[Skip to Navigation]
Sign In
Figure 1.  Progression-Free Survival (PFS) and Overall Survival (OS)
Progression-Free Survival (PFS) and Overall Survival (OS)

Patients in the group receiving concomitant proton pump inhibitors (PPI concomitant group) were propensity score matched with the patients who were not receiving PPIs (nonconcomitant PPI group).

Figure 2.  Subgroup Analysis of Progression-Free Survival (PFS)
Subgroup Analysis of Progression-Free Survival (PFS)

Patients receiving concomitant proton pump inhibitors (PPI concomitant group) and those who were not (nonconcomitant PPI group) were stratified between those who received endocrine-sensitive or endocrine-resistant therapy.

Figure 3.  Subgroup Analysis of Overall Survival (OS)
Subgroup Analysis of Overall Survival (OS)

Patients receiving concomitant proton pump inhibitors (PPI concomitant group) and those who were not (nonconcomitant PPI group) were stratified between those who received endocrine-sensitive or endocrine-resistant therapy.

Table 1.  Baseline Characteristics Before and After 1:3 Propensity Score Matchinga
Baseline Characteristics Before and After 1:3 Propensity Score Matchinga
Table 2.  HRs of Clinical PFS and OS
HRs of Clinical PFS and OS
1.
Budha  NR, Frymoyer  A, Smelick  GS,  et al.  Drug absorption interactions between oral targeted anticancer agents and PPIs: is pH-dependent solubility the Achilles heel of targeted therapy?   Clin Pharmacol Ther. 2012;92(2):203-213. doi:10.1038/clpt.2012.73 PubMedGoogle ScholarCrossref
2.
Ha  VH, Ngo  M, Chu  MP, Ghosh  S, Sawyer  MB, Chambers  CR.  Does gastric acid suppression affect sunitinib efficacy in patients with advanced or metastatic renal cell cancer?   J Oncol Pharm Pract. 2015;21(3):194-200. doi:10.1177/1078155214527145 PubMedGoogle ScholarCrossref
3.
Meriggi  F.  Controversial link between proton pump inhibitors and anticancer agents: review of the literature.   Tumori. 2022;108(3):204-212. doi:10.1177/03008916211025091 PubMedGoogle ScholarCrossref
4.
Smelick  GS, Heffron  TP, Chu  L,  et al.  Prevalence of acid-reducing agents (ARA) in cancer populations and ARA drug-drug interaction potential for molecular targeted agents in clinical development.   Mol Pharm. 2013;10(11):4055-4062. doi:10.1021/mp400403s PubMedGoogle ScholarCrossref
5.
Raoul  JL, Edeline  J, Simmet  V, Moreau-Bachelard  C, Gilabert  M, Frénel  JS.  Long-term use of proton pump inhibitors in cancer patients: an opinion paper.   Cancers (Basel). 2022;14(5):1156. doi:10.3390/cancers14051156 PubMedGoogle ScholarCrossref
6.
Spring  LM, Wander  SA, Andre  F, Moy  B, Turner  NC, Bardia  A.  Cyclin-dependent kinase 4 and 6 inhibitors for hormone receptor-positive breast cancer: past, present, and future.   Lancet. 2020;395(10226):817-827. doi:10.1016/S0140-6736(20)30165-3 PubMedGoogle ScholarCrossref
7.
Kwapisz  D.  Cyclin-dependent kinase 4/6 inhibitors in breast cancer: palbociclib, ribociclib, and abemaciclib.   Breast Cancer Res Treat. 2017;166(1):41-54. doi:10.1007/s10549-017-4385-3 PubMedGoogle ScholarCrossref
8.
Del Re  M, Omarini  C, Diodati  L,  et al.  Drug-drug interactions between palbociclib and proton pump inhibitors may significantly affect clinical outcome of metastatic breast cancer patients.   ESMO Open. 2021;6(5):100231. doi:10.1016/j.esmoop.2021.100231 PubMedGoogle ScholarCrossref
9.
Eser  K, Önder  AH, Sezer  E,  et al.  Proton pump inhibitors may reduce the efficacy of ribociclib and palbociclib in metastatic breast cancer patients based on an observational study.   BMC Cancer. 2022;22(1):516. doi:10.1186/s12885-022-09624-y PubMedGoogle ScholarCrossref
10.
Reis  J, Costa  I, Costa  M,  et al.  Impact of drug-drug interaction between CDK4/6 inhibitors and proton pump inhibitors on survival outcomes in the treatment of metastatic breast cancer—real world data from a Portuguese center.   J Cancer Ther. 2022;13:266-274. doi:10.4236/jct.2022.135022 Google ScholarCrossref
11.
Çağlayan  D, Koçak  MZ, Geredeli  Ç,  et al.  The effect of concomitant use of proton pump inhibitors with CDK 4/6 inhibitors on survival in metastatic breast cancer.   Eur J Clin Pharmacol. 2023;79(2):243-248. doi:10.1007/s00228-022-03435-7 PubMedGoogle ScholarCrossref
12.
Peabody  JW, Lee  SW, Bickel  SR.  Health for all in the Republic of Korea: one country’s experience with implementing universal health care.   Health Policy. 1995;31(1):29-42. doi:10.1016/0168-8510(94)00669-6 PubMedGoogle ScholarCrossref
13.
Suissa  S.  Immortal time bias in pharmaco-epidemiology.   Am J Epidemiol. 2008;167(4):492-499. doi:10.1093/aje/kwm324 PubMedGoogle ScholarCrossref
14.
Mahtani  RL, Parisi  M, Glück  S,  et al.  Comparative effectiveness of early-line nab-paclitaxel vs paclitaxel in patients with metastatic breast cancer: a US community-based real-world analysis.   Cancer Manag Res. 2018;10:249-256. doi:10.2147/CMAR.S150960 PubMedGoogle ScholarCrossref
15.
Campbell  BA, Scarisbrick  JJ, Kim  YH, Wilcox  RA, McCormack  C, Prince  HM.  Time to next treatment as a meaningful endpoint for trials of primary cutaneous lymphoma.   Cancers (Basel). 2020;12(8):2311. doi:10.3390/cancers12082311 PubMedGoogle ScholarCrossref
16.
Labaki  C, Bakouny  Z, Sanglier  T,  et al.  Real-world progression-free survival (rwPFS) and time to next line of therapy (TTNT) as intermediate endpoints for survival in metastatic breast cancer: a real-world experience.   J Clin Oncol. 2022;40(16)(suppl):6520. doi:10.1200/JCO.2022.40.16_suppl.6520 Google ScholarCrossref
17.
Morgan  CJ.  Landmark analysis: a primer.   J Nucl Cardiol. 2019;26(2):391-393. doi:10.1007/s12350-019-01624-z PubMedGoogle ScholarCrossref
18.
Freedberg  DE, Lebwohl  B, Abrams  JA.  The impact of proton pump inhibitors on the human gastrointestinal microbiome.   Clin Lab Med. 2014;34(4):771-785. doi:10.1016/j.cll.2014.08.008 PubMedGoogle ScholarCrossref
19.
Moreau-Bachelard  C, Letailleur  V, Bompas  E, Soulié  P, Paul  J, Raoul  JL.  Effect of concomitant proton pump inhibitors with pazopanib on cancer patients: a retrospective analysis.   Cancers (Basel). 2022;14(19):4721. doi:10.3390/cancers14194721 PubMedGoogle ScholarCrossref
20.
Sharma  M, Holmes  HM, Mehta  HB,  et al.  The concomitant use of tyrosine kinase inhibitors and proton pump inhibitors: prevalence, predictors, and impact on survival and discontinuation of therapy in older adults with cancer.   Cancer. 2019;125(7):1155-1162. doi:10.1002/cncr.31917 PubMedGoogle ScholarCrossref
21.
Raoul  JL, Moreau-Bachelard  C, Gilabert  M, Edeline  J, Frénel  JS.  Drug-drug interactions with proton pump inhibitors in cancer patients: an underrecognized cause of treatment failure.   ESMO Open. 2023;8(1):100880. doi:10.1016/j.esmoop.2023.100880 PubMedGoogle ScholarCrossref
22.
Del Re  M, Crucitta  S, Omarini  C,  et al.  Concomitant administration of proton pump inhibitors does not significantly affect clinical outcomes in metastatic breast cancer patients treated with ribociclib.   Breast. 2022;66:157-161. doi:10.1016/j.breast.2022.10.005 PubMedGoogle ScholarCrossref
4 Comments for this article
EXPAND ALL
Palbociclib in research articles should always specify a formulation due to differing pharmacokinetics
Calum Polwart, BSc(Hons) MSc | South Tees NHS Foundation Trust
I read with interest the recent observations of Lee et al(1) on concomitant use of Proton Pump Inhibitors (PPIs) with palbociclib. While their observation is interesting, given the reformulation of palbociclib, the finding is alarmist and probably should not influence contemporaneous clinical practice. Additionally, as with all retrospective analysis, there are significant limitations which in this case the authors do not appear to have fully discussed.

