Fact-Finding Survey and Exploration of Preventive Drugs for Antineoplastic Drug-Induced Oral Mucositis Using the Japanese Adverse Drug Event Report Database

: Oral mucositis (OM) is one of the most common adverse events associated with antineoplastic drug treatment. Studies on the risk of antineoplastic drug-induced OM and its prevention are limited. We, therefore, conducted a disproportionality analysis of antineoplastic drug-induced OM and explored candidate preventive drugs for OM using the Japanese Adverse Drug Event Report (JADER) database. The JADER database showed that between April 2004 and March 2022, antineoplastic drug-related adverse events were reported in 210,822 cases, of which 2922 were OM. Forty-two drugs appeared to be associated with OM. The weibull distribution showed different patterns of time-to-onset depending on the type of antineoplastic drug administered. Cluster analyses classified antineoplastic drugs according to the typical symptoms of OM. These findings suggest that antineoplastic drug-induced OM should be monitored based on expression patterns of symptoms. Upon analyzing the inverse association, several concomitant drugs, including lenalidomide hydrate and febuxostat, were expected to be candidate preventive drugs for antineoplastic drug-induced OM. Concomitant drugs that showed an inverse association with antineoplastic drug-induced OM differed within the Anatomical Therapeutic Chemical classification. These findings could serve as a reference when considering drugs that should be prioritized to validate their prophylactic effect against antineoplastic-induced OM in the future.


Introduction
Oral mucositis (OM) is an inflammatory disease of the oral cavity mucosa and is one of the most common adverse events (AEs) associated with antineoplastic drugs [1][2][3].It affects the quality of life of patients and their ability to tolerate treatment.The main symptoms of OM include oral pain, hemorrhage, dryness, redness, swelling of the oral mucosa, restricted mouth opening, dysarthria, dysphagia, and taste disorders, which cause extreme pain interfering with oral ingestion in severe cases.The frequency of antineoplastic drug-induced OM varies according to the type of drug administered.For example, a frequency of 65% has been documented with alkylating drugs (e.g., cyclophosphamide), plant alkaloids and other natural products (e.g., docetaxel), cytotoxic antibiotics and related substances (e.g., doxorubicin) [4]; 14-35% with antimetabolites (e.g., fluorouracil) [4]; 67% with everolimus [5]; 5-33% with sorafenib [6]; 40.5% with bevacizumab [7]; and approximately 40% with platinum drugs [4].Chemotherapy-induced OM generally occurs 3-4 days after administration, peaks approximately 2 weeks after administration, and improves almost 3 weeks after administration [8,9].The onset of OM induced by protein kinase inhibitors and other molecularly targeted drugs varies; however, with everolimus Sci.Pharm.2024, 92, 34 2 of 20 and other mTOR inhibitors, OM occurs several days after administration, with a median time to onset of approximately 10 days, improving in approximately 1 week [10].These results suggest that antineoplastic drug-induced OM should be monitored based on the expression pattern of symptoms.The primary sources of information for managing OM are package inserts of prescription drugs and clinical trials [11].However, recent reports [12,13] have primarily included only certain antineoplastic drugs or investigated these based on meta-analyses [14]; thus, comprehensive reporting on OM induced by all antineoplastic drugs is lacking.
Effective drugs to prevent drug-induced OM are yet to be developed.In Japan, sodium gualenate hydrate, sodium bicarbonate, and other drugs are recommended by the Ministry of Health, Labor, and Welfare for prophylaxis against OM [15], but these do not include specialized drugs that can specifically prevent OM.Recent studies have reported potential preventive drugs against antineoplastic drug-induced OM.In a meta-analysis of 10,514 randomized subjects in 131 trials, aloe vera, amifostine, cryotherapy, granulocytecolony stimulating factor, intravenous glutamine, honey, keratinocyte growth factor, laser, polymyxin/tobramycin/amphotericin antibiotic pastille/paste, and sucralfate provided significant prevention against OM [11].There is also reported that composed herbal extract, applied in the form of a mouthwash and in the form for topical application with benzocaine content showed remission of oral mucositis [16].In a multicenter, single-arm phase II study (SWISH study) of patients with breast cancer who received everolimus and exemestane, a dexamethasone mouthwash reduced the incidence of grade 2 or higher OM [12].However, this evidence is based on limited interventions with specific antineoplastic drugs and is not universally applicable to all antineoplastic drugs.The number of reports regarding OM induced by new types of antineoplastic drugs, such as molecular targeted drugs and immune checkpoint inhibitors, has also increased [17][18][19].Therefore, there is a growing need to develop drugs capable of preventing OM.
Currently, studies using pharmacovigilance databases are underway worldwide.These studies evaluate the association between drugs and AEs, identify the time from the onset of AE, and investigate drug repositioning [20].The Japanese Adverse Drug Event Report (JADER) published by the Pharmaceuticals and Medical Devices Agency (PMDA) has been used to investigate the relationship between drugs and AEs [21,22].
For successful antineoplastic drug therapy, evaluating the risk of OM for each antineoplastic drug and identifying preventive drugs is crucial.In the current study, we aimed to measure the disproportionality of antineoplastic drug-induced OM and analyze the expression pattern of OM.The results suggest that antineoplastic drug-induced OM should be monitored based on the expression pattern of symptoms.These findings will provide useful reference information for the management of OM.We also explored candidate preventive drugs against OM using the JADER database.The result showed that several concomitant drugs, including lenalidomide hydrate and febuxostat, are potential candidates for antineoplastic-induced OM prevention.The findings serve as a reference when considering drugs that should be prioritized to validate their prophylactic effect against OM in patients treated with antineoplastic drugs.

