Interactions between epidermal growth factor receptor tyrosine kinase inhibitors and proton-pump inhibitors/histamine type-2 receptor antagonists in non-small cell lung cancer: a systematic review and meta-analysis.

BACKGROUND: Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) are increasingly used for advanced non-small cell lung cancer (NSCLC) as first-line therapy. The bioavailability and efficacy of oral EGFR-TKIs could be affected by acid suppression (AS) therapy such as PPIs and H2RAs which are reported to be over-prescribed. Hence, there is a need to investigate the effect of AS on the overall survival (OS), progression-free survival (PFS) and adverse effect profile in patients treated with EGFR TKIs.
METHODS: An electronic database search of Medline and Embase was performed following PRISMA guidelines on 17 January 2021. Studies analyzing interactions between EGFR TKIs and PPIs/H2RAs in NSCLC patients were included. Abstracts, non-English or non-Japanese studies or studies using non-EGFR TKIs were excluded. Hazard ratios (HRs) were pooled using generic inverse variance random effects model. Effect sizes for dichotomous variables were pooled using Mantel-Haenszel random effects model. Significance was considered at P≤0.05. Heterogeneity was assessed with Cochran Q-test and I2 test. Publication bias was assessed with funnel plots. The assessment of quality and risk of bias of randomized and non-randomized studies were undertaken with RoB 2 and the ROBINS-I tool respectively.
RESULTS: Out of 1,173 potentially relevant articles, 14 articles were included in the final analysis. The pooled prevalence of AS in patients taking EGFR TKI was 30.71% in 4,010 individuals. Patients who were treated only with EGFR TKI had significantly better OS (HR =1.46, 95% CI: 1.27-1.72; P<0.00001) and PFS (HR =1.63, 95% CI: 1.35-1.98; P<0.00001). The OS for EGFR mutation positive patients only was as similarly significant as the OS in all patients taking EGFR TKI, while the PFS in mutation positive patients was significantly worsened with AS. PPIs resulted in a significantly worsened OS and PFS but H2RAs did not produce significantly different OS and PFS between AS and non-AS users. There were no significant differences in the incidence of rash (OR =0.81, 95% CI: 0.50-1.32; P=0.40), diarrhoea (OR =1.03, 95% CI: 0.63-1.67; P=0.91) or other adverse effects.
CONCLUSIONS: Co-administration of AS medications with first-generation EGFR-TKIs in NSCLC worsens survival outcomes. Physicians should only prescribe AS medications when absolutely clinically indicated. 2021 Translational Lung Cancer Research. All rights reserved.

Introduction

With lung cancer being the leading cause of cancer related deaths worldwide (1), the epidermal growth factor receptor (EGFR) has been found to play a significant role in the development of non-small cell lung cancer (NSCLC) (2). Recent guidelines have put forth systemic treatment regimens involving EGFR-tyrosine kinase inhibitors (TKIs), including erlotinib, gefitinib, afatinib, dacomitinib and osimertinib, as the first line therapy for advanced NSCLC harbouring sensitizing EGFR mutations (exon 19 deletions, exon 21p.L858R point mutation) (3). The use of EGFR TKIs has also been indicated as an adjuvant therapy for stage IIB, IIIA or high risk stage IB, IIA EGFR mutation positive NSCLC patients (3). The superiority of EGFR-TKIs over the conventional platinum-based doublet chemotherapy have been demonstrated by an improved response and progression-free survival (PFS) in large, randomized trials (4,5).

The bioavailability of the most widely used first-generation EGFR TKIs (erlotinib and gefitinib) are dependent on the gastric acidity and absorption in the stomach, raising questions about the possible effect of co-administering medications that raise gastric pH (6,7). Concurrent omeprazole administration has been shown to reduce erlotinib area under curve (AUC) and maximum concentration (Cmax) by about 46% and 61%, respectively (8). Similarly, gefitinib has a pH-dependent solubility and AUC and Cmax could decrease by up to 44% and 70% respectively (8). The pharmacokinetics of another widely used third-generation EGFR TKI osimertinib has not been shown to be affected by AS in an open-label study of healthy male volunteers (6).

