Edited by: Tao Li, Sichuan Cancer Hospital, China
Reviewed by: Chunyan Hua, Wenzhou Medical University, China; Luigi Cavanna, Ospedaliera di Piacenza, Italy
*Correspondence: Jiancheng Li,
†These authors have contributed equally to this work
This article was submitted to Radiation Oncology, a section of the journal Frontiers in Oncology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
To systematically evaluate the safety and adverse event profiles of immune checkpoint inhibitors (ICIs) in patients with esophageal cancer (EPC) or gastroesophageal junction cancer (GEJC).
PubMed, Web of Science, Cochrane Library, and major conference proceedings were systematically searched for all phase II or phase III randomized controlled trials (RCTs) in EPC or GEJC using ICIs. Safety outcomes including treatment-related adverse events (trAEs), immune-related adverse events (irAEs), and serious trAEs were evaluated by network meta-analysis or dichotomous meta-analysis based on the random-effects model.
Eleven RCTs involving EPC (five RCTs) and GEJC (six RCTs) were included in the final meta-analysis. NMA showed that placebo was associated with the best safety ranking for grade 3–5 trAEs (SUCRA = 96.0%), followed by avelumab (78.6%), nivolumab (73.9%), ipilimumab (57.0%), and pembrolizumab (56.6%). Conventional pairwise meta-analysis (CPM) showed that ICIs have similar grade 3–5 trAE risk compared with chemotherapy (RR = 0.764, 95% CI: 0.574 to 1.016,
Different ICIs had different toxicity manifestations and should not be considered as an entity. Compared with chemotherapy, ICIs were more prone to irAEs, but the overall rates remained low and acceptable. For clinicians, it is important to recognize and monitor the adverse events caused by ICIs for patients with EPC or GEJC.
Worldwide, esophageal cancer (EPC) still remains one of the most commonly diagnosed cancers and the leading cause of cancer-related death (
In recent years, cancer immunotherapies based on immune checkpoint inhibitors (ICIs) have become the fifth largest tumor treatment after surgery, chemotherapy, radiotherapy, and small molecules targeted therapy in oncology and have revolutionized the treatment landscape and made major breakthroughs in the treatment of tumors, especially for advanced or metastatic cancer (
The current study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) (
Relevant clinical trials published in various databases such as PubMed, Web of Science, and Cochrane Library were searched. Major conference proceedings including the Clinicaltrial.gov, American Society of Clinical Oncology (ASCO), and European Society for Medical Oncology (ESMO) databases were also searched for recent conference abstracts.
Relevant search terms relating to the present study were composed of various combinations of the medical subject headings (MeSH) and free-text terms. Search terms were combined by the Boolean operator “AND” or “OR” if necessary. A PubMed search was conducted using the following search terms: 1) search terms related to disease were “esophageal neoplasm,” “esophagus cancer,” “esophageal cancer,” “gastro-oesophageal junction cancer,” “gastro-esophageal junction cancer,” “cancer of the gastroesophageal junction,” “adenocarcinoma of the esophagus and gastroesophageal junction,” etc. (
The selection criteria for clinical trials were organized according to the guidelines of the participants, interventions, comparisons, outcomes, and study design (PICOS) recommended by the Cochrane Collaboration. The inclusion criteria were as follows: i) the included patients were all pathologically diagnosed esophageal cancer or gastroesophageal junction cancer (GEJC) patients (P); ii) interventions of concern referred to immunotherapy with ICIs alone or in combination with chemotherapy (I); iii) controlled treatment regimens included chemotherapy alone (ChT) or best supportive treatment (BST), but there were no restrictions related to the chemotherapy regimens and chemotherapy cycles (C); iv) five safety outcomes included rates of treatment-related adverse events (trAEs), immune-related adverse events (irAEs), death, discontinuation of therapy, and grades 3–5 organ-special adverse events (O); and v) all randomized, open-label, controlled clinical trials with efficacy and safety data of ICIs were included. Although priority was given to phase III clinical trials, phase II clinical trials with a control group would be also included.
The exclusion criteria were as follows: i) phase I clinical trials and non-RCT studies, ii) participants with other tumors, iii) case reports and reviews, iv) incomplete data or non-original research, and v) repeated publications.
