If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Early Immune-Related Adverse Events and Association with Outcome in Advanced Non–Small Cell Lung Cancer Patients Treated with Nivolumab: A Prospective Cohort Study
Corresponding author. Address for correspondence: Daichi Fujimoto, MD, Department of Respiratory Medicine, Kobe City Medical Center General Hospital, 2-1-1 Minatojima-minamimachi, Chuo-ku, Kobe, Hyogo 650-0047, Japan.
Clinical Research Center, Kobe City Medical Center General Hospital, Hyogo, JapanDepartment of Clinical Epidemiology, Hyogo College of Medicine, Hyogo, Japan
Retrospective studies have shown immune-related adverse events (irAEs) to be associated with better prognosis. However, no prospective clinical trials have been conducted, and little is known regarding the association between irAEs and the outcome of patients with NSCLC after treatment with immunotherapy.
Methods
We conducted a prospective cohort study of patients with advanced NSCLC who were treated with nivolumab between January and December 2016. The association between clinical outcome and irAEs 2 to 6 weeks after commencement of nivolumab treatment was investigated. IrAEs were assessed by at least three independent medical professionals.
Results
A total of 43 patients were enrolled, including 19 patients with irAEs 2 weeks after commencement of nivolumab treatment. Common irAEs included rash, pyrexia, and diarrhea. Programmed cell death ligand 1–positive tumor cell expression was not significantly different between patients with and without irAEs. The objective response and disease control rates were higher in patients with irAEs than in those without irAEs (37% versus 17% and 74% versus 29% [p = 0.17 and p < 0.01], respectively]). Patients with irAEs were associated with a significantly longer median progression-free survival than those without (6.4 months [95% confidence interval: 2.5–not reached] versus 1.5 months [95% confidence interval: 1.2–2.3] [p = 0.01]). These findings were comparable to those for patients with and without irAEs 6 weeks after commencement of nivolumab treatment.
Conclusions
Early irAEs are associated with a better outcome after treatment with immunotherapy. We predicted responses to nivolumab by using early irAEs. Further research is needed to elucidate the mechanisms of these associations.
Recently, programmed cell death protein 1 (PD-1)/programmed death ligand 1 (PD-L1) checkpoint inhibitors have exhibited potent antitumor activity in metastatic NSCLC, and PD-1 axis inhibitors are a standard anticancer treatment for these patients.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
Immune checkpoint inhibitors (ICIs), including PD-1 axis inhibitors, are molecules that cause imbalances in immunological tolerance that result in an unchecked immune response. These can clinically manifest with inflammatory side effects, so-called immune-related adverse events (irAEs), which differ from those of conventional systemic therapy.
These toxicities have varying times to onset and affect different organs. Examples include dermatological irAEs (rash and pruritus), gastrointestinal irAEs (diarrhea and colitis), hepatitis, endocrinopathies, pneumonitis, and renal insufficiencies.
irAEs were associated with a durable response and clinical benefit to ICIs in patients with melanoma. Recently, dermatological irAEs and immunotherapy-related thyroid dysfunction have been associated with a clinical benefit to ICIs in patients with NSCLC.
However, no prospective clinical trials have been conducted and little is known regarding the association between irAEs and the outcome of patients with NSCLC after treatment with ICIs.
The aim of this study was to investigate the association between early irAEs and the efficacy of nivolumab treatment for NSCLC and provide scope for improving therapy with ICIs.
Patients and Methods
We conducted a prospective cohort study. The primary objective was to investigate the association between clinical outcome and irAEs 2 to 6 weeks after commencement of nivolumab treatment. We designed this assessment schedule to specifically evaluate irAEs after the first treatment dose and before the first response evaluation because the response data could potentially be biased. The primary end point was progression-free survival (PFS). Secondary end points included overall survival (OS) and the objective response rate (ORR) and disease control rate (DCR).
