 T cell receptor (TCR) repertoire, both at diagnosis and centers in Spain between April 26, 2017 and Aug 25, after neoadjuvant treatment. TCR Library preparation 2018. All of them were treated with neoadjuvant chemo- and sequencing was carried out as previously described.24 immunotherapy according to protocol described in Mate- Briefly, RNA input for PBMCs-­ derived libraries using rial and Methods. the Oncomine TCR Beta–LR Assay was 25 ng, whereas At the time of the data cut-­off, with a median follow-­up for formalin-­ fixed, paraffin-­ embedded (FFPE) derived of 24.0 months (IQR 21.4–28.1), 12/46 patients (26.1%) libraries using Oncomine TCR Beta–SR Assay, was 100 ng. developed a clinical and/or radiological treatment For PBMCs-­derived libraries, equal volumes from eight related Pn within 180 days after receiving the first dose of samples at 25 pM were pooled for sequencing on an Ion nivolumab. The median to Pn development was 109.5 day 530 chip. For FFPE-­derived libraries, equal volumes from (IQR 58–130). up to 32 samples at 25 pM were combined for sequencing We found no difference between patients who devel- on an Ion 540 chip. oped Pn and those who did not regarding sex (p=1.000), Once the libraries were templated on an Ion Chef age (p=0.910), smoking status (p=0.749), packs-­ year System, they were sequenced in the Ion GeneStudio S5 (p=0.196), histological subtype (p=0.316), or clinical Series and analysis was done via Ion Reporter V.5.12. response (p=0.836). No significant differences were Convergence, Shannon’s diversity index and evenness found between patients who had complete pathological were provided by Ion Reporter Software as a standard response (CPR) and those who had not (p=0.480) or output. Dynamics of expanded and contracted clones those who had CPR or major pathological response and in blood and tissue were calculated as percentages and those with incomplete pathological response (p=0.398). clonal space of clones that increased or decreased their We found no differences in progression-­ free survival frequency after treatment, relative to pretreatment (p=0.761) or overall survival (p=0.726) between patients repertoire. who developed Pn and those who did not. No association Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 3 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by Table 1 Clinical characteristics of patients and statistical significance related to pneumonitis development Pneumonitis Non-­pneumonitis Characteristic (n=12) (n=34) P value Age (years)—median (range) 61.0 (47.0–76.0) 63.5 (41.0–76.0) 0.910 Sex—n (%) 1.000  Male 9 (75.0) 25 (73.5)  Female 3 (25.0) 9 (26.5) Histology—n (%)* 0.316  Adenocarcinoma 9 (75.0) 17 (50.0)  Squamous 3 (25.0) 13 (38.2)  No otherwise specified 0 (0.0) 4 (11.8) Smoking status—n (%) 0.749    Former 6 (50.0) 19 (55.9)    Current 6 (50.0) 15 (44.1) Packs-­year—median (range) 57.0 (20.0–100.0) 45.0 (22.0–114.0) 0.196 Clinical response—n (%) 0.836    Complete 0 (0.0) 2 (5.9)    Partial 10 (83.3) 23 (67.6)    Stable 2 (16.7) 9 (26.5) Pathological response—n (%)† 0.480‡, 0.398§    Complete 9 (75.0) 17 (58.6)    Major 0 (0.0) 8 (27.6)    Incomplete 3 (25.0) 4 (13.8) copyright. PFS (months)—median (95% CI) NR NR 0.761 OS (months)—median (95% CI) NR NR 0.726 No cycles—N (%) 0.409 ≤3 cycles    1 (8.3) 8 (23.5) >3 cycles    11 (91.7) 26 (76.5) *Four patients had no specified histology †Five patients did not undergo surgery ‡Complete responses versus other §Incomplete responses versus other NR, not reported; OS, overall survival; PFS, progression-­free survival. between number of nivolumab cycles received and Pn Pn (66.7%), while only three cases of Pn were described development was found (p=0.409) (table 1). in the 26 non-­mutated patients (11.5%) (figure 1B and We also wanted to investigate any relationship between table 2). Pn development and molecular characteristics of the Other less frequent mutations found in our cohort of tumors, specifically, PD-­L1 TPS, TMB or specific baseline patients (n<3 cases) were STK11 (one mutated patient mutations. 28 of 46 patients had valid data for PD-­L1 TPS who did not develop Pn), EGFR (one mutated patient (six developed Pn), and 29 had valid data for baseline who did not develop Pn) and BRAF (one mutated patient mutation analysis (five developed Pn). who developed Pn). No mutations were found in RET, No association between PD-­ L1 TPS (Pn 45%, IQR ROS, MET, or ALK. 0%–100%; Non-­ Pn 32.5%, IQR 0%–75%; p=0.491) or We further sought to identify Pn associated biomarkers TMB (Pn 7.6, IQR 4.2–55.8; Non-­Pn 5.9, IQR 3.6–9.0; in patient cell blood counts at baseline (n=46) or postneo- p=0.298) and Pn development were observed (figure 1A). adjuvant treatment (n=45). Different standard parame- Mutations in KEAP1, ARID1A, RB1, HNF1A, TP53 or ters in the hemograms from all patients were analyzed, as KRAS genes did not show significant association to Pn well as the ratios and scores derived from this data. development. However, despite the low number of cases, No significant differences were observed for hemo- the presence of ARID1A mutations showed a trend to be globin, platelets, eosinophils, basophils, lymphocytes associated to Pn development (p=0.068). Out of three and LDH. (online supplemental figure S1A and online patients who presented the mutation, two developed supplemental table 1). At diagnosis, only a trend to 4 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by Figure 1 PD-­L1, TMB and specific mutations. PD-­L1 TPS levels (n=28; p=0.491), TMB (n=29; p=0.298) and specific mutations (n=29; p=0.127 for KEAP1; p=0.068 for ARID1A; p=1 for RB1; p=1 for HNF1A; p=0.622 for TP53 and p=0.553 for KRAS). *P<0.05; **p<0.01; ***p<0.001. n.s., not significant; PD-­L1, programmed death ligand 1; TMB, tumor mutational burden. higher monocyte counts in patients who develop Pn, was increased levels of neutrophils (Pn 4.31, IQR 3.38–4.84; copyright. observed (Pn 0.75, IQR 0.63–1.15; non-­Pn 0.70, IQR 0.50– non-­Pn 3.1, IQR 2.56–4.15; p=0.041), as well as, a trend to 0.84; p=0.079) (figure 2A). However, in postneoadjuvant higher leukocyte counts (Pn 7.61, IQR 6.16–8.27; non-­Pn treatment samples, patients who developed Pn showed 5.95, IQR 4.8–7.11; p=0.063) (figure 2A). Table 2 Association of specific mutations and pneumonitis development using statistical Fisher’s exact values Gene Non-­pneumonitis (n=24) Pneumonitis (n=5) Total (n=29) P value KEAP1—n (%) 0.127 Wt 22 (88) 3 (12) 25 Mut 2 (50) 2 (50) 4 ARID1A—n (%) 0.068 Wt 23 (88.5) 3 (11.5) 26 Mut 1 (33.3) 2 (66.7) 3 RB1—n (%) 1 Wt 21 (80.8) 5 (19.2) 26 Mut 3 (100) 0 (0) 3 HNF1A—n (%) 1 Wt 19 (82.6) 4 (17.4) 23 Mut 5 (83.3) 1 (16.7) 6 TP53—n (%) 0.622 Wt 15 (88.2) 2 (11.8) 17 Mut 9 (75) 3 (25) 12 KRAS—n (%) 0.553 Wt 21 (84) 4 (16) 25 Mut 3 (75) 1 (25) 4 Mut, mutated; Wt, wild-­type. Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 5 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by Figure 2 Total blood counts and blood ratios. (A) Total blood count parameters (n=46 in pretreatment and n=45 in postneoadjuvant treatment samples; p=0.940 and p=0.063 for leucocytes; p=0.881 and p=0.041 for neutrophils; and p=0.079 and p=0.275 for monocytes; in preneoadjuvant and postneoadjuvant samples, respectively). (B) Ratios derived from hemograms (n=46 in pretreatment and n=45 in postneoadjuvant treatment samples; p=0.054 and p=0.270 for PLR, and p=0.012 and p=0.095 for PMR). *P<0.05; **p<0.01; ***p<0.001. n.s., not significant; PLR, platelet to lymphocyte ratio; PMR, platelet to monocyte ratio. Related to blood counts ratios, there were no signifi- Pn 2.8, IQR 1.7–5.7; non-­Pn 2.0, IQR 1.2–4.6; p=0.283) cant differences in NLR, dNLR, MLR, and LIPI between and non-­ classical (CD14-­ CD16+; Pn 2.1, IQR 1.0–6.2; copyright. patients with and without Pn (online supplemental figure non-­Pn 1.5, IQR 0.6–2.9; p=0.317) monocytes (online S1B). However, patients who developed Pn showed a trend supplemental figure S2A). In postneoadjuvant treatment to lower platelet-­to-­lymphocyte ratio (PLR; Pn 105.3, IQR samples, there was no statistical significance, neither 70.8–161.8; non-­Pn 146.9, IQR 103.8–214.2; p=0.054) and in CD14 + general monocytes (figure 3A), nor in their statistically significant lower platelet-­ to-­ monocyte ratio, different subpopulations (online supplemental figure at baseline(PMR; Pn 317.3, IQR 204.6–443.8; non-­ Pn S2A). 436.0, IQR 353.9–597.0; p=0.012). These differences were Regarding B cells, Pn patients had a trend to lower reduced after neo-­adjuvant treatment (figure 2B). percentage of CD3-­CD19+B cell population in pretreat- ment samples (Pn 12.6, IQR 7.9–23.7; non-­Pn 22.5, IQR Immunophenotyping and TCR repertoire 15.1–38.8; p=0.064), that reached statistical significance To identify candidate immune populations to serve as after neoadjuvant treatment (Pn 5.44, IQR 3.43–11.94; peripheral blood biomarkers, we performed flow cytom- non-­Pn 24, IQR 14.15–35.15; p=0.002) (figure 3B). etry analysis in preneoadjuvant and postneoadjuvant samples from 29 patients, of which 8 developed Pn. We However, we did not see any differences regarding to characterized general monocytes, NK, T and B cells CD25 + and CTLA4 + B cells subpopulations in both as well as additional immune subpopulations (online pretreatment and posttreatment samples (online supple- supplemental table 2). mental figure S2B). Similarly, as described above for blood counts, we Patients who had developed Pn showed a significant saw higher percentage of monocytes in pretreatment higher percentage of CD3-­CD56+ NK cells at diagnosis samples, characterized by CD14 expression detected (Pn 11.2, IQR 2.8–15.0; non-­Pn 3.3, 1.4–6.9; p=0.019), by flow cytometry, in patients who had developed Pn and after neoadjuvant treatment (Pn 6.57, IQR 4.66– (Pn 36.6, IQR 26.6–43.4; non-­Pn 25.7, IQR 17.4–35.7; 15.3; non-­Pn 2.74, IQR 1.33–4.79; p=0.005) (figure 3C). p=0.017) (figure 3A). The percentage of CD14 + popu- Among NK cells, we detected that PD1− percentages lation at diagnosis could serve as a good predictive were elevated in patients who develop Pn (Pn 76.8, IQR biomarker since its area under the ROC curve value 75.5–80.9; non-­Pn 73.6, IQR 66.3–76.7; p=0.032), but no was 0.791 (95% CI 0.613 to 0.970) (figure 3A). Never- differences in PD1 + cells were observed in these patients theless, deepening in the different monocyte subsets, (Pn 2.5, IQR 0.8–5.2; non-­Pn 1.4, IQR 0.4–3.1; p=0.407) we did not see significant differences between classical (figure 3C). Their AUC values for ROC curves were 0.786 (CD14 +CD16−; Pn 84.0, IQR 60.7–88.6; non-­Pn 78.9, (95% CI 0.581 to 0.990) for general NK cells and 0.762 IQR 57.7–88.1; p=0.591), intermediate (CD14 +CD16+; (95% CI 0.577 to 0.947) for PD1− subset (figure 3D). 6 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by copyright. Figure 3 Flow cytometry immunophenotyping of peripheral mononuclear cells (PMBCs). (n=29 in preneoadjuvant and postneoadjuvant treatment samples) (A) CD14 + cells (total monocytes, p=0.017 and p=0.329) and AUC curve to predict pneumonitis at baseline. (B) CD3-­CD19+ (total B cells, p=0.064 and p=0.002). (C) CD3-­CD56+ (total NK cells, p=0.019 and p=0.005), and positive and negative PD1 subpopulations in pretreatment samples (p=0.407 and p=0.032). (D) AUC curve of total NK and PD1—subpopulation to predict pneumonitis in pretreatment samples. *P<0.05; **p<0.01; ***p<0.001. AUC, area under the curve; NK, natural killer; n.s., not significant; PD1, programmed death-1. Related to T cells, we found a trend to lower percentage −4.9–16.63; p=0.045) (figure 4A). Looking deeper in T of CD3 + general T cells at diagnosis in patients with Pn cells subpopulations, we found no significant differences (Pn 54.6, IQR 51.0–66.3; non-­Pn 67.9, IQR 57.6–74.0; neither in CD3 +CD4+ helper T cells nor CD3 +CD8+ p=0.064), that reached statistical significance after pretreatment or posttreatment samples. However, in neoadjuvant treatment (Pn 50.3, IQR 38.21–65.10; pretreatment samples Pn patients had a trend to lower non-­Pn 70.2, IQR 63.45–76.03; p=0.001) (figure 4A). In percentage of CD3 +CD4+PD1+ T helper subpopulation