Data Sources
JADER is the Adverse Drug Reaction Database provided by the PMDA, documenting case reports of adverse drug reactions.Adverse drug reaction reports, which can be downloaded and used freely, are updated monthly and published on the PMDA website (http://www.pmda.go.jp/; accessed on 15 July 2022).JADER data are divided into four tables: demographic information pertaining to the patient, sex, age, and reporting year (demo); drug information, including prescribed drug, dosage, route, the reason for use, and start and end dates of administration (drug); AE, including the type of AE, outcome, and date of onset (reac); and primary disease (hist).Each reported item could be associated with an identification number.Data was downloaded on 15 July 2022 and included data Sci.Pharm.2024, 92, 34 3 of 20 reported from April 2004 to March 2022.Drugs listed in the dataset table were classified into suspected, concomitant, or interacting drugs.Herein, we used all drug classifications to investigate the actual relevance of drugs and OM and to evaluate the inverse association between potentially preventive drugs and antineoplastic-induced OM.Overlapping drugs with the same case identification number (ID) were excluded from the analysis.In the "reac" table, the adverse drug reactions were registered based on the Preferred Terms of Medical Dictionary for Regulatory Activities/Japan version; MedDRA/J.OM-related AEs are listed in Supplementary Table S1 online.Data were extracted using MedDRA/J ver.25.1 (https://www.jmo.pmrj.jp/;accessed on 15 July 2022).Overlapping cases of OM with the same ID were excluded from the analysis.

Selection of Antineoplastic Drugs
We included antineoplastic drugs classified as antineoplastic drugs (L01) in the Anatomical Therapeutic Chemical (ATC) classification system (https://www.kegg.jp/brite/jp08303/A10BK02; accessed on 15 July 2022).From the following standardized Med-DRA Querie (SMQ)s, we determined the antineoplastic indication using primary disease data [23