Eligibility criteria included (1) NSCLC histologically or cytologically proved to be stage IIIB to IV on the basis of the TNM staging system (seventh edition),
The IASLC Lung Cancer Staging Project: proposals for the revision of the N descriptors in the forthcoming seventh edition of the TNM classification for lung cancer.
which includes postoperative recurrence; (2) prior treatment with one or more chemotherapy regimens; (3) measurable disease lesions; (4) an age of 20 years or older; and (5) an Eastern Cooperative Oncology Group performance status of 0 to 2. Patients were excluded if they were treated with systemic glucocorticoids or other immunosuppressive agents or had active autoimmune disease. All study participants provided informed consent. The study was approved by the Ethics Committee of Kobe City Medical Center General Hospital (Hyogo, Japan). The research was conducted in accordance with the Declaration of Helsinki.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
as AEs with a potential immunological basis that require potential intervention with immunosuppressive or endocrine therapy. These include pyrexia as well as cutaneous, endocrine, gastrointestinal, hepatic, lymph node, neurological, optic, pancreatic, pulmonary, and renal irAEs. In addition, to reduce bias, we focused only on irAEs that medical professionals could recognize objectively. All patients were admitted within the first week of commencement of nivolumab treatment, and the details of the irAEs before and after treatment were explained. In an outpatient setting, patient symptoms and laboratory data were assessed at 1-week intervals by using a diary card. Thyroid function was evaluated before commencement of nivolumab treatment and every 2 to 4 weeks thereafter. Patients were advised to consult us by telephone if they suspected irAEs. IrAEs of any grade according to the Common Terminology Criteria for Adverse Events, version 4.0, were independently assessed by at least three medical professionals, including a doctor, nurse, and pharmacist.
Tumor assessment was performed according to the Response Evaluation Criteria in Solid Tumors (version 1.1).
Computed tomography scans were performed every 8 to 12 weeks to assess tumor response. The time interval between date of commencement of nivolumab treatment and date of disease progression or death (PFS) or death alone (OS) was calculated for each patient. The data cutoff date was March 1, 2017.
PD-L1 expression in human NSCLC specimens was analyzed by immunohistochemical staining using the PD-L1 IHC 22C3 pharmDx antibody (clone 22C3 [Dako North America, Inc., Carpinteria, CA]). The antibody was applied according to DAKO-recommended detection methods. PD-L1 expression in tumor cells was scored as the percentage of stained cells in each section, which was estimated in increments of 5%, except for a 1% positivity value. Patients in whom at least 1% of the tumor cells were stained for PD-L1 were considered positive, as is consistent with other studies using this antibody.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Samples were anonymized and independently scored by at least four pathologists. In cases in which there was a discrepancy, the slides were reexamined until a consensus was reached.
Statistical Analyses
Age was analyzed using the Wilcoxon rank sum test. Dichotomous variables were analyzed using a chi-square or Fisher’s exact test, as appropriate. Times to events were estimated using the Kaplan-Meier method and compared by the log-rank test. Statistical analyses were conducted with JMP software (version 11 [SAS Institute, Cary, NC]). A two-tailed p value less than 0.05 was considered significant.
Results
Patient Characteristics
In total, 43 patients with NSCLC who were treated with nivolumab at Kobe City Medical Center General Hospital (Hyogo, Japan) between January and December 2016 were enrolled (Fig. 1). The baseline characteristics and comparisons between patients with and without early irAEs are summarized in Table 1. The cohort comprised 27 men and 16 women, with a median age of 70 years (range 50−82).
Patients with squamous and nonsquamous cell carcinoma, with and without EGFR mutations, with and without ALK rearrangement, PD-L1 expression of 50% or more and between 1% and less than 50% were compared by using Fisher’s exact tests.
Patients with squamous and nonsquamous cell carcinoma, with and without EGFR mutations, with and without ALK rearrangement, PD-L1 expression of 50% or more and between 1% and less than 50% were compared by using Fisher’s exact tests.
Patients with squamous and nonsquamous cell carcinoma, with and without EGFR mutations, with and without ALK rearrangement, PD-L1 expression of 50% or more and between 1% and less than 50% were compared by using Fisher’s exact tests.