Acid suppression (AS) medications are among the most common drug classes used in the world and are also available as over-the-counter medications (9). Among which, proton-pump inhibitors (PPIs) and histamine type-2 receptor antagonists (H2RAs) are the most commonly used in AS therapy which aims to maintain an intragastric pH above 4 (10-12). Consequently, the absorption, AUC and Cmax of erlotinib and gefitinib could be drastically reduced.

AS therapy is often prescribed as prophylaxis in patients with NSCLC treated with corticosteroids and has reported to be over-prescribed for therapeutic or prophylactic purposes (13), with a given prevalence of 33.2–46.3% of lung cancer patients concurrently using AS (14). Given the widespread AS use, a significant proportion of NSCLC patients may be receiving EGFR-TKI and AS therapy concurrently (14). Hence, there is a need to investigate the effect of AS on EGFR TKI efficacy, investigating its impact on overall survival (OS), PFS and its side effect profile.

We present the following article in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) reporting checklist (available at https://dx.doi.org/10.21037/tlcr-21-378).

Methods

Search strategy

The Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines was used in the synthesis of this review (15) with the PRISMA checklist completed. An electronic database search of Medline and Embase was performed on 17 January 2021 using keywords and terms synonymous with ‘EGFR, ‘TKI’, ‘Acid Suppression’ and ‘Drug Interaction’. An example of the search strategy can be found in the Table S1. The references of included articles were also assessed for suitability for inclusion.

Study selection and data extraction

The inclusion criteria comprised comparative studies, that focused on the effects of EGFR TKIs (predominately first-generation erlotinib and gefitinib) in NSCLC patients on PPIs/H2RAs versus NSCLC patients not on AS therapy. A variety of quantitative methodology, such as randomized controlled trials (RCT) and cohort studies, was selected. Exclusion criteria included the use of non-EGFR TKIs, abstracts and studies that were non-English or non-Japanese. Article selection was performed by two authors, using the inclusion and exclusion criteria and any discrepancies were resolved based on the consensus with a third author. The blinded pair then extracted details such as the author, year of publication, title, country of origin, study design and duration, number and demographics of subjects, OS, PFS and adverse effects of the two drugs. When mean and standard deviation data were not reported, transformation of existing values was performed using existing methods (16,17). If hazard ratios (HRs) were not provided, they were estimated from the log-rank p value, the median time-to-event, and time-point survival rates, using methods from Parmar et al. (18).

Statistical analysis and quality assessment

Analysis was done in Revman 5.4 and R studio (Version 1.3.1093). A single arm analysis was used to pool the proportion of NSCLC patients on PPI and H2RA using the generalized linear mixed model (GLMM) with Clopper-Pearson intervals to stabilize the variance (19,20). For time-to-event variables, hazard ratios (HRs) were pooled using the generic inverse variance method with a random effects model for both OS and PFS. A sensitivity analysis based on the EGFR mutation status was performed to observe the effects of only mutant NSCLC. Next, a subgroup analysis based on the type of acid suppression (AS) used was performed to compare the effect size between PPI and H2RAs. Effect sizes for dichotomous variables were pooled using Mantel-Haenszel random effects model. Significance was considered at P≤0.05. When there was insufficient data amount for meta-analysis, a descriptive approach was undertaken for the presentation of findings (this was performed for AEs). Heterogeneity was assessed with Cochran Q-test and I2 test, with a significance value of at P≤0.10 or I2≥40 respectively (21,22). Publication bias was assessed with funnel plots when sufficient studies were available (n>10) (23,24). Publication bias was assessed using visual inspection of funnel plots.

The assessment of quality and risk of bias of randomized and non-randomized studies were undertaken with Risk of Bias 2 (RoB 2) (25), and the ROBINS-I tool (26) respectively. RoB 2 assesses the risk of bias from the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome and in selection of the reported result. The ROBINS-I tool assesses bias due to confounding, selection, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes and in reporting results.