Statistical Analysis
The reporting odds ratio (ROR) was calculated to evaluate the association between OM and antineoplastic drugs.The ROR was calculated using a two-by-two contingency table (Supplementary Figure S1 online), which is widely used as a signal indicator in the Adverse Drug Reaction Database [24].The two-by-two contingency table cannot be calculated with zero cells, and the estimation becomes unstable when the cell frequency is small.Therefore, to correct this bias, 0.5 was added to all cells (Haldane-Anscombe 1/2 correction) [25,26].In the current study, we determined the association between antineoplastic drugs and OM when the ROR showed a lower limit of 95% confidence interval (CI) > 1 and a p-value < 0.05, as determined by Fisher's exact test [24].We also created a volcano plot using the natural logarithm of the ROR (ln ROR) and the normal logarithm of the p-value (−log[p]) to visually evaluate the association between OM and antineoplastic drugs [23,27].
For the time to onset analysis, distinct from the other analysis, we extracted the "suspected drug" from the "drug" table.The start date of administration was defined as the date of the first administration before the onset of OM.The number of days until the appearance of OM was calculated as follows: [(onset date)−(start date of administration)] plus 1 day.If the number of days to the onset was greater than 365, it was calculated as 365 days.Additionally, the pattern of onset time was classified using the shape parameter β of the Weibull distribution.Weibull parameters were used for the time-to-onset analysis to investigate the time-course hazards of AEs.When β is equal to 1 (random failure type), the hazard is estimated to remain constant over time.When β < 1 and the 95% CI of β excluded the value 1 (early failure type), the hazard is estimated to decrease over time.When β > 1 and the 95% CI of β excluded the value of 1 (wear-out failure type), the hazard is estimated to increase over time [28].
Hierarchical cluster analyses were performed to classify OM-related AEs induced by antineoplastic drugs.We examined 42 antineoplastic drugs with OM and 11 OM-related AEs with ≥10 cases reported.OM-related AEs with <10 cases were excluded for the cluster analysis.We calculated the ROR from two contingency tables for 11 AEs and 42 drugs.We then created a dendrogram based on the presence or absence of disproportionality and evaluated the distribution of drugs that showed disproportionality to each AE using the Ward method with Euclidean distance [29].
Among drugs administered concomitantly with antineoplastic drugs, we examined those with total reports of ≥100 AEs.The ROR and p-values were calculated for each drug using Fisher's exact test to establish a volcano plot.Herein, we defined an inverse association between concomitant drugs and antineoplastic-induced OM when the ROR showed an upper limit of 95% CI < 1 and a p-value < 0.05, according to Fisher's exact test [24].
All statistical analyses were performed using JMP Pro17 software (SAS Institute Inc., Cary, NC, USA).

Results
The total number of cases registered in JADER between April 2004 and March 2022 was 775,555.In total, 210,822 cases involved antineoplastic drug-induced AEs, of which 2922 were OM (Figure 1).
. 2024, 92, x FOR PEER REVIEW 4 of 19 analysis.We calculated the ROR from two contingency tables for 11 AEs and 42 drugs.We then created a dendrogram based on the presence or absence of disproportionality and evaluated the distribution of drugs that showed disproportionality to each AE using the Ward method with Euclidean distance [29].Among drugs administered concomitantly with antineoplastic drugs, we examined those with total reports of ≥100 AEs.The ROR and p-values were calculated for each drug using Fisher's exact test to establish a volcano plot.Herein, we defined an inverse association between concomitant drugs and antineoplastic-induced OM when the ROR showed an upper limit of 95% CI < 1 and a p-value < 0.05, according to Fisher's exact test [24].
All statistical analyses were performed using JMP Pro17 software (SAS Institute Inc., Cary, NC, USA).