addition, patients who developed Pn suffered a greater (Pn 1.2, IQR 0.4–4.9; non-­Pn 5.2, IQR 1.3–16.6; p=0.088) reduction in the percentage of CD3 + T cells with treat- that was completely absent for CD3 +CD8+PD1+ T cyto- ment (Pn −8.18, IQR −13.68–5.73; non-­ Pn 3.9, IQR toxic subpopulation (Pn 1.8, IQR 0.0–8.0; non-­Pn 2.9, Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 7 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by Figure 4 T cells immunophenotyping and TCR repertoire. (A) CD3+ (total T cells, n=29, p=0.064 in pretreatment and p=0.001 in postneoadjuvant treatment samples, respectively) and shift in CD3 + population calculated by POST-­PRE differences (p=0.045). (B) T cell receptor repertoire evenness at diagnosis and postneoadjuvant treatment in both tissue (n=22; p=0.865 for pretreatment and n=38; p=0.028 for postneoadjuvant treatment) and blood (n=30; p=0.815 for pretreatment and n=35; p=0.827 for posttreatment samples). *P<0.05; **p<0.01; ***p<0.001. n.s., not significant; TCR, T cell receptor. IQR 0.3–10.4; p=0.329) (online supplemental figure 251.0–653.0; p=0.017) and E-­cadherin (Pn 1321.5, IQR copyright. S3A). 843.9–2076.7; non-­Pn 683.2, IQR 555.5–947.3; p=0.022) In addition to immunophenotyping, we characterized at diagnosis. Also, a trend for higher CCL22 levels, also the T cell repertoire by massive TCR sequencing, in both, known as macrophage-­derived chemokine (MDC), was tumor (22 samples at diagnosis, 4 Pn; and 38 samples at found in patients who developed pneumonitis (Pn 2281.8, surgery, 12 Pn) and blood (30 pretreatment samples, 8 IQR 1342.3–3168.0; non-­ Pn 941.7, IQR 612.5–1970.2; Pn; and 35 postneoadjuvant samples, 10 Pn). We found p=0.055) (figure 5A). However, these differences were no differences in TCR diversity, convergence, and lost after neoadjuvant treatment, only a trend to a lower expanded or contracted clones, between patients who TDGF1 (Pn 339.4, IQR 81.1–1821.6; non-­Pn 1969.2; IQR developed Pn and those who did not, for any of the time- 691.8–6402.1; p=0.076) and a trend to higher MDC (Pn points or compartments (online supplemental figure S3B 1784.3, IQR 1243.9–3051.0; non-­Pn 1132.8, IQR 606.1– and online supplemental table 3). However, we found 1778.7; p=0.089) were observed (figure 5A). Referring to significant lower TCR evenness levels in postneoadjuvant postneoadjuvant treatment samples, Pn patients showed tissue samples (Pn 0.89: IQR 0.84–0.90; non-­Pn 0.90, IQR significant higher relative levels of proinflammatory cyto- 0.88–0.94; p=0.028) (figure 4B). kines such as angiopoietin-1 (ANG-1), epithelial growth Cytokine analysis factor (EGF), Chemokine (C-­C motif) ligand 16 (CCL16), In order to describe additional blood biomarkers, we neutrophil gelatinase-­ associated lipocalin (NGAL), performed further protein analysis in baseline (30 macrophage migration inhibitory factor (MIF), platelet-­ patients, 8 developed Pn) and postneoadjuvant (34 derived growth factor (PDGF-­AA), bone morphogenetic patients, 10 developed Pn) plasma samples. Among protein 4 (BMP-4), beta nerve growth factor (b-­NGF), protein measured (online supplemental table 4), signifi- hepatocyte growth factor (HGF), neurotrophin-4 (NT-4), cantly lower baseline levels of teratocarcinoma-­derived urokinase-­ type plasminogen activator receptor (uPAR) growth factor 1 (TDGF1) (Pn 582.5, IQR 272.0–1150.4; and vascular endothelial growth factor (VEGF) (online non-­Pn 2174.6, IQR 775.8–3289.8; p=0.013) were found supplemental figure S4A). Likewise, a differential impact in patients who develop Pn (figure 5A). On the contrary, of treatment on the plasma levels of latency-­associated patients who develop Pn showed significantly higher levels peptide (LAP), chemokine (C-­X-­C motif) ligand 1, 5, 6 of plasmatic macrophage stimulating protein (MSP) (CXCL1, CXCL5, CXCL6), Cathepsin S, CCL4 (MIP-­1b), (Pn 1564.1, IQR 1477.2–1599.9; non-­ Pn 1358.6, IQR EGF, and metalloproteinase inhibitor 1 (TIMP-1) was 1144.4–1521.2; p=0.006), Poly(A)-­ specific ribonuclease observed between patients with and without Pn (online (PARN) (Pn 774.5, IQR 482.0–1031.3; non-­Pn 480.6, IQR supplemental figure S5A). 8 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by copyright. Figure 5 Cytokine levels and pneumonitis development. (n=30 pretreatment; n=34 posttreatment). (A) Relative levels of TDGF1 (p=0.013 and p=0.076), MSP (p=0.006 and p=0.104), MDC (p=0.055 and p=0.089), E-­cadherin (p=0.022 and p=0.364) and PARN (Poly(A)-­specific ribonuclease, p=0.017 and p=0.496). (B) AUC curves to predict pneumonitis in pretreatment samples. *P<0.05; **p<0.01; ***p<0.001. AUC, area under the curve; MDC, macrophage-­derived chemokine; MSP, macrophage stimulating protein; n.s., not significant; PARN, poly(A)-­specific ribonuclease; TDGF1, teratocarcinoma-­derived growth factor 1. AUC values for ROC curves to predict Pn in baseline include dysregulated immune cells, cytokines levels, and samples were 0.801 (95% CI 0.645 to 0.958) for TDGF1; neoantigen cross-­reactivity, areas that could also repre- 0.838 (95% CI 0.695 to 0.982) for MSP; 0.790 (95% CI sent a source of predictive biomarkers. 0.604 to 0.976) for PARN; and 0.788 (95% CI 0.566 to Most hemogram parameters were not associated to Pn 0.991) for E-­Cadherin (figure 5B). development in our cohort, including the NLR ratio that has been associated with anti-­PD-­(L)1 irAEs in previous studies.26 27 Nevertheless, two of the platelet-­related ratios DISCUSSION appear to be relevant in this setting at diagnosis. Platelets PN development after anti-­PD-­(L)1 could lead to treat- are crucial in hemostasis and thrombosis. However, their ment discontinuation, and in the worst cases, to death.6 role as modulators of immune responses has emerged With the increasing use of anti-­PD(L)1 and their intro- in the last years. In fact, the lungs could be a source duction in the early stage first-­line scenario, the absolute burden and mortality of Pn will certainly rise in NSCLC of platelet generation, with resident megakaryocytes patients. Even more so when the rate of