Results

The search strategy yielded 1,173 potentially relevant articles. After titles and abstracts screened, 35 full texts were reviewed, of which 14 were included in the final analysis (). Chu et al. (27) was excluded as the study evaluated patients with advanced gastroesophageal cancer and patients were under concurrent capecitabine and oxaliplatin (CapeOx) chemotherapy and capecitabine may increase plasma concentration of erlotinib (28), hence complicating the potential effects of acid suppression (AS) on EGFR TKIs. In total, 1,197 patients were treated with EGFR TKI together with AS and 3,298 patients were treated with EGFR TKI only. Of the included studies, 13 were retrospective cohort studies while one was a retrospective analysis of the BR.21 phase III clinical trial (29). Four studies analyzed patients treated with first-line therapy, 6 studies analyzed patients with at least 1 prior treatment, while the remaining 4 did not report the line of therapy. The characteristics of included articles including patient demographics, details of AS and EGFR TKI treatment and risk of bias assessment can be found in .

Figure 1

PRISMA flow diagram of included articles.

Table 1

Characteristics of included studies

Study (year)	Approach	N	Patient characteristics	Number of patients with metastasis	Number of smokers	EGFR mutations	Type of AS and EGFR TKI and overlap between AS and EGFR TKI therapy	First-line EGFR TKI	ROBINS-I or RoB 2 overall risk of bias
Hilton (2013)	AS	190	119 male, 71 female; Age >60: 118 (62%)	–	Active smoker at baseline: AS 108 (57%), non-AS 155 (53%); never smoked: AS 46 (24%), non-AS 58 (20%)	–	PPI or H2RA; Erlotinib. Controlled for using proportional hazards model. All patients either took AS at baseline or midway through EGFR therapy	After failure of standard chemotherapy	Low
	Non-AS	295	194 male 101 female; Age >60: 161 (55%)	–	–	–	Erlotinib	After failure of standard chemotherapy
Chu (2015)	AS	124	56 male, 68 female; Median age: 64	–	–	–	PPI or H2RA; Erlotinib. 100 100% overlap (81%), 5 80–99% (4%), 4 60–79% (3%), 3 40–59% (2%), 12 20–39% (0%)	5 (4%)	Low
	Non-AS	383	179 male, 204 female; Median age: 65	–	–	–	Erlotinib	16 (4%)
Lam (2016)	AS	24	12 male, 12 female; Median age (range): 66 (33–84)	Any site: 15 (62.5%); Brain: 7 (29.2%)	–	13 activating known status (29.2%) mutation (54.2%), 5 wild type (20.8%), 7 unknown status (29.2%);	PPI or H2RA; Erlotinib. Patients who were prescribed acid-suppression therapy within the date range of erlotinib therapy or within 90 days of the initiation or discontinuation of erlotinib therapy were included in the concurrent acid-suppression group	8 (33.3%)	Moderate
	Non-AS	52	31 male, 21 female; Median age (range): 67 (46–90)	Any site: 29 (55.7%); Brain: 17 (32.7%)	–	30 activating mutation (57.7%), 11 wild type (21.2%), 7 unknown status (29.2%)	Erlotinib	22 (42.3%)
Miyazaki (2016)	AS	11	2 male, 9 female; Median age (range): 74 (52–88)	–	AS: 11 never smokers, 0 current/former smoker; non-AS: 31 never smokers, 4 current/former smokers	9 exon 19 deletion, 2 exon 21 L858R(100% EGFR-mutation positive)	Erlotinib or gefitinib	–	Moderate
	Non-AS	35	9 male, 26 female; Median age (range): 77 (52–89)	–	–	15 Exon 19 deletion, 18 exon 21 L858R (94.3% EGFR-mutation positive), 2 others	Erlotinib or gefitinib	–
Zenke (2016)	AS	47	17 male, 30 female; Median age (range): 62 (45–84)	Bone: 25 (53%), Lungs: 10 (21%), Pleural: 11 (23%), Brain: 9 (19%), Liver: 5 (10%)	30 never smokers (64%), ever smokers 17 (36%);	25 exon 19 deletion (53%), 19 exon 21 L858R mutation (40%), 3 other mutation (7%)	PPI or H2RA; Erlotinib or gefitinib. Duration of AS use (months): median 12.1 (1.0–32.1); AS combination time: concurrent 40 (85.1%); sequential 7 (14.9%)	21 (45%)	Moderate
	Non-AS	83	31 male, 52 female; Median age (range): 64 (36–87)	Bone: 17 (20%), Lungs: 26 (31%), Pleural: 22 (25%), Brain: 14 (14%), Liver: 4 (4%)	53 never smokers (64%), 30 ever smokers (36%)	35 exon 19 deletion (42%), 45 exon 21 L858R mutation (54%), 3 other mutation (4%)	Erlotinib or gefitinib.	33 (40%)