Results
The total number of cases registered in JADER between April 2004 and March 2022 was 775,555.In total, 210,822 cases involved antineoplastic drug-induced AEs, of which 2922 were OM (Figure 1).S2 (online).The sex distribution was as follows: 40.3% were females, 54.6% were males, and 5.1% were unknown.Regarding age, 29.1%, of patients were 70-79 years of age, and 28.0% were 60-69 years of age.According to the ATC classification, monoclonal antibodies and antibody-drug conjugates (L01F) were the most prevalent (36.3%).Among the patients with OM, 47.1% were females, 49.9% were males, and 3.1% were others or unknown.Regarding age, 29.1% of patients were 60-69 years of age, and 28.9% were 70-79 years of age.According to the ATC classification, antimetabolites (L01B) were the most prevalent (51.2%).
The number of cases and ROR for antineoplastic drug-induced OM are shown in Table 1. Figure 2 presents a volcano plot of the association between OM and antineoplastic drugs.As shown in Table 1 and Figure 2, 42 drugs were associated with OM.The most frequently reported drugs with OM were tegafur/gimeracil/oteracil potassium (523 reports; 17.9%), fluorouracil (333 reports; 11.4%), cisplatin (305 reports; 10.4%), everolimus (264 reports; 9.0%), capecitabine (222 reports; 7.6%), and oxaliplatin (209 reports; 7.2%).Among the 42 antineoplastic drugs associated with OM, 39 drugs with two or more usable data points for both the starting date of administration and the onset date of AEs were included in the time-to-onset analysis.Figure 3 presents box plots of the time to onset.Table 2 shows the median time to onset and pattern of OM classified using the Weibull distribution.The median time to OM onset ranged from approximately 7.0 days to 23.0 days.Considering the Weibull distribution, tegafur/gimeracil/oteracil potassium, fluorouracil, capecitabine, methotrexate, tegafur and uracil, pralatrexate, irinotecan, erlotinib, bevacizumab, trastuzumab, cisplatin, and eribulin were estimated to be early failures.Melphalan, thiotepa, ranimustine, dactinomycin, palbociclib, and avelumab were considered wear-out failures.
of the plot shows the antineoplastic drugs associated with oral mucositis-related AEs.AEs, adverse events; ROR, reporting odds ratio.
Among the 42 antineoplastic drugs associated with OM, 39 drugs with two or more usable data points for both the starting date of administration and the onset date of AEs were included in the time-to-onset analysis.Figure 3 presents box plots of the time to onset.Table 2 shows the median time to onset and pa ern of OM classified using the Weibull distribution.The median time to OM onset ranged from approximately 7.0 days to 23.0 days.Considering the Weibull distribution, tegafur/gimeracil/oteracil potassium, fluorouracil, capecitabine, methotrexate, tegafur and uracil, pralatrexate, irinotecan, erlotinib, bevacizumab, trastuzumab, cisplatin, and eribulin were estimated to be early failures.Melphalan, thiotepa, ranimustine, dactinomycin, palbociclib, and avelumab were considered wear-out failures.Figure 4 presents the hierarchical cluster analyses classifying antineoplastic drugs based on the presence or absence of disproportionality of OM-related AEs.Based on the dendrogram, the clusters were divided into four types.Association with stomatitis was observed in all clusters.The first cluster included fluorouracil and other drugs, and the second cluster included docetaxel and other drugs.Among alkylating agents, cyclophosphamide was included in the first cluster, and other agents were included in the second cluster.All drugs classified as cytotoxic antibiotics and related substances were included in the third cluster.These clusters showed association with pharyngeal ulceration and mouth ulceration.The fourth cluster mainly included protein kinase inhibitors (e.g., afatinib).These clusters were associated with lip erosion, stomatitis, hemorrhagic, and glossitis.Figure 4 presents the hierarchical cluster analyses classifying antineoplastic drugs based on the presence or absence of disproportionality of OM-related AEs.Based on the dendrogram, the clusters were divided into four types.Association with stomatitis was observed in all clusters.The first cluster included fluorouracil and other drugs, and the second cluster included docetaxel and other drugs.Among alkylating agents, cyclophosphamide was included in the first cluster, and other agents were included in the second cluster.All drugs classified as cytotoxic antibiotics and related substances were included in the third cluster.These clusters showed association with pharyngeal ulceration and mouth ulceration.The fourth cluster mainly included protein kinase inhibitors (e.g., afatinib).These clusters were associated with lip erosion, stomatitis, hemorrhagic, and glossitis.Treatment with concomitantly administered drugs showing an inverse association with antineoplastic-induced OM are summarized in Table 3. Twelve drug combination treatments showed inverse associations with all antineoplastic drug-induced OM. Figure 5a presents a volcano plot of the association between concomitantly administered drugs and antineoplastic drug-induced OM.Dexamethasone, prednisolone, methylprednisolone, sulfamethoxazole/trimethoprim, voriconazole, febuxostat, and lenalidomide hydrate exhibited inverse associations.However, drugs such as sodium gualenate hydrate, which is used to prevent or treat OM, did not display an inverse association with OM. Figure 5b-h present volcano plots showing the associations between concomitant medications and antineoplastic-induced OM by ATC classification.Lenalidomide hydrate showed an inverse association with OM induced by monoclonal antibodies, antibody-drug conjugates, and other antineoplastic drugs.Febuxostat showed an inverse association with OM induced by antimetabolites and other antineoplastic drugs.Dexamethasone showed an inverse association with OM induced by other antineoplastic drugs, and prednisolone showed an inverse association with OM induced by alkylating drugs, plant alkaloids, and other natural products, and cytotoxic antibiotics and related substances.

Discussion
In the present study, 42 drugs were associated with OM.All the 42 drugs associated with OM has been reported that they are likely to cause OM in the package inserts of prescription drugs or textbook.The most frequently reported drugs associated with OM were tegafur/gimeracil/oteracil potassium, fluorouracil, cisplatin, everolimus, capecitabine, and oxaliplatin.These drugs are known to have a high incidence of OM in previous clinical trials [4,[30][31][32]; hence, these antineoplastic drugs may have a high potential to induce OM.Cytotoxic antineoplastic drugs, such as alkylating drugs and antimetabolites, inhibit the growth of actively dividing cells with abundant blood flow [33].Cells in the