As stated in the article, palbociclib was originally formulated as a capsule but is now formulated as a tablet. The absorption kinetics of the tablet are different(2) and so the postulated
interaction is unlikely to remain clinically relevant. The formulation should be specifically mentioned.

Further, food can influence  the absorption of palbociclib capsules(3), which is a confounder that could not be accounted for in the study, but was specifically included in the product label for the capsule but not tablet formulation(2,4). It was also surprising that the authors failed to make comparison between their cohort and the cohort in the pivotal trials(5), who had a higher proportion of patients with metastatic disease but are otherwise difficult to compare due to choices in descriptive statistics.

It is important to remember that this article shows association not causation. Patients who require a PPI may have a higher probability of death unrelated to interaction; this may merely be a predictor of poor prognosis. The authors have used the Charlson Comorbidity Index (CCI) to try to match clinical characteristics. CCI includes age, local versus metastatic cancer and peptic ulcer disease as inputs to predict 10 year mortality. However, this may result in potential double counting of these factors which they already were case matching for, at least to some extent. ECOG Performance Status is not reported.

The paper uses an index point (the point of commencing PPI) which makes comparison to general published literature on the outcomes seen with palbociclib impossible, even though it would have been useful. The linked cases were indexed at the same point; although this methodology is not referenced to any peer reviewed statistical methodology.

Finally, the paper also fails to describe subsequent treatment. When the outcome being measured is overall survival, the management of patients after study drug remains as important as the management prior to study drug; which was described. Nor was overall dose intensity described which may be important.

I would encourage the authors to revise the title and conclusion of their article to specifically mention capsules.

References
1. Lee JE, Kwon SH, Kwon S, Jung HI, Nam JH, Lee EK. Concomitant Use of Proton Pump Inhibitors and Palbociclib Among Patients With Breast Cancer. JAMA Netw Open. 2023 Jul 21;6(7):e2324852.
2. Prescribing information: IBRANCE® (palbociclib) tablets. [Internet]. Food and Drug Administration, USA; 2019 [cited 2023 Jul 29].
3. Sun W, Klamerus KJ, Yuhas LM, Pawlak S, Plotka A, O’Gorman M, et al. Impact of Acid-Reducing Agents on the Pharmacokinetics of Palbociclib, a Weak Base With pH-Dependent Solubility, With Different Food Intake Conditions. Clin Pharmacol Drug Dev. 2017;6(6):614–26.
4. Prescribing information: IBRANCE® (palbociclib) capsules [Internet]. Food and Drug Administration, USA; 2019 [cited 2023 Jul 29].
5. Finn RS, Martin M, Rugo HS, Jones S, Im SA, Gelmon K, et al. Palbociclib and Letrozo

CONFLICT OF INTEREST: None Reported
READ MORE
Author Response: Concomitant use of proton pump inhibitors and palbociclib capsule
Sun-Hong Kwon, PhD | Sungkyunkwan University
As Polwart’ commented, we observed the capsule of palbociclib, and it was mentioned in Lee et al. (2023) as shown below:

[Methods] “During the study period, we did not observe the use of Palbociclib as a tablet, but a capsule. This is because in South Korea, the palbociclib tablet was approved only in February 2022.”, “In the study period, the palbociclib taken by all patients was in capsule formulation.”

