Temporal offset association between the number of irinotecan-related adverse reactions and pharmacogenomic studies: A cross-correlation analysis

Abstract

Objectives

Studies have proved that UGT1A1 (*6, *28 and *93) gene polymorphism was closely related to the side effects of irinotecan. This study intends to perform a correlation analysis on the relationship between pharmacogenomic studies and ADRs based on time series.

Methods

The ADRs related to irinotecan were derived through the FAERS; searched all pharmacogenomic studies in PubMed and Web of Science; then analyzed the sequence of correlation coefficients between total ADRs, fatal ADRs and pharmacogenomic studies under different time offset.

Results

There is a positive correlation between the number of total ADRs and pharmacogenomic studies, of which the maximum correlation coefficient was 0.78 (95 % CI: 0.58–0.90), with a lag of 1 year. There is also a positive correlation between the number of fatal ADRs and pharmacogenomic studies, with the maximum correlation coefficient of 0.87 (95 % CI: 0.73–0.94) and a offset of − 4 years.

Conclusion

It was found that both the total ADRs and fatal ADRs were significantly positively correlated with change trend of published pharmacogenomic literatures, which confirmed the role of pharmacogenomic research in promoting the safe use of irinotecan, and have a faster response time in reducing fatal ADRs during clinical application.

1. Introduction

Irinotecan is an essential component of first and second line treatment chemotherapy regimens for advanced colorectal cancer, such as FOLFIRI, CapIRI, FOLFOXIRI and monotherapy (Benson et al 2021). It can also be used for gastric cancer, esophageal cancer and lung cancer, etc (Ajani et al., 2022, Ettinger et al., 2022). However, diarrhea and severe neutropenia are common adverse drug reactions (ADRs), which compromises the treatment outcomes by significantly impacting patient compliance and causing delay in chemotherapy cycles, and even lead to death (de Man et al 2018). Numerous studies have proven that UDP-glucuronosyltransferase 1A1 (UGT1A1, *6, *28 and *93) gene polymorphism was closely related to the occurrence of irinotecan side effects, adjusting its dose under the guidance of UGT1A1 genotype can greatly reduce the incidence and severity of adverse reactions, showed economic advantages and consequently save cost on healthcare systems. (Argevani et al., 2020, Carrato, 2022, Hulshof et al., 2022, Kong et al., 2022).

Irinotecan was first approved by the FDA in 1994 for the treatment of advanced colorectal cancer, but the association between its ADRs and UGT1A1 gene polymorphism was unknown until 1998 where this association was reported (Ando et al 1998). Nowadays, with the advancements in personalized medicine and the increase in its applicability, there are more published studies related to pharmacogenomics, including mechanism studies, case reports, clinical trials, and reviews. However, the data assessing the impact of pharmacogenomics on overall treatment safety in clinical practice are lacking.

The aim of this study is to perform a correlation analysis on the relationship between the timing of ADRs reporting and pharmacogenomic studies publication and to highlight the lag time from basic research to clinical research, and eventually, to clinical application. In the correlation analysis, considering the lag from basic / clinical research to clinical application, a time-delay model is introduced (Muthuchellappan et al., 2018, Shao et al., 2022).

2. Methodology

2.1. Adverse drug reactions data collection

From January 1994 to December 2021, the ADRs related to irinotecan were derived through the adverse event reporting system of Food and Drug Administration (FAERS), with a total of 29, 994 cases by using the search term of irinotecan and irinotecan hydrochloride. After removing incomplete information such as age, sex, purpose, event date and those from non-health professionals (it is considered that the authenticity and accuracy of these cases were questionable), 13, 428 cases were retained for analysis (Fig. 1).

Fig. 1.

The exclusion process of included ADRs in this study.

2.2. Pharmacogenomic studies collection

Taking irinotecan and UGT1A1 as terms in the abstract, we searched all literatures in PubMed and Web of Science. After removing the duplicate results, a total of 582 literatures were included in the follow-up analysis.

