Utilizing Mobile Networks for the Detection of Clinically Relevant Interactions Between Chemotherapy Regimens and Complementary and Alternative Medicines

Abstract

Objectives:

Patients with cancer who use complementary and alternative medicines (CAMs) in conjunction with chemotherapy treatment are at risk of manifesting anticancer drug–CAM interactions (DCIs), which may lead to negative therapeutic outcomes. This article describes a novel iPhone application developed for the Mobile Internet, called OncoRx-MI, which identifies DCIs of single-agent and multiple-agent chemotherapy regimen (CReg) prescriptions.

Methods:

Drug-, CAM-, and DCI-related information was compiled from various hardcopy and softcopy sources, and published literature from PubMed. Overall management plans for the CRegs were then developed. The iPhone Web documents were constructed using Adobe software and programming scripts, and mounted onto a third-party server. DCI searches are based on CReg acronyms, and OncoRx-MI is designed to fit the iPhone screen configuration for improved usability. A small usability study was also carried out and the user feedback presented.

Results:

OncoRx-MI is able to detect over 2700 interactions between 256 CRegs and 166 CAMs, making up a total of over 4400 DCI pairs. The CAMs are classified into seven categories based on their uses in supportive care, and non-cancer-related CAMs are also included. The majority of the DCIs are pharmacokinetic in nature (79%), involving the induction and inhibition of the cytochrome P450 isozymes and p-glycoprotein. Pharmacodynamic DCIs include hepatotoxicity (39%), altered corticosteroid efficacies (30%), and increased risks of hypoglycemia (4%), hypertensive crisis (2%), bleeding, and serotonin syndrome (1% each).

Conclusions:

OncoRx-MI is the first mobile application of its kind that allows searching of DCIs for CRegs through 3G networks, and is intended to improve pharmaceutical care of patients with cancer by assisting health care practitioners in managing CReg interactions in their clinical practices.

Introduction

T he use of complementary and alternative medicines (CAMs), which includes Traditional Chinese Medicines, has become increasingly popular among patients with cancer.¹ A review of 26 worldwide surveys suggested that the use of CAMs in adult cancer populations ranged from 7% to 64%, with an average prevalence of 31.4%.² Furthermore, in the local oncology practice setting, it was reported that more than half (56%) of Singaporean adult patients with cancer used CAMs concomitantly with their chemotherapy.³ In general, patients with cancer use CAMs for many purposes, such as for their anticancer and cancer-preventive properties, as well as for supportive care.^4,5 These patients are highly susceptible to drug–CAM interactions (DCIs) due to the complex pharmacological profiles, steep dose–toxicity curves, and narrow therapeutic indices of the anticancer drugs (ACDs),⁶ which may lead to increased toxicities or subtherapeutic outcomes and may ultimately result in undesirable clinical consequences.⁷ Hence, it is critical that resources on DCIs be available to clinicians in their daily practice so as to help them make a more informed decision with regard to managing DCIs in their patients with cancer.

The advent of third-generation (3G) networks and mobile handheld devices has enabled numerous health care applications to be developed, such as Epocrates Rx⁸ and Lexi-Comp ON-HAND.⁹ Although drug information databases on mobile devices tend to be less comprehensive in terms of completeness and scope compared to their online counterparts,^10,11 the use of such mobile applications among health care professionals will nevertheless continue to grow as there is a need for these applications in the health care setting.¹² However, to the authors' knowledge, there is currently no application that has been designed to identify DCIs of chemotherapy regimen (CReg) prescriptions. A mobile application that draws on the utility of 3G and Wi-Fi networks will be useful in both the hospital and clinic settings for health care professionals who may need to check for DCIs on the spot during patient consultations. Thus, the objective of this study was to develop a novel database application for the Mobile Internet called OncoRx-MI, the purpose of which is to detect DCIs of both single-agent and multiple-agent CReg prescriptions. The initial phase concentrates on developing this application using the iPhone as a prototype, since there is a demand from clinicians for health care applications specific to the iPhone, due to its superior functionality and ease of use.¹³

Methods

Development of the OncoRx-MI application

The OncoRx-MI application is a further development of the CAM module in its parent OncoRx database, which was created for Internet browsers on computers. A summary process of developing the OncoRx-MI application is shown in Figure 1. The list of ACDs and CAMs was compiled based on the methods previously described for OncoRx.¹⁴ Information on the pharmacokinetics of ACDs and the characteristics and uses of CAMs was collated from various hardcopy and softcopy resources.^{15

