<p>Incidence risk of PD-1/PD-L1 related diarrhea in non-small cell lung cancer (NSCLC) patients: a systematic review and meta-analysis</p>

(1)

O R I G I N A L R E S E A R C H

Incidence risk of PD-1/PD-L1 related diarrhea in

non-small cell lung cancer (NSCLC) patients:

a systematic review and meta-analysis

This article was published in the following Dove Press journal: Cancer Management and Research

Caiqing Zhang1,*

Shuisheng Zhang2,*

Deguo Xu3,*

Rujun Liu4,*

Qingshan Zhu5

Yi Zhao6

Yantao Mao7

Yuan Tian3

1_{Department of Respiratory, Shandong}

Provincial Qianfoshan Hospital Afﬁliated

to Shandong University, Jinan, Shandong

250014, People’s Republic of China;

2_{Department of General Surgery, Peking}

University Third Hospital, Beijing 100191,

People’s Republic of China;3_Department

of Radiotherapy Oncology, Shandong

Provincial Qianfoshan Hospital Afﬁliated

to Shandong University, Jinan, Shandong

4_{Department of Oncology, LongKou}

People,s Hospital, Yantai, Shandong

5_{Department of Radiotherapy oncology,}

Anyang Cancer Hospital of Henan Province, Anyang, Henan 455000,

People’s Republic of China;6_Department

of Oncology, The First Afﬁliated Hospital

of Dalian Medical University, Dalian, Liaoning 116000, People’s Republic of

China;7_{Department of Oncology,}

Yantaishan Hospital of Shandong

Province, Yantai, 264000, People’s

Republic of China

*These authors contributed equally to this work

Purpose:We designed the study to illustrate the OR of programmed cell death-1 (PD-1) or ligand 1 (PD-L1) inhibitor-related diarrhea in patients with non-small cell lung cancer.

Method:This systematic review and meta-analysis were put into practice according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. Incidence of all grades for PD-1/PD-L1 inhibitor-related diarrhea in NSCLC was taken into account.

Results:After screening and eligibility assessment of 57 articles, a total of 12 clinical trials involving 6,659 participants were collected for theﬁnal meta-analysis. The incidence risk of diarrhea for all grades was lower in PD-1 inhibitor monotherapy compared to monoche-motherapy of docetaxel (OR=0.31, 95% CI [0.24, 0.41]; I2=0%, Z=8.23 (p<0.00001)), while a similar result could also be seen in PD-L1 inhibitor monotherapy group (OR=0.41, 95% CI [0.27, 0.64]; I2=59%, Z=3.92 [p<0.00001]). The opposite result can be seen when PD-1/PD-L1 inhibitor combined chemotherapy was compared to chemotherapy alone (OR=1.51, 95% CI [1.22, 1.87]; I2=0%, Z=3.77 [p<0.00001]). Similar incidence trend could also be seen in the meta-analysis of diarrhea for grade 1–2 and grade 3–5.

Conclusion:The incidence risk of diarrhea associated with PD-1/-PD-L1 inhibitor

mono-therapy was signiﬁcantly lower than that of docetaxel monotherapy group. However it was higher in PD-1/PD-L1 inhibitor combined with chemotherapy group compared with the chemotherapy alone group.

Keywords:diarrhea, PD-1/PD-L1, NSCLC, meta-analysis

Introduction

The programmed cell death 1 (PD-1) receptor has emerged as a dominant negative regulator of antitumor T-cell effector function when engaged by its ligand pro-grammed cell death ligand 1 (PD-L1), expressed on the surface of cells within a tumor.1Therapies that target the PD-1 receptor have shown unprecedented rates of durable clinical responses in patients with various cancer types. One mechanism by which cancer tissues limit the host immune response is via playing a key role in the maintenance of immunological tolerance to self-antigens, preventing autoim-mune disorders, while imautoim-mune checkpoint inhibitors (ICIs), including CTLA-4 and PD-1, were able to unleash T cells to fight cancer.2–6 Nivolumab showed its antitumor efficacy in non-small cell lung cancer (NSCLC) when it was first administered to a patient in October 2006 in a Phase I trial.7With the development of clinical research, more and more PD-1/PD-L1 inhibitors have been tried in Correspondence: Yuan Tian

Department of Radiotherapy Oncology, Shandong Provincial Qianfoshan Hospital, NO.16766, Jingshi Road, LiXia District,

Jinan, Shandong 250014, People’s Republic

China

Tel +86 139 5419 7326 Email [email protected]

Cancer Management and Research

Dovepress

open access to scientiﬁc and medical research

Open Access Full Text Article

Cancer Management and Research downloaded from https://www.dovepress.com/ by 118.70.13.36 on 20-Aug-2020

(2)

clinical trials for antilung cancer treatment and have shown good efﬁcacy,8–10especially for NSCLC.11–22

The toxic effects associated with PD-1/PD-L1 inhibi-tors may affect any organ and result from the activation of autoreactive T cells, thereby damaging the host tissue and even jeopardizing the patient’s life.11–23 Diarrhea is a common side effect of antitumor medications. Moderate or severe diarrhea can cause electrolyte imbal-ance in patients, further leading to the interruption of antitumor treatment.11–22 There was no signiﬁcant differ-ence in the inciddiffer-ence of diarrhea between PD-L1 inhibitors and placebo.8Compared with other antilung cancer treat-ment programs, there is no systematic analysis and report on the incidence risk of PD-1/PD-L1-related diarrhea. In order to clarify the incidence risk of diarrhea in the treat-ment of NSCLC with PD-1/PD-L1 inhibitors, we con-ducted a systematic review and meta-analysis.

Methods

This systematic review and meta-analysis was put into practice according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.24

Types of studies

We paid our attention mainly to randomized clinical trials, especially for Phase III clinical trials related to NSCLC. The criteria for the selected data: 1) PD-1/PD-L1-chemotherapy combinations compared with chemotherapy alone; 2) PD-1/ PD-L1 monotherapy compared with chemotherapy alone; 3) PD-1/PD-L1 plus antitumor therapy compared with PD-1/ PD-L1 inhibitor alone; 4) safety or toxicity was for evaluat-ing OR with 95% CI or other evaluable indicators such as RR, HR and so on; and 5) subjects were diagnosed with NSCLC but SCLC. Some studies without useful informa-tion, such as observational studies, editorials, commentaries, and review articles, were excluded.

Outcome and exposure of interest

The outcome of interest was incidence rate of PD-1/PD-L1 inhibitor-related diarrhea, including any grade of diarrhea. Diarrhea-related death was considered to be secondary outcome. Study data that reported measures of colitis, including colitis-related diarrhea, was not taken into account. We sought to examine the difference of incidence risk between PD-1/PD-L1 group and control group.

Search strategy and review method

Original articles, related to results of prospective clinical trials of PD1/PD-L1 inhibitor regimens for NSCLC patients, including monotherapy and combination therapy, were ver-iﬁed by a Pubmed search. The date range was limited from January 22, 2013, to January 22, 2019. Keywords were displayed just as the following:“PD1/PD-L1”,“nivolumab”, “BMS-963558”, “pembrolizumab/MK-3475”, “ atezolizu-mab/MPDL3280A”, “Avelumab”, “Durvalumab”, “lung cancer”,“NSCLC”,“SCLC”,“safety”, and“toxicity”.

We selected studies limited in human beings which were shown in full text, abstract, or poster form. Two members of our team were asked to identify their eligibility indepen-dently. References from review articles, editorials, and included studies were reviewed and cross-referenced to check completeness. If no useful data about toxicities of related drugs were found, we would try to get in touch with the corresponding author for more information, or it was precluded from the meta-analysis. The characteristics of enrolled studies, includingﬁrst author, year of publication, drug name, treatment regimen, study design, phase, number of patients, number of PD-1/PD-L1-related diarrhea and baseline demographic characteristics, were extracted. Risk of bias was evaluated by the Cochrane Collaboration tool for assessing the risk of bias in randomized trials.25

Assessment of heterogeneity and

statistical analysis

We took Newcastle–Ottawa scale, proposed by the Cochrane Collaboration,26 to check the quality of enrolled studies. Cochrane’s Q statistic and the I2 statistic were used for accessing the heterogeneity among studies just as proposed by Higgins et al,27while Harbord’s test was taken to check publication bias for all studies. OR value was considered to be a much more conservative and sensitive evaluation para-meter and might be more inclined to reveal a safety signal, as the method by which an OR is calculated provided a point estimate farther from unity than that provided by an HR. OR, and 95% CI would be calculated by random effect (RE) for the heterogeneity inherent in the data..28 p＜0.05 was deemed to be statistically signiﬁcant for all the results of meta-analysis. Statistical tests were all two-sided. Meta-analysis was performed using Review Manager, version 5.3.