[Discussion] “As mentioned, the Palbociclib taken by the patients in this study was limited to the capsule formulation.”

According to the prescribing information of IBRANCE® (palbociclib) tablets,
which Polwart (2023) refers to, an effect of acid-reducing agents on palbociclib exposure is not expected. In contrast, the capsule reported the drug interaction with gastric pH elevation medications regarding pharmacokinetics. Our study is the first large population-based approach for measuring the association between palbociclib capsule and proton-pump inhibitor (PPI). In this study, the association between palbociclib tablets and PPI was not confirmed, and it was also stated in Lee et al. (2023). This result should be interpreted as the association of palbociclib capsule.

We confirmed the effect of the concomitant use of PPI on the outcome by matching the factors expected to affect the outcome as much as possible between the two groups. CCI corrects severity by using the weight of comorbidity or the presence of disease. Variables such as age or metastatic disease, which were mentioned as being at risk of double counting with CCI, did not potentially affect the results, as the proportions before and after matching were not significantly different. Also, we checked the multicollinearity of all variables when using a logistic regression to estimate propensity score. There was no multicollinearity among the variables we used. It would have been better if variables like ECOG performance, etc. were used for propensity-matching. However, we conducted the analysis using claims data without clinical information. It is a known limitation that such clinical characteristics cannot be identified in the claims data.

Our study aimed to identify the association between “PPI intake” and treatment outcomes. Therefore, in the design of this study, the two groups were compared according to whether or not PPI was used together, and subsequent treatment after the concomitant use of medication was not specifically considered. There was no evidence that PPI affects the choice of following treatments for breast cancer. Instead, it seems possible to say that including subsequent treatments in analyses may cause confounding since the selection of subsequent treatments is based on the response of the previous treatment. Also, this study did not examine the clinical outcomes of palbociclib but examined the association with PPI. To achieve this purpose, the index date with initiating PPI is more proper than the date of administration of palbociclib as the index date. When exploring the association, results from both index dates were robust.
CONFLICT OF INTEREST: None Reported
READ MORE
The importance of taking palbociclib capsules with food
Grace Foley, MPharm | Grace Foley Global Medical Affairs Pfizer Inc.
Lee and colleagues (1) evaluated the association of palbociclib with concomitant use of proton pump inhibitors (PPIs) on real-world clinical outcomes in women with breast cancer using data from a South Korean claims database. The retrospective analysis, which evaluated claims data from November 2016 to July 2021, found shorter progression-free survival (PFS) and overall survival with concomitant PPIs and palbociclib versus palbociclib alone.

We would like to note an important educational point about palbociclib capsules which provides relevant clinical treatment context about the formulation taken by the patients in this analysis. While the authors state that only palbociclib
capsules were available during the study period, they make no specific reference to the drug-drug interaction between palbociclib capsules and PPIs being a known phenomenon that is explicitly addressed in the palbociclib capsule label. Under fasted conditions, the concomitant use of PPIs with palbociclib capsules decreased palbociclib exposure (AUCinf) by 62%, but only decreased exposure by 13% when given with food, which was not expected to have any clinically relevant effects as stated in the capsule label (2,3). Consequently, a recommendation to take palbociclib capsules with food is included in the label. A limitation of the study by Lee and colleagues not mentioned in the manuscript is that it is impossible to know whether patients were counselled appropriately on the importance of taking palbociclib capsules with food, or if they actually took palbociclib with food, as aspects around assessing adherence (including adherence to administration instructions) were not stated to be a component of this real-world study. As such, it is not possible to determine how many patients deviated from the recommendation to take the capsules with food during the course of treatment.

To address the authors’ comment related to the unknown relationship between concomitant use of PPIs with palbociclib tablets, we can confirm that subjects receiving concomitant PPIs with palbociclib tablets showed no impact on palbociclib exposure (4,5). Furthermore, another study reported that palbociclib tablets administered with a high fat meal compared with fasted conditions had no clinically meaningful impact on palbociclib exposure (4). Since palbociclib tablets have no food or PPI drug interaction effects, palbociclib tablets may be taken with or without food in all patients, including those receiving concomitant PPIs. A recent small retrospective study reported no significant difference in PFS between patients with HR+/HER2– metastatic breast cancer treated with first-line palbociclib tablets with or without a PPI (20.6 vs 21.0 months, respectively; P = 0.95) (6).

While the authors concluded that physicians should be cautious when prescribing concomitant PPIs and palbociclib, we would like to emphasize that this pertains particularly to palbociclib capsules (and not tablets), and that physicians should counsel their patients on the importance of taking palbociclib capsules with food.

Grace Foley, on behalf of the Pfizer Global Breast Cancer Team

1. Lee JE, et al. JAMA Netw Open. 2023;6:e2324852.
2. IBRANCE (palbociclib) capsules, for oral use, PI. In. New York: Pfizer Inc.
3. Sun W, et al. Clin Pharmacol Drug Dev. 2017;6:614–626.
4. IBRANCE (palbociclib) tablets, for oral use, PI. In. New York: Pfizer Inc.
5. Yu Y, et al. Clin Pharmacol Drug Dev. 2019;8(SUPPL 1):10–11.
6. Schieber T, et al. Clin Breast Cancer. 2023;23:658–663.
CONFLICT OF INTEREST: Grace Foley is an employee and stockholder in Pfizer Inc. Medical writing support was provided by Dominique J. Verlaan, PhD, of Oxford PharmaGenesis (Newtown, PA, USA) and funded by Pfizer Inc.
READ MORE
Author Response: Palbociclib included in the study is a capsule formulation.
Ju eun Lee, MS, PharmD | Sungkyunkwan University
We conducted our research including only the palbociclib capsules. As Foley mentioned, the research findings and key conclusions are specific to the capsule formulation of palbociclib.