–25} In addition, a list of single-agent and multiple-agent CRegs that were approved by the U.S. Food and Drug Administration or positively evaluated in Phase II and III trials and published in English-language journals was compiled from PubMed using the keywords anticancer drugs, chemotherapy regimens, and generic names of the ACDs. Information was gathered regarding the acronyms of the CRegs, constituting drugs, types of cancers they are used for, and references substantiating the use of these CRegs in particular types of cancers. Information on the doses and frequencies of administration were not included in the database because these parameters could vary among different institutions. The DCIs of the individual drugs in each CReg were identified using information collated from a total of six hardcopy and softcopy resources.^23

–28 Searches for relevant literature on DCIs were also performed on PubMed using the keywords anticancer drugs, chemotherapy, complementary and alternative medicines, herb–drug/drug–herb interactions, and the common and scientific names of the CAMs. Information on the interaction effects, severities, interacting mechanisms, supporting evidences, relevant references, and management plans was compiled. An overall management plan was then proposed for each interacting CReg. All collected data were organized and tabulated in Microsoft Excel 2007 and uploaded to a Structured Query Language (MySQL 5.0) database.

FIG. 1.

Summary workflow for developing the OncoRx-MI application.

The iPhone Web documents were created using Adobe Dreamweaver CS4 and a combination of hypertext markup languages, javascripts, and hypertext preprocessor (PHP5) scripts. The application was specifically designed to provide a 320×480-pixel Web-browsing environment via the iPhone's Multi-Touch interface.²⁹ The documents were then uploaded and hosted from a third-party domain.³⁰

Search interface of OncoRx-MI

OncoRx-MI is an oncology-specific application that detects DCIs between CRegs and CAMs, and is coupled to an online database located at www.onco-informatics.com/OncoRxMI. Users are required to register for a free account at the Web site, which can be accessed using the iPhone via 3G or Wi-Fi connection. Users can search for interactions based on the acronyms of the CRegs and the common name of the CAMs from the iPhone Safari browser. The relevant information is extracted from the database and presented on the results page, which consists of four sections: information about the selected CReg, pharmacokinetic profiles of the ACDs in the CReg, information about the selected CAM, as well as data on the detected DCIs (Fig. 2).

FIG. 2.

Interaction search between the chemotherapy regimen DHAP (consisting of the drugs dexamethasone, cisplatin, and cytarabine) and licorice in OncoRx-MI.

Under the section on detected DCIs, the information is organized into interaction effects, severities, mechanisms of interactions, substantiating evidences and references, as well as overall management plans for the CRegs. As an example, the interaction between the chemotherapy regimen DHAP, used for the treatment of non-Hodgkin's lymphoma,³¹ and licorice is described (Fig. 3). The regimen DHAP consists of three drugs (dexamethasone, cisplatin, and cytarabine), of which 2 DCI pairs are detected by OncoRx-MI with licorice. The DCI between dexamethasone and licorice, substantiated by a patient case report, leads to an increased duration of corticosteroid activity and an increased risk of potassium depletion.^32

–35 Another DCI detected with cisplatin leads to a decreased therapeutic efficacy of the ACD based on an animal study.³⁶ OncoRx-MI does not detect an interaction with cytarabine. Consequently, OncoRx-MI recommends that a suitable noninteracting alternative herb be considered by the clinician in consultation with a registered herbalist, naturopath, and/or Traditional Chinese Medicine (TCM) practitioner, but if concomitant use of the herb and the regimen is warranted, the patient should be monitored for altered therapeutic efficacies and safety of the two regimen drugs. Additionally, a nonexhaustive list of other noninteracting CAMs that are used for cancer treatment or prevention, and inflammation, is also provided to complement the knowledge and experiences of herbalists, naturopaths, and TCM practitioners, so that they can make better-informed decisions as to which CAMs to prescribe.

FIG. 3.

Screenshots of results page showing the interaction between the chemotherapy regimen DHAP (consisting of dexamethasone, cisplatin and cytarabine) and licorice.

Usability study on the OncoRx-MI database

During the period from July 21, 2010 to January 31, 2011, a specific online form was created to seek feedback from users regarding the OncoRx-MI database. Users who registered for an account to the database during this period were asked to provide their feedback via the link to the online form together with their login information in the e-mails sent to them. However, completion of the survey was voluntary, and users who successfully completed the online form were considered as consenting to the survey. The form asked the users to comment about the usefulness of the OncoRx-MI database in relation to their current sources of oncology DCI information, as well as its relevance to their clinical practice. For usefulness, they were asked to rate the database in terms of information on the drugs, CRegs and CAMs, interaction-related content, management plans, noninteracting alternatives, and usability with 5 scale points (a lot better, a little better, makes no difference, a little worse, a lot worse). For clinical relevance, they were asked whether this database increased their knowledge on managing oncology DCIs, improved their drug-related decision making, and whether it was a useful tool in practice. Users could agree, disagree, or remain neutral to these questions.