Chen (2016)	AS	57	29 male, 28 female; Mean age (SD): 66.6 (14.2)	Bone: 25 (43.9%), Pleural: 33 (57.9%), Brain: 18 (31.6%), Liver: 12 (21.1%)	–	Deletions in exon 19 and the L858R mutations 51 (89.5%), uncommon 6 (10.5%)	PPI or H2RA; Erlotinib or gefitinib. Patients who exhibited >30% overlap between the use of TKIs and antacids were considered antacid users	All	Low
	Non-AS	212	84 male, 128 female; Mean age (SD): 64.6 (11.7)	Bone: 94 (44.3%), Pleural: 96 (45.3%), Brain: 46 (21.7%), Liver: 23 (10.8%)	--	Deletions in exon 19 and the L858R mutations 191 (90.1%), uncommon 21 (9.9%)	Erlotinib or gefitinib.	All
Kumaraku-langsinghe (2016)	AS	56	27 male, 21 female; Mean age (SD): 61.7 (9.8)	Brain: 34 (61.8%), Liver: 11 (20.0%)	37 never smokers, 6 ever smokers	All	PPI or H2RA; Erlotinib or gefitinib. Patients were classified as AS-users if the periods of AS and anti-EGFR therapy overlapped by ≥30%	–	Low
	Non-AS	102	48 male, 45 female; Mean age (SD): 62.0 (10.8)	Brain: 28 (27.5%), Liver: 19 (18.6%)	67 never smokers, 22 ever smokers	All	Erlotinib or gefitinib.	–
Lizuka(2017)	AS	29	–	–	–	All	Gefitinib	–	Moderate
	Non-AS	34	–	–	–	All	Gefitinib	–
Sedano (2018)	AS	118	75 male, 43 female; Mean age (SD): 70.61 (10.97)	Brain: 24 (20.3%)	–	51 EGFR mutation (43.2%), 28 no mutation (23.7%), 39 NA (33.1%)	PPI or H2RA; Erlotinib or gefitinib. Patients were classified as AS-users if the periods of As and TKI therapy overlapped by ≥20%	26 (22.03%)	Low
	Non-AS	45	30 male, 15 female; Mean age (SD): 67.13 (10.84)	Brain: 5 (11.1%)	–	13 EGFR mutation (18.1%), 14 no mutation (31.1%), 18 NA (40.0%)	Erlotinib or gefitinib	11 (24.44%)
Fang (2019)	AS	309	105 males, 204 females; Age ≤65: 182, Age >65: 127	–	–	All	PPI only; Gefitinib. Duration of PPI treatment in days divided by the duration of TKI treatment in days. Patients who exhibited an overlap of >20% between PPI and TKI usage days were defined as having a high coverage ratio	All	Moderate
	Non-AS	969	346 males, 623 females; Age ≤65: 558, Age >65: 411	–	–	All	Gefitinib	All
Guo (2020)	AS	49	19 males, 30 females; Median age (range): 63 (49–72)	Any site: 47 (95.9%), Brain: 11 (22.4%), Liver: 3 (10.1%),	34 never smokers, 15 ever smokers	EGFR19 24 (49.0%), EGFR 21 19 (38.8%), other 6 (12.2%)	PPI or H2RA; Gefitinib. Duration of overlap between AS and EGFR TKI: 76–100%, 13 (26.5%); 51–75%, 18 (36.7%); 26–50%, 12 (24.5%), 0–25%, 6 (12.2%)	–	Moderate
	Non-AS	139	60 males, 79 females; Median age (range): 60 (51–75)	Any site: 131 (96.4%), Brain: 30 (21.6%), Liver: 14 (10.1%)	90 never smokers, 49 ever smokers	EGFR19 72 (51.8%), EGFR21 59 (42.4%), other 8 (5.8%)	Gefitinib.	–
Saito (2020)	AS	31	18 males, 13 females; Median age (range): 61 (37–87)	–	11 never smokers, 20 ever smokers	All	H2RA only; Gefitinib. Concurrent administration with H2RAs	20 (64.5%)	Moderate
	Non-AS	56	21 males, 35 females; Median age: 64 (51–82)	–	30 never smokers, 26 ever smokers	All	Gefitinib.	42 (75%)
Kwok (2020)	AS	61	PPI: 10 males, 17 females; H2RA: 14 males, 20 females; PPI Median age (range): 71.9 (40–88); H2RA Median age (range): 73.3 (57–88)	–	PPI: 22 never smoker 22 (81.5%), 5 ever smoker 5 (18.5%); H2RA 26 never smoker, 8 ever smoker	PPI: Exon 19 deletion (56%); L858R mutation (44%); H2RA: Exon 19 deletion (56%); L858R mutation (44%)	PPI or H2RA; Gefitinib. Patients who took H2RA and PPI for more than 75% of the duration of their overall gefitinib treatment period, satisfied the definition of “co-administration”	All	Moderate
	Non-AS	132	31 males, 101 females; Median age (range): 66.3 (35–92)	–	108 never smoker (82%), 24 ever smoker (18%)	Exon 19 deletion (54%); L858R mutation (46%)	Gefitinib	All
Su (2020)	AS	92	310 male, 543 female; Mean age (SD) =65.5 (12.6)	332	193	All	PPI or H2RA; Erlotinib, gefitinib or afatinib. Patients prescribed ≥28 cumulative defined daily doses (cDDDs) of each group of co-medication within the first 3 months of receiving EGFR-TKI therapy were assigned to respective co-medication groups	All	Low
	Non-AS	761				All	–	All