any-­grade or showing a potential role in lung immunity.28 Remarkably, high-­grade irAEs to immunotherapy is higher in first-­ the physical interaction of platelets with lymphocytes line compared with successive lines.25 Despite its clinical and monocytes induces these cells to have an anti-­ relevance, Pn development remains an unpredictable inflammatory profile, with higher levels of interleukin treatment adverse effect, whose mechanisms are mainly 10 (IL-10) and lower levels of TNF, which could explain unknown. This is even more true for chemoimmuno- the observed higher incidence of Pn in our patients with therapy, which is associated to higher Pn incidence; low PMR ratio.29 30 In addition, it has been observed how and due to its novelty, it completely lacks biomarkers lung injury caused by chemotherapy may involve inflam- or studies describing the possible mechanisms involved. mation, T cells, monocytes, and the balance of cytokines Probable mechanisms of chemoimmunotherapy Pn such as TNF or MCP1.17 18 Besides, low levels of PLR ratio Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 9 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by prior to anti PD-­(L)1 treatment have been associated to elevated in posttreatment plasma of Pn patients, and may irAEs development in stage III–IV NSCLC patients.26 27 indicate inflammation and lung tissue repair in response Concerning PMR, to our knowledge, there are no data to injured epithelial cells.44 45 regarding its association with irAES to anti PD-­(L)1, but On lymphocyte populations, we observed a trend for a lower PMR has been described in patients with lupus lower baseline levels of B (CD3−CD19+) and T (CD3+) nephritis compared with healthy controls.31 cells, but statistically higher levels of NKs (CD3−CD56+) Using hemogram data, baseline monocyte levels cells, in patients developing Pn. Interestingly, these differ- showed a tendency to be increased in patients who ences were increased after neoadjuvant treatment, which developed Pn. Importantly, this difference in monocyte may indicate their involvement in mechanisms main- levels reached statistical significance when monocytes tained during chemoimmunotherapy. Although B cells populations were analyzed by flow cytometry using CD14 in tumors and tertiary lymphoid structures are key to the antibody positivity. Monocytes are important regulators antitumor response,46 it has been described how B cells of the immune response, and their levels are increased at the peripheral level may have an immunosuppressive in several autoimmune diseases.32 However, no associa- function through the production of cytokines,47 48 playing tion of basal peripheral monocytes levels to anti-­PD(L)1 an important role in autoimmune disorders.49 Thus, low irAEs development has been reported, although its levels of baseline peripheral B cells have been associated involvement in the inflammatory component of bron- with better responses to anti-­PD-1 therapy.48 50 choalveolar lavage has been described.33 In addition, we T cells have been described as pivotal regulators have observed elevated levels of baseline MSP (macro- of irAEs.51 We observed a trend for lower CD3 + levels phage stimulating protein) associated with the develop- in Pn patients, with a similar behavior for CD4 + and ment of chemoimmunotherapy-­related Pn. One possible CD8+ subpopulations. However, we observed a trend for explanation is that MSP seems to inhibit inducible lower PD1 + cells percentage in CD4 + subpopulation in nitric oxide (NO) synthase, blocking the generation of patients with Pn. PD1-­deficient mice develop lymphocyte-­ NO in macrophages.34 NO is known to induce macro- dependent myocarditis,52 and showed an increased phage apoptosis, downregulate superoxide activity, inflammatory and necrotic response to Mycobacterium and generate an immunosuppressive microenviron- tuberculosis, indicating an essential role for this coinhib- ment recruiting MDSCs and Tregs.35 Similarly, a trend itory receptor in controlling inflammatory responses.53 for higher levels of MDC was observed in patients who Thus, lower levels of PD1 + cells may be associated with a copyright. developed Pn, and MDC has been associated with lung higher frequency of Pn. inflammation in different models, including smoking Apart from that, there are data supporting that the injure, idiopathic pulmonary fibrosis and acute asthma.36 presence of common antigens between tumor cells and Conversely, low baseline levels of the immunosuppres- lung tissues could cause T cell dependent damage to the sive TDGF1 (teratocarcinoma-­derived growth factor 1) lungs.54 55 An indirect measure to assess this process can were found in patients who developed Pn. Exposure to be achieved through the study of the T cell repertoire TDGF1 enriched supernatants to macrophages in vitro, or the TMB. In fact, both parameters have been associ- increased their immunosuppressive potential through ated with the development of irAEs.51 56 57 We do not see increased expression of IL-10 and IL1-­b.37 Therefore, any association between these parameters at baseline and a situation with high monocytes, MSP and MDC levels the development of Pn. However, in surgical specimens (in addition to low TDGF1 levels), may lead to a basal of patients who developed Pn, we have described a lower proinflammatory profile of lungs macrophages that TCR repertoire evenness, that is, a higher clonality of T increases the susceptibility to Pn development. Besides, cells, which could reflect a higher activation and prolif- the increased levels of the nuclear protein PARN and the eration of antitumor clones after neoadjuvant treatment, membrane protein E-­cadherin, (proteins that are not that could present cross-­reactivity with lung tissue.54 55 actively secreted into the bloodstream), in the plasma of Anyhow, a study focused on specific clones with cross-­ patients who developed Pn, may be indirect indicators specificities to neoantigens would be necessary, which is of increased inflammatory damage at diagnosis in these beyond the scope