Results

OncoRx-MI database statistics

The OncoRx-MI database consists of a total of 256 single-agent and multiple-agent CRegs and 166 CAMs. The CRegs are classified based on the types of cancer they treat (Table 1). On the other hand, the CAMs are classified into seven categories according to their uses in cancer therapy (Table 2). Other non-cancer-related CAMs used by patients with cancer are included as well.

Table 1.

List of Chemotherapy Regimens in OncoRx-MI Classified According to Cancer Types

List of chemotherapy regimens classified based on types of cancer they treat ^a
Bone and tissue cancers Single-agent: doxorubicin, trabectedin Combination: AD, DI/AIM, ICE, IFoVP, MAID, MTX-CDDP/Adria, P6, Topo/CTX, VAC, vincristine/doxorubicin/cyclophosphamide
Breast cancers Single-agent: Abraxane, anastrazole, capecitabine, docetaxel, exemestane, fulvestrant, gemcitabine, ixabepilone, letrozole, megestrol, paclitaxel, tamoxifen, toremifene, trastuzumab, vinorelbine Combination: AC, ACTH, AT, A-T-C, CAF, CEF, CMF, FNC, GT, HEC, ixabepilone/capecitabine, NA, NFL, paclitaxel/vinorelbine, sequential AC-T, sequential Dox-CMF, TAC, tamoxifen/epirubicin, TCH, TH, X+T
Central nervous system and brain cancers Single-agent: carboplatin, temozolomide Combination: A1, A3, BCNU, CAV+P/VP, CDDP/VP-16, COPE (Baby Brain I), high-risk neuroblastoma regimen, PCV, POC, vincristine/cisplatin/cyclophosphamide, vincristine/etoposide/carboplatin/cyclophosphamide/mesna
Gastrointestinal, colorectal, liver, and pancreatic cancers Single-agent: capecitabine, cetuximab, fluorouracil, gemcitabine, imatinib, irinotecan Combination: 5-FU/LV, cetuximab/irinotecan, Douillard regimen, ECF, ELF, FAMTX, FLOX, FOLFIRI, FOLFOX, FUP, GEMOX, IFL/BV, Mayo regimen, OXAFI, Roswell Park regimen, Saltz regimen, TC, XELOX
Genitourinary cancers Single-agent: estramustine, goserelin, interferon α-2b, interleukin-2, leuprolide, nilutamide, prednisone, sorafenib, sunitinib, temsirolimus, triptorelin Combination: BEP, bicalutamide with leuprolide/goserelin, cisplatin/docetaxel, CMV, docetaxel/estramustine, docetaxel/prednisone, EP, estramustine/vinblastine, FL, FZ, gemcitabine/cisplatin, interleukin-2/interferon α-2a, mitoxantrone/prednisone, MVAC, PC, PE, PVB, vinblastine/interferon alfa, VIP
Leukemias Single-agent: alemtuzumab, arsenic trioxide (induction, post-remission), chlorambucil, cyclophosphamide, cytarabine, fludarabine, gemtuzumab, interferon α-2a, imatinib, prednisone Combination: 5+2, 6-MP/CPM/Ara-C, 6-MP/MTX, 7+3, 7+3+7, ATRA, CA, CALGB 8811 (induction, early intensification, central nervous system prophylaxis and interim maintenance, late intensification, maintenance), DA, DAT, DAV, DCTER, double-delayed intensification regimen, HI-C DAZE, HDAC, Hyper-CVAD (central nervous system prophylaxis, even cycles, odd cycles), idarubicin/cytarabine/etoposide, IDMTX, interferon α-2b/hydroxyurea/cytarabine, MTX/6-MP, MTX/6-MP/VP, OD, PVA, PVDA, TIT, vincristine/prednisone/pegaspargase/cytarabine/methotrexate
Lung cancers Single-agent: docetaxel, erlotinib, etoposide, gefitinib, gemcitabine, paclitaxel, topotecan, vinorelbine Combination: CAE, carboplatin/paclitaxel/etoposide, Carbo-Tax, CAV, CAV/EP, CDDP+CPT-11, docetaxel/cisplatin, EC, EP, gemcitabine/cisplatin, gemcitabine/docetaxel/filgrastim, gemcitabine/paclitaxel, gemcitabine/vinorelbine, PC, pemetrexed/cisplatin, vinorelbine/cisplatin, VIP
Lymphomas and myelomas Single-agent: bexarotene, bortezomib, denileukin, dexamethasone, fludarabine, interferon α-2b, melphalan, rituximab, thalidomide, tositumomab, vorinostat Combination: ABVD (adult and pediatric regimens), BEACOPP, CDE, CHOP (AIDs-related, adult and pediatric regimens for non-Hodgkin's lymphoma), CNOP, CODOX-M/IVAC, COP, CVP, DeAngelis protocol, DHAP, EPOCH, ESHAP, Hyper-CVAD/MTX-Ara-C (courses 1 and 2), ibritumomab/rituximab, m-BACOD, MINE-ESHAP, MOPP, MOPP/ABVD, MP, Ommaya regimen, RCHOP, RICE, SNCCL/B-ALL, Stanford V, thalidomide/dexamethasone, VAD
Other cancers Adenocarcinomas: CaT, EP, paclitaxel/carboplatin/etoposide AIDs-related Kaposi's sarcoma: alitretinoin, liposomal daunorubicin, pegylated liposomal doxorubicin, paclitaxel Gynecologic cancers: altretamine, cisplatin, doxorubicin, pegylated liposomal doxorubicin, gemcitabine, medroxyprogesterone, paclitaxel, topotecan, Carbo-docetaxel, Carbo-Tax, cisplatin/fluorouracil, cisplatin/vinorelbine, C-P, CT, TAP Head, neck, and esophageal cancers: carboplatin/fluorouracil, cisplatin/fluorouracil Melanomas: aldesleukin, dacarbazine, interferons α-2a and −2b, temozolomide, cisplatin/dacarbazine/carmustine/tamoxifen, CVD Wilms' tumor: VAD