Also, as mentioned in the limitations, the claims data cannot confirm the patient's exact medication compliance, including whether the medication was taken with food.
CONFLICT OF INTEREST: None Reported
Original Investigation
Oncology
July 21, 2023

Concomitant Use of Proton Pump Inhibitors and Palbociclib Among Patients With Breast Cancer

Author Affiliations
  • 1School of Pharmacy, Sungkyunkwan University, Gyeonggi-do, Republic of Korea
  • 2Division of Big Data Science, Korea University Sejong Campus, Sejong, Republic of Korea
JAMA Netw Open. 2023;6(7):e2324852. doi:10.1001/jamanetworkopen.2023.24852
Key Points

Question  Is concomitant use of palbociclib with a proton pump inhibitor (PPI) associated with a higher risk of progression among patients with advanced or metastatic breast cancer?

Findings  In this cohort study of 1310 South Korean women with breast cancer identified using nationwide claims data, progression-free survival and overall survival in the concomitant PPI group were shorter than those in the nonconcomitant PPI group.

Meaning  These findings suggest that taking PPIs with palbociclib may interrupt the full therapeutic benefits of palbociclib.

Abstract

Importance  Proton pump inhibitors (PPIs) are commonly used drugs to relieve gastrointestinal tract symptoms, but their acid-inhibitory action negatively affects the bioavailability and clinical outcomes of orally administered concomitant drugs.

Objective  To identify the clinical outcomes of patients with advanced breast cancer who concomitantly use PPIs and palbociclib.

Design, Setting, and Participants  This retrospective cohort study used nationwide claims data between November 1, 2016, and July 31, 2021, in South Korea. Patients with breast cancer receiving palbociclib between November 1, 2017, and July 31, 2020, were identified. Patients whose prescriptions for palbociclib and PPI overlapped by at least 33% were classified into a concomitant PPI group. Patients who never received PPI during the palbociclib treatment period were classified into a nonconcomitant PPI group. Patients were selected through 1:3 propensity score matching for analyses.

Exposures  Concomitant use of PPIs with palbociclib.

Main Outcomes and Measures  Time to progression and death. These outcomes were presented as progression-free survival (PFS) and overall survival (OS) and were analyzed using the Kaplan-Meier method and log-rank test. Cox proportional hazards regression was used to estimate the hazard ratio (HR) of concomitant PPI use associated with clinical PFS and/or OS.

Results  A total of 344 women were included in the concomitant PPI group and 966 in the nonconcomitant PPI group. Among 1310 patients identified after matching, 1108 (84.6%) were older than 50 years; 1111 (84.8%) were treated with letrozole and anastrozole (endocrine sensitive); and 199 (15.2%) were treated with fulvestrant (endocrine resistant). The median clinical PFS in the concomitant PPI group was shorter than that of the nonconcomitant PPI group (25.3 [95% CI, 19.6-33.0] vs 39.8 [95% CI, 34.9 to not applicable] months; P < .001), and the HR was 1.76 (95% CI, 1.46-2.13). Concomitant use of PPI was also associated with shorter OS (HR, 2.71 [95% CI, 2.07-3.53]). Both clinical PFS and OS in the concomitant PPI group were consistently poor in patients receiving endocrine-sensitive and endocrine-resistant treatment.

Conclusions and Relevance  These findings suggest that concomitant use of PPIs with palbociclib may hinder the complete therapeutic benefits of palbociclib in patients with breast cancer.

Introduction

Most patients with cancer use proton pump inhibitors (PPIs) to mitigate anticancer drug–related gastrointestinal symptoms, such as gastroesophageal reflux disease.1 Proton pump inhibitors irreversibly bind to and inhibit the hydrogen-potassium adenosine triphosphatase pump located on the luminal surface of the parietal cell membrane, reducing the secretion of gastric acid.2,3 However, acid suppression negatively affects the oral bioavailability, pharmacokinetics, and clinical effects of orally administered anticancer medications.4,5 For this reason, PPIs could be considered to have a high risk of drug-drug interaction with other anticancer drugs.

Cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors for hormone receptor (HR)–positive and ERBB2 (previously known as HER2)-negative advanced or metastatic breast cancer have changed the landscape of treatment for breast oncology.6 Palbociclib, an oral small molecular inhibitor of CDK4/6, has been recommended in combination with aromatase inhibitors or fulvestrant as a treatment for patients with HR-positive and ERBB2-negative advanced or metastatic breast cancer.7 Palbociclib is a weak base medication whose absorption and blood concentration can also be decreased by coadministering a PPI, resulting in poor clinical efficacy. The first US Food and Drug Administration approval of palbociclib as an oral capsule noted that concurrent PPI use reduced exposures of palbociclib based on 62% of the area under the plasma concentration time curve from time 0 to infinity and 80% of the maximum plasma palbociclib concentration. With regard to these associations, several related prior studies8-11 have reported that a combination of a PPI and palbociclib, a CDK4/6 inhibitor, changes the therapeutic effect (with shorter progression-free survival [PFS]) of patients with HR-positive, ERBB2-negative advanced and metastatic breast cancer.

To our knowledge, insufficient data and small sample size have limited retrospective studies reporting drug-drug interactions between PPIs and palbociclib. In addition, some studies10,11 have shown no statistically significant differences and have reported the small number of patients as a limitation. Therefore, studies targeting more patients are needed to investigate the effects of concurrent PPI and palbociclib administration. In this study, we aimed to identify the clinical outcomes of patients with HR-positive and ERBB2-negative advanced or metastatic breast cancer who concomitantly use PPIs and palbociclib, based on South Korean claims data.