Patients with squamous and nonsquamous cell carcinoma, with and without EGFR mutations, with and without ALK rearrangement, PD-L1 expression of 50% or more and between 1% and less than 50% were compared by using Fisher’s exact tests.
≥50
8 (19)
5 (26)
3 (13)
0.44
≥1
18 (42)
10 (53)
8 (33)
0.51
NK
8 (19)
2 (11)
6 (25)
IrAE, n (%)
Rash
12 (28)
—
—
—
Pyrexia
6 (14)
—
—
—
Diarrhea
4 (9)
—
—
—
Increased hepatic enzyme levels
1 (2)
—
—
—
irAE, immune-related adverse event; ECOG PS, Eastern Cooperative Oncology Group performance status; ADC, adenocarcinoma; SCC, squamous cell carcinoma; NK, not known; ALK, anaplastic lymphoma kinase; PD-L1, programmed death ligand 1.
a Patients with squamous and nonsquamous cell carcinoma, with and without EGFR mutations, with and without ALK rearrangement, PD-L1 expression of 50% or more and between 1% and less than 50% were compared by using Fisher’s exact tests.
In patients evaluated for irAEs 2 weeks after commencement of nivolumab treatment, no significant differences in clinical profiles were observed between those with and without irAEs. PD-L1−positive tumor cell expression was not significantly different between patients with and without irAEs. The ORR and median PFS were higher in patients with PD-L1−positive tumor cell expression than in patients with PD-L1–negative tumor cell expression (ORR 50% versus 6% [p < 0.01] and median PFS = 7.7 months [95% confidence interval (CI): 1.5–not reached] versus 2.1 months [95% CI: 1.2–5.5] [p = 0.01]). The ORR and median PFS were also considerably higher in patients with PD-L1 expression on 50% or more of tumor cells (ORR 75% and median PFS not reached). IrAEs developed in 19 patients within 2 weeks of commencement of nivolumab treatment, including rash (28%), pyrexia (14%), diarrhea (9%), and increased hepatic enzyme levels (2%). Grade 3 to 4 irAEs did not develop in any of the patients, and none of them discontinued treatment within the first 2 weeks of commencement of nivolumab treatment (see Table 1).
We also investigated differences between patients with and without irAEs 6 weeks after commencement of nivolumab treatment and compared patient clinical profiles between the two groups (Supplementary Table 1). Patients with irAEs were significantly younger than those without irAEs (p = 0.03). IrAEs developed in 27 patients within 6 weeks of commencement of nivolumab treatment, including rash (35%), diarrhea (21%), pyrexia (16%), increased hepatic enzyme levels (9%), hyperthyroidism (2%), and stomatitis (2%). Grade 3 rash that was successfully treated with steroids developed in two patients (5%).
Association between irAEs and Clinical Outcome
The DCR and ORR for patients with and without irAEs are displayed in Table 2. In patients evaluated for irAEs 2 weeks after commencement of nivolumab treatment, the DCR and ORR were higher in patients with than without irAEs (74% versus 29% [p < 0.01] and 37% versus 17% [p = 0.17], respectively). A similar trend was observed in patients with and without irAEs 6 weeks after commencement of nivolumab treatment (59% versus 31% [p = 0.12] and 33% versus 13% [p = 0.17], respectively).
Table 2Association between the Presence of Immune-Related Adverse Events and Treatment Response
The Kaplan-Meier curves of PFS are displayed in Figure 2A through D and Figure 3A through D. The PFS was significantly longer in patients in whom irAEs developed within 2 weeks of commencement of nivolumab treatment than in those in whom it did not develop (6.4 months [95% CI: 2.5–not reached] versus 1.5 months [95% CI: 1.2–2.3] [p = 0.01]). We also analyzed patients without EGFR mutations, because there could have been an imbalance in patients with EGFR mutations. In patients without EGFR mutations, there was also a significant difference in the median PFS between those with and without irAEs 2 weeks after commencement of nivolumab treatment (not reached [95% CI: 2.5–not reached] versus 1.5 [95% CI: 1.2–4.1] months; [p = 0.01]). A similar trend was observed in patients with and without irAEs 6 weeks after commencement of nivolumab treatment (4.1 [95% CI: 1.8–not reached] versus 1.5 [95% CI: 1.2–2.3] months [p = 0.06]).