Subgroup analysis

We divided all the data into three groups according to the degree of diarrhea, including all grades of diarrhea, grade 1–2,

Zhang et al Dovepress

(3)

and grade 3–5. We performed a number of subgroup analyses to assess the potential association between PD-1/PD-L1 inhi-bitor and chemotherapy in the incidence risk of diarrhea. We considered subgroups, including drug type (PD-1 or PD-L1 inhibitor), combination with chemotherapy (combination with chemotherapy or mono-therapy), and drug name.

Results

Literature search results

Among all the citations identified by our electronic and manual searches, 124 articles met the preliminary inclusion criteria. After screening and eligibility assessment of 57 articles, a total of 12 clinical trials involving 6,659 partici-pants were collected for the final meta-analysis.11–22 The study flow diagram was shown in Figure 1. All studies included a subgroup analysis that compared the intervention group with the control group, with an OR for incidence risk of diarrhea (Figure 3–5).

Characteristics of identi

ﬁ

ed trials

The main characteristics of all enrolled 12 studies are summarized in (Table 1). In total, 6,659 patients were included, of which 1,537 (23.1%) experienced diarrhea. Study types included 9 Phase III studies;11–16,18,21,22 1 Phase II/III;20 and 2 Phase II;17,19 All included studies had a control chemotherapy group. 5 of enrolled NSCLC studies received previous therapy before PD-1/ PD-L1 inhibitor.11,16,17,20–22 The drug used in 4 studies was PD-L1 inhibitor,11,12,16,17 while the PD-1 inhibitor was used in the other 8 studies.13–15,18–22 11 studies had the results of diarrhea for grade 3–5.11–16,18–22 In 6 studies, the drug used in the control group was mono-therapy with docetaxel.11,16,17,20–22 In the other 6 trials, the control treatment regimen was combined chemotherapy.12–15,18,19

Risk of bias

We used Newcastle–Ottawa scale to evaluate study quality and risk of bias in both comparison and noncomparison studies. All selected clinical studies were randomized and double-blinded. They were of high quality and at a low risk for selection bias. The risk of bias graph and summary are summarized in Figure 2A and B. The risk for performance and detection bias was too low to affect the outcome. Harbord's test statistic did not suggest obvious publication bias in funnel plot (Figure S1–S3).

OR of diarrhea for all grades and grade 1-2

The data of 12 studies involving NSCLC patients with PD-1/PD-L1 inhibitors data taken to make further analysis for diarrhea by grade.11–22Diarrhea of all grades was taken to be evaluated ﬁrst. All the enrolled studies were divided into 2 groups according to the treatment regimen. The treatment regimen of the experimental group was PD-1/ PD-L1 inhibitor monotherapy, named group A,11,13,16–

18,20–22

while the treatment regimen of the experimental group was PD-1/PD-L1 inhibitor combined with che-motherapy, named group B.12,14,15,19

The overall outcome of meta-analysis for group A is sum-marized at the bottom of (Figure 3A) (OR=0.45, 95% CI [0.31, 0.66], I2=79%, Z=4.17 [p<0.00001]),11,13,16–18,20–22while the opposite result of group B is displayed in Figure 3B (OR=1.51, 95% CI [1.22, 1.87], I2=0%, Z=3.77 [p=0.0002]).12,14,15,19 Funnel plot is provided in Figure S1. The OR of PD-1 sub-group was a little higher (OR=0.48, 95% CI [0.27, 0.86]) than that of PD-L1 subgroup (OR=0.41, 95% CI [0.27, 0.64]) (Figure 3A1). Due to the existence of heterogeneity, we

82 of records 42 of additional records identified through other sources (other databases

or documents covered in related meta-analysis

articles)

120 of records after duplicates removed

120 of records screened

57 of full-text articles

13 of studeis included in qualitative synthesis

12 of studeis included in quantitative synthesis

(meta-analysis) assessed for eligibility

63 records were excluded

42 articles without useful information were

1 of full-text articles were excluded because the control group was placebo rather than chemotherapy

or other anti-tumor therapies excluded. 2 of full-text articles were excluded for the tumor type was SCLC.

because they were observational studies,editorials, and review articles without useful information related to

PD-1/PD-L1 for lung cancer identified by pubmed

searching

Figure 1Studyﬂow diagram of inclusion.

Abbreviations:PD-1, programmed cell death 1; PD-L1, Programmed death-ligand 1; SCLC, small cell lung cancer; NSCLC, non-small cell lung cancer.

Dovepress Zhang et al

(4)

T able 1 Basic characteristics of enr olled studies Name of study Artic le sour ce Drug name PD-1 or PD-L1 T reatment regimen P atients Diarrhea of all grades Diarrhea of grade 3 – 5 Pr e vious therap y Phase T umor type Barlesi F et al, 11 2018 Lancet Oncol A velumab PD-L1 A velumab vs docetax el 758 80 5 Y es III Stage IIIB or IV or recurr ent NSCLC Socinski MA et al, 12 2018 N Engl J Med Atezolizumab PD-L1 Atezolizumab+BCP vs BCP 692 141 13 No III Metastatic nonsqua-mous NSCLC Hellmann MD et al, 13 2018A N Engl J Med Nivolu mab PD-1 Niv olumab vs niv olumab +ipilimumab 299 138 12 No III Stage IV or recurr ent NSCLC Hellmann MD et al, 13 2018B N Engl J Med Nivolu mab PD-1 Niv olumab vs chemotherap y 99 7 No Hellmann MD et al, 13 2018C N Engl J Med Ipilimumab CTLA-4 Niv olumab+ipilimumab vs chemotherap y 149 13 No Paz-Ar es L et al, 14 2018 N Engl J Med P embr olizumab PD-1 P embr olizumab+CP vs placebo +CP 558 148 17 No III Squamous NSCLC Gandhi L et al, 15 2018 N Engl J Med P embr olizumab PD-1 P embr olizumab+PC vs placebo+PC 616 168 27 No III Metastatic nonsqua-mous NSCLC Rittme ye r A et al, 16 2017 Lancet Atezolizumab PD-L1 Atezolizumab vs docetax el 1,187 235 15 Y es III Squamous or nonsqua-mous NSCLC F ehr enbacher L et al, 17 2016 Lancet Atezolizumab PD-L1 Atezolizumab vs doctax el 277 40 NA Y es II Advanced or metastatic NSCLC Reck M et al, 18 2016 N Engl J Med. P embr olizumab PD-1 P embr olizumab vs chemotherap y 304 42 8 No III Advanced NSCLC Langer CJ et al, 19 2016 Lancet Oncol P embr olizumab PD-1 P embr olizumab+PC vs placebo+PC 123 19 1 No II Stage IIIB or IV , non-squamous NSCLC Herbst RS et al, 20 2016A Lancet P embr olizumab PD-1 P embr olizumab 2 vs doctax el 991 80 9 Y es II/III NSCLC Herbst RS et al, 20 2016B Lancet P embr olizumab PD-1 P embr olizumab 10 vs doctax el 78 7 Y es II/III Brahmer J et al, 21 2015 N Engl J Med Nivolu mab PD-1 Niv olumab vs docetax el 272 36 3 Y es III Advanced squamous NSCLC Borghaei H et al, 22 2015 N Engl J Med Nivolu mab PD-1 Niv olumab vs docetax el 582 84 5 Y es III Nonsquamous NSCLC Abbre viations: NSCLC , non – small cell lung cancer ; PD-1, pr ogrammed cell death 1; PD-L1, pr ogrammed cell death ligand 1; CTLA-4, cytoto xic T-lymphocyte-associated pr otein 4; SCLC , small cell lung cancer ; CP , carboplatin plus paclitax el; NA, No A vailable; PC , pemetr ex ed and a platinum-based drug; chemotherap y, carboplatin plus pemetr ex ed, cisplatin plus pemetr ex ed, ca rboplatin plus gemcitabine, cisplatin plus gemcitabine, or carboplatin plus paclitax el.