Articles were only included if they were published in English, but there was no restriction related to publication year. Two researchers (JZ and BH) were assigned to independently review all the data. If there were repeated articles in the selected clinical trials, only the latest published articles will be used for the final analysis. After the discussion according to the inclusion criteria and reaching a consensus, a decision was made to finally include or exclude the eligible articles. If a consensus cannot be reached, the corresponding author (JL) of this article is responsible for the final ruling.
After reading the full text, two researchers (JZ and Tingting Li) extracted and cross-checked the data, including the following: 1) basic information: such as the title of the trial, author’s name, year of publication, source of literature, etc.; 2) methodological information of the trial: the sample size of the study included, the basic information of the study population, including the entry time and number of participants, disease stages, etc.; the randomization method of the trial, the evaluation method of important outcome indicators; median follow-up duration, death, and withdrawal, etc.; 3) detailed information on intervention measures: ICI medication, medication in the control group, etc.; and 4) detailed information for safety outcome indicators mentioned above. Disagreements were resolved by consensus.
Two independent researchers evaluated the included RCTs according to the bias risk assessment method recommended by the Cochrane Assistance Network. The evaluation methodological criteria and items were as follows: 1) generation of random allocation sequence, 2) the method of allocation concealment, 3) the method of blinding the patient, 4) the method of blinding the doctor or the therapist, 5) the method of blinding the data collection and analysis personnel, 6) the incomplete data reported, 7) selective reporting bias, and 8) other potential bias affecting authenticity.
We evaluate the risk of bias for each RCT according to the following criteria: “Yes” indicates a low risk of bias, “No” indicates a high risk of bias, and “Unclear” indicates that the literature does not provide sufficient information for bias assessment. The two researchers discussed according to the above standards and methods and, if necessary, reached a consensus according to the opinions of the third researcher.
Adverse events including trAEs and irAEs were evaluated from two different perspectives: overview and detail. An overview analysis involved all kinds of AEs observed in ≥ grade 3–5 or all grade of the study population, and a detailed analysis involved some prespecified AEs of interest observed in ≥ grade 3–5 or all grades of the study population. The detailed information of related safety was extracted from the original literature and recorded as the number of events reported and no events for each specific treatment, respectively. If enough data were available to achieve network meta-analysis, a random-effects NMA was conducted in the frequency framework, using the command of “network” in Stata 16.0. Direct or indirect safety effects were combined into some summary statistics, that is, risk ratios (RRs) and 95% credibility intervals, to quantify the effect of adverse events in the network meta-analysis. Risk ratios less than 1 represented a beneficial effect favoring the ICI group. Two-sided
The pooled rates of grade 3–5 or all-grade adverse events for treatments were meta-analyzed by the command of “metan” in STATA 16.0. Subgroup analyses for RRs between the ICI-treated group and the control group were performed based on the panoramic analysis, and prespecified, exploratory stratification factors for subgroup analyses involved the phase of the study (phase II versus phase III), treatment lines (first line, second line, and third line), ICI drug type (anti-PD-1, anti-CTLA-4, anti-PD-L1), treatment mode (ICIs alone versus ICIs combined with ChT), sample size (<500 versus ≥500), etc.
In the literature retrieval stage, a total of 459 articles were obtained through preliminary screening. After reading the titles and abstracts, 422 articles including duplicate reports, irrelevant articles, non-randomized controlled trials, review articles, and phase I trials were excluded. The remaining 20 articles were excluded based on the selection criteria after reading the full text. Finally, a total of 11 trials reported in 17 articles met the inclusion criteria, of which 6 articles were updated or subgroup reports (
Flowchart of the study selection and design.
Characteristics of the included studies.