Discussion
In the present study, 42 drugs were associated with OM.All the 42 drugs associated with OM has been reported that they are likely to cause OM in the package inserts of prescription drugs or textbook.The most frequently reported drugs associated with OM were tegafur/gimeracil/oteracil potassium, fluorouracil, cisplatin, everolimus, capecitabine, and oxaliplatin.These drugs are known to have a high incidence of OM in previous clinical trials [4,[30][31][32]; hence, these antineoplastic drugs may have a high potential to induce OM.Cytotoxic antineoplastic drugs, such as alkylating drugs and antimetabolites, inhibit the growth of actively dividing cells with abundant blood flow [33].Cells in the oral mucosa are believed to be susceptible to the cytotoxic effects of antineoplastic drugs because they are actively dividing.Molecularly targeted drugs, such as protein kinase inhibitors, are likely to have a lower incidence of OM than cytotoxic anticancer drugs, as they act only on specific target molecules in cancer cells [34]; however, everolimus and other mTOR inhibitors, which act on tumor growth, are associated with a higher incidence of OM than other molecularly targeted drugs [10].
In the present study, the median time to onset ranged from approximately 7.0 days to 23.0 days.This finding is consistent with that of previous studies, which showed that antineoplastic drug-induced OM occurs between a few days and 2-3 weeks after administration [35].The Weibull distribution suggested that OM expression patterns may differ for each type of antineoplastic drug.Early failure was estimated for antimetabolites (e.g., tegafur/gimeracil/oteracil potassium), monoclonal antibodies, and antibody-drug conjugates (e.g., bevacizumab).Wear-out failure was estimated for alkylating drugs (e.g., melphalan).Antineoplastic drug-induced OM is caused via direct effects, such as destruction of the oral mucosa, inflammation, and secondary infections in the oral cavity associated with bone marrow suppression, including leukopenia.OM caused by secondary infections develops later than that caused by the direct action of antineoplastic drugs [36].These findings explain why more than one drug exhibited wear-out failure among the alkylating drugs in this study.The Weibull distribution could be valuable in establishing the specific safety monitoring period for AEs based on expression patterns of OM.Therefore, OM attributed to antineoplastic drugs that exhibit wear-out failure of use may require longer-term monitoring than other antineoplastic drugs.However, the use of avelumab should be carefully considered owing to the lack of data (n = 2).
Hierarchical cluster analysis classified antineoplastic drugs into four clusters based on the presence or absence of disproportionality of OM-related AEs.Stomatitis was associated with all clusters; therefore, it was expected to be associated with all antineoplastic drugs.The third cluster, which included all drugs classified as cytotoxic antibiotics and related substances, was associated with pharyngeal ulceration and mouth ulceration; this cluster also included dactinomycin, everolimus, and methotrexate.These drugs are known to be associated with mouth ulceration [6,37,38].Therefore, the risk of mouth ulceration with these drugs needs to be carefully considered.The fourth cluster, which included protein kinase inhibitors such as afatinib, lapatinib, and dacomitinib, was associated with lip erosion, hemorrhagic stomatitis, and glossitis.The incidence of lip erosion and glossitis was less than 10% or unknown on the package inserts of these drugs.However, molecularly targeted drugs, such as protein kinase inhibitors, are known to induce aphthous OM, and aphthous OM localized in non-excursion and keratinized mucosal areas are typical because these areas can be difficult to stimulate physically [19,39].Further investigations are needed to evaluate the correlation between protein kinase inhibitors and these AEs.
Twelve of the concomitantly administered drugs were inversely associated with antineoplastic drug-induced OM by all antineoplastic drugs.In clinical trials, dexamethasone [12], prednisolone [40], and methylprednisolone [41] were shown to prevent OM.Lenalidomide hydrate, a thalidomide analog, improves aphthous OM [42].Considering other drugs, evidence from clinical trials is lacking; however, febuxostat was found to exert antioxidant and anti-inflammatory effects [43][44][45].In a study using a rat model, pretreatment with febuxostat reduced parotid salivary gland injury induced by 5-fluorouracil through potent antioxidant effects and inhibition of interleukin (IL)-1β and tumor necrosis factor (TNF)-α activity [46].Cytotoxic drugs directly inhibit DNA synthesis and cellular biochemical metabolic pathways, leading to the production of inflammatory cytokines such as reactive oxygen species and TNF-α, tissue damage in the oral mucosa, and induction of apoptosis [47].These findings indicate that febuxostat might reduce OM by exerting antioxidant effects and inhibiting IL-1β and TNF-α.Moreover, antineoplastic-induced OM can occur via secondary infection due to bone marrow suppression.Sulfamethoxazole/trimethoprim [48] and voriconazole [49] exert preventive effects against infections; hence, these drugs may prevent secondary OM induced by antineoplastic drugs.In Japan, sodium gualenate hydrate is recommended as a prophylactic drug for OM [15] .However, in the current study, a combination of the drugs showed an association with OM, which could be attributed to the fact that these drugs were used for patients at high risk of OM, and the number of cases was higher than expected.Furthermore, some other drugs that lacked an inverse association with OM were found to afford notable preventative effects against OM [11].Hence, drugs that did not exhibit disproportionality could be candidates for OM prevention.Herein, concomitantly administered drugs displaying an inverse association with antineoplastic drug-induced mucositis differed within the ATC classification.Accordingly, it is important to determine the prophylactic effect against antineoplastic-induced OM by drug or ATC classification.Recently, the pathobiology of mucositis has been explored, and potential druggable targets that inhibit mucositis have been identified [50].Our findings on drugs that showed an inverse association with OM could provide insights for further basic research elucidating the mechanisms of action leading to drug repositioning.The findings of this study could serve as a guide when considering drugs that should be prioritized when validating their prophylactic effect against antineoplastic-induced OM.
This study had several limitations.First, because the JADER database is a spontaneous reporting database, it lacks denominator information, and the frequency of AEs cannot be calculated.Second, in studies that utilize spontaneous reporting databases for adverse drug reactions, such as JADER, mild and well-known adverse reactions are less likely to be reported, and severe adverse reactions are more frequently reported, which may result in reporting biases [51].Third, the data retrieved from the JADER database included inadequate information, such as blank data and the start date of administration; therefore, these data were excluded from the analysis.Fourth, for recently launched drugs, the number of cases is limited but occasionally higher than expected owing to aggressive reporting, a phenomenon known as the Weber effect [52].Fifth, it was not possible to eliminate the impact of concomitantly administered medications on our evaluation [53,54].Reportedly, concomitant therapy of antineoplastic drugs is associated with a higher incidence of OM than monotherapy [1,55].In the current study, we did not investigate the incidence of OM by monotherapy and combination regimens.Therefore, the signal indicators observed in the current study must be interpreted carefully, considering the effects of concomitant use.Sixth, spontaneous reports may lack information on concomitant medications, especially in terms of common AEs [56].Moreover, AEs will decrease if appropriate medical care is afforded in clinical settings, which may provide signals that oppose the pharmacological basis [57].This may affect the signals detected with and without concomitant medication.Finally, only the JADER database was used in this study.AEs reported in the JADER and FDA Adverse Event Reporting System FAERS databases are known to differ [58].Therefore, we hope that the gaps caused by these limitations will be addressed in future research using other methods, such as a combination of database-based studies of patients with certain adverse reactions.