(5)

conducted a further subgroup analysis (I2=79%), three differ-ent subgroup analysis methods were used for group A (Figure 3A1-A3).11,13,16–18,20–22

Regardless of whether the experimental group was single-agent PD-1 or PD-L1 inhibitor, the incidence risk of subgroup meta-analysis for diarrhea was signiﬁcantly lower than that of the control group related to docetaxel (OR=0.31, 95% CI [0.24, 0.41], I2=0%; OR=0.41, 95% CI [0.27, 0.64], I2=59%; Figure 3A2). The results of the two subgroup meta-analysis were statistically signiﬁcant. Heterogeneity was only found in the subgroup of PD-L1 inhibitor (I2=59%, Figure 3A2).11,16,17However, the simi-lar trend could not be seen when PD-1/PD-L1 inhibitor was compared with combination chemotherapy (OR=1.16, 95% CI [0.81, 1.64], I2=0%; Figure 3A2).13,18

When another subgroup analysis was performed based on the speciﬁc drug name of the PD-1/PD-L1 inhibitor, compared

with the control group, it was found that except for nivolumab, the incidence of diarrhea in the PD-1/PD-L1 group was lower than that of the control group, and the results were statistically signiﬁcant (Figure 3A3). In the subgroup of nivolumab (I2=91%) and atezolizumab (I2=70%), there was signiﬁcant heterogeneity in the analysis results (Figure 3A3) .

After subgroup analysis of group B data, we found that PD-1 inhibitor combined with chemotherapy signiﬁcantly increased the incidence risk of diarrhea, and the results were statistically signiﬁcant (OR=1.55, 95% CI [1.19, 2.02], I2=0%, Z=3.24 [p=0.001]; Figure 3B).14,15,19

Then, the incidence rates of grade 1–2 diarrhea were taken into account. The incidence risk of diarrhea for all grades (Figure 3) and grade 1–2 (Figure 4) were almost in the same trend. The details of the meta-analysis for grade 1–2 related to group A are summarized in (Figure 4A1–3) (OR=0.50, 95% CI [0.35, 0.71], I2=75%; Z=3.86

Figure 2(A): Risk of bias summary: review authors’judgments about each risk of bias item for each included study. (B): Risk of bias graph: review authors’judgments about each risk of bias item presented as percentages across all included studies.

(6)

[p=0.0001]), while the opposite result of group B is displayed in Figure 4B (OR=1.39, 95% CI [1.11, 1.74], I2=0%, Z=2.87 [p=0.004]). The OR of PD-1 subgroup was a little lower (OR=0.49, 95% CI [0.29, 0.84]) than that of PD-L1 subgroup (OR=0.52, 95% CI [0.35, 0.75]) in (Figure 4A1). Regardless of whether the experimental group was single-agent PD-1 or PD-L1 inhibitor, the incidence risk of subgroup meta-analysis for diarrhea related to grade 1–2 was significantly lower than that of the control group related to docetaxel (OR=0.34, 95% CI [0.25, 0.45], I2=0%; OR=0.52, 95% CI [0.35, 0.75], I2=44%; Figure 4A2). When the subgroup analysis was per-formed based on the specific drug name of the PD-1/PD-L1 inhibitor, compared with the control group, it was found that except for nivolumab, the incidence rate of diarrhea in the PD-1/PD-L1 group was lower than that of the control group, and the results were statistically significant (Figure 4A3). Funnel plot is provided in Figure S2.

OR of diarrhea for grade 3-5

The incidence of diarrhea related to grade 3–5 was reported in 11 studies.11–16,18–22The results of the meta-analysis were gathered at the bottom of Figure 5. Different from the above results of all grades and grade 1–2, no statistical signiﬁcance

could be found when we put group A (PD-1/PD-L1 VS chemotherapy) into meta-analysis for the incidence risk of diarrhea (OR=0.45, 95% CI[0.20, 1.04], I2=39%, Z=1.88 [p=0.06]; Figure 5A).11,13,16,18,20–22 The incidence risk of diarrhea in the PD-1/PD-L1 combined chemotherapy group was higher than that in the control group, and the results were statistically signiﬁcant (OR=2.07, 95% CI [1.12, 3.82], I2=1%, Z=2.32 [p=0.02]; Figure 5B).12,14,15,19Funnel plot is provided in Figure S3.

As the existence of heterogeneity (I2=39%) in group A, stratified subgroup analysis was taken into account (Figure 5A1-A3).11,13,16,18,20–22 When the PD-1 inhibitor was compared with docetaxel (Figure 5A2), the incidence of diarrhea for grade 3–5 was significantly lower than that of the control group (OR=0.26, 95% CI [0.09, 0.73], I2=0%, Z=2.55 [p=0.01]).20–22 The same trend of results can also be seen when the PD-L1 inhibitor was compared to docetaxel (OR=0.28, 95% CI [0.10, 0.82], I2=0%, Z=2.33 [p=0.02]; Figure 5A2).11,16 When we mixed the single-agent docetaxel and the combination chemotherapy together as a control group, we did not find statistically significant results in the subgroup analysis (PD-1 vs chemotherapy) (OR=0.55, 95% CI [0.19, 1.59], I2=46%, Study or subgroup

A1

A2

A3

B

2.7.1 PD-1 VS chemotherapy

2.7.2 PD-1 VS chemotherapy

1.1.1 PD-1+Chemotherapy VS chemotherapy

1.1.2 PD-1+Chemotherapy VS chemotherapy 2.1.1 PD-1 VS Docetaxel

2.1.2 PD-1 VS Docetaxel

2.1.3 PD-1/PD-L1 VS Combined chemotherapy 22 22 339 343 131 1100 287 62 26 268 11.4% 9.0% 11.5% 11.4% 43.3% 129 309 309 1015 200 78 56 56 10 24 Borghaei H et al 2015 Brahmer J et al 2015 Reck M et al 2016

Reck M et al 2016

Reck M et al 2016 Herbst RS et al 2016A

Herbst RS et al 2016A Herbst RS et al 2016B

Herbst RS et al 2016B Hellmann MD et al 2018B

Hellmann MD et al 2018B 22 44

545 391 154 20 150

570 720 75 66 272 2789 502 2813 100% 55 144 10 94 24 128 227 272 272 2789 502 2813 100.0% 1144 393 609 142 30 135

578 365 1078 141 56 275 Events Total Events Total WeightPD-1/PD-L1 Chemotherapy

Chemotherapy

Odds ratio Odds ratio

Year

M-H, Random, 95% CI M-H, Random, 95% CI

Study or subgroup Events Total Events Total Weight

PD-1/PD-L1 Chemotherapy Odds ratio Odds ratio Year

Study or subgroupPD-1/PD-L1+ChemotherapyEvents Total Events Total Weight Odds ratio

Odds ratio M-H, Random, 95% CI M-H, Random, 95% CI Study or subgroup Events Total Events Total WeightPD-1/PD-L1 Chemotherapy M-H, Random, 95% CIOdds ratio Year M-H, Random, 95% CIOdds ratio

Chemotherapy SubTotal (95% CI)

SubTotal (95% CI)

Total (95% CI) SubTotal (95% CI) Fehrenbacher L et al 2016 Rittmeyer A et al 2017

Fehrenbacher L et al 2016 Rittmeyer A et al 2017 Barlesi F et al 2018

Total events

Total (95% CI) SubTotal (95% CI) Total events

Total events

SubTotal (95% CI) Socinski MA et al 2018

Total (95% CI) Total events

Total events

Total events Total events

Langer CJ et al 2016 Gandhi L et al 2018 Paz-Ares L et al 2018 Hellmann MD et al 2018B SubTotal (95% CI) Total events

Total events

Borghaei H et al 2015 Brahmer J et al 2015

2.2.1 Nivalumab VS Chemotherapy

2.2.2 Pembrolizumab VS Chemotherapy

2.2.4 Avelumab VS Docetaxel Barlesi F et al 2018 24

393 393

393 393 60 394

394 60 56 56 365 365 12.9% 4.5% 29.2% 32.2% 65.9% 34.1% 34.1% 12.9% 24 404 2655 632 12 125 83 220 81 81 301 1135 175 938 100.0% 59 7 43 65 115 62 202 280 544 405 278 742 2663 100.00% Borghaei H et al 2015

Brahmer J et al 2015 10 22 44

131 26 129 268 570 967 62 55 287 391 809 76 10 94 104 751 142 609 30 141 171 713 578 135 20 24 22 339343

702 262 56 56 150 0 309 309 618 143

Herbst RS et al 2016A Herbst RS et al 2016B SubTotal (95% CI) Total events

Test for overall effect: Z=2.47 (P=0.01)