SD, standard deviation; AS, acid suppression; PPI, proton pump inhibitor; H2RA, histamine type-2 receptor antagonist; EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor; ECOG, eastern cooperative oncology group; DM, diabetes mellitus.

Proportion of patients with concurrent AS

The overall pooled prevalence of AS in patients taking EGFR TKI was 30.71% (95% CI: 23.28–39.31%; ) in 4,010 individuals. A sensitivity analysis was conducted to observe the rate of PPI and H2RA in NSCLC. The prevalence of PPI in EGFR TKI patients was 19.33% (95% CI: 13.01–27.73%), while the prevalence of H2RA was 25.13% (95% CI: 14.76–39.42%).

Figure 2

Overall pooled prevalence of AS in patients taking EGFR TKI. AS, acid suppression; EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor.

Survival outcomes

Patients who were treated only with EGFR TKI had significantly better OS (HR =1.46, 95% CI: 1.27–1.72; P<0.00001; ) in 3,694 patients. The funnel plot for OS was symmetrical (Figure S1). Similarly, PFS favored patients who were treated with EGFR TKI only (HR =1.63, 95% CI: 1.35–1.98; P<0.00001; ) in 2,433 patients. The funnel plot for PFS was symmetrical revealing no publication bias (Figure S1). A sensitivity analysis was conducted to include only EGFR mutation positive patients. The OS for EGFR mutation positive patients among 2,544 patients was as similarly significant as the OS for all patients taking EGFR TKI (HR =1.50, 95% CI: 1.13–1.99, P=0.005; ). However, AS significantly worsened the PFS among 350 EGFR mutation positive patients (HR =2.19, 95% CI: 1.34–3.59, P=0.002; ).

Figure 3

Forest plot for OS. OS, overall survival; AS, acid suppression.

Figure 4

Forest plot for PFS. PFS, progression-free survival; AS, acid suppression.

Figure 5

Forest plot for OS sensitivity analysis of EGFR-mutation positive patients. OS, overall survival; EGFR, epidermal growth factor receptor; AS, acid suppression.