Methods
Data Source and Ethics

For this cohort study, we obtained nationwide claims data from the Health Insurance Review and Assessment Service in South Korea. As health insurance in South Korea is provided as a single-payer system, the database contains information regarding medical treatment, medicines, and medical resources for a total population of 50 million people.12 We collected demographic characteristics such as age and sex and medical information such as disease diagnosis and medical drug use from the database to identify eligible patients. We used data from November 1, 2016, to July 31, 2021 (the study period). During the study period, we did not observe the use of palbociclib as a tablet, but a capsule. This is because in South Korea, the palbociclib tablet was approved only in February 2022. This study was approved by the institutional review board of Sungkyunkwan University, which did not require informed consent for the use of retrospective deidentifed data. The study followed the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) reporting guideline.

Study Design and Patient Selection

In this retrospective cohort study, we assessed the clinical outcomes of patients with breast cancer and concomitant use of PPIs with palbociclib. The patient identification period was from November 1, 2017—the reimbursement date for palbociclib—to July 31, 2020. In the study period, the palbociclib taken by all patients was in capsule formulation. We identified women with breast cancer based on the presence of at least 2 claims with code C50 from the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), during the study period. They were required to use palbociclib for at least 1 cycle, during which the drug was administered continuously for 21 days. We excluded patients who used drugs targeting ERBB2 such as trastuzumab, trastuzumab emtansine, pertuzumab, and lapatinib ditosylate during the study period. A detailed patient selection flow is shown in eFigure 1 in Supplement 1. Cohort entry was defined as the date of the first palbociclib prescription (eFigure 2 in Supplement 1). The date of first PPI use was defined as the index date. Patients were assessed from the index date for outcomes.

Baseline characteristics of the patients were assessed by index date (age, menopause, and treatment combination) or from 1 year prior to the index date (metastases, Charlson Comorbidity Index score, prior chemotherapy, and prior endocrine therapy) (eFigure 2 in Supplement 1). Palbociclib is approved in combination with an aromatase inhibitor as a first-line endocrine therapy and in combination with fulvestrant for disease progression after endocrine therapy. Both regimens are applicable to postmenopausal patients; however, they are only applicable for reimbursement to premenopausal women when both regimens are administered with ovarian function suppressants. As the claims data could not confirm the menopausal status of the patient, this was classified based on the use of ovarian function suppressants (leuprorelin or goserelin acetate).

Exposure

The concomitant PPI group was defined as those who were coadministered PPI for more than one-third of the palbociclib treatment duration, while those who did not take any PPI in this period were classified as the nonconcomitant PPI group. We assumed patient medication adherence based on the date of prescription and the number of days supplied for each prescription. The PPIs included in this study were dexlansoprazole (Anatomical Therapeutic Chemical [ACT] code A02BC06), esomeprazole magnesium (ACT code A02BC05), ilaprazole (ACT code A02BC), lansoprazole (ACT code A02BC03), omeprazole (ACT code A02BC01), pantoprazole sodium (ACT code A02BC02), and rabeprazole sodium (ACT code A02BC04).

We performed 1:3 propensity score matching by matching 1 patient from the concomitant PPI group with 3 patients from the nonconcomitant PPI group to balance the characteristics of both groups. Using a logistic regression model, we estimated propensity scores of patients, including age, menopause, treatment combination, Charlson Comorbidity Index, whether prior chemotherapy or endocrine therapy was administered, and whether metastases had occurred. We additionally conducted analyses in 1:5 matched patients to see the robustness of results with reducing the loss of patients.

Patients were followed up after the index date to avoid immortal time bias,13 including events that occurred before the first prescription of PPI from the cohort entry. The index date for non-PPI users was specified as the interval between their cohort entry date and the index date of matched PPI users. Patients were excluded if the index date assigned by the matched PPI users was after the last palbociclib prescription date (eFigure 3 in Supplement 1).

Outcomes

The outcomes measured were clinical PFS and overall survival (OS). As a measure of clinical PFS, we used the time to next treatment (TTNT), which estimates the time from the index date to the start of the next line of treatment or death. Time to next treatment has been used as a proxy for disease progression in patients since it is impossible to determine whether the patient’s disease has progressed in the claims data. As one might conclude that a patient commenced a new therapy because they experienced progression during their previous therapy, TTNT has been used as a proxy for time to progression in several analyses.14 By taking into account the progression of medication tolerability and patient adherence over time, TTNT provides a more accurate depiction of patient treatment experiences than traditional disease-related end points.15 It is also validated that TTNT may serve as a significant interim objective for the OS of patients with metastatic breast cancer.16 Therefore, TTNT was measured to represent clinical PFS in the study. Chemotherapy, immunotherapy, or endocrine therapy, such as tamoxifen, exemestane, and megestrol acetate, were defined as the next line of treatment if they were prescribed after the index date.

We estimated OS from the index date to last follow-up or death. We identified death if any one of ICD-10 codes I461, R96, R98, or R99 was recorded or when the result of treatment was coded as death.

Statistical Analysis
Main Analysis

Baseline characteristics are presented as mean (SD) for continuous variables and frequency for all categorical variables. A standardized mean difference was obtained to confirm whether the characteristics of the 2 groups were well balanced. In addition, the statistical significance of the difference between the concomitant and nonconcomitant groups was determined using unpaired t tests and χ2 tests, with a 2-sided P < .05 indicating statistical significance. We used the Kaplan-Meier method to generate survival curves and estimate patient median survival time. The log-rank test was used to compare the 2 groups. Hazard ratios (HRs) were determined using the Cox proportional hazard model, adjusting for other covariates. All statistical analyses were conducted using SAS Enterprise Guide, version 6.1 (SAS Institute Inc) and R, version 3.5 (R Project for Statistical Computing).

Subgroup and Sensitivity Analysis

We conducted subgroup analysis to compare the associations of concomitant PPI use among patients with different disease progression status. Since palbociclib is reimbursed in combination with an aromatase inhibitor as an initial endocrine-based therapy and with fulvestrant in patients with disease progression following initial endocrine therapy, disease in patients treated with fulvestrant is more advanced. To identify the outcomes of PPI after excluding patient disease factors, we assessed outcomes by classifying the patient groups according to whether they were sensitive to palbociclib plus letrozole or anastrozole or resistant to endocrine therapy consisting of palbociclib plus fulvestrant.