Study name | References | Trial phase | Treatment line | Cancer type | Treatment | ICI type | Treatment mode | No. of patients | Analysis population for AEs |
---|---|---|---|---|---|---|---|---|---|
ATTRACTION-3 |
|
III | Second-line | EPC | ChT | 209 | 208 | ||
III | Second-line | EPC | Nivo | PD-1 | ICIs | 210 | 209 | ||
ESCORT |
|
III | Second-line | EPC | ChT | 220 | 220 | ||
III | Second-line | EPC | Camr | PD-1 | ICIs | 228 | 228 | ||
KEYNOTE-181 |
|
III | Second-line | EPC | ChT | 314 | 296 | ||
III | Second-line | EPC | Pemb | PD-1 | ICIs | 314 | 314 | ||
KEYNOTE-590 |
|
III | First-line | EPC | ChT | 376 | 370 | ||
III | First-line | EPC | Pemb + ChT | PD-1 | ICIs + ChT | 373 | 370 | ||
CheckMate-648 |
|
III | First-line | EPC | ChT | 324 | 304 | ||
III | First-line | EPC | Nivo + ChT | PD-1 | ICIs + ChT | 321 | 310 | ||
III | First-line | EPC | Nivo + Ipil + ChT | CTLA-4 | ICIs + ChT | 325 | 322 | ||
CheckMate-649 | Moehler et al., 2020s2202020202020 ( |
III | First-line | GEJC | ChT | 792 | 792 | ||
III | First-line | GEJC | Nivo + ChT | PD-1 | ICIs + ChT | 789 | 789 | ||
Study name | References | Trial phase | Treatment line | Cancer type | Treatment | ICI type | Treatment mode | No. of patients | Analysis population for AEs |
ATTRACTION-4 |
|
III | First-line | GEJC | ChT | 362 | 362 | ||
III | First-line | GEJC | Nivo + ChT | PD-1 | ICIs + ChT | 362 | 362 | ||
KEYNOTE-061 |
|
III | Second-line | GEJC | ChT | 296 | 276 | ||
III | Second-line | GEJC | Pemb + ChT | PD-1 | ICIs + ChT | 296 | 294 | ||
NCT01585987 |
|
II | Third-line | GEJC | Placebo | 57 | 57 | ||
II | Third-line | GEJC | Ipil | CTLA-4 | ICIs | 57 | 57 | ||
JAVELIN Gastric 300 |
|
III | Third-line | GEJC | ChT | 186 | 177 | ||
III | Third-line | GEJC | Avel | PD-L1 | ICIs | 185 | 184 | ||
ATTRACTION-2 |
|
III | Third-line | GEJC | Placebo | 163 | 161 | ||
III | Third-line | GEJC | Nivo | PD-1 | ICIs | 330 | 330 |
EPC, esophageal cancer; GEJC, gastroesophageal junction cancer; ChT, chemotherapy; Nivo, nivolumab; Camr, camrelizumab; Pemb, pembrolizumab; Ipil, ipilimumab; Avel, avelumab; ICIs, immune checkpoint inhibitors; PD-1, programmed cell death-1; PD-L1, programmed cell death ligand 1; CTLA-4, cytotoxic T lymphocyte associate protein-4.
The risk of bias assessment for the included studies involving the 11 articles is summarized and shown in
Risk of bias assessment for the included studies. Green for low risk of bias, yellow for unclear risk of bias, and red for high risk of bias.
Only one trial had not reported the results of grade 3–5 trAE (
Network plots of comparisons with
In the consistency model, for the rates of grade 3–5 trAEs, the results with significant benefits for different pairwise comparisons could be found in avelumab versus ChT, nivolumab versus ChT, pembrolizumab versus ChT, placebo versus ChT, placebo versus nivolumab + ChT, and placebo versus pembrolizumab + ChT. The results with significant increasing risk could be found in nivolumab + ChT versus avelumab, nivolumab + ipilimumab + ChT versus avelumab, pembrolizumab + ChT versus avelumab, nivolumab + ChT versus camrelizumab, nivolumab + ChT versus nivolumab, and nivolumab + ipilimumab + ChT versus nivolumab.
For the rates of all-grade trAEs, the results with significant benefits for different pairwise comparisons could be found in ipilimumab versus ChT, placebo versus ChT, ipilimumab versus avelumab, placebo versus avelumab, ipilimumab versus camrelizumab, placebo versus camrelizumab, placebo versus nivolumab, placebo versus nivolumab + ChT, placebo versus nivolumab + ipilimumab + ChT, placebo versus pembrolizumab, and placebo versus pembrolizumab + ChT. The results with significant increasing risk could be found in nivolumab versus ipilimumab, nivolumab + ChT versus ipilimumab, nivolumab + ipilimumab + ChT versus ipilimumab, pembrolizumab versus ipilimumab, and pembrolizumab + ChT versus ipilimumab. The details of all comparisons are indicated in
Results of the network meta-analysis for 10 treatment regimens in terms of treatment-related adverse events(trAEs) with grade 3–5 trAEs and all-grade trAEs.