Chemotherapy regimens that are indicated for multiple conditions are classified into their respective categories.

Table 2.

List of Complementary and Alternative Medicines in OncoRx-MI Classified According to Their Uses

List of complementary and alternative medicines classified based on their uses in cancer therapy^a
Cancer treatment and prevention 714X (trimethylaminohydroxybicycloheptane chloride), agrimony grass (Agrimonia pilosa), α-lipoic acid (1,2-dithiolane-3-pentanoic acid), amygdalin (d-mandelonitrile-b- d-glucosido-6-b- d-glucoside), anvirzel (oleandrin) (Nerium oleander), arisaema rhizome (Arisaema consanguineum, Arisaema heterophyllum, Arisaema amurense), astragalus (Astragalus membranaceus), barley (Hordeum vulgare), bee pollen, bilberry fruit (Vaccinium myrtillus), bitter melon (Momordica charantia), black nightshade (Solanum nigrum), bloodroot (Sanguinaria canadensis), blue-green algae (Spirulina sp., Aphanizomenon flos-aquae), bupleurum (Bupleurum chinense, Bupleurum scorzoneraefolium), burdock (Arctium majus), calcium glucarate (d-glucaro-1,4-lactone), cascara (Rhamnus purshiana), cassia bark (Cinnamomum aromaticum, Cinnamomum cassia), cat's claw (Uncaria tomentosa), cephalotaxaceae (Cephalotaxus fortunei), chaparral (Larrea tridentate, Larrea divaricata), chinese anemone root pulsatilla (Pulsatilla chinensis), Chinese asparagus (Asparagus cochinchinensis), cordyceps (Cordyceps sinensis), coriolus versicolor (Coriolus versicolor, Trametes versicolor, Polyporus versicolor, Polystictus versicolor), dong quai (Angelica sinensis), evening primrose oil (Oenothera biennis), evodia (Evodia rutaecarpa, Evodia officinalis), fenugreek (Trigonella foenum-graecum), forskolin (Coleus forskohlii), fourstamen stephania root (Stephania tetrandra), garlic (Allium sativum), American ginseng (Panax quinquefolius), siberian ginseng (Eleutherococcus senticosus, Acanthopanax senticosus), glossy privet fruit (Ligustrum lucidum Ait), goldenseal (Hydrastis canadensis), gotu kola (Centella asiatica, Hydrocotyle asiatica), green tea (Camellia sinensis), indigo (Isatis tinctoria), isatis leaf (Isatidis baphicacanthi, Isatis tinctoria, Isatis indigotica), isatis root (Isatidis baphicacanthi, Isatis