of this work. Additionally, we have seen patients.38 In this sense, patients who developed Pn had a possible association between ARID1A and the develop- elevated neutrophil levels after neoadjuvant treatment, ment of Pn, not described to date. ARID1A acts as a tumor which may reflect a pro-­inflammatory state at systemic suppressor gene controlling DNA damage repair. It has level, or even a compensatory response to repair damaged been shown that mutations in ARID1A are associated with lung tissue.39 40 Indeed, NGAL, and soluble uPAR, both better responses to immunotherapy, probably due to a related to neutrophilic inflammation,41 42 were higher greater presence of neoantigens in these tumors,58 which after treatment in patients with Pn. Similarly, Pn patients could imply a greater predisposition to Pn. showed a differential increase in CXCL1, CXCL5, Beyond B and T cells, NK cells have a role stimulating CXCL6 levels, also related to neutrophil response.43 or suppressing autoimmunity.59 The NKs profile of the Additionally, beyond neutrophils, numerous proinflam- lungs is very similar to that found in peripheral blood,60 matory cytokines and growth factors, related to angio- and although there are no data on NKs association with genesis, epithelial proliferation and wound healing, were irAEs, elevated levels of NKs have been described in the 10 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by 19 lungs of patients with hypersensitivity Pn or organizing Flow Cytometry Core Facility, Instituto de Investigación Sanitaria Puerta de Hierro-­ pneumonitis not related to anti-­PD(L)-1 treatment.61 Segovia de Arana (IDIPHISA), Majadahonda, Spain To conclude, this is the first study describing parameters Twitter Alberto Cruz-­Bermúdez @CruzAPhD, Margarita Majem @margamajem, associated with the development of Pn after neoadjuvant Alex Martinez-­Marti @AlxMtzMarti, Atocha Romero @LiquidBiopsyLab and Mariano chemoimmunotherapy treatment. Remarkably, without Provencio @MARIANOPROVENCI ruling out the involvement of adaptive immunity in pneu- Acknowledgements We thank the patients, their families, all the participating monitis predisposition, our results indicate a relevant clinical teams, and all the Spanish Lung Cancer Group, BMS and ThermoFisher role for innate immunity parameters, with several AUC staff, for making this study possible. We also would like to thanks to Maria del Rocio curve values higher than 0.75. Specifically, patients who Moreno Villa for their technical assistance. developed Pn had decreased baseline levels of PMR ratio Contributors AC-­B and MP conceived and design the study. EN, AI, JM, JC-­R, and TDGF1, but elevated baseline levels of NKs, mono- MD, MM, DR-­A, AM-­M, JDCC, MCo, GLV, EDB, RC, NV, IBA, SV, BM, FF and MP recruited and treated patients. BS-­R, AC-­B, YG, RB, MCa and AG-­G carried out the cytes, and MSP, E-­cadherin and PARN plasma levels. In experiments and analyzed the data. BS-­R, AC-­B, YG, AR and MP, interpreted the addition, we have described postneoadjuvant treatment data. All the authors read and contributed to the final version of the manuscript and differences that may help to elucidate the inflammatory approve its submission for publication. BS-­R and AC-­B contributed equally. AC-­B scenario of chemoimmunotherapy related pneumonitis, and MP are joint senior authors. involving: high levels of neutrophils and NK cells, but low Funding Work in the authors’ laboratories was supported by ‘‘Instituto de Salud Carlos III’’ (ISCIII) PI19/01652 grant cofunded by European Regional Development levels of B and T cells in peripheral blood; increased clon- Fund (ERDF), Bristol-­Myers Squibb (BMS), Ministry of Science and Innovation ality of intratumoral T cell clones; and elevated plasma RTC2017-6502-1 'INmunoSIGHT', RTC2019-007359-1 'BLI-­O' and European levels of several growth factors (EGF, HGF, VEGF, ANG-1, Union’s Horizon 2020 research and innovation programme, CLARIFY 875160 grant, PDGF, NGF, and NT4) and inflammatory cytokines (MIF, to M.P. ThermoFisher provided reagents for TCR sequencing. A.C-­B, received a CCL16, NGAL, BMP-4, and u-­PAR). Spanish Lung Cancer Group (SLCG) grant and is supported by a ISCIII-'Sara Borrell' contract CD19/00170. MC is supported by PEJD-2019-­PRE/BMD-17006 contract This study has several limitations, such as the post hoc granted to AC-­B. RL-­B was supported by PEJ16/MED/AI-1972 and PEJD-2018-­PRE/ nature of the analyses, the reduced number of patients, SAL-8641 from European Social Fund (ESF) and Comunidad de Madrid, both the elevated number of variables, and the lack of control granted to MP. arms to evaluate the impact of adjuvant treatment or Competing interests EN reports personal fees from Bristol Myers Squibb, personal differentiate the effect of chemotherapy or immuno- fees from Merck Sharpe & Dohme, personal fees from AstraZeneca, grants and therapy alone. All this makes it necessary to confirm the personal fees from Roche, grants and personal fees from Pfizer, personal fees from Lilly, personal fees from Amgen, personal fees from Boehringer Ingelheim, proposed biomarkers in independent cohorts of patients copyright. outside the submitted work; AI reports personal fees from Bristol, personal fees with appropriate statistical power. from BOERINGHER, personal fees from MSD, personal fees from PFIZER, personal However, our results open the possibility to predict fees from Roche, personal fees from ASTRA ZENECA, outside the submitted work; MD reports personal fees from Astra-­Zeneca, personal fees from BMS, personal patients at high risk of Pn, allowing the personaliza- fees from Boehringer Ingelheim, personal fees from MSD, personal fees from tion of the surveillance strategy. Moreover, the precise Pfizer, personal fees from Roche, outside the submitted work; MM reports grants knowledge of the mechanisms involved may allow new and personal fees from BMS, personal fees and non-­financial support from MSD, personalized therapies to control the harmful effects of personal fees and non-­financial support from BOEHRINGER