Figure 2Kaplan-Meier curves of progression-free survival (PFS) in patients with or without immune-related adverse events (irAEs) (A), rash (B), pyrexia (C), and diarrhea (D) 2 weeks after commencement of nivolumab treatment.
Figure 3Kaplan-Meier curves of progression-free survival (PFS) in patients with or without immune-related adverse events (irAEs) (A), rash (B), pyrexia (C), and diarrhea (D) 6 weeks after commencement of nivolumab treatment.
OS was inconclusive at the time of data cutoff (Supplementary Figs. 1 and 2). Twelve of 43 patients (4 of 19 patients with irAEs and 8 of 24 patients without irAEs 2 weeks after commencement of nivolumab treatment) died before the data cutoff; follow-up is ongoing.
Association between Type of IrAE and Clinical Outcome
Patients with and without each type of common irAE were subsequently analyzed. In patients evaluated for irAEs 2 weeks after commencement of nivolumab treatment, rash and pyrexia were associated with a longer PFS (not reached [95% CI: 2.1 months–not reached] versus 1.6 months [95% CI: 1.2–3.3] and not reached [95% CI: 3.3 months–not reached] versus 2.1 months [95% CI: 1.4–5.5] [p = 0.01 and p = 0.12], respectively). In contrast, diarrhea was not a predictive factor for a longer PFS (1.9 months [95% CI: 0.9–not reached] versus 2.5 months [95% CI: 1.5–6.4] [p = 0.63]) (see Fig. 2A–D). Similar trends were observed in patients with and without irAEs 6 weeks after commencement of nivolumab treatment. Rash and pyrexia were associated with a longer PFS (5.5 months [95% CI: 1.2–not reached] versus 1.9 months [95% CI: 1.3–4.1] and 6.4 months [95% CI: 1.0–not reached] versus 2.1 months [95% CI: 1.4–5.5] [p = 0.08 and p = 0.26], respectively). Additionally, diarrhea and increased hepatic enzyme levels were not predictive factors for a longer PFS (2.6 months [95% CI: 0.9–not reached] versus 2.4 months [95% CI: 1.5–7.7] and 1.4 months [95% CI: 0.9–7.7] versus 2.6 months [95% CI: 1.5–6.4] [p = 0.86 and p = 0.10], respectively) (see Fig. 3A–D). In only one patient did hyperthyroidism develop within 6 weeks of commencement of nivolumab treatment. This patient exhibited a partial response.
Moreover, the presence of rash or pyrexia 2 weeks after commencement of nivolumab treatment was associated a significantly better outcome (DCR and ORR 81% versus 30% and 44% versus 15% [p < 0.01 and p = 0.07], respectively) (median PFS not reached [95% CI: 3.3 months–not reached] versus 1.5 months [95% CI: 1.2–2.3] [p < 0.001] and median OS not reached in either arm [95% CIs: 9.7 months–not reached and 5.4 months–not reached] [p = 0.04]) (Supplementary Figs. 3 and 4). Similar but weaker trends were observed in patients with rash or pyrexia 6 weeks after commencement of nivolumab treatment (DCR and ORR 65% versus 35% and 35% versus 17% [p = 0.07 and p = 0.29, respectively]) (median PFS 5.5 months [95% CI: 1.8–not reached] versus 1.5 months [95% CI: 1.2–2.6] [p = 0.02] and median OS: not reached in either arm [95% CIs: 9.7 months–not reached and 2.6 months–not reached] [p = 0.05]) (Supplementary Figs. 5 and 6).
Discussion
To the best of our knowledge, this is the first prospective study to demonstrate an association between irAEs and the efficacy of nivolumab treatment for NSCLC. Moreover, rash and pyrexia were strong early predictive factors of efficacy.