2.3. Time series correlation analysis

We defined X_i, Y_i and Z_i as the time series of total ADRs, fatal ADRs and pharmacogenomic studies, respectively. The cross-correlation coefficients (r) under different time offsets were:

$${\text{r}}_{\text{k}}(X,Z)=\frac{{\sum }_{i=1}^{n-k}(X_{\text{i}}-\overline{X}\text{i})(Z_{\text{i}+\text{k}}-\overline{Z_{\text{i}+\text{k}}})}{\sqrt{\sum _{i=1}^{n-k}(Xi-\overline{X\text{i}})·\sum _{i=1}^{n-k}(Z_{\text{i}+\text{k}}-\overline{Z_{\text{i}+\text{k}}})}}$$

$${\text{r}}_{\text{k}}(Y,Z)=\frac{{\sum }_{i=1}^{n-k}(Y_{\text{i}}-\overline{Y\text{i}})(Z_{\text{i}+\text{k}}-\overline{Z_{\text{i}+\text{k}}})}{\sqrt{\sum _{i=1}^{n-k}(Yi-\overline{Y\text{i}})·\sum _{i=1}^{n-k}(Z_{\text{i}+\text{k}}-\overline{Z_{\text{i}+\text{k}}})}}$$

Where, r_k (X, Z) and r_k (Y, Z) are the sequence of correlation coefficients between total ADRs, fatal ADRs and pharmacogenomic studies under different time offset k, respectively; n is the sequence length; k is the offset time (0, ± 1, ± 2, …). According to experience, k value is less than or equal to n/4, so the maximum value of k is 7.

3. Result

As shown in Fig. 2, the number of total ADRs gradually increased after irinotecan was approved, and began to decrease after 2018, while the number of fatal ADRs showed a downward trend after 2013. Additionally, the number of pharmacogenomic studies reached the peak in 2017.

Fig. 2.

The annual change trends of total ADRs, fatal ADRs and pharmacogenomic studies.

When the time offset ranging from 0 to ± 7, Fig. 3 shows that there is a positive correlation between the number of total ADRs and pharmacogenomic studies (0.52–0.78), of which the maximum correlation coefficient was 0.78 (95 % CI: 0.58–0.90), with a lag of 1 year. At the same time, there is also a positive correlation between the number of fatal ADRs and pharmacogenomic studies (0.39–0.87), with the maximum correlation coefficient of 0.87 (95 % CI: 0.73–0.94), with an offset of − 4 years, indicating that before the peak of pharmacogenomic studies, the resulting fatal ADRs began to decline 4 years in advance (Fig. 4).

Fig. 3.

The time series cross-correlation between the number of total ADRs and pharmacogenomic studies.

Fig. 4.

The time series cross-correlation between the number of fatal ADRs and pharmacogenomic studies.

4. Discussion

To our knowledge, this is the first study to investigate the relationship between the ADRs and pharmacogenomic studies using a time-offset cross-correlation method. It was found that both total ADRs and fatal ADRs were significantly positively correlated with trends in the published pharmacogenomic literatures. Interestingly, the time-offset of total ADRs and fatal ADRs were 1 year and − 4 years, respectively, suggesting that pharmacogenomic studies have a faster response time in reducing fatal ADRs during clinical application.

It is undeniable that the number of ADRs is not only affected by the single factor of UGT1A1 gene polymorphism, but also by various factors such as changes in the recommended treatment regimen by the guideline, liposome-encapsulated form and so on (Assiri and Noor, 2020, Karas et al., 2022, Milano et al., 2022). However, this study provides large-scale real world data support for the important role of UGT1A1 genotype in guiding irinotecan dose in clinical practice, which helps to promote the inclusion of mandatory genetic testing in guideline (Karas & Innocenti 2022), and is worthy of reference for other drugs.

5. Conclusion

Overall, this is the first study to investigate the relationship between the number of ADRs and pharmacogenomic studies using a time-offset cross-correlation method. It was found that the change trend of ADRs was significantly positively correlated with published pharmacogenomic literatures, which confirmed the role of pharmacogenomic research in promoting the safe use of irinotecan in clinical practice.

Funding

This work was supported by the key research and development program project of Anhui Province (No. 202004j07020008) and talent training plan of Bengbu Medical College (No. by 51201316).

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.