Test for overall effect: Z=3.92 (P<0.0001)

Test for overall effect: Z=3.77 (P=0.0002) Test for overall effect: Z=4.17 (P<0.0001)

Test for overall effect: Z=8.23 (P<0.0001) Heterogeneity: Tau2

=0.43; Chi2

=33.26, df=5 (P<0.00001); I2_=85%

Heterogeneity: Tau2 =0.00; Chi2

=0.08, df=1 (P=0.77); I2_=0%

=3.34, df=1 (P=0.07); I2_=70%

Heterogeneity: Not applicable

Heterogeneity: Not applicable Heterogeneity: Tau2

=0.22; Chi2

=32.47, df=7(P<0.0001); I2_=78%

=0.60, df=2 (P=0.74); I2_=0%

=0.68, df=3 (P=0.88); I2_=0% Heterogeneity: Tau2

=0.75; Chi2

=21.18, df=2 (P<0.00001); I2_=91%

=4.86, df=2 (P=0.09); I2_=59%

Total (95% CI) Fehrenbacher L et al 2016 Rittmeyer A et al 2017 Barlesi F et al 2018 Total events

Total events

Test for overall effect: Z=0.81 (P=0.42) Heterogeneity: Tau2

=0.09; Chi2

=4.86, df=2 (P=0.09); I2 =59%

=0.05, df=1 (P=0.82); I2 =0% Heterogeneity: Tau2

=0.24; Chi2

=38.36, df=8 (P<0.00001); I2 =79%

=0.38, df=3 (P=0.95); I2_=0% Test for subgroup differences: Chi2

=0.18, df=1 (P=0.68). I2_=0%

Test for overall effect: Z=4.17 (P<0.0001) Heterogeneity: Tau2

=0.24; Chi2

=38.36, df=8 (P<0.00001); I2 =79% Test for subgroup differences: Chi2

=33.53, df=2 (P=0.00001). I2_=94.0%

Test for subgroup differences: Chi2

=0.77, df=3 (P=0.86). I2_=0%

=0.09, df=1 (P=0.77). I2 =0% 22 10 22 24 22 44 287 131 154 339 343 391 62 26 20 56 56 55 268 129 150 309 309 570 1735 1645 11.4% 2015 2016 2016A 2016B 2018B 2015 2015 2018B 2015 2016A 2016B 2016 2016 2017 2018 2018 2018 2018 2016 2015 2016A 2016B

0.1 0.2 0.5 1 2 5 10 2015

2017 2018

0.05 0.2 1 5 20 2016

0.28 [0.16, 0.46] 0.33 [0.15, 0.71] 1.08 [0.56, 2.08] 0.34 [0.21, 0.57] 0.31 [0.18, 0.52] 1.19 [0.78, 1.81] 0.48 [0.27, 0.86]

0.27 [0.12, 0.57] 0.57 [0.42, 0.76] 0.36 [0.22, 0.59] 0.41 [0.27, 0.64]

0.45 [0.31, 0.66]

0.28 [0.16, 0.46] 0.33 [0.15, 0.71] 0.34 [0.21, 0.57] 0.31 [0.18, 0.52] 0.31 [0.24, 0.41] 9.0% 10.1% 11.5% 11.4% 12.4% 65.8%

0.33 [0.15, 0.71] 0.28 [0.16, 0.46] 1.19 [0.78, 1.81] 0.49 [0.17, 1.37]

0.34 [0.21, 0.57] 0.31 [0.18, 0.52] 0.33 [0.23, 0.47]

0.27 [0.12, 0.57] 0.57 [0.42, 0.76] 0.42 [0.20, 0.87]

0.36 [0.22, 0.59] 0.36 [0.22, 0.59]

2.01 [0.73, 5.51] 1.65 [0.11, 2.46] 1.41 [0.96, 2.05] 1.55 [1.19, 2.02]

1.45 [1.00, 2.09] 1.45 [1.00, 2.09]

1.51 [1.22,1.87] 0.41 [0.28, 0.60] 9.8% 12.7% 13.9% 36.4% 12.9% Not estimable 12.7% 25.5% 25.2% 10.0% 15.2% 9.1% 13.5% 34.2% 11.6% 10 94 24 128 227 1144 393 609 142 30 135

578 365 1078 141

56 20172018 2016

2016 2018B 0.27 [0.12, 0.57] 0.57 [0.42, 0.76] 0.36 [0.22, 0.59] 0.41 [0.27, 0.64]

1.08 [0.56, 2.08] 1.19 [0.78, 1.81] 1.16 [0.81, 1.64]

0.45 [0.31, 0.66] 9.1% 13.5% 34.2% 10.1% 12.4% 22.5% 11.6% PD-1/PD-L1 Chemotherapy 0.01 0.1 1 10 100

PD-1/PD-L1

0.05 0.2 1 5 20 PD-1/PD-L1+Chemotherapy Chemotherapy

PD-1/PD-L1

Chemotherapy

Figure 3Forest plots of the incidence risk for all-grade diarrhea. (A1): OR of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 vs chemotherapy). (A2): OR of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 vs docetaxel/combined chemotherapy), the data included were assigned to the corresponding subgroup according to control group (docetaxel or combined chemotherapy). (A3): OR of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 vs chemotherapy), the data included were assigned to the corresponding subgroup according to the name of PD-1/PD-L1 inhibitor and the control group. (B): OR of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 + chemotherapy vs chemotherapy).

Abbreviation:RE, random effect; PD-1, programmed cell death 1; PD-L1, programmed cell death ligand 1;.

(7)

Z=1.11 [p=0.27]; Figure 5A1).13,18,20–22 Moreover, mod-erate heterogeneity results had emerged (I2=46%), so the heterogeneity was considered to be derived from the two newly included data.13,18 When we performed a subgroup analysis based on the speciﬁc drug name of PD-1/PD-L1 inhibitor, no statistically signiﬁcant analysis results could be found (Figure 5A3).

Discussion

Therapies that target the PD-1 receptor have shown unpre-cedented rates of durable clinical responses in patients with various cancer types. It was reported that PD-1 or PD-L1 inhibitor improved overall survival compared with standard-of-care chemotherapy in patients with metastatic NSCLC, and several of them have received regulatory approvals.29–31 With the development of clinical research, more and more PD-1/PD-L1 inhibitors have been tried in clinical trials for antilung cancer treatment and have shown good efﬁcacy,8–10 especially for NSCLC.11–22 Similar to other antitumor drugs, good antitumor efﬁcacy was along with many therapeutic side effects, especially for combination with chemotherapy.

Diarrhea is a common side effect of antitumor medica-tions, such as docetaxel, irinotecan, and topotecan. Moderate or severe diarrhea can cause electrolyte imbal-ance in patients, further leading to the interruption of antitumor treatment.11–22 Docetaxel was used to be taken as a better choice for second-line treatment for advanced NSCLC based on improvement in survival versus best supportive care.32,33 Other treatment regimens, such as pemetrexed and erlotinib, did not show better efﬁcacy than docetaxel in patients with NSCLC.34,35 In order to clarify the incidence of diarrhea in the treatment of NSCLC with PD-1/PD-L1 inhibitors, we conducted a systematic review and meta-analysis.