We have clearly demonstrated that early irAEs are associated with the efficacy of nivolumab treatment for NSCLC. Several retrospective studies
have shown that dermatological irAEs are associated with a better prognosis in patients with melanoma and NSCLC treated with ICIs. More recently, another retrospective study
demonstrated a similar association with thyroid dysfunction in patients with NSCLC. Considering these findings and our present data, there is an association between irAEs and the efficacy of nivolumab treatment for NSCLC.
In the present study, rash, pyrexia, and diarrhea occurred 6 weeks after commencement of nivolumab treatment in 35%, 16%, and 21% of patients, respectively. These proportions are marginally higher than those reported previously (4%–26%, 3%−14.5%, and 6%−19%, respectively).
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
We considered two explanations for this discrepancy. The first is our strict detection method in which irAEs are independently assessed by at least three medical professionals, including a doctor, nurse, and pharmacist. Additionally, all patients were admitted within the first week of commencement of nivolumab treatment and were provided with an explanation of the details of the irAEs before and after treatment. Therefore, our patients may have been more sensitive to irAEs. The second explanation is the difference in ethnicity. The CheckMate 017 and 057 trials
Our study demonstrates that early detection of irAEs (within 2−6 weeks of commencement of nivolumab treatment) contributes to the prediction of nivolumab response in a clinical setting. Furthermore, our findings suggest that irAEs detected 2 weeks after commencement of nivolumab treatment may be stronger predictive factors for treatment outcome than those detected after 6 weeks. Although sample sizes were limited, comparisons were made between patients with (n = 18) and without (n = 25) irAEs 2 to 6 weeks after commencement of nivolumab treatment. The DCR (44% versus 52% [p = 0.76]), ORR (33% versus 20 [p = 0.48]), and median PFS (3.3 months [95% CI: 1.2–7.7] versus 2.3 months [95% CI: 1.4–9.5] [p = 0.92]) were comparable between the two groups (Supplementary Fig. 7). IrAEs were assessed after 2 weeks in patients treated with one cycle of nivolumab therapy, since nivolumab was administered biweekly in patients with NSCLC. Moreover, responses were assessed 8 or more weeks after the commencement of nivolumab treatment in clinical trials of patients with NSCLC.
Considering these findings, irAEs occurring within 2 weeks of commencement of nivolumab treatment may be more effective predictors of nivolumab response in a clinical setting.
Our study also reveals differences in the characteristics of patients with and without early irAEs. In patients with irAEs 2 weeks after commencement of nivolumab treatment, no significant differences in patient characteristics were detected. However, patients with irAEs 6 weeks after commencement of nivolumab treatment were considerably younger than those without. In the CheckMate 017 trial,
reported that a significant benefit was not observed in patients aged 75 or older years who were treated with PD-1 axis inhibitors. Therefore, PD-1 axis inhibitors may be more likely to induce an unchecked immune response in younger patients.
PD-L1 expression was also investigated, which did not differ significantly between patients with and without irAEs. This finding is comparable to that of the CheckMate 057 trial,
which also reported similar frequencies of irAEs in patient subgroups with different levels of PD-L1 expression. A comparison between the KEYNOTE 010 and 024 trials
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
also revealed similar frequencies of irAEs (63% and 73%, respectively), although the cutoff for inclusion for PD-L1 expression differed (≥1% versus ≥50%, respectively).
The precise mechanism of the association between irAEs and the efficacy of nivolumab treatment remains unclear. However, it is possible that healthy tissues express antigens identical or similar to those expressed by the tumor, a concept known as antigen sharing or molecular mimicry.
demonstrated that the lymphocytic skin infiltration patterns differed according to the histological subtype of the tumor in patients with NSCLC with dermatological irAEs, which correlated with tumor response. In view of this, antigen sharing was considered the most likely cause of this association. As demonstrated in the present study, the occurrence of irAEs after 2 weeks correlated with outcome better than that after 6 weeks. Therefore, we hypothesized that the unchecked immune activity of the tumor may be enhanced by nivolumab in the early period, which results in T-cell recognition of antigens expressed by healthy tissues. We believe that these associations provide scope for improving therapy with ICIs, as well as important clues as to the mechanisms of PD-1–mediated toxicity and antitumor efficacy.