In the sensitivity analysis, we redefined the definition of the concomitant PPI group by changing the concomitant period to 50%, 67% (two-thirds of the period), and 80%. Matching with the nonconcomitant PPI group was repeated whenever the coverage ratio defining the concomitant group was changed. This analysis demonstrated how the results could vary depending on assumptions of the operational definition of the concomitant PPI group. In addition, we conducted a landmark analysis in which only patients who had survived until the landmark time were analyzed from those time points.17 The landmarks included were 3, 6, and 12 months after the initiation of the PPI. The robustness of the results was confirmed through landmark analyses.

Results

Table 1 presents the baseline demographic and clinical characteristics of the patients before and after propensity score matching. After matching the 1310 selected patients, 344 women were determined to be in the concomitant PPI group and 966 in the nonconcomitant PPI group (eFigure 1 in Supplement 1). The standardized mean differences for all patient characteristics, except for lung metastasis, were less than 0.1. Most of the patients (1108 [84.6%]) were older than 50 years and had attained menopause (1289 [98.4%]). Palbociclib was combined with nonsteroidal aromatase inhibitors such as anastrozole or letrozole in 1111 patients (84.8%). Most patients had not received chemotherapy (1298 [99.1%]) or endocrine therapy (1240 [94.7%]) prior to palbociclib. Half of the patients in both groups had metastases. Nearly one-third of patients in both groups (368 [28.1%]) developed bone metastasis, the most frequent site of metastasis in both groups.

The median clinical PFS of the concomitant PPI group was 25.3 [95% CI, 19.6-33.0] months, significantly shorter compared with 39.8 [95% CI, 34.9 to not applicable] months for the nonconcomitant PPI group (P < .001) (Figure 1A). The OS was also shorter in the concomitant PPI group than in the nonconcomitant PPI group. The difference between the 2 groups was statistically significant (1-year OS, 83.1% vs 94.0%; 2-year OS, 69.5% vs 89.3%; P < .001) (Figure 1B), even though the median OS in both the groups was not reached. Absolute risk differences were 15 per 100 person-years in clinical PFS and 11 per 100 person-years in OS (eTable 1 in Supplement 1). When other variables were adjusted, the HR for concomitant PPI use associated with clinical PFS was 1.76 (95% CI, 1.46-2.13) and the HR for OS was 2.71 (95% CI, 2.07-3.53) (Table 2 and eTable 2 in Supplement 1).

In the subgroup analysis (n = 344), 292 patients were classified as receiving endocrine-sensitive treatment and the remaining 52 patients were found to be receiving endocrine-resistant treatment. The median clinical PFS for the nonconcomitant PPI group of patients with endocrine-sensitive treatment was 40.4 [95% CI, 34.9 to not applicable] months, whereas the concomitant PPI group with endocrine-sensitive treatment had a median clinical PFS of 27.2 [95% CI, 20.6-34.0] months (P < .001). The HR for the endocrine-sensitive subgroup associated with clinical PFS was 1.75 (95% CI, 1.42-2.15) (Figure 2A). The endocrine-resistant subgroup also exhibited a substantial difference in clinical PFS between the 2 groups (adjusted HR, 1.82 [95% CI, 1.12-2.94]; P = .03) (Figure 2B). None of the groups reached a median OS. However, the difference between the 2 groups was statistically significant in all subgroups, and the HRs for taking PPIs with palbociclib associated with OS were 2.68 (95% CI, 2.01-3.58; P < .001) and 2.98 (95% CI, 1.49-5.96; P = .006) in the endocrine-sensitive and endocrine-resistant subgroups, respectively (Figure 3).

When we defined the concomitant PPI group by coverage ratios of 50%, 67%, and 80% as sensitivity analyses, the increased risk associated with clinical PFS and OS for taking PPI was robust in the sensitivity analysis (Table 2). A significant difference between the 2 groups was also consistently demonstrated. In patients who took a PPI for 80% or more of the duration of palbociclib treatment, concomitant PPI use increased the HR for clinical PFS to 1.91 (95% CI, 1.48-2.47) and for OS to 1.31 (95% CI, 2.23-4.90). The trend was also observed in the cohorts constructed with 3-, 6-, and 12-month landmarks (Table 2). In 1:5 matching that was performed to secure a somewhat larger number of patients, concomitant PPI use was associated with an increased risk to clinical PFS and OS (eTables 3-4 and eFigures 4-6 in Supplement 1).

Discussion

Several studies8-11 have demonstrated that using palbociclib concurrently with PPIs leads to inferior clinical results for patients, particularly with regard to PFS. We attempted to obtain a greater number of patients for analysis by using nationwide claims data to address the issue of limited sample sizes in previous studies. As an indicator of the results, we attempted to determine when the disease actually progressed and the next treatment was started. Additionally, we analyzed OS, which has been relatively unexplored in previous research.

We found that in patients with advanced or metastatic breast cancer, taking PPIs concomitantly with palbociclib was associated with considerably reduced clinical PFS and OS compared with those who did not take PPIs during the entire palbociclib treatment period, supporting the findings of earlier studies.8-11 In this study, the clinical PFS of patients with breast cancer receiving palbociclib and PPIs in combination was approximately 15 months shorter. After adjustment for other risk factors, the HR associated with taking PPI concomitantly appeared to be 1.76 for clinical PFS. Del Re et al8 reported that concomitant PPI increases progression by an HR of 2.77 (95% CI, 1.62-4.75), and Eser et al (95% CI, 2.67-23.05)9 reported an HR of 7.85. Compared with earlier research, the HR in our study was more conservative, but the results that concomitant use of PPIs was associated with the patient clinical outcomes were mostly consistent. While the early separation of PFS and OS were observed in our study, the trends were similar with PFS from the previous studies.8-10

Reis et al10 (2022) concluded that survival outcomes were not statistically significantly different between the 2 groups. However, the literature also implies that if a larger number of patients are followed up for a longer period of time, potential differences may become more apparent. From the previous literature, the nonconcomitant PPI group was as few as 40 patients, of whom fewer than 10 were included in each variable for subgroup analysis, making it difficult to determine whether the variables were significantly adjusted. Our study, which analyzed a larger number of patients, identified a statistically significant difference in survival outcomes (HR, 2.71). Results from landmark and sensitivity analyses also demonstrated conclusive evidence that the combination of PPI with palbociclib was associated with a higher risk of disease progression and death. We confirmed that concomitant PPI use was associated with an increase in the risk to clinical PFS and OS by performing subgroup analysis and classifying patients according to the combined treatment. Even when palbociclib was administered with fulvestrant, it appeared to significantly increase the risk of disease progression and death. This was an expected result because the treatment lines of the patients using palbociclib were different.