Through our meta-analysis of the included clinical trials, the incidence of diarrhea for all grades was signiﬁcantly lower in the PD-1/PD-L1 inhibitor monotherapy group than in the chemotherapy group, while it was higher in PD-1/PD-L1 combined with chemotherapy compared with chemotherapy group alone (Figure 3).11–22Similar incidence trend could also be seen in meta-analysis of diarrhea for grade 1–2 (Figure

4).11–16,18–22Due to the existence of heterogeneity, we

con-ducted a further subgroup analysis (I2=79%), three different Study or subgroup

A1

A2

A3

B

Events Total Events Total Weight

PD-1/PD-L1 Chemotherapy Odds ratio Odds ratio M-H, Random, 95% CI M-H, Random, 95% CI

Study or subgroup Events Total EventsTotal Weight

PD-1/PD-L1 Chemotherapy Odds ratio Odds ratio M-H, Random, 95% CI Year

2015 2015 2015 2018B 2016 2018 2017 2016A 2016B 2015 2017 2018 2016 2018B 2016A 2016B

M-H, Random, 95% CI

Study or subgroup Events Total EventsTotal Weight

PD-1/PD-L1+Chemotherapy Chemotherapy Odds ratio Odds ratio M-H, Random, 95% CI Year M-H, Random, 95% CI Study or subgroup Events Total Events Total Weight

PD-1/PD-L1 Chemotherapy Odds ratio Year

Odds ratio M-H, Random, 95% CI M-H, Random, 95% CI 2.8.1 PD-1 VS chemotherapy

2.8.2 PD-L1 VS chemotherapy

2.3.1 PD-1 VS Docetaxel 10 339 343 1100 74 90 24 609 393 1002 130 578 365 94312.9% 15.6% 28.5% 51 181 16 41 57 154 391 545 18 51 150570

720 10.3% 13.9% 24.2% 69 2647 245 430 2678 100.0% 180

23 129 9.5% 12.5% 12.6% 12.6% 47.3% 268 309 309 1015 59 49 49 287 131 20 22 22

Borghaei H et al 2015 20 287 Borghaei H et al 2015 339 339 131 154 391 1641 131 249 49 59 268 309 309 129 150 570 1735 24 90

393 51 365 578 943

0.10.2 0.5 1 2 5 10 181 114 2643 245 430 2678 100.0% 130 609 1002 71.5% 12.9% 15.6% 28.5% 13.9% 10.3% 9.4% 12.6% 12.6% 12.5% 49 23 18 51 22 22 10 16 41 Brahmer J et al 2015

Brahmer J et al 2015

Reck M et al 2016

Reck M et al 2016 Herbst RS et al 2016A

Herbst RS et al 2016B

Hellmann MD et al 2018B

Total events

SubTotal (95% CI)

Total (95% CI) SubTotal (95% CI) Total events

Total events

Total events Test for overall effect: Z=2.63 (P=0.09)

Test for overall effect: Z=3.86 (P=0.0001) Test for overall effect: Z=3.40 (P=0.0007) Test for overall effect: Z=3.86 (P=0.0001)

=0.34; Chi2

=25.82, df=5 (P<0.0001); I2_=81%

=19.81, df=2 (P<0.0001); I2_=90%

=4.20, df=2 (P=0.12); I2_=52%

=27.76, df=7(P=0.0002); I2 =75% Heterogeneity: Tau2

=0.19; Chi2

=27.63, df=7 (P=0.0003); I2 =75% Heterogeneity: Tau2

=0.04; Chi2

=1.78, df=1 (P=0.18); I2_=44%

0.27 [0.15, 0.45] 0.37 [0.22, 0.63] 0.37 [0.22, 0.63] 0.38 [0.17, 0.84] 0.85 [0.42, 1.74] 1.19 [0.77, 1.84] 0.49 [0.29, 0.84]

0.27 [0.15, 0.45] 0.38 [0.17, 0.84] 1.19 [0.77, 1.84] 0.50 [0.18, 1.41]

0.85 [0.42, 1.74] 0.37 [0.22, 0.63] 0.36 [0.21, 0.62] 0.46 [0.28, 0.75]

0.40 [0.24, 0.67] 0.40 [0.24, 0.67]

0.60 [0.44, 0.80] 0.60 [0.44, 0.80]

0.50 [0.35, 0.71]

1.54 [1.01, 2.35] 2.38 [0.83, 6.84] 1.31 [0.88, 1.94] 1.46 [1.11, 1.93]

1.26 [0.86, 1.84] 1.26 [0.86, 1.84]

1.39 [1.11, 1.74] 0.40 [0.24, 0.67]

0.60 [0.44, 0.80] 0.52 [0.35, 0.75]

0.50 [0.35, 0.71]

0.38 [0.17, 0.84] 0.27 [0.15, 0.45] 0.37 [0.22, 0.63] 0.36 [0.21, 0.62] 0.34 [0.25, 0.45]

0.60 [0.44, 0.80] 0.40 [0.24, 0.67] 0.52 [0.35, 0.75]

0.85 [0.42, 1.74] 1.19 [0.77, 1.84] 1.09 [0.75, 1.58]

0.50 [0.35, 1.71] Rittmeyer A et al 2017

Barlesi F et al 2018

Barlesi F et al 2018 SubTotal (95% CI)

Total (95% CI) Total events

Total events

=0.02 df=1 (P=0.89). I2_=0% PD-1/PD-L1Chemotherapy

0.01 0.1 1 10 100 Chemotherapy PD-1/PD-L1

0.1 0.2 0.5 1 2 5 10 Chemotherapy PD-1/PD-L1

0.1 0.2 0.5 1 2 5 10 Chemotherapy PD-1/PD-L1+Chemotherapy Borghaei H et al 2015

Brahmer J et al 2015 Herbst RS et al 2016A Herbst RS et al 2016B SubTotal (95% CI)

Total (95% CI) SubTotal (95% CI) Total events

Total events 114

Total events Total events

=0.00; Chi2

=1.04, df=3 (P=0.79); I2_=0%

=1.78, df=1 (P=0.18); I2 =44%

=0.63, df=1 (P=0.43); I2 =0%

=27.76, df=7 (P=0.0002); I2_=75% Rittmeyer A et al 2017

2.3.2 PD-L1 VS Docetaxel

2.3.3 PD-1 VS chemotherapy Reck M et al 2016 Hellmann MD et al 2018B

=24.25, df=2 (P<0.00001). I2_=91.8%

2.4.1 Nivalumab(PD-1) VS Chemotherapy 20 10 41 71 16 22 22 18 49 49 150 309 309 768 154 339 343 836 60 116 24 24 90 90 2647 430 2678 100.0% 245 609 609 130 578

57815.6%15.6% 130 393 393 51 51 365 12.9% 12.9% 365 133 287 131 391 809 59 268 9.5% 12.5% 13.9% 35.9% 10.3% 12.6% 12.6% 35.6% 129 570 967 23 51

2.4.2 Pembrolizumab(PD-1) VS Chemotherapy Herbst RS et al 2016A

Herbst RS et al 2016B

2.4.3 Avelumab(PD-L1) VS Docetaxel

2.4.4 Atezolizumab(PD-L1) VS Docetaxel Barlesi F et al 2018

Heterogeneity: Not applicable

Heterogeneity: Not applicable Rittmeyer A et al 2017

=2.11, df=3 (P=0.55). I2_=0%

=0.41, df=1 (P=0.52). I2 =0% 1.2.1 PD-1+Chemotherapy VS chemotherapy

SubTotal (95% CI) Total events Gandhi L et al 2018 104

12 72 188

70 70

393 58 394 34.7% 34.7% 100.0% 394 938 58 160 1135 258 393 102 405 37 6 59 202 62 280 544 65.3% 32.4%4.5% 28.3% 59 278 742 Paz-Ares L et al 2018

=0.00; Chi2

=1.19, df=2 (P=0.55); I2_=0%

=0.00; Chi2

=1.60, df=3 (P=0.66); I2 =0% Langer CJ et al 2016

1.1.2 PD-1+Chemotherapy VS chemotherapy SubTotal (95% CI)

Total (95% CI) Socinski MA et al 2018 Total events

Total events

Test for overall effect: Z=1.17 (P=0.24) Heterogeneity: Not applicable

Figure 4Forest plots of the incidence risk for grade 1–2 diarrhea. (A1): OR of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy) (A2): OR of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 vs docetaxel/combined chemotherapy), the data included were assigned to the corresponding subgroup according to control group (docetaxel or combined chemotherapy). (A3): OR of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy), the data included were assigned to the corresponding subgroup according to the name of PD-1/PD-L1 inhibitor and the control group. (B): OR of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 + chemotherapy VS chemotherapy).

Abbreviation:PD-1, programmed cell death 1; PD-L1, programmed cell death ligand 1; RE, random effect.

(8)

subgroup analysis methods were used for group A (Figure 3A1-A3).11,13,16–18,20–22When we mixed the single-agent doc-etaxel and the combination chemotherapy together as a control group (Figures 3A1, 4A1 and 5A1),11,13,16–18,20–22 heteroge-neity results emerged. So the heterogeheteroge-neity was considered to be derived from the two newly included data (Reck et al, 2016 and Hellmann et al, 2018B).13,18

In order to eliminate the influence of heterogeneity on the analysis results, we conducted a comprehensive analy-sis of the results. Regardless of the degree of diarrhea (all grade, grade 1–2 and grade 3–5), the incidence risk of diarrhea in PD-1 monotherapy group was significantly lower than that of docetaxel monotherapy group, and the difference was of statistical significance. There is no obvious heterogeneity and publication bias among the included data (Figures 3A2, 4A2 and 5A2, Figures 1– 3).20–22 When PD-1/PD-L1 inhibitor was combined with chemotherapy, the incidence risk of diarrhea was signifi -cantly increased.12,14,15,19

Although diarrhea was reported in 20% of the patients, some of grade 3–5 treatment-related diarrhea was reported.8–22,36–39 They could be managed with treatment of the symptoms, and the patient recovered promptly with-out glucocorticoid treatment.36–39 Although diarrhea was

reported as a common toxic side effect associated with PD-1/PD-L1 inhibitor appeared in clinical trials, no standard systematic treatment regimen for it was proposed.8–22,36–39 Therefore, for the fully understanding of PD-1/PD-L1-related diarrhea, more clinical trials and mechanism research are needed.