Figure 6

Forest plot for PFS sensitivity analysis of EGFR-mutation positive patients. PFS, progression-free survival; EGFR, epidermal growth factor receptor; AS, acid suppression.

A subgroup analysis was conducted to compare OS and PFS between PPI and H2RA groups. The OS HR for PPI was 1.98 (95% CI: 1.33–2.94, P=0.0007) in 1,114 patients and the OS HR for H2RA was 1.04 (95% CI: 0.70–1.55, P=0.28) in 253 patients. There was significant difference between the two groups (P=0.03). Similarly, the PFS for PPI in 159 patients was significantly different (HR =3.39, 95% CI: 2.18–5.26, P<0.00001) The PFS for H2RA was however, not significantly different in 253 patients (HR =1.48, 95% CI: 0.63–3.49, P=0.37). The subgroup difference was not statistically significant (P=0.09).

Adverse effects

There were no significant differences in the incidence of rash (OR =0.81, 95% CI: 0.50–1.32; P=0.40; ) and diarrhoea (OR =1.03, 95% CI: 0.63–1.67; P=0.91; ) between AS and non-AS users. Other adverse effects reported include vomiting (30), loss of appetite (30), oral ulcers (30), stomatitis (31), elevated aminotransferase (AST/ALT) (30,31), interstitial lung disease (30-32), were not significantly different. However, Hilton et al. reported that the rate of infection was higher in patients with AS (33.7% AS vs. 20% non-AS, P=0.0008) (29). The summary of adverse effects can be found in Table S2.

Figure 7

Forest plot for adverse effects. AS, acid suppression.

Discussion

With the prevalence of AS use in cancer patients to treat gastric irritation (14) and the effect of AS in EGFR-TKI therapy (8), it is vital to study the interactions between AS drugs and EGFR-TKIs in NSCLC patients. This meta-analysis of high quality retrospective studies comparing the use of EGFR-TKIs with and without AS has shown that AS was strongly associated with poor OS and PFS outcomes. In our study, 30.71% of predominately NSCLC patients were concurrently treated with AS drugs. OS (HR =1.46, 95% CI: 1.27–1.67; P<0.00001; ) and PFS (HR =1.63, 95% CI: 1.35–1.98; P<0.00001; ) both favored patients treated with EGFR TKI only.

Similar to a previous meta-analysis on the impact of concurrent administration of EGFR TKIs and AS on various cancers (33), this meta-analysis found better survival outcomes for OS and PFS when only EGFR-TKIs were used. When combined with AS medications that lower the gastric pH, the alterations in pharmacokinetics of oral EGFR-TKIs led to a decrease in absorption and bioavailability of these drugs, possibly increasing the risk of disease progression and eventually poorer survival. This is despite no significant differences in the adverse reactions including rash (OR =0.81, 95% CI: 0.50–1.32; P=0.40) and diarrhoea (OR =1.03, 95% CI: 0.63–1.67; P=0.91). However, it is to be noted that those on AS might concurrently be on other medications due to other comorbidities that have an unknown impact on the EGFR TKIs in the NSCLC patient. These comorbidities may also have an adverse effect on the survival of the patient, leading to a decreased OS and PFS in those on AS. With the frequency of over-prescribing AS for gastric issues, it might prove to be valuable to practice caution when prescribing these medications for patients on EGFR-TKI for NSCLC.

However, the effect sizes in this study may be underestimated as some of the included studies analyzed patients with wild-type EGFR or EGFR of unknown mutational status (34-36). EGFR mutation positive patients have been reported to be 10 to 50-fold more sensitive to gefitinib (37,38). Moreover, some studies included patients receiving afatinib (32,39), which has an absorption that has not been shown to be affected by AS. The sensitivity analysis performed for EGFR mutation positive patients suggests a larger effect size for PFS but there was a lack of appreciable difference in OS potentially due to the use of erlotinib and afatinib in Su et al. (39). Additionally, most of the included studies analyzed patients receiving gefitinib which could be less affected than erlotinib since erlotinib is recommended to be taken with food and AS medications are usually taken before food as well (8). However, this point is not to be overstated since AS medications such as PPI achieve long duration of suppression.