Conclusions
We investigated the disproportionality of OM caused by antineoplastic drugs using the JADER database and potential concomitant drug treatment to prevent OM.In total, 42 drugs were associated with OM.The Weibull distribution showed a distinctive time-toonset pattern depending on the type of antineoplastic drug administered, which could be valuable in monitoring OM.Cluster analyses classified antineoplastic drugs according to the typical symptoms of OM.These findings suggest that antineoplastic drug-induced OM should be monitored based on expression patterns.Considering the inverse association analysis, several concomitant drugs were expected to be the candidate drugs for preventing antineoplastic drug-induced OM.Additional investigations are required to validate the preventive mechanisms and clinical evaluation of OM.

Figure 2 .
Figure 2. Antineoplastic drugs associated with oral mucositis-related adverse events.The X-axis shows the natural logarithm of the reported odds ratios (ln ROR), and the y-axis shows the common logarithm of the inverse p-value (−log (p-value)) from Fisher's exact test.The do ed line on the yaxis represents p = 0.05.The plot colors represent the number of reports in the ATC category.Plot size indicates the total number of reports.As the RORs became more positive, the tendency toward AEs increased; decreasing p-values indicated greater statistical significance.The upper-right portion

Figure 2 .
Figure 2. Antineoplastic drugs associated with oral mucositis-related adverse events.The X-axis shows the natural logarithm of the reported odds ratios (ln ROR), and the y-axis shows the common logarithm of the inverse p-value (−log (p-value)) from Fisher's exact test.The dotted line on the y-axis represents p = 0.05.The plot colors represent the number of reports in the ATC category.Plot size indicates the total number of reports.As the RORs became more positive, the tendency toward AEs increased; decreasing p-values indicated greater statistical significance.The upper-right portion of the plot shows the antineoplastic drugs associated with oral mucositis-related AEs.AEs, adverse events; ROR, reporting odds ratio.

