Influence of Microbiome and Antibiotics on the Efficacy of Immune Checkpoint Inhibitors

The human microbiome mainly consists of bacteria and interacts closely with the immune system. Immune checkpoint inhibitors (ICI) are used to treat several types of cancers. Recently, it has been identified that the gut microbiome plays a role in the effectiveness of immunotherapy. This study aims to analyze the effect of microbiome and antibiotics on the effectiveness of ICI in cancer patients and the measures to improve efficacy based on that. A detailed review was conducted on articles published in PubMed and Science Direct in the last five years i.e., 2016 to 2021. A total of 16 articles involving 1293 patients with cancer who were receiving immunotherapy, were deemed eligible to be included in the final review. Data were extracted from the eligible articles and were checked for quality appraisal. All 16 articles revealed the effect of either gut microbiome or antibiotics or both on ICI. Based on our findings, we found that the microbiome enriched in different microorganisms responded differently to the ICI and that antibiotics negatively impacted the effectiveness of ICI. The time at which patients receiving ICI were prescribed antibiotics influenced the effect of ICI. Antibiotics and different microbiome also affected progression-free survival (PFS) and overall survival (OS).

The response rate of a patient to immunotherapy depends on immune competency, diversity, antigen specificity variation, antigen expression, and the recent identification of the role of gut microbiota [10]. The microbiome plays a role in developing inflammation, the integrity of mucosal immunity, and protecting against pathogens [11]. Side effects of ICI include severe diarrhea and colitis, which suggests the role of the gut microbiome and its influence on therapeutic purpose and toxicity [11].
This systematic review aims to learn about how the gut microbiome influences the effectiveness of ICI and identify strategies to improve the efficacy of immunotherapy.

Review Method
Two independent reviewers carried out a rigorous literature review using electronic databases. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines were followed for carrying out this systematic review. PubMed and Science Direct were screened for peer-reviewed articles published in the last five years i.e., 2016 to 2021. A search was performed using medical subject headings (MeSH) and keywords combined with Boolean connectors from April 2021 to May 2021.
The search strategy for PubMed was as follows: ("Microbiota/drug effects"[Majr] OR "Microbiota/immunology"[Majr] OR "Microbiota/physiology"[Majr]) OR Microbiome OR "Gastrointestinal microbiome" OR "intestinal flora" OR "bacterial flora" AND ("Immunotherapy/drug effects"[Majr] OR "Immunotherapy/immunology"[Majr] OR "Immunotherapy/pharmacology"[Majr] OR "Immunotherapy/physiology" [Majr] OR "Immunotherapy/therapeutic use"[Majr]) OR Immunotherapy OR "molecular targeted therapy" OR "immunological therapy" AND ("Neoplasms/drug effects" For Science Direct, the keywords 'immunotherapy' and 'microbiome' were used. Both authors screened each article by title initially. The articles selected based on their title were further evaluated and were included or excluded based on abstract review. A complete article review was done following the abstract review to exclude irrelevant articles. Any duplicate articles that coincided with multiple keywords or MeSH terms were also excluded. Figure 1 shows how the search was conducted as per PRISMA 2020 guidelines [12].

FIGURE 1: Flowchart of literature review search as per Preferred
Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines

Inclusion Criteria
The search was limited to full-text publications between the years 2016 and 2021, articles published in English, and studies recruiting the adult population (aged more than or equal to 18). We ensured that the studies chosen featured patients who had received at least one ICI (anti-PD-1, anti-PD-L1, or anti-CTLA-4). And our study in turn looked at the effect of either or both microbiome and antibiotics on ICI.

Exclusion Criteria
All studies that were conducted on the children population and animals, case reports and case series, and inaccessible full-text articles were excluded.

Data Extraction and Study Selection
Two researchers carried out an individual review of the literature. After compiling the results, both researchers compared the data and resolved any conflicts through mutual understanding.

Results
Initial screening of the electronic database PubMed yielded 300 records, and Science Direct yielded 438 records. After 40 duplicates were removed, the remaining 698 articles were screened by title for relevance.
Following that 474 non-relevant articles were excluded. From the remaining 224 articles, 26 articles couldn't be retrieved. Abstracts and full-text of the remaining 198 articles were thoroughly reviewed, and 170 were excluded due to irrelevance, two articles were not in the English language, and 10 were animal studies. Finally, 16 articles were included in the review. Out of the 16, seven were retrospective cohort studies, five were prospective cohort studies, and four were review articles.

