Table 2 Recent studies investigating the association between antibiotics use and ICI efficacy in cancer patients

NSCLC(n = 74)

Nivolumab

Those receiving ATB 3 months before the first nivolumab injection or during treatment

15 (20.3%) patients received ATB. ATB medication has no impact on either response rate to PD-1 blockade or PFS.

Kaderbhai et al. 2017 [112]

NSCLC (n = 74)

PD-1 inhibitors

Those receiving ATB within 6 weeks before initiation of PD-1 inhibitors

18 (24%) patients received ATB. ATB use did not impact ORR but was associated with worse OS and PFS even in multivariate analysis.

Thompson et al. 2017 [113]

Melanoma(n = 39)

Ipilimumab (n = 1), nivolumab (n = 1), ipilimumab plus nivolumab(n = 24), or pembrolizumab (n = 13)

Those receiving ATB or probiotics before or during ICI treatment.

3 (8%) patients received ATB and 1 (3%) received probiotics. Neither clinical response nor toxicity was associated with antibiotic or probiotic use.

Frankel et al. 2017 [95]

Advanced cancer(n = 60)

PD-1 inhibitors

(nivolumab or pembrolizumab) or PD-L1 inhibitor (atezolizumab)

Those receiving ATB within 2 weeks prior to and after ICI initiation and within 10 weeks prior to disease progression.

17 (28%) patients received systemic antibiotics. They had a lower RR and shorter PFS. Multivariate analysis identified antibiotics as the only factor affecting RR and PFS. Patients who received broad-spectrum antibiotics experienced shorter OS.

Ahmed et al. 2018 [114]

NSCLC(n = 239), RCC(n = 121)

Patients with RCC received anti-PD-(L)1 mAb alone (n = 106) or in combination with anti-CTLA-4 mAb (n = 10) or bevacizumab (n = 5). Patients with NSCLC received anti-PD-(L)1 mAb alone (n = 205) or combined with anti-CTLA-4 mAb (n = 34)

Those receiving ATB within 30 days of beginning ICI

16 (13%) RCC patients and 48 (20%) NSCLC patients received ATB. In multivariate analyses, ATB was associated with shorter PFS in RCC and shorter OS in NSCLC.

Derosa et al. 2018 [115]

Non-squamous NSCLC (n = 30)

Nivolumab (n = 25) or pembrolizumab (n = 5)

Those receiving ATB within 1 month before and 1 month after ICI initiation.

11 (36.7%) patients received ATB. Median PFS and OS were significantly shorter in ATB group. In a multivariate analysis, ATB use was identified as the only parameter significantly associated with PFS and OS.

Huemer et al. 2018 [116]

Nivolumab (92.3%) or pembrolizumab (7.7%)

Those receiving ATB within 2 months before and 1 month after ICI initiation

47.9% patients received ATB. Patients who received ATB had shorter OS and PFS. The patients receiving ATB intravenously had a shorter OS and PFS than orally.

Mielgo-Rubio et al. 2018 [117]

NSCLC (n = 90)

Nivolumab

Those receiving ATB for ≥3 days within 30 days prior to nivolumab

13 (14.4%) patients received ATB. In multivariate analysis, no significant association was observed between survival and previous antibiotic use.

Hakozaki et al. 2019 [118]

Melanoma(n = 74)

Anti-PD-1 mAb alone (n = 54) or anti-CTLA-4 mAb alone (n = 5) or anti-CTLA-4 mAb plus chemotherapy (n = 15)

Those receiving ATB within 30 days before ICI initiation

10 (13.5%) patients received ATB. Patients who received ATB experienced more PD and shorter PFS.

Elkrief et al. 2019 [119]

NSCLC(n = 109)

Anti-PD-1 mAb alone (n = 57) or anti-PD-1 mAb plus chemotherapy (n = 33) or anti-PD-1 mAb plus anti-angiogenic agent (n = 19)

Those receiving ATB within 1 month before or after the first administration of PD-1 blockade

20 (18.3%) patients received ATB. In multivariable analysis, ATB treatment was markedly associated with worse PFS and OS.

Zhao et al. 2019 [120]

NSCLC(n = 119), melanoma(n = 38), other types(n = 39)

ICI

Those receiving ATB within 30 days prior to (pATB) or concurrent with (cATB) ICI therapy

pATB therapy, but not cATB therapy, was associated with worse OS and a higher likelihood of primary disease refractory to ICI therapy. Multivariate analyses confirmed the association between pATB therapy and OS.

Pinato et al. 2019 [121]

NSCLC (n = 37)

Nivolumab

Those receiving ATB within 2 months before clinical assessment

11 (29.7%) patients received ATB within 2 months. However, the R/NR ratio was similar in ATB and no-ATB groups.

Jin et al. 2019 [29]

Urothelial carcinoma (n = 101)

PD-1/PD-L1 inhibitors

Those receiving ATB within 1 months before or during ICI treatment

26 (25.7%) patients received ATB. Antibiotics compromised clinical outcomes significantly.