In a word, the incidence risk of diarrhea associated with single-agent PD-1/-PD-L1 inhibitors was signiﬁcantly lower than that of the single-agent docetaxel group, while it was higher in PD-1/PD-L1 combined with chemotherapy group compared with chemotherapy group alone.

Conclusions

The incidence risk of diarrhea associated with single-agent PD-1/-PD-L1 inhibitors was signiﬁcantly lower than that of docetaxel monotherapy group, while it was higher in PD-1/PD-L1 inhibitor combined with chemotherapy group compared with chemotherapy group.

Statement of Ethics

This study belongs to the type of data analysis and rear-rangement and does not involve human- or animal-related ethical issues.

Study or subgroup

A1

A2

A3

B

Events Total Events Total Weight

Study or subgroup Events Total Events Total Weight

Study or subgroupPD-1/PD-L1+Chemotherapy ChemotherapyEvents Total Events Total Weight

Odds ratio Odds ratio

Year

M-H, Random, 95% CI M-H, Random, 95% CI 2015 2016 2016A 2016B 2018B 2017 2018 2015 2015 2017 2018 2016A 2016B 2018B2016 2018B 2015 2015 2016B 2018 2017 2016 2018 2018 2018 2016 2016A

0.001 0.1 1 10 1000 2015

2.9.1 PD-1 VS chemotherapy Borghaei H et al 2015 2 287

131 154

3 3 268 13.2%6.4%

14.9% 15.3% 13.2% 6.4% 16.2% 35.7% 14.9% 15.3% 6.8% 37.0% 3.6% 36.5% 43.5% 83.6% 6.8% 16.2% 72.7% 129 150 309 309 370 1735 26 2 7 7 4 339 343 391 1645 0 6 2 0 3 6 2 0 0 4 4 609 11 578 578 20.6% 20.6% 11 609 2647 17 0 11 21 11 393 393 2 2 394 394 16.4% 16.4% 11 43 1135 938 15 100.0% 59 278 405 742 62 280 202 544 1 6 6 13 32 42 2678 100.0% 2 7 7 8 16 287 131 391 3 3 4 809 154 339 343 836 393 393

5 365 6.7% 6.7% 365 5 150 309 309 768 5 10 268 129 570 967 2 0 2 2 0 131 287 339 343 4 4 0 6 3 154 391 545 2 4 6 720 150 570 609 393 11 5 578365

943 16 20.6% 27.3%6.7% 14.9% 16.2% 31.1% 100.0% 2678 42 2647 17 1002 4 1100 20

3 129 6.4% 13.2% 15.3% 6.8% 41.6% 268 309 309 1015 3 7 7 0 3 13 Brahmer J et al 2015 Reck M et al 2016 Herbst RS et al 2016A Herbst RS et al 2016B Hellmann MD et al 2018B SubTotal (95% CI) Total events

=0.78; Chi2

=9.24, df=5 (P=0.10); I2 =46%

0.62 [0.10, 3.74] 0.14 [0.01, 2.69] 3.00 [0.60, 15.11] 0.26 [0.05, 1.24] 0.06 [0.00, 10.3] 1.09 [0.24, 4.92] 0.55 [0.19, 1.59]

3.00 [0.60, 15.11] 0.26 [0.05, 1.24] 0.06 [0.00, 1.03] 0.44 [0.05, 3.84]

0.45 [0.20, 1.04] 0.08 [0.00, 1.51] 0.08 [0.00, 1.51]

0.34 [0.11, 1.08] 0.34 [0.11, 1.08]

0.34 [0.01, 8.63] 1.88 [0.69, 5.16] 1.79 [0.71, 4.50] 1.70 [0.87, 3.32]

5.64 [1.24, 25.63] 5.64 [1.24, 25.63] 1.70 [0.87, 3.32] 0.34 [0.11, 1.08]

0.08 [0.00, 1.51] 0.28 [0.10, 0.82]

0.62 [0.10, 3.74] 0.14 [0.01, 2.69] 1.09 [0.24, 4.92] 0.68 [0.23, 1.99]

0.14 [0.01, 2.69] 0.62 [0.10, 3.74] 0.26 [0.05, 1.24] 0.06 [0.00, 1.03] 0.26 [0.09, 0.73]

0.34 [0.11, 1.08] 0.08 [0.00, 1.51] 0.28 [0.10, 0.82]

3.00 [0.60, 15.11]

2.07 [1.12, 3.82] 1.09 [0.24, 4.92]

1.75 [0.58, 5.25]

0.45 [0.20, 1.04] 0.45 [0.20, 1.04] 2.9.2 PD-L1 VS chemotherapy

4 609

2647 2678 100.0% 393 1002 578 365 943 20.6% 6.7% 27.3% 11 5 4 16 0 SubTotal (95% CI) Rittmeyer A et al 2017 Barlesi F et al 2018 Total events

=0.00; Chi2

=0.82, df=1 (P=0.37); I2_=0%

Total (95% CI)

Total events 17 42 Test for overall effect: Z=1.88 (P=0.06) Heterogeneity: Tau2

=0.53; Chi2

=11.48, df=7 (P=0.12); I2_=39% Test for subgroup differences: Chi2

=0.75, df=1 (P=0.39). I2_=0% PD-1/PD-L1Chemotherapy

0.001 0.1 1 10 1000 Chemotherapy PD-1/PD-L1+Chemotherapy 0.001 0.1 1 10 1000

Chemotherapy PD-1/PD-L1

0.001 0.1 1 10 1000 Chemotherapy PD-1/PD-L1 2.5.1 PD-1 VS Docetaxel

Borghaei H et al 2015 Brahmer J et al 2015 Herbst RS et al 2016A Herbst RS et al 2016B SubTotal (95% CI) Total events

=0.00; Chi2

=2.24, df=3 (P=0.52); I2_=0%

2.5.2 PD-L1 VS Docetaxel

Total (95% CI) Rittmeyer A et al 2017 Barlesi F et al 2018 Total events

Total events

=0.00; Chi2

=0.82, df=1 (P=0.37); I2 =0%

=0.81, df=1 (P=0.37); I2_=0%

=11.48, df=7 (P=0.12); I2_=39% 2.5.3 PD-1 VS chemotherapy

Reck M et al 2016 Hellmann MD et al 2018B Total events

=7.57, df=2 (P=0.02). I2 =73.6%

Borghaei H et al 2015 Brahmer J et al 2015 Hellmann MD et al 2018B SubTotal (95% CI) Total events

=0.00; Chi2

=1.55, df=2 (P=0.46); I2_=0% 2.6.1 Nivalumab(PD-1) VS Chemotherapy

2.6.3 Avelumab(PD-L1) VS Chemotherapy

2.6.4 Atezolizumab(PD-L1) VS Chemotherapy Reck M et al 2016

Herbst RS et al 2016A Herbst RS et al 2016B SubTotal (95% CI) Total events

=2.66; Chi2

=7.62, df=2 (P=0.02); I2_=74%

=11.48, df=7 (P=0.12); I2 =39% SubTotal (95% CI)

Total (95% CI) Total events Barlesi F et al 2018

Heterogeneity: Not applicable Test for overall effect: Z=1.68 (P=0.09)

Test for overall effect: Z=2.24 (P=0.02) SubTotal (95% CI)

SubTotal (95% CI) Rittmeyer A et al 2017 Total events

Total events

Total events Heterogeneity: Not applicable

=2.09, df=3 (P=0.55). I2 =0%

=2.02, df=1 (P=0.16). I2_=50.4% Total (95% CI)

Total events

Total events Langer CJ et al 2016 Gandhi L et al 2018 Paz-Ares L et al 2018

Test for overall effect: Z=2.32 (P=0.0.02) Heterogeneity: Tau2

=0.00; Chi2

=99, df=2 (P=61); I2_=0%

=3.02, df=3 (P=0.39); I2_=1% 1.3.1 PD-1+Chemotherapy VS chemotherapy

Heterogeneity: Not applicable Socinski MA et al 2018 1.3.2 PD-1+Chemotherapy

Figure 5Forest plots of the incidence risk for grade 3–5 diarrhea. (A1): OR of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy) (A2): OR of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 vs docetaxel/combined chemotherapy), the data included were assigned to the corresponding subgroup according to control group (docetaxel or combined chemotherapy). (A3): OR of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy), the data included were assigned to the corresponding subgroup according to the name of PD-1/PD-L1 inhibitor and the control group. (B): OR of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 + chemotherapy vs chemotherapy).