Quality Assessment
We carried out a thorough quality assessment for 12 cohort studies using the Newcastle-Ottawa scale and four review articles using a quality appraisal checklist for qualitative studies.  Articles with a Newcastle-Ottawa scale score more than or equal to 6/8 and articles with a quality appraisal checklist for qualitative studies score more than or equal to 12/14 were considered high-quality articles.
Articles with a Newcastle-Ottawa scale score between 4-6/8 and articles with a quality appraisal checklist for qualitative studies score between 8-12/14 were considered intermediate quality articles.
Articles with a Newcastle-Ottawa scale score between 1-4/8 and articles with a quality appraisal checklist for qualitative studies score between 1-8/14 were considered low-quality articles.
Out of 12 cohort studies, 10 scored 8/8, and two scored 7/8 on the Newcastle-Ottawa scale. Therefore, all 12 cohort articles were high-quality articles. Out of four review articles, two scored 12/14, and the other two scored 11/14 on the quality appraisal checklist for qualitative studies. And so, two review articles were highquality articles and the remaining two review articles were intermediate quality articles.

Study Characteristics
The finalized articles were all peer-reviewed articles published in the last five years i.e., 2016 to 2021, with full texts freely available. All articles were in English, and articles in other languages were excluded. Only human studies were included. Table 2 shows the data extracted from the studies. Out of 12 cohort studies, seven studies looked for the influence of antibiotics and the gut microbiome on ICI [13, 14, 17-19, 22, 23], and five studies checked for the effect of the gut microbiome on ICI [15,16,20,21,24]. Intestinal microbiota might be a potential biomarker to help predict diarrhea in patients treated with anti-PD-1 antibodies.

5.
Pinato et al. Antibiotic treatment before ICI therapy is associated with worse treatment response and OS.

6.
Derosa et al. Antibiotics were associated with decreased benefits from ICI in RCC and NSCLC.

7.
Ahmed et al. The use of broad-spectrum antibiotics while receiving ICI treatment is associated with a poor response rate and shorter PFS.

8.
Gopalakrishnan et al. [20] 2018 Prospective study Adult Metastatic melanoma patients -89 Patients with favorable gut microbiome have a better response to anti-PD-1 immunotherapy due to increased antigen presentation and improved effector T cell function.

9.
Frankel et al. Antibiotic use within six weeks of starting ICI leads to inferior PFS and OS, which suggests a link between antibiotic altering gut microbiome and efficacy of ICI.

Effect of Antibiotics on the Microbiome
Antibiotics have helped in the morbidity and mortality of various fatal infections, but they have a speedy and long-lasting effect on gut microbiota composition [13,25]. Antibiotics alter the gut microbiome by causing loss of diversity, metabolic capacity changes, loss of vital taxa, and decreasing colonization resistance against invasive pathogens [25]. These changes are named gut dysbiosis [25]. Gut dysbiosis occurs within days of exposure, and it may take >6 weeks to recover after broad-spectrum antibiotic treatment [14]. A decrease in gut microbiota was noted lasting four years after the seven-day treatment of Helicobacter pylori with clarithromycin, metronidazole, and omeprazole [14]. Antibiotics also impair the response of cytotoxic T-cells against cancer [17]. Due to the gut dysbiosis caused by antibiotics, their use negatively influences immune response against cancer cells [13].
Antibiotics help in curing mild and severe infections but at the same time cause dysfunction in the gut microbiome. It has a chronic harmful effect on the microbiome by decreasing its diversity. Gut microbiome dysfunction, in turn, causes a damaging effect on the immune response.