INGELHEIM, personal fees, non-­financial support and other from ASTRA ZENENCA, personal fees, chemoimmunotherapy. non-­financial support and other from ROCHE, personal fees from KYOWA KYRIN, personal fees from PIERRE FABRE, outside the submitted work; DRA reports grants Author affiliations and personal fees from Bristol-­Myers-­Squibb, personal fees from GENENTECH/ 1 Oncología Médica, Instituto de Investigación Sanitaria Puerta de Hierro-­Segovia ROCHE, personal fees from MSD, personal fees from ASTRA ZENECA, personal fees de Arana (IDIPHISA), Hospital Universitario Puerta de Hierro, Majadahonda, Madrid, from BOEHRINGER INGELHEIM, personal fees from Novartis, personal fees from Spain Lilly, outside the submitted work; AM-­M reports personal fees and non-­financial 2 Department of Medical Oncology, Catalan Institute of Oncology, Oncobell Program, support from Bristol-­Myers Squibb, personal fees and non-­financial support from IDIBELL, L’Hospitalet de Llobregat, L'Hospitalet, Barcelona, Spain F. Hoffmann La Roche AG, personal fees and non-­financial support from Merck 3 Medical Oncology, INCLIVA, Valencia, Valencia, Spain Sharp & Dohme, personal fees and non-­financial support from Pfizer, personal 4 Medical Oncology, Hospital Universitario A Coruña, A Coruña, Spain fees and non-­financial support from Boehringer Ingelheim, personal fees and 5 Medical Oncology, Hospital Universitario de Vigo, Pontevedra, Spain non-­financial support from MSD Oncology, personal fees, non-­financial support and 6 Medical Oncology, Hospital Universitario Fundación Jiménez Díaz, Madrid, Spain other from AstraZeneca, outside the submitted work; JDCC reports personal fees 7 Medical Oncology, Hospital de la Santa Creu i Sant Pau, Barcelona, Catalunya, from Astra Zeneca, personal fees from Boehringer Ingelheim, personal fees from Spain Merck Sharp and Dohme, personal fees from Hoffmann-­la Roche, personal fees 8 Medical Oncology, Hospital Universitario Insular de Gran Canaria, Las Palmas, from Bristol-­Myers Squibb, personal fees from Takeda, personal fees from Pfizer, Canarias, Spain personal fees from Novartis, outside the submitted work; EDB reports non-­financial 9 VHIO, Barcelona, Spain support from ROCHE, BMS, PFIZER, ASTRA-­ZENECA, MERCK, during the conduct 10 of the study; IBA reports consulting or advisory board for Bristol Myers, Takeda, Medical Oncology, Hospital Universitario La Paz, Madrid, Spain 11 Roche, Astra Zeneca, Behringer Inngelheim; SV reports personal fees and non-­ Medical Oncology, Hospital Regional Universitario de Málaga, Malaga, Andalucía, financial support from BMS, personal fees and non-­financial support from ROCHE, Spain 12 personal fees from MSD, personal fees from ABBVIE, non-­financial support from Medical Oncology, Hospital Universitario Cruces, Barakaldo, Spain OSE IMMUNOTHERAPEUTICS, non-­financial support from MERCK SERONO, outside 13 Medical Oncology, Hospital Universitario de Salamanca, Salamanca, Spain the submitted work; BM reports grants and personal fees from Roche, personal 14 Medical Oncology, Hospital U. Virgen Rocio, Seville, Spain fees and other from BMS, personal fees from Takeda, other from MSD, personal 15 Medical Oncology, Hospital Clínic de Barcelona, Barcelona, Catalunya, Spain fees from Boehringer, other from Takeda, outside the submitted work; AT.R. reports 16 Medical Oncology, Hospital Universitario Reina Sofia, Cordoba, Spain personal fees from Boehringer Ingelheim, outside the submitted work;MP reports 17 Instituto Oncológico Dr. Rosell. Hospital Universitario Quiron Dexeus, Barcelona, grants, personal fees and non-­financial support from BMS, grants, personal fees Spain and non-­financial support from ROCHE, grants, personal fees and non-­financial 18 Medical Oncology, Alicante General University Hospital, Alicante, Valencia, Spain support from ASTRAZENECA, personal fees from MSD, personal fees from TAKEDA, Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 11 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by outside the submitted work; BS-­R, AC-­B, YG, JM, JC-­R, MCo, GLV, RB, NV, RL-­B, MC, 12 Johnson DB, Taylor KB, Cohen JV, et al. Anti-­PD-1-­induced AG-­G and FF declare no conflicts of interest. pneumonitis is associated with persistent imaging abnormalities in melanoma patients. Cancer Immunol Res 2019;7:1755–9. Patient consent for publication Not required. 13 Provencio M, Nadal E, Insa A, et al. Neoadjuvant chemotherapy Ethics approval Informed consent for the collection of research samples and and nivolumab in resectable non-­small-­cell lung cancer (NADIM): an open-­label, multicentre, single-­arm, phase 2 trial. Lancet Oncol study protocol were approved by the clinical research ethics committee of Hospital 2020;21:1413–22 https://​linkinghub.​elsevier.​com/​retrieve/​pii/​ Puerta de Hierro and the Spanish Lung Cancer Group (SLCG) Board in accordance S1470204520304538 with the International Conference on Harmonization Guidelines on Good Clinical 14 Shu CA, Gainor JF, Awad MM, et al. Neoadjuvant atezolizumab and Practice and the Declaration of Helsinki. chemotherapy in patients with resectable non-­small-­cell lung cancer: an open-­label, multicentre, single-­arm, phase 2 trial. Lancet Oncol Provenance and peer review Not commissioned; externally peer reviewed. 