Agarwal et al. 2019 [122]

NSCLC (n = 157)

PD-1/PD-L1 inhibitor (n = 150) or CTLA-4 inhibitor (n = 1) or combination (n = 6)

1 months before or 3 months after ICI treatment was defined early immunotherapy period (EIOP). Antibiotic-immunotherapy exposure ratio (AIER) defined as “days of antibiotic/days of immunotherapy” during the whole immunotherapy period (WIOP) was also calculated.

46 (29.3%) patients received ATB during WIOP, 27 (17.2%) patients received ATB during EIOP. ATB use during EIOP has no impact on either PFS or OS. But the patients with a higher AIER had worse PFS and OS.

Galli et al. 2019 [123]

Esophagogastric cancer (n = 161)

Anti-PD-1/PD-L1 (n = 110) or anti-PD-1/PD-L1 combined with anti-CTLA-4 (n = 51)

Those receiving ATB within 2 months before or during ICI treatment

62 (38%) patients received ATB. No difference in PFS or OS between those patients treated with antibiotics versus those who were not.

Greally et al. 2019 [124]

Solid cancer (n = 234)

ICI alone or ICI combination or ICI combined with chemotherapy

Those receiving ATB within 60 days before ICI initiation

108 (46.2%) patients received ATB. ATB use was associated with a decreased OR, shorter PFS and OS. In the multivariate analysis, antibiotics use was a significant predictor of patient survival.

Kim et al. 2019 [125]

NSCLC (n = 72)

Nivolumab

Those receiving ATB within 2 months before and 1 month after ICI initiation

30 (42%) patients received ATB. ATB use was associated with shorter OS.

Krief et al. 2019 [126]

RCC (n = 146)

PD-1/PD-L1 inhibitors

Those receiving ATB within 8 weeks before and 4 weeks after ICI initiation

31 (21%) patients received ATB. ATB use was associated with a lower objective response rate and shorter PFS.

Lalani et al. 2020 [127]

NSCLC (n = 218)

Anti-PD-1 mAb alone (n = 207) or anti-PD-1 mAb plus chemotherapy (n = 5) or investigational immunotherapy^a (n = 6)

Those receiving ATB within 2 months before ICI treatment

33 (15.1%) patients received ATB. PFS and OS were significantly shorter in patients receiving ATB.

Schett et al. 2020 [128]

NSCLC(n = 1512)

Randomly assigned to receive atezolizumab (n = 757) or docetaxel (n = 755)

Those receiving ATB within 30 days before and 30 days after the first treatment

169 (22.3%) patients in the atezolizumab group received ATB. Multivariate analysis in all patients revealed that ATB were associated with shorter OS. Within the atezolizumab population, OS was significantly shorter in patients who received ATB.

Chalabi et al.2020 [129]

RCC(n = 69)

Nivolumab (3 mg/kg i.v. q2w)

Those receiving ATB use within 60 days of nivolumab

11 (16%) patients received ATB. Patients who received ATBs had a lower ORR, PFS and OS.

Derosa et al. 2020 [99]

NSCLC(n = 140), RCC(n = 55)

Single-agent ICI

Those receiving antibiotics within 4 weeks before and 6 weeks after the ICI initiation

54 (39%) NSCLC and 24 (44%) RCC patients received ATB. In multivariable analysis, PFS and OS were shorter in NSCLC patients who received broad-spectrum anti-anaerobes or ‘other’ antibiotics (vancomycin predominant). In RCC, patients who received penicillins /penicillin-class/early-generation cephalosporins had shorter PFS.

Kulkarni et al. 2020 [130]

Advanced cancer (n = 291, including 179 melanoma, 64 NSCLC and 48 RCC)

ICI

Those receiving ATB within 2 weeks before and 6 weeks after ICI initiation

92 (32%) patients received ATB. ATB use was associated with shorter PFS and OS in multivariate analysis. Administration of a single course of ATB had non-significant impact on PFS and OS while patients who received cumulative ATB for>7 days had significantly worse PFS and OS.

Tinsley et al. 2020 [131]

A meta-analysis included 19 eligible studies comprising 2740 cancer patients

Anti-PD-1/PD-L1 mAb (n = 14), anti-PD-1/PD-L1 mAb and/or anti-CTLA-4 mAb (n = 3), no information of the type of ICI drug (n = 2)

Pre-therapy ATB use (n = 11), post-therapy ATB use (n = 1), pre- or post-therapy ATB use (n = 9)

ATB use was negatively associated with OS and PFS in cancer patients. Similar results were obtained in the subgroup analyses stratified by the time of ATB use and cancer type.

Huang et al. 2019 [132]

A meta-analysis included 33 eligible studies comprising 5565 cancer patients

ICI (anti-PD−/PD-L1 or anti-CTLA-4) alone or combined with chemotherapy/targeted therapy.

ATB use prior to or within therapy

ATB use was significantly correlated with worse OS and PFS. The similar results were also found in subgroup analysis for lung cancer (both OS and PFS), RCC (only significant in PFS) and other cancers. The ICI efficacy was more likely to be diminished by ATB administration within a time frame from 60 days before to 60 days after ICI initiation.

Yang et al. 2020 [133]