As shown in
Forest plots for pairwise comparisons of all individual regimens with each other with
For the rates of grade 1–2 trAEs, nivolumab + ChT was associated with the best safety ranking for grade 1–2 trAEs (88.2%), followed by pembrolizumab + ChT (85.1%), nivolumab + ipilimumab + ChT (82.4%), ChT (53.5%), and placebo (53.3%). The results of NMA are indicated in
Stratification factors used for subgroup analyses included treatment lines (first line, second line, and third line), ICI drug type (anti-PD-1, anti-CTLA-4, anti-PD-L1), treatment mode (ICIs alone versus ICIs combined with ChT), and sample size (<500 versus ≥500). Based on the panoramic analysis of whether ICI treatment was applied, although the overall rates of grade 3–5 and all-grade trAEs were similar between the two groups, there were statistical differences in the rates of trAEs in some subgroups. For first-line treatment, ICIs were usually applied in combination with chemotherapy; consequently, the additional ICIs had significantly increased the rates of grade 3–5 trAEs (RR = 1.159, 95% CI = 1.012 to 1.327). However, for second-line treatment, ICIs had significantly decreased the rates of grade 3–5 trAEs (RR = 0.395, 95% CI = 0.317 to 0.491). In the case of ICIs alone, compared with chemotherapy, ICIs significantly reduced the rates of grade 3–5 trAEs (RR = 0.584, 95% CI = 0.350 to 0.974). The detailed results for subgroup analyses are listed in
Subgroup analysis of risk ratios for treatment-related adverse events (trAEs) comparing ICI therapy with chemotherapy.
Subgroup | Grade 3–5 trAEs | All-grade trAEs | ||||
---|---|---|---|---|---|---|
analysis |
|
RR (95% CI) |
|
|
RR (95% CI) |
|
Overall | 95.7% (0.000) | 0.764 (0.574, 1.016) |
|
96.7% (0.000) | 0.916 (0.831, 1.010) |
|
Subgroup | ||||||
Treatment lines | ||||||
Second-line | 52.1% (0.100) | 0.395 (0.317, 0.491) |
|
97.6% (0.000) | 0.762 (0.570, 1.019) |
|
First-line | 80.1% (0.000) | 1.159 (1.012, 1.327) |
|
89.5% (0.000) | 1.006 (0.952, 1.062) |
|
Third-line | 94.2% (0.000) | 1.198 (0.199, 7.200) |
|
95.2% (0.000) | 1.190 (0.600, 2.358) |
|
ICI drug type | ||||||
PD-1 | 96.2% (0.000) | 0.773 (0.566, 1.057) |
|
97% (0.000) | 0.919 (0.827, 1.020) |
|
CTLA-4 | 82.1% (0.000) | 1.531 (0.434, 5.404) |
|
93.1% (0.000) | 1.177 (0.614, 2.258) |
|
PD-L1 | – | 0.251 (0.156, 0.404) |
|
– | 0.661 (0.557, 0.785) |
|
Treatment mode | ||||||
ICIs alone | 88.7% (0.000) | 0.584 (0.350, 0.974) |
|
95.40% (0.000) | 0.952 (0.755, 1.200) |
|
ICIs + ChT | 91.6% (0.000) | 1.007 (0.818, 1.239) |
|
96.80% (0.000) | 0.926 (0.836, 1.025) |
|
Sample size | ||||||
<500 | 90.9% (0.000) | 0.663 (0.327, 1.344) |
|
95.50% (0.000) | 1.012 (0.760, 1.348) |
|
≥500 | 94.4% (0.000) | 0.892 (0.697, 1.142) |
|
97.60% (0.000) | 0.891 (0.795, 0.998) |
|
Subgroup analyses were conducted based on the pairwise comparisons of all individual trials.