Figure 3 .Table 2 .
Figure 3. Box plot for time-to-onset of antineoplastic drugs associated with oral mucositis.Box plots show the 25th and 75th percentiles and medians.The whiskers present the maximum and minimum values within 1.5 times the inner quartile point's length.The values outside the box represent outliers.All data for which the number of days until the onset of oral mucositis exceeded 365 days were calculated as 365 days.Table 2. Median values and Weibull parameters of antineoplastic drug-induced oral mucositis.Antineoplastic Agents (ATC a Code) Case Median (25-75%) (Day)

Figure 3 .
Figure 3. Box plot for time-to-onset of antineoplastic drugs associated with oral mucositis.Box plots show the 25th and 75th percentiles and medians.The whiskers present the maximum and minimum values within 1.5 times the inner quartile point's length.The values outside the box represent outliers.All data for which the number of days until the onset of oral mucositis exceeded 365 days were calculated as 365 days.

Figure 4 .
Figure 4. Classification of oral mucositis-related AEs by hierarchical cluster analysis.The dendrogram shows the relationships between 11 types of oral mucositis-related AEs.The do ed line indicates cluster separation.The color map shows the presence or absence of associated signals with AEs.Red implies "presence" and blue implies "absence".AEs, adverse events.

Figure 4 .
Figure 4. Classification of oral mucositis-related AEs by hierarchical cluster analysis.The dendrogram shows the relationships between 11 types of oral mucositis-related AEs.The dotted line indicates cluster separation.The color map shows the presence or absence of associated signals with AEs.Red implies "presence" and blue implies "absence".AEs, adverse events.

Figure 5 .
Figure 5. Drugs associated with oral mucositis-related AEs when combined with antineoplastic drugs.Combination with (a) all antineoplastic drugs, (b) alkylating drugs, (c) antimetabolites, (d) with plant alkaloids and other natural products, (e) with cytotoxic antibiotics and related substances, (f) protein kinase inhibitors, (g) monoclonal antibodies and antibody-drug conjugates, and (h) other antineoplastic drugs.The X-axis shows the natural logarithm of the reporting odds ratios (ln ROR), and the y-axis shows the common logarithm of the inverse p-value (−log (p-value)) from Fisher's exact test.The do ed line on the y-axis represents p = 0.05.The plot colors represent the total number of reports.As the RORs became more negative, the tendency toward AEs decreased, and decreasing p-values indicated greater statistical significance.The upper-left portion of the plot shows the concomitant drugs that are less frequently associated with oral mucositis-related AEs.AEs, adverse events; ROR, reporting odds ratio.

Figure 5 .
Figure 5. Drugs associated with oral mucositis-related AEs when combined with antineoplastic drugs.Combination with (a) all antineoplastic drugs, (b) alkylating drugs, (c) antimetabolites, (d) with plant alkaloids and other natural products, (e) with cytotoxic antibiotics and related substances, (f) protein kinase inhibitors, (g) monoclonal antibodies and antibody-drug conjugates, and (h) other antineoplastic drugs.The X-axis shows the natural logarithm of the reporting odds ratios (ln ROR), and the y-axis shows the common logarithm of the inverse p-value (−log (p-value)) from Fisher's exact test.The dotted line on the y-axis represents p = 0.05.The plot colors represent the total number of reports.As the RORs became more negative, the tendency toward AEs decreased, and decreasing p-values indicated greater statistical significance.The upper-left portion of the plot shows the concomitant drugs that are less frequently associated with oral mucositis-related AEs.AEs, adverse events; ROR, reporting odds ratio.

Table 1 .
Number of reports and reporting odds ratios for oral mucositis.
a ATC: Anatomical Therapeutic Chemical.b ROR: Reporting odds ratio.c

Table 2 .
Median values and Weibull parameters of antineoplastic drug-induced oral mucositis.
a ATC: Anatomical Therapeutic Chemical.b CI: Confidence interval.

Table 3 .
Number of reports and reporting odds ratios for concomitant drugs with antineoplastic drugs that show a low association with oral mucositis.