Abbreviation:PD-1, programmed cell death 1; PD-L1, programmed cell death ligand 1; RE, random effect.

(9)

Abbreviation list

PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; PD-1, programmed cell death-1; PD-L1, programmed cell death ligand 1; SCLC, small cell cancer; NSCLC, non-small cell lung cancer; RR, risk ratio; RE, ran-dom effect; FE,ﬁxed effect.

Acknowledgments

This study was funded by the Natural Science Foundation of Shandong Province (ZR2015HL078), which were in charge by Yuan Tian.

Author contributions

All authors contributed to data analysis, drafting and revis-ing the article, gave ﬁnal approval of the version to be published, and agree to be accountable for all aspects of the work.

Disclosure

The authors have no conﬂicts of interest to report in relation to this paper.

References

1. Ribas A, Wolchok JD. Cancer immunotherapy using checkpoint blockade.

Science. 2018;359(6382):1350–1355. doi:10.1126/science.aar4060 2. Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces

responses by inhibiting adaptive immune resistance. Nature. 2014;515(7528):568–571. doi:10.1038/nature13954

3. Waterhouse P, Penninger JM, Timms E, et al. Lymphoproliferative disorders with early lethality in mice deﬁcient in Ctla-4. Science. 1995;270(5238):985–988.

4. Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA, Sharpe AH. Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4.Immunity. 1995;3(5):541–547.

5. Nishimura H, Nose M, Hiai H, Minato N, Honjo T. Development of lupus-like autoimmune diseases by disruption of the PD-1 gene encoding an ITIM motif-carrying immunoreceptor.Immunity. 1999;11(2):141–151. 6. Nishimura H, Okazaki T, Tanaka Y, et al. Autoimmune dilated cardiomyopathy in PD-1 receptor-deﬁcient mice.Science. 2001;291 (5502):319–322. doi:10.1126/science.291.5502.319

7. Brahmer JR, Drake CG, Wollner I, et al. Phase I study of single-agent anti-programmed death-1 (MDX-1106) in refractory solid tumors: safety, clinical activity, pharmacodynamics, and immunologic correlates.J Clin Oncol. 2010;28(19):3167–3175. doi:10.1200/JCO.2009.26.7609 8. Antonia SJ, Villegas A, Daniel D, et al. PACIFIC investigators. Overall

survival with durvalumab after chemoradiotherapy in stage III NSCLC.

N Engl J Med. 2018;379(24):2342–2350. doi:10.1056/NEJMoa1809697 9. Horn L, Mansﬁeld AS, Szczęsna A, et al. IMpower133 study group. First-line atezolizumab plus chemotherapy in extensive-stage small-cell lung cancer. N Engl J Med. 2018;379(23):2220–2229. doi:10.1056/NEJMoa1809064

10. Kowanetz M, Zou W, Gettinger SN, et al. Differential regulation of PD-L1 expression by immune and tumor cells in NSCLC and the response to treatment with atezolizumab (anti-PD-L1).Proc Natl Acad Sci USA. 2018;115(43):E10119–E10126. doi:10.1073/pnas.1802166115

11. Barlesi F, Vansteenkiste J, Spigel D, et al. Avelumab versus docetaxel in patients with platinum-treated advanced non-small-cell lung cancer (JAVELIN Lung 200): an open-label, randomised, phase 3 study.Lancet Oncol. 2018;19(11):1468–1479. doi:10.1016/S1470-2045(18)30673-9 12. Socinski MA, Jotte RM, Cappuzzo F, et al. IMpower150 study group.

atezolizumab forﬁrst-line treatment of metastatic nonsquamous NSCLC.

N Engl J Med. 2018;378(24):2288–2301. doi:10.1056/NEJMoa1716948 13. Hellmann MD, Ciuleanu TE, Pluzanski A, et al. Nivolumab plus ipili-mumab in lung cancer with a high tumor mutational burden.N Engl J Med. 2018;378(22):2093–2104. doi:10.1056/NEJMoa1801946 14. Paz-Ares L, Luft A, Vicente D, et al. KEYNOTE-407 investigators.

pembrolizumab plus chemotherapy for squamous non-small-cell lung cancer. N Engl J Med. 2018;379(21):2040–2051. doi:10.1056/ NEJMoa1810865

15. Gandhi L, Rodríguez-Abreu D, Gadgeel S, et al. KEYNOTE-189 investigators. Pembrolizumab plus chemotherapy in metastatic non-small-cell lung cancer. N Engl J Med. 2018;378 (22):2078–2092. doi:10.1056/NEJMoa1801005

16. Rittmeyer A, Barlesi F, Waterkamp D, et al. Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised con-trolled trial.Lancet. 2017;389(10066):255–265. doi:10.1016/S0140-6736(16)32517-X

17. Fehrenbacher L, Spira A, Ballinger M, et al. Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial. Lancet. 2016;387(10030):1837–1846. doi:10.1016/ S0140-6736(16)00587-0

18. Reck M, Rodríguez-Abreu D, Robinson AG, et al. KEYNOTE-024 investigators. pembrolizumab versus chemotherapy for PD-L1-positive non-small-cell lung cancer. N Engl J Med. 2016;375 (19):1823–1833. Epub 2016 Oct 8. doi:10.1056/NEJMoa1606774 19. Langer CJ, Gadgeel SM, Borghaei H, et al. KEYNOTE-021 investigators.

Carboplatin and pemetrexed with or without pembrolizumab for advanced, non-squamous non-small-cell lung cancer: a randomised, phase 2 cohort of the open-label KEYNOTE-021 study.Lancet Oncol. 2016;17(11):1497–1508. doi:10.1016/S1470-2045(16)30498-3 20. Herbst RS, Baas P, Kim DW, et al. Pembrolizumab versus docetaxel

for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet. 2016;387(10027):1540–1550. doi:10.1016/S0140-6736(15)01281-7 21. Brahmer J, Reckamp KL, Baas P, et al. Nivolumab versus docetaxel

in advanced squamous-cell non-small-cell lung cancer.N Engl J Med. 2015;373(2):123–135. doi:10.1056/NEJMoa1504627

22. Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer.N Engl J Med. 2015;373(17):1627–1639. doi:10.1056/NEJMoa1507643

23. Wang DY, Salem JE, Cohen JV, et al. Fatal toxic effects associated with immune checkpoint inhibitors: a systematic review and meta-analysis.

JAMA Oncol. 2018;4(12):1721–1728. doi:10.1001/jamaoncol.2018.3923 24. Moher D, Liberati A, Tetzlaff J, Altman DG. PRISMA group pre-ferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–69.W64. doi:10.7326/0003-4819-151-4-200908180-00135

25. Higgins JP, Altman DG, Gøtzsche PC, et al. Cochrane Statistical Methods Group. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials.BMJ. 2011;343:d5928.