Proton pump inhibitors alter the gastrointestinal environment by decreasing gastric acid secretion, consequently influencing drug absorption.18 Based on the research that reported that patient clinical outcomes deteriorate when oral weakly basic anticancer medications such as tyrosine kinase inhibitors are used in combination with PPIs,19-21 it appears that complete absorption of palbociclib has been achieved, so it has not shown sufficient therapeutic performance. Even though the type of cancer studied was different from ours, this association was also confirmed from the previous studies.2,20 Ha et al2 reported a significant difference in PFS and OS between PPI and non-PPI groups in patients with advanced or metastatic renal cell cancer (40.9 vs 62.4 weeks; P = .02). Eser at al9 concluded that increasing gastric pH immediately induced by PPIs may occur through lowering palbociclib plasma concentrations, which affects treatment efficacy and results in shorter PFS. Likewise, it seems that the increase in gastric pH by PPI reduces the absorption and efficacy of weakly basic drugs such as palbociclib, which have pH-dependent solubility.

Similarly, we wondered whether PPI administration had similar associations with clinical outcomes for other weakly basic CDK4/6 inhibitors, such as ribociclib succinate and abemaciclib. Previous studies concerning palbociclib have also investigated the PPI effect of ribociclib, but the conclusions are inconsistent.22 Studies with larger numbers of patients are required to obtain reliable findings. As mentioned, the palbociclib taken by the patients in this study was limited to the capsule formulation. The relationship between concomitant PPI and palbociclib tablet formulation requires confirmation by further studies.

Limitations

Our study has certain limitations. Owing to the characteristics of the claims data used in our analysis, we could not confirm whether the patients actually took the medication. However, patients who received multiple prescriptions for palbociclib together with a PPI for more than a certain period of time were selected, which means that patients who received concomitant PPI only once were not selected. Moreover, in claims data, it is difficult to measure how much PPI should be taken together for drug-drug interactions to interfere with the treatment outcome. We attempted to overcome this uncertainty to some extent by performing a sensitivity analysis by varying the definition of the PPI group.

Conclusions

The findings of this cohort study suggest that concomitant PPI use with palbociclib in patients with breast cancer was associated with poorer treatment outcomes than among those not using PPIs. Physicians should be cautious when prescribing PPIs to patients who are receiving palbociclib and should inform patients about the risks of interaction to prevent inadequate prescription of PPI by others.

Back to top
Article Information

Accepted for Publication: June 8, 2023.

Published: July 21, 2023. doi:10.1001/jamanetworkopen.2023.24852

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2023 Lee JE et al. JAMA Network Open.

Corresponding Author: Eui-Kyung Lee, PhD, School of Pharmacy, Sungkyunkwan University, 2066, Seobu-ro, Jangan-gu, Suwon-si, Gyeonggi-do 16419, South Korea (ekyung@skku.edu).

Author Contributions: Ms J.-E. Lee and Dr S.-H. Kwon had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Ms J.-E. Lee and Dr S.-H. Kwon contributed equally as first authors.

Concept and design: All authors.

Acquisition, analysis, or interpretation of data: J.-E. Lee, S.-H. Kwon, Nam.

Drafting of the manuscript: J.-E. Lee, S.-H. Kwon, S. Kwon, Nam.

Critical review of the manuscript for important intellectual content: All authors.

Statistical analysis: J.-E. Lee, S.-H. Kwon, Jung, Nam.

Administrative, technical, or material support: J.-E. Lee, S.-H. Kwon, S. Kwon, E.-K. Lee.

Supervision: E.-K. Lee.

Conflict of Interest Disclosures: None reported.

Funding/Support: Support for the article publication fee was provided by Sungkyunkwan University and the Brain Korea 21 Program for Leading Universities & Students (Graduate School Innovation), funded by the Ministry of Education and National Research Foundation of Korea.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 2.