2020;21:786–95. Data availability statement All data relevant to the study are included in the 15 Zinner R, Axelrod R, Solomides CC, et al. Neoadjuvant nivolumab article or uploaded as online supplemental information. (N) plus cisplatin (C)/pemetrexed (P) or cisplatin /gemcitabine (G) in resectable NSCLC. JCO 2020;38:9051. Supplemental material This content has been supplied by the author(s). It has 16 Rothschild S, Zippelius A, Eboulet EI, et al. SAKK 16/14: Anti-­PD-­L1 not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been antibody durvalumab in addition to neoadjuvant chemotherapy in peer-­reviewed. Any opinions or recommendations discussed are solely those patients with stage IIIA(N2) non-­small cell lung cancer (NSCLC)—A of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and multicenter single-­arm phase II trial. JCO 2020;38:9016. responsibility arising from any reliance placed on the content. Where the content 17 Fujimori K, Yokoyama A, Kurita Y, et al. Paclitaxel-­Induced cell-­ includes any translated material, BMJ does not warrant the accuracy and reliability mediated hypersensitivity pneumonitis. diagnosis using leukocyte migration test, bronchoalveolar lavage and transbronchial lung of the translations (including but not limited to local regulations, clinical guidelines, biopsy. Oncology 1998;55:340–4. terminology, drug names and drug dosages), and is not responsible for any error 18 Li L, Mok H, Jhaveri P, et al. Anticancer therapy and lung and/or omissions arising from translation and adaptation or otherwise. injury: molecular mechanisms. Expert Rev Anticancer Ther Open access This is an open access article distributed in accordance with the 2018;18:1041–57. 19 Reuss JE, Anagnostou V, Cottrell TR, et al. Neoadjuvant nivolumab Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which plus ipilimumab in resectable non-­small cell lung cancer. J permits others to distribute, remix, adapt, build upon this work non-­commercially, Immunother Cancer 2020;8. and license their derivative works on different terms, provided the original work is 20 Cui P, Liu Z, Wang G, et al. Risk factors for pneumonitis in patients properly cited, appropriate credit is given, any changes made indicated, and the use treated with anti-­programmed death-1 therapy: a case-­control study. is non-­commercial. See http://c​ reativecommons.​org/​licenses/​by-​nc/​4.0​ /. Cancer Med 2018;7:4115–20 https://​onlinelibrary.​wiley.​com/​doi/ 21 Sears CR, Peikert T, Possick JD, et al. Knowledge gaps and research ORCID iDs priorities in immune checkpoint Inhibitor-­related pneumonitis. An Belén Sierra-­Rodero http://​orcid.​org/​0000-​0003-​0276-​6088 official American thoracic Society research statement. Am J Respir Alberto Cruz-­Bermúdez http://​orcid.​org/​0000-​0001-​5136-​7011 Crit Care Med 2019;200:e31–43 https://www.​atsjournals.​org/​doi/ Ernest Nadal http://o​ rcid.​org/​0000-​0002-​9674-​5554 22 Mezquita L, Auclin E, Ferrara R, et al. Association of the lung immune prognostic index with immune checkpoint inhibitor outcomes in Margarita Majem http://o​ rcid.​org/​0000-​0002-​9919-​7485 patients with advanced non-­small cell lung cancer. JAMA Oncol copyright. Alex Martinez-­Marti http://​orcid.​org/​0000-​0002-​6714-​9526 2018;4:351–7. Mariano Provencio http://​orcid.​org/​0000-​0001-​6315-​7919 23 Laza-­Briviesca R, Cruz-­Bermúdez A, Nadal E, et al. Blood biomarkers associated to complete pathological response on NSCLC patients treated with neoadjuvant chemoimmunotherapy included in REFERENCES NADIM clinical trial. Clin Transl Med 2021;11:e491. 1 Finn OJ. Immuno-­oncology: understanding the function and 24 Casarrubios M, Cruz-­Bermúdez A, Nadal E, et al. Pre-­Treatment dysfunction of the immune system in cancer. Ann Oncol 2012:86–9. tissue TCR repertoire evenness is associated with complete 2 Sharma P, Allison JP. Immune checkpoint targeting in cancer pathological response in patients with NSCLC receiving neoadjuvant therapy: toward combination strategies with curative potential. Cell chemoimmunotherapy. Clin Cancer Res 2021:clincanres.1200.2021 2015;161:205–14. https://​clincancerres.​aacrjournals.​org/​content/​early/​2021/​08/​09/​ 3 Postow MA, Sidlow R, Hellmann MD. Immune-­related adverse 1078-​0432.​CCR-​21-​1200 events associated with immune checkpoint blockade. N Engl J Med 25 Yang Y, Pang P, Xie Z, et al. The safety of first and subsequent lines 2018;378:158–68. of PD-1/PD-­L1 inhibitors monotherapy in non-­small cell lung cancer 4 Forde P, Spicer J, Lu S. Abstract CT003 - Nivolumab (NIVO) + patients: a meta-­analysis. Transl Cancer Res 2020;9:3231–41. platinum-­doublet chemotherapy (chemo) vs chemo as neoadjuvant 26 Pavan A, Calvetti L, Dal Maso A, et al. Peripheral blood markers treatment (tx) for resectable (IB-­IIIA) non-­small cell lung cancer identify risk of immune-­related toxicity in advanced non-­small cell (NSCLC) in the phase 3 CheckMate 816 trial, 2021. AACR Annu lung cancer treated with Immune-­Checkpoint inhibitors. Oncologist Meet 2021. Available: https://www.​abstractsonline.​com/​pp8/#!/​9325/​ 2019;24:1128–36. presentation/​5134 27 Liu W, Liu Y, Ma F, et al. Peripheral blood markers associated with 5 Martins F, Sofiya L, Sykiotis GP, et al. Adverse effects of immune-­ immune-­related adverse effects in patients who had advanced non-­ checkpoint inhibitors: epidemiology, management and surveillance. small cell lung cancer treated with PD-1 inhibitors. Cancer Manag Nat Rev Clin Oncol 2019;16:563–80. Res 2021;13:765–71. 6 Wang DY, Salem J-­E, Cohen JV, et al. Fatal toxic effects associated 28 Lefrançais E, Ortiz-­Muñoz G, Caudrillier A, et al. The lung is a site of with immune checkpoint inhibitors: a systematic review and meta-­ platelet biogenesis and a reservoir for haematopoietic progenitors. analysis. JAMA Oncol 2018;4:1721–8. Nature 2017;544:105–9. 7 Shannon VR. Pneumonitis associated with immune checkpoint 29 Zamora C, Cantó E, Nieto JC, et al. Binding of platelets to inhibitors among patients with non-­small cell lung cancer. Curr Opin lymphocytes: a potential anti-­inflammatory therapy in rheumatoid Pulm Med 2020;26:326–40. arthritis. J Immunol 2017;198:3099–108. 8 Paz-­Ares L, Luft A, Vicente D, et al. Pembrolizumab plus 30 Linke B, Schreiber Y, Picard-­Willems B, et al. Activated platelets chemotherapy for squamous non-­small-­cell lung cancer. N Engl J induce an anti-­inflammatory response of monocytes/macrophages Med 2018;379:2040–51 http://www.​nejm.​org/​doi/ through cross-­regulation of PGE2 and cytokines. Mediators Inflamm 9 Gandhi L, Rodríguez-­Abreu D, Gadgeel S, et al. Pembrolizumab plus 2017;2017:1–14. chemotherapy in metastatic non-­small-­cell lung cancer. N Engl J 31 Liu P, Li P, Peng Z, et al. Predictive value of the neutrophil-­to-­ Med 2018;378:2078–92 http://www.​nejm.