Our study demonstrates the impact of individual irAEs on the efficacy of nivolumab. Rash and pyrexia were associated with efficacy whereas diarrhea was not. According to previous studies, the impact of individual irAEs may differ depending on the cancer type. A previous study of malignant melanoma
reported that dermatological irAEs were associated with a prolonged survival, whereas hypothyroidism or hyperthyroidism, diarrhea, and pneumonia were not. However, thyroid dysfunction was associated with the efficacy of pembrolizumab in patients with NSCLC.
reported that vitiligo was a rare irAE. Because our subgroups included very small numbers of patients, we were unable to draw definitive conclusions. However, our findings do reveal the different impact of each irAE. Future large-scale studies are needed to confirm our findings.
This study has several limitations. The first is its small sample size. However, all previous studies investigating these associations
have been retrospective with similar sample sizes. Second, only Japanese patients were included, and third, we were unable to distinguish fully between irAEs and non-irAEs. The latter remains a significant challenge, as attributing causality is an unavoidable problem in drug-related studies given the presence of multiple confounders and the absence of definitive standard diagnostic tests. However, we designed this prospective study weighing on the objectivity of irAEs and strictly checked irAEs by the same method. Therefore, we believe that the quality of our study is high.
In conclusion, this is the first prospective study to show that early irAEs are associated with the efficacy of nivolumab treatment in patients with NSCLC. In clinical cases, we predicted the response to nivolumab by using early irAEs. We believe that these associations provide scope for improving therapy with ICIs, as well as important clues as to the mechanisms of PD-1–mediated toxicity and antitumor efficacy.
Acknowledgments
This work was supported by internal funding from Kobe City Medical Center General Hospital (Hyogo, Japan). We wish to thank T. Tanaka, H. Kisoi, and S. Nakanishi for their assistance and Y. Imaoka, K. Hayashi, and M. Sugahara for their analysis of the pathological data in this trial.
Supplementary Data
Supplementary Figure S1Kaplan-Meier curves of overall survival (OS) in patients with or without immune-related adverse events (irAEs) 2 weeks after commencing nivolumab treatment.
Supplementary Figure S2Kaplan-Meier curves of overall survival (OS) in patients with or without immune-related adverse events (irAEs) 6 weeks after commencing nivolumab treatment.
Supplementary Figure S3Kaplan-Meier curves of progression-free survival (PFS) in patients with or without a rash or pyrexia 2 weeks after commencing nivolumab treatment.
Supplementary Figure S4Kaplan-Meier curves of overall survival (OS) in patients with or without a rash or pyrexia 2 weeks after commencing nivolumab treatment.
Supplementary Figure S5Kaplan-Meier curves of progression-free survival (PFS) in patients with or without a rash or pyrexia 6 weeks after commencing nivolumab treatment.
Supplementary Figure S6Kaplan-Meier curves of overall survival (OS) in patients with or without a rash or pyrexia 6 weeks after commencing nivolumab treatment.
Supplementary Figure S7Kaplan-Meier curves of progression-free survival (PFS) in patients with or without immune-related adverse events (irAEs) from 2 to 6 weeks after commencing nivolumab treatment.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
The IASLC Lung Cancer Staging Project: proposals for the revision of the N descriptors in the forthcoming seventh edition of the TNM classification for lung cancer.
Drs. Teraoka and Fujimoto contributed equally to this work.
Disclosure: Dr. Fujimoto reports personal fees from Ono Pharmaceutical Co., Ltd. and Bristol-Myers Squibb K.K. outside the submitted work. The remaining authors declare no conflict of interest.
30825-0&id=gr1.jpg){kind=link}
30825-0&id=gr2.jpg){kind=link}
30825-0&id=gr3.jpg){kind=link}
30825-0&id=figs1.jpg){kind=link}
30825-0&id=figs2.jpg){kind=link}
30825-0&id=figs3.jpg){kind=link}
30825-0&id=figs4.jpg){kind=link}
30825-0&id=figs5.jpg){kind=link}
30825-0&id=figs6.jpg){kind=link}
30825-0&id=figs7.jpg){kind=link}