References
1.
Budha  NR, Frymoyer  A, Smelick  GS,  et al.  Drug absorption interactions between oral targeted anticancer agents and PPIs: is pH-dependent solubility the Achilles heel of targeted therapy?   Clin Pharmacol Ther. 2012;92(2):203-213. doi:10.1038/clpt.2012.73 PubMedGoogle ScholarCrossref
2.
Ha  VH, Ngo  M, Chu  MP, Ghosh  S, Sawyer  MB, Chambers  CR.  Does gastric acid suppression affect sunitinib efficacy in patients with advanced or metastatic renal cell cancer?   J Oncol Pharm Pract. 2015;21(3):194-200. doi:10.1177/1078155214527145 PubMedGoogle ScholarCrossref
3.
Meriggi  F.  Controversial link between proton pump inhibitors and anticancer agents: review of the literature.   Tumori. 2022;108(3):204-212. doi:10.1177/03008916211025091 PubMedGoogle ScholarCrossref
4.
Smelick  GS, Heffron  TP, Chu  L,  et al.  Prevalence of acid-reducing agents (ARA) in cancer populations and ARA drug-drug interaction potential for molecular targeted agents in clinical development.   Mol Pharm. 2013;10(11):4055-4062. doi:10.1021/mp400403s PubMedGoogle ScholarCrossref
5.
Raoul  JL, Edeline  J, Simmet  V, Moreau-Bachelard  C, Gilabert  M, Frénel  JS.  Long-term use of proton pump inhibitors in cancer patients: an opinion paper.   Cancers (Basel). 2022;14(5):1156. doi:10.3390/cancers14051156 PubMedGoogle ScholarCrossref
6.
Spring  LM, Wander  SA, Andre  F, Moy  B, Turner  NC, Bardia  A.  Cyclin-dependent kinase 4 and 6 inhibitors for hormone receptor-positive breast cancer: past, present, and future.   Lancet. 2020;395(10226):817-827. doi:10.1016/S0140-6736(20)30165-3 PubMedGoogle ScholarCrossref
7.
Kwapisz  D.  Cyclin-dependent kinase 4/6 inhibitors in breast cancer: palbociclib, ribociclib, and abemaciclib.   Breast Cancer Res Treat. 2017;166(1):41-54. doi:10.1007/s10549-017-4385-3 PubMedGoogle ScholarCrossref
8.
Del Re  M, Omarini  C, Diodati  L,  et al.  Drug-drug interactions between palbociclib and proton pump inhibitors may significantly affect clinical outcome of metastatic breast cancer patients.   ESMO Open. 2021;6(5):100231. doi:10.1016/j.esmoop.2021.100231 PubMedGoogle ScholarCrossref
9.
Eser  K, Önder  AH, Sezer  E,  et al.  Proton pump inhibitors may reduce the efficacy of ribociclib and palbociclib in metastatic breast cancer patients based on an observational study.   BMC Cancer. 2022;22(1):516. doi:10.1186/s12885-022-09624-y PubMedGoogle ScholarCrossref
10.
Reis  J, Costa  I, Costa  M,  et al.  Impact of drug-drug interaction between CDK4/6 inhibitors and proton pump inhibitors on survival outcomes in the treatment of metastatic breast cancer—real world data from a Portuguese center.   J Cancer Ther. 2022;13:266-274. doi:10.4236/jct.2022.135022 Google ScholarCrossref
11.
Çağlayan  D, Koçak  MZ, Geredeli  Ç,  et al.  The effect of concomitant use of proton pump inhibitors with CDK 4/6 inhibitors on survival in metastatic breast cancer.   Eur J Clin Pharmacol. 2023;79(2):243-248. doi:10.1007/s00228-022-03435-7 PubMedGoogle ScholarCrossref
12.
Peabody  JW, Lee  SW, Bickel  SR.  Health for all in the Republic of Korea: one country’s experience with implementing universal health care.   Health Policy. 1995;31(1):29-42. doi:10.1016/0168-8510(94)00669-6 PubMedGoogle ScholarCrossref
13.
Suissa  S.  Immortal time bias in pharmaco-epidemiology.   Am J Epidemiol. 2008;167(4):492-499. doi:10.1093/aje/kwm324 PubMedGoogle ScholarCrossref
14.
Mahtani  RL, Parisi  M, Glück  S,  et al.  Comparative effectiveness of early-line nab-paclitaxel vs paclitaxel in patients with metastatic breast cancer: a US community-based real-world analysis.   Cancer Manag Res. 2018;10:249-256. doi:10.2147/CMAR.S150960 PubMedGoogle ScholarCrossref
15.
Campbell  BA, Scarisbrick  JJ, Kim  YH, Wilcox  RA, McCormack  C, Prince  HM.  Time to next treatment as a meaningful endpoint for trials of primary cutaneous lymphoma.   Cancers (Basel). 2020;12(8):2311. doi:10.3390/cancers12082311 PubMedGoogle ScholarCrossref
16.
Labaki  C, Bakouny  Z, Sanglier  T,  et al.  Real-world progression-free survival (rwPFS) and time to next line of therapy (TTNT) as intermediate endpoints for survival in metastatic breast cancer: a real-world experience.   J Clin Oncol. 2022;40(16)(suppl):6520. doi:10.1200/JCO.2022.40.16_suppl.6520 Google ScholarCrossref
17.
Morgan  CJ.  Landmark analysis: a primer.   J Nucl Cardiol. 2019;26(2):391-393. doi:10.1007/s12350-019-01624-z PubMedGoogle ScholarCrossref
18.
Freedberg  DE, Lebwohl  B, Abrams  JA.  The impact of proton pump inhibitors on the human gastrointestinal microbiome.   Clin Lab Med. 2014;34(4):771-785. doi:10.1016/j.cll.2014.08.008 PubMedGoogle ScholarCrossref
19.
Moreau-Bachelard  C, Letailleur  V, Bompas  E, Soulié  P, Paul  J, Raoul  JL.  Effect of concomitant proton pump inhibitors with pazopanib on cancer patients: a retrospective analysis.   Cancers (Basel). 2022;14(19):4721. doi:10.3390/cancers14194721 PubMedGoogle ScholarCrossref
20.
Sharma  M, Holmes  HM, Mehta  HB,  et al.  The concomitant use of tyrosine kinase inhibitors and proton pump inhibitors: prevalence, predictors, and impact on survival and discontinuation of therapy in older adults with cancer.   Cancer. 2019;125(7):1155-1162. doi:10.1002/cncr.31917 PubMedGoogle ScholarCrossref
21.
Raoul  JL, Moreau-Bachelard  C, Gilabert  M, Edeline  J, Frénel  JS.  Drug-drug interactions with proton pump inhibitors in cancer patients: an underrecognized cause of treatment failure.   ESMO Open. 2023;8(1):100880. doi:10.1016/j.esmoop.2023.100880 PubMedGoogle ScholarCrossref
22.
Del Re  M, Crucitta  S, Omarini  C,  et al.  Concomitant administration of proton pump inhibitors does not significantly affect clinical outcomes in metastatic breast cancer patients treated with ribociclib.   Breast. 2022;66:157-161. doi:10.1016/j.breast.2022.10.005 PubMedGoogle ScholarCrossref
×