Interestingly, clinical outcomes were observed to be dependent on the inverse relationship between levels of AS and the plasma levels of EGFR TKIs. van Leeuwen et al. found that AUC of erlotinib decreased by 15% to 33% when ranitidine dosage was increased from 150 to 300 mg and the AUC decreased even further by 46% when omeprazole was given (40). In our subgroup analysis of PPI and H2RA, PPI resulted in significantly worse OS and PFS and there was also a significant difference between PPI’s and H2RA’s effect on OS. H2RA did not significantly worsen both OS and PFS and seems to have a limited impact on the efficacy of EGFR TKI. In the study by Fang et al. (41), higher coverage of PPI resulted in poorer OS and PFS (high coverage HR: 1.67, P<0.001; low coverage ratio HR: 1.29, P=0.027). Similarly, Kwok et al. (42) also found that PPI had stronger acid suppressive effects had a negative impact on OS and PFS and H2RA only had appreciable negative effect on PFS. Potentially, the strength of AS and duration of action has a linear relationship with efficacy of EGFR TKI. Correspondingly, H2RAs which has a shorter duration of action and achieves a lower intragastric pH than PPIs has not been shown to negatively affect EGFR TKI efficacy in the study by Saito et al. (43). While it can be hypothesised that only PPI has an impact on the efficacy of EGFR TKI, the subgroup analysis is limited by its small sample size and warrants further investigation into how the strength of AS affects EGFR TKI efficacy.

As EGFR TKIs continue to be used as first line in the treatment of NSCLC and AS drugs continue to be prescribed for various conditions, it is vital to explore the impact on long-term outcomes and safety profile of these drug-drug interactions. There is also limited data available on the pharmacokinetics of the dose-dependent reactions of AS on patients taking EGFR-TKIs, hence such data is necessary to find out whether a lower dose of AS may reduce gastrointestinal discomfort while also having an insignificant impact on therapeutic outcomes.

The findings of this study are limited to NSCLC patients receiving first-generation EGFR TKIs (erlotinib and gefitinib). The clinical effect of AS on afatinib efficacy has not been delineated and past studies have also reported that afatinib remained soluble across a wide pH range of 1–7 (33). Other EGFR TKIs such as second-generation dacomitinib and third generation osimertinib were also not analyzed in this study. Osimertinib is part of the current first-line therapy for advanced NSCLC patients with EGFR-sensitizing mutations and the pharmacokinetics of osimertinib has not been shown to be affected by AS. When compared to taking osimertinib alone, the AUC and Cmax of taking with omeprazole gave a geometric least-squares mean ratio of 106.05% and 92.75% respectively with the confidence intervals falling within the equivalence limits of 80–125% (6). Despite the potential effectiveness of osimertinib in patients concurrently treated with AS medications, the use of 1st generation EGFR TKIs (erlotinib and gefitinib) as initial treatment is still ubiquitous in many parts of the world due to its comparative affordability and accessibility. There was also an inability to perform subgroup analysis based on the type of EGFR TKI due to insufficient granularity on the oncological agent adopted in the included studies. Furthermore, as an intrinsic limitation of retrospective studies, AS therapy was not randomized and heterogeneity could be introduced in terms of patient baseline characteristics such as performance status, smoking history, histology of NSCLC and presence of metastasis which were not fully controlled for since slightly different covariates were being adjusted for across the studies.

In summary, this meta-analysis delineates that co-administration of 1st generation EGFR-TKIs in NSCLC and AS medications worsens therapeutic outcomes, with a significantly better OS and PFS with only EGFR-TKIs. However, more data is needed regarding specific AS drugs, the characteristics of the cancer, and possible dose dependent reactions that AS medications may have on survival outcomes and the side effects. In current practice, physicians should err on the side of caution when prescribing patients undergoing anti-cancer treatment with EGFR-TKIs for NSCLC, and only prescribe such medications when absolutely clinically indicated and opt for a lower strength of AS.

The article’s supplementary files as