Effect of the Microbiome on ICI Efficacy
Baseline gut microbiota is an important marker for the response of treatment and anti-CTLA-4 treatmentinduced colitis, according to a study by Chaput et al. [24]. Their study showed gut microbiome rich in the Bacteroides genus had an inadequate response to anti-CTLA-4 treatment and remained free of colitis [24]. Patients who had a gut microbiome rich in Firmicutes had longer progression-free survival (PFS) and overall survival (OS) and had increased occurrence of anti-CTLA-4 induced colitis [24]. This evidence is also supported by a study conducted by Gopalakrishnan et al., which showed patients having enhanced antitumor immune response if they have a favorable gut microbiome consisting of high diversity and enriched in Ruminococcaceae/Faecalibacterium [20]. This enhanced response is due to increased antigen presentation and enhanced effector T cell function [20]. Patients with an unfavorable gut microbiome i.e. a microbiome with less diversity and a high abundance of Bacteroidales, have decreased anti-tumor immune response [20]. This reduced response is due to weak antigen presentation and inadequate intratumoral lymphoid and myeloid infiltration [20]. A metagenomic shotgun sequencing and metabolomic profiling performed by Frankel et al. showed that different ICI responders have different microbiomes [21]. This study found that patients responding to a combination of ipilimumab plus nivolumab had a gut microbiome enriched in Faecalibacterium prausnitzii, Bacteroides thetaiotamicron, and Holdemania filiformis [21]. Pembrolizumab responders had gut microbiome enriched in Dorea formicigenerans [21]. ICI responders for all types of treatment were found to have microbiome enriched in Bacteroides caccae [21].
Patients who respond to anti-PD-1 immunotherapy have increased taxonomic diversity and enriched genes than the non-responders of anti-PD-1 immunotherapy [15]. Before immunotherapy treatment, both the responders and non-responders have an abundance of Bacteroidetes, followed by Firmicutes and Proteobacteria similar to a healthy adult [15]. After the treatment initiation, responders have stable microbial composition, while non-responders have increased Proteobacteria [15]. This suggests that gut microbial diversity is associated with drug efficacy and disease prognosis [15]. It was also found that oral administration of Bifidobacterium can improve the efficacy of anti-PD-L1 immunotherapy [15].
It is very common for patients receiving immunotherapy to have diarrhea as an immune-related side effect [16]. Intestinal microbiota plays a vital role in immune-related diarrhea in patients receiving immunotherapy [16].
In summary, the intestinal microbiome plays a vital role in response to ICI. And ICI responders have microbiome enriched in different microbes compared to non-responders. Microbiome enriched in Firmicutes, Ruminococcaceae/Faecalibacterium, Holdemania filiformis, Dorea formicogenerans, Bacteroides thetaiotamicron, and Bacteroides caccae has a beneficial effect on the efficacy of ICI due to increased antigen expression. An enriched microbiome led to longer PFS and OS, but at the same time, it increased ICI-related colitis. Microbiome rich in Bacteroides genus has an inappropriate response to ICI due to decreased antigen expression and has a lesser occurrence of ICI-related colitis. Non-responders of ICI have increased Proteobacteria compared to stable microbial composition in responders.

Effect of Antibiotics on ICI Efficacy
Antibiotics modify the gut microbiome, and clinical studies have emphasized the negative effect of antibiotics in patients treated with ICI [13]. Antibiotics are associated with decreased PFS in patients on ICI [13]. A study conducted by Tinsley et al. showed a decrease in PFS and OS in patients on ICI receiving antibiotics [14]. Multiple and lengthy antibiotic use results in inadequate clinical outcomes [14]. Antibiotic therapy 30 days before initiation of ICI has shown poor response rate and survival in cancer patients on ICI [17]. However, concurrent use of antibiotics and ICI was not associated with poor response or decreased OS [17]. A study conducted by Derosa et al. showed there was an absence of immune-related colitis when the fecal sample was enriched in Bacteroidetes and depleted in Firmicutes [18]. This study also showed increased progressive disease and poor PFS and OS in patients receiving ICI who were treated with antibiotics [18]. The use of narrow-spectrum antibiotics did not affect the response rate, while broad-spectrum antibiotics decreased the response rate [19]. It was also found that antibiotics received two weeks before and after starting ICI had extended time to respond, lower response rate and PFS than patients who didn't receive antibiotics during that period [19]. Thompson et al. conducted a study that showed patients who received antibiotics within six weeks of starting ICI had shorter PFS and OS [23]. Contrary to these findings, a study conducted by Kaderbhai et al. showed antibiotics didn't have any effect on nivolumab in lung cancer patients [22].
Overall, antibiotics have a damaging effect on the response of ICI due to their negative impact on the gut microbiome. The use of several and prolonged broad-spectrum antibiotics leads to an inadequate ICI response. Antibiotic use is also associated with inferior PFS and OS in some cancer patients receiving ICI. The administration of antibiotics during different phases of ICI also plays a role in clinical outcomes in patients receiving ICI.
An intact gut microbiome is required for an adequate immunotherapy effect. Antibiotics should be used very carefully in cancer patients on immunotherapy to prevent drug resistance and inadequate response to ICI. Manipulation of the microbiome should be considered for ICI therapy.

Limitations
Some limitations of the study are that there have not been enough studies evaluating the efficacy of ICI by other products altering the gut microbiome, such as probiotics. We didn't take other factors affecting gut microbiomes like diet, exercise, and age into consideration. The search was limited to articles in the English language, which might have missed some relevant articles. The search was limited to two databases. However, on performing a thorough literature review of these two databases, studies that feature over 1000 patients were included and this was deemed an adequate sample size to derive a conclusion.

Conclusions
In this systematic review, we evaluated how the microbiome and antibiotics affect the efficacy of immunotherapy. The effectiveness of immunotherapy depends on the type of microbiome present in the gut. Antibiotics were found to influence the gut microbiome negatively, hence decreasing the efficacy of immunotherapy. This proves that a stable microbiome needs to be maintained for increased effectiveness of immunotherapy. It is recommended that antibiotics be used very judiciously in cancer patients on ICI. The use of microbiome sequencing in the future can be a helpful tool in predicting ICI response due to variation in the microbiome in responders and non-responders of ICI.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.