26. Wells G, Shea B, O’Connell D, et al. The Newcastle–Ottawa Scale (NOS) for assessing the quality if nonrandomized studies in meta-analyses. 2009 http://www.ohri. ca/programs/clinical _epidemiology/ oxford.aspaccesse July 6, 2012. Doi:10.1094/PDIS-11-11-0999-PDN 27. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsis-tency in meta-analyses. BMJ. 2003;327:557–560. doi:10.1136/ bmj.327.7420.895

28. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188. doi:10.1016/0197-2456(86)90046-2

(10)

29. Kazandjian D, Suzman DL, Blumenthal G, et al. FDA approval summary: nivolumab for the treatment of metastatic non-small cell lung cancer with progression on or after platinum-based chemotherapy. Oncologist. 2016;21:634–642. doi:10.1634/theoncologist.2015-0507

30. Sul J, Blumenthal GM, Jiang X, He K, Keegan P, Pazdur R. FDA approval summary: pembrolizumab for the treatment of patients with metastatic non-small cell lung cancer whose tumors express pro-grammed death-ligand 1. Oncologist. 2016;21:643–650. doi:10.1634/theoncologist.2015-0498

31. Weinstock C, Khozin S, Suzman D, et al. US food and drug admin-istration approval summary: atezolizumab for metastatic non-small cell lung cancer.Clin Cancer Res. 2017;23:4534–4539. doi:10.1158/ 1078-0432.CCR-17-0540

32. Fossella FV, DeVore R, Kerr RN, et al. Randomized phase III trial of docetaxel versus vinorelbine or ifosfamide in patients with advanced non-small-cell lung cancer previously treated with platinum-containing chemotherapy regimens. The TAX 320 non-small cell lung cancer study group. J Clin Oncol. 2000;18:2354–2362. doi:10.1200/JCO.2000.18.10.2059

33. Shepherd FA, Dancey J, Ramlau R, et al. Prospective randomized trial of docetaxel versus best supportive care in patients with non-small-cell lung cancer previously treated with platinum-based chemotherapy. J Clin Oncol. 2000;18:2095–2103. doi:10.1200/ JCO.2000.18.10.2059

34. Hanna N, Shepherd FA, Fossella FV, et al. Randomized phase III trial of pemetrexed versus docetaxel in patients with non-small-cell lung cancer previously treated with chemotherapy. J Clin Oncol. 2004;22:1589–1597. doi:10.1200/JCO.2004.08.040

35. Garassino MC, Martelli O, Broggini M, et al. Erlotinib versus doc-etaxel as second-line treatment of patients with advanced non-small-cell lung cancer and wild-type EGFR tumours (TAILOR): a randomised controlled trial. Lancet Oncol. 2013;14:981–988. doi:10.1016/S1470-2045(13)70310-3

36. Hamid O, Robert C, Daud A, et al. Safety and tumor responses with lambrolizumab (anti-PD-1) in melanoma.N Engl J Med. 2013;369 (2):134–144. doi:10.1056/NEJMoa1305133

37. Rizvi NA, Mazières J, Planchard D, et al. Activity and safety of nivolumab, an anti-PD-1 immune checkpoint inhibitor, for patients with advanced, refractory squamous non-small-cell lung cancer (CheckMate 063): a phase 2, single-arm trial. Lancet Oncol. 2015;16(3):257–265. doi:10.1016/S1470-2045 (15)70054-9

38. Hodi FS, Chesney J, Pavlick AC, et al. Combined nivolumab and ipilimumab versus ipilimumab alone in patients with advanced mel-anoma: 2-year overall survival outcomes in a multicentre, rando-mised, controlled, phase 2 trial. Lancet Oncol. 2016;17 (11):1558–1568. doi:10.1016/S1470-2045(16)30366-7

39. Migden MR, Rischin D, Schmults CD, et al. PD-1 blockade with cemiplimab in advanced cutaneous squamous-cell carcinoma.N Engl J Med. 2018;379(4):341–351. doi:10.1056/NEJMoa1805131

(11)

Supplementary materials

0

0.1

0.2

0.3

0.4

0.5

0.1

0.2

0.3

0.4

0.5

0.1

0.2

0.3

0.4

0.5

0.05 0.2 1 5 20

OR SE(log[OR])

SE(log[RR])

SE(log[RR]) 0

0.1 0.2 0.5 1 2 5 10

RR

0

0.1

0.2

0.3

0.4

0.5 0

RR Subgroups

Subgroups

Subgroups Subgroups

PD-L1 VS chemotherapy PD-1 VS chemotherapy

PD-L1 VS Docetaxel

PD-1/PD-L1 VS Combined chemotherapy PD-1 VS Docetaxel

PD-1+chemotherapy VS chemotherapy PD-L1+chemotherapy VS chemotherapy Nivolumab VS chemotherapy Atezolizumab VS chemotherapy Pembrolizumab VS chemotherapy Avelumab VS Docetaxel

RR

A1

A3

A2

_B

Figure S1Funnel plots of the incidence risk for all-grade diarrhea. (A1): Funnel plots of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 vs chemotherapy). (A2): Funnel plots of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 vs docetaxel/combined chemotherapy), the data included were assigned to the corresponding subgroup according to control group (docetaxel or combined chemotherapy). (A3): Funnel plots of diarrhea for all grades in the subgroup analysis (PD-1/ PD-L1 vs chemotherapy), the data included were assigned to the corresponding subgroup according to the name of PD-1/PD-L1 inhibitor and the control group. (B): Funnel plots of diarrhea for all grades in the subgroup analysis (PD-1/PD-L1 + chemotherapy vs chemotherapy).

Abbreviation:PD-1, programmed cell death 1; PD-L1, programmed cell death ligand 1; OR, odds ratio; RR, risk ratio.

(12)

0

0.1

0.2

0.3

0.4

0.5

0

0.1

0.2

0.3

0.4

0.5

0.1 0.2 0.5 1 2 5 10

SE(log[RR])

0.1

0.2

0.3

0.4

0.5

SE(log[OR]) SE(log[RR])

SE(log[RR]) 0

0.2

0.5 1 5 20

0.1 0.2 0.5 1 2 5 10

RR

0.1

0.2

0.3

0.4

0.5 0

RR Subgroups

Subgroups Subgroups

Subgroups

PD-L1 VS chemotherapy PD-1 VS chemotherapy

PD-L1 VS Docetaxel PD-1 VS chemotherapy

PD-1 VS Docetaxel PD-1+chemotherapy VS chemotherapy PD-L1+chemotherapy VS chemotherapy Nivolumab(PD-1 VS chemotherapy

Atezolizumab(PD-1) VS Docetaxel Pembrolizumab(PD-1) VS chemotherapy

Avelumab(PD-1) VS

A1

A3

A2

B

RR OR

Figure S2Funnel plots of the incidence risk for grade 1–2 diarrhea. (A1): Funnel plots of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy). (A2): Funnel plots of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 vs docetaxel/combined chemotherapy), the data included were assigned to the corresponding subgroup according to control group (docetaxel or combined chemotherapy).(A3): Funnel plots of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy), the data included were assigned to the corresponding subgroup according to the name of PD-1/PD-(PD-1/PD-L1 inhibitor and the control group. (B): Funnel plots of diarrhea for grade 1–2 in the subgroup analysis (PD-1/PD-L1 + chemotherapy vs chemotherapy).

(13)

Cancer Management and Research

Dovepress

Publish your work in this journal

Cancer Management and Research is an international, peer-reviewed open access journal focusing on cancer research and the optimal use of preventative and integrated treatment interventions to achieve improved outcomes, enhanced survival and quality of life for the cancer patient.

The manuscript management system is completely online and includes a very quick and fair peer-review system, which is all easy to use. Visit http://www.dovepress.com/testimonials.php to read real quotes from published authors.

Submit your manuscript here:https://www.dovepress.com/cancer-management-and-research-journal 0

0.5

1

1.5

2

0

0.5

1

1.5

2

0

0.5

1

1.5

2 0.5

1

1.5

2

0.1 1 10

SE(log[RR])

SE(log[OR]) SE(log[RR])

SE(log[RR]) 0

RR

RR Subgroups

Subgroups Subgroups

Subgroups PD-L1 VS chemotherapy

PD-1 VS chemotherapy

PD-L1 VS Docetaxel PD-1 VS chemotherapy PD-1 VS Docetaxel

Nivolumab(PD-1) VS chemotherapy

Atezolizumab(PD-L1) VS chemotherapy Pembrolizumab(PD-1) VS chemotherapy

Avelumab(PD-1) VS chemotherapy

A1

A3

A2

B

RR 1000 0.001

0.1 1 10 1000

0.001 0.001 0.1 1 10 1000

0.1 1 10 1000

0.001 OR

Figure S3Funnel plots of the incidence risk for grade 3–5 diarrhea. (A1): Funnel plots of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 vs chemotherapy). (A2): Funnel plots of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 vs docetaxel/combined chemotherapy), the data included were assigned to the corresponding subgroup according to control group (docetaxel or combined chemotherapy). (A3): Funnel plots of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 VS chemotherapy), the data included were assigned to the corresponding subgroup according to the name of PD-1/PD-(PD-1/PD-L1 inhibitor and the control group. (B): Funnel plots of diarrhea for grade 3–5 in the subgroup analysis (PD-1/PD-L1 + chemotherapy vs chemotherapy).