Only five clinical trials had provided detailed data comparing the rates of serious trAEs between ICIs and chemotherapy (
meta-analysis shows that the rates of events leading to discontinuation in the ICI group and the chemotherapy group were 22.42% (570/2542) and 11.59% (289/2,494), respectively, without statistical significance (RR = 1.447, 95% CI = 0.908 to 2.307,
The meta-analysis for some specific treatment-related adverse events of interest is listed in
Summary forest plots for specific treatment-related adverse events with
Except for diarrhea and rash, ICIs had significantly reduced the rates of specific treatment-related adverse events. The most common all-grade trAEs were alopecia (33.44%), followed by decreased white blood cell count (27.19%), anemia (23.63%), decreased neutrophil count (23.6%), and nausea (21.31%) in the chemotherapy group and diarrhea (9.84%) in the ICI group, followed by fatigue (9.34%), asthenia (7.26%), rash (6.43%), and decreased appetite (6.25%).
Only seven and eight trials had provided data on the rates of grade 3–5 irAEs (
The NMA results of the consistency model for the rates of grade 3–5 irAEs and all-grade irAEs are indicated in
Results of the network meta-analysis for 10 treatment regimens in terms of immune-related adverse events (irAEs) with grade 3–5 irAEs and all-grade irAEs.
Conventional pairwise meta-analysis was used to integrate all available data of irAEs. Seven clinical trials had provided detailed data comparing the rates of grade 3–5 irAEs between ICIs and chemotherapy (
Forest plots for irAEs with
The stratification factors of irAEs were the same as those of trAEs. The detailed results for the subgroup analyses are listed in
Subgroup analysis of risk ratios for immune-related adverse events (irAEs) comparing ICI therapy with chemotherapy.
Subgroup analysis |
Grade 3–5 irAEs | All-grade irAEs | ||||
---|---|---|---|---|---|---|
|
RR (95% CI) |
|
|
RR (95% CI) |
|
|
Overall | 95.7% (0.000) | 3.151 (2.175, 4.563) |
|
80.1% (0.000) | 3.851 (2.767, 5.359) |
|
Subgroup | ||||||
Treatment lines | ||||||
Second-line | 51.6% (0.151) | 3.387 (0.690, 16.635) |
|
89.8% (0.000) | 4.036 (1.833, 8.888) |
|
First-line | 37.6% (0.201) | 3.011 (1.880, 4.823) |
|
87.0% (0.000) | 3.653 (2.430, 5.493) |
|
Third-line | 0.0% (0.823) | 9.716 (1.849, 51.060) |
|
60.8% (0.078) | 7.513 (1.096, 51.516) |
|
ICI drugs | ||||||
PD-1 | 0.0% (0.504) | 2.484 (1.620, 3.807) |
|
85.3% (0.000) | 3.464 (2.280, 5.262) |
|
CTLA-4 | 9.4% (0.293) | 4.729 (2.071, 10.798) |
|
19.0% (0.267) | 5.562 (2.176, 14.215) |
|
PD-L1 | – | 8.659 (0.470, 159.675) |
|
– | 24.054 (1.435, 403.229) |
|
Treatment mode | ||||||
ICIs alone | 0.0% (0.943) | 9.690 (2.670, 35.166) |
|
76.9% (0.002) | 5.099 (2.396, 10.850) |
|
ICIs + ChT | 23.4% (0.270) | 2.839 (1.892, 4.261) |
|
84.5% (0.000) | 3.357 (2.320, 4.858) |
|
Sample size | ||||||
<500 | 0.0% (0.943) | 9.690 (2.670, 35.166) |
|
68.8% (0.022) | 6.573 (2.383, 18.128) |
|
≥500 | 23.4% (0.270) | 2.839 (1.892, 4.261) |
|
80.0% (0.000) | 3.329 (2.432, 4.559) |
|
Subgroup analyses were conducted based on the pairwise comparisons of all individual trials.
Some specific immune-related adverse events of interest are listed in
Immunotherapy based on ICIs has currently become one of the most promising treatment regimens for cancer, which plays an encouraging role in the treatment of advanced cancer (
In this review, we have included a total of 11 studies, including 7,089 patients, of which 6,992 cases can be used for adverse event analysis. As far as we know, the current meta-analysis may be the study with the largest sample size to explore the possible adverse events of immunotherapy in esophageal cancer and gastroesophageal junction cancer. Based on the results of our NMA analysis of different lines of immunotherapy for esophageal/gastroesophageal junction cancer, we can draw five main conclusions that may affect clinical practice.