​org/​doi/ lymphocyte ratio, monocyte-­to-­lymphocyte ratio, platelet-­to-­ 10 West H, McCleod M, Hussein M, et al. Atezolizumab in combination neutrophil ratio, and neutrophil-­to-­monocyte ratio in lupus nephritis. with carboplatin plus nab-­paclitaxel chemotherapy compared with Lupus 2020;29:1031–9. chemotherapy alone as first-­line treatment for metastatic non-­ 32 Ma W-­T, Gao F, Gu K, Chen D-­K, et al. The role of monocytes and squamous non-­small-­cell lung cancer (IMpower130): a multicentre, macrophages in autoimmune diseases: a comprehensive review. randomised, open-­label, phase 3 trial. Lancet Oncol 2019;20:924–37. Front Immunol 2019;10:1140. 11 Socinski MA, Jotte RM, Cappuzzo F, et al. Atezolizumab for first-­ 33 Suresh K, Naidoo J, Zhong Q, et al. The alveolar immune cell line treatment of metastatic Nonsquamous NSCLC. N Engl J Med landscape is dysregulated in checkpoint inhibitor pneumonitis. J Clin 2018;378:2288–301. Invest 2019;129:4305–15. 12 Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 Open access J Immunother Cancer: first published as 10.1136/jitc-2021-002804 on 26 August 2021. Downloaded from http://jitc.bmj.com/ on April 11, 2023 at Sant Paul Hospital Consortia. Protected by 34 Chen YQ, Fisher JH, Wang MH. Activation of the RON receptor 48 de Jonge K, Tillé L, Lourenco J, et al. Inflammatory B cells tyrosine kinase inhibits inducible nitric oxide synthase (iNOS) correlate with failure to checkpoint blockade in melanoma patients. expression by murine peritoneal exudate macrophages: Oncoimmunology 2021;10:1873585. phosphatidylinositol-3 kinase is required for RON-­mediated inhibition 49 Lampropoulou V, Calderon-­Gomez E, Roch T, et al. Suppressive of iNOS expression. J Immunol 1998;161:4950–9. functions of activated B cells in autoimmune diseases reveal 35 Ekmekcioglu S, Grimm EA, Roszik J. Targeting iNOS to the dual roles of Toll-­like receptors in immunity. Immunol Rev increase efficacy of immunotherapies. Hum Vaccin Immunother 2010;233:146–61. 2017;13:1105–8. 50 Yuan S, Liu Y, Till B, et al. Pretreatment peripheral B cells are 36 Ritter M, Göggel R, Chaudhary N, et al. Elevated expression of associated with tumor response to Anti-­PD-1-­Based immunotherapy. TARC (CCL17) and MDC (CCL22) in models of cigarette smoke-­ Front Immunol 2020;11:563653. induced pulmonary inflammation. Biochem Biophys Res Commun 51 Jing Y, Liu J, Ye Y, et al. Multi-­omics prediction of immune-­related 2005;334:254–62. adverse events during checkpoint immunotherapy. Nat Commun 37 Zhang D-­mei, Bao Y-­L, Yu C-­L, et al. Cripto-1 modulates 2020;11. macrophage cytokine secretion and phagocytic activity via NF-κB 52 Nishimura H, Okazaki T, Tanaka Y, et al. Autoimmune dilated signaling. Immunol Res 2016;64:104–14. cardiomyopathy in PD-1 receptor-­deficient mice. Science 38 McGuire JK, Li Q, Parks WC. Matrilysin (matrix metalloproteinase-7) 2001;291:319–22. mediates E-­cadherin ectodomain shedding in injured lung epithelium. 53 Lázár-­Molnár E, Chen B, Sweeney KA, et al. Programmed death-1 Am J Pathol 2003;162:1831–43. (PD-1)-­deficient mice are extraordinarily sensitive to tuberculosis. 39 Blázquez-­Prieto J, López-­Alonso I, Huidobro C, et al. The emerging Proc Natl Acad Sci U S A 2010;107:13402–7. role of neutrophils in repair after acute lung injury. Am J Respir Cell 54 Berner F, Bomze D, Diem S, et al. Association of checkpoint inhibitor-­ Mol Biol 2018;59:289–94. induced toxic effects with shared cancer and tissue antigens in non-­ 40 Potey PM, Rossi AG, Lucas CD, et al. Neutrophils in the initiation and small cell lung cancer. JAMA Oncol 2019;5:1043–7 http://​oncology.​ resolution of acute pulmonary inflammation: understanding biological jamanetwork.​com/​article.​aspx?​doi= function and therapeutic potential. J Pathol 2019;247:672–85. 55 Läubli H, Koelzer VH, Matter MS, et al. The T cell repertoire in tumors 41 Ikezoe K, Handa T, Mori K, et al. Neutrophil gelatinase-­ overlaps with pulmonary inflammatory lesions in patients treated with associated lipocalin in idiopathic pulmonary fibrosis. Eur Respir J checkpoint inhibitors. Oncoimmunology 2018;7:e1386362–6. 2014;43:1807–9 http://​erj.​ersjournals.​com/​cgi/​doi/ 56 Johnson DB, Balko JM, Compton ML, et al. Fulminant myocarditis 42 Pliyev BK. Activated human neutrophils rapidly release the with combination immune checkpoint blockade. N Engl J Med chemotactically active D2D3 form of the urokinase-­type 2016;375:1749–55. plasminogen activator receptor (uPAR/CD87). Mol Cell Biochem 57 Bomze D, Hasan Ali O, Bate A, et al. Association between immune-­ 2009;321:111–22. related adverse events during anti-­PD-1 therapy and tumor 43 Capucetti A, Albano F, Bonecchi R. Multiple roles for chemokines in mutational burden. JAMA Oncol 2019;5:1633–5. neutrophil biology. Front Immunol 2020;11:1–9. 58 Li Z, Lin J, Zhang L, et al. Comprehensive analysis of multiple 44 Croasdell Lucchini A, Gachanja NN, Rossi AG, et al. Epithelial parameters associated with tumor immune microenvironment in cells and inflammation in pulmonary wound repair. Cells 2021;10. ARID1A mutant cancers. Future Oncol 2020;16:2295–306. doi:10.3390/cells10020339. [Epub ahead of print: 05 Feb 2021]. 59 Yang Y, Day J, Souza-­Fonseca Guimaraes F, et al. Natural killer 45 Barrientos S, Stojadinovic O, Golinko MS, et al. Growth factors and cells in inflammatory autoimmune diseases. Clin Transl Immunology cytokines in wound healing. Wound Repair Regen 2008;16:585–601. 2021;10:e1250. 46 Helmink BA, Reddy SM, Gao J, et al. B cells and tertiary 60 Hervier B, Russick J, Cremer I, et al. NK cells in the human lungs. copyright. lymphoid structures promote immunotherapy response. Nature . Front Immunol 2019;10:1263. 2020;577:549–55 http://www.​ncbi.​nlm.​nih.​gov/​pubmed/​31942075 61 Sokhatska O, Padrão E, Sousa-­Pinto B, et al. NK and NKT cells in 47 Fillatreau S. Regulatory functions of B cells and regulatory plasma the diagnosis of diffuse lung diseases presenting with a lymphocytic cells. Biomed J 2019;42:233–42. alveolitis. BMC Pulm Med 2019;19:39. Sierra-­Rodero B, et al. J Immunother Cancer 2021;9:e002804. doi:10.1136/jitc-2021-002804 13