First of all, from the point of view of different treatment modalities, different combinations of treatment modalities had obviously distinct safety outcomes in trAEs and irAEs. Similar to the results of practice in lung cancer, ICIs were generally less toxic in monotherapy than in chemotherapy, and the combination of ICIs and chemotherapy would increase the rates of grade 3–5 trAEs and grade 3–5 irAEs (
Secondly, the application of ICI drugs in esophageal cancer involved first-line, second-line, third-line, or later-line treatment (
Third, previous studies had shown that different types of ICIs have different toxicity profiles because of their different mechanisms of action (
Fourth, the spectrum of trAEs caused by ICIs was also significantly different from that caused by ChT. Our meta-analysis based on specific treatment-related adverse events showed that ICIs were safer and had a significantly different spectrum of grade 3–5 trAEs and all-grade trAEs from chemotherapy. Hematological toxicity was the main adverse event for chemotherapy, while systemic symptoms such as fatigue, asthenia, and decreased appetite were the main adverse events for ICIs (
Finally, there was no consensus on whether the rates of irAEs were related to the primary site of the tumor. One review found that the rates of several specific AEs of interest varied among different cancer types (
It should be pointed out from the results of our meta-analysis that, although ICIs increased the adverse events, the rates were actually low and acceptable. Although immunotherapy had increased the rates of irAEs, to a certain extent, the occurrence of immune-related events may be positively correlated with the therapy’s efficacy and the patient’s prognosis (
There are some limitations in our review that need to be mentioned. First, the network meta-analysis assumes that the estimates of the study effects between the various trials have commonality, transferability, and exchangeability, which means that the similarities of population characteristics, interventions, chemotherapy regimen, and other features among different trials are required. However, as the conditions of the trials may affect the study results, this assumption is very unrealistic. In our meta-analysis, heterogeneity was detected in the results of grade 3–5 trAEs and all-grade trAEs. Subgroup meta-analyses revealed that trials with treatment line = second line, treatment line = first line, treatment mode, and a sample size ≥500 patients were potential sources of heterogeneity. Second, some specific irAEs and trAEs may be selectively reported in most trials because the rates of these adverse events were lower than a preset threshold, such as 1% or 5%. In this case, we cannot obtain the pooled estimates of rates for these rare adverse events, so it is inevitable to underestimate the overall mean rates of some adverse events. Third, in order to catch the latest data from newly published trials, some recent conference abstracts were enrolled in our meta-analysis, from which some summary data were extracted. However, this may lead to another selection bias because the comprehensive toxicity data might not be reported in these abstracts. Furthermore, some previous meta-analyses on this topic had shown the influence of different drug doses on the occurrence of adverse events (
Monotherapy with immune checkpoint inhibitors displayed better safety profiles in terms of trAEs than chemotherapy alone; however, combinational treatment regimens involving ICIs increased the risk of trAEs. Different ICIs had different toxicity manifestations and should not be considered as an entity. Compared with chemotherapy, ICIs were more prone to irAEs, but the overall rates remained low and acceptable. For clinicians, it is important to recognize and monitor the adverse events caused by ICIs for patients with esophageal cancer or gastroesophageal junction cancer.
The original contributions presented in the study are included in the article/
JZ and LX collected the data. JZ and MW performed data cleaning and analysis. JZ and BH performed the systematic review. JZ and BH evaluated the data. JL drafted and reviewed the manuscript for scientific soundness. All authors contributed to the article and approved the submitted version.
This study was supported in part by the National Clinical Key Specialty Construction Program (Grant No. 2021 to JL), the Fujian Provincial Clinical Research Center for Cancer Radiotherapy and Immunotherapy (Grant No. 2020Y2012 to JL), the Fujian Provincial Health Technology Project (Youth Scientific Research Project, 2019-1-50 to JZ), and the Nursery Fund Project of the Second Affiliated Hospital of Fujian Medical University (Grant No. 2021MP05 to JZ).
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
The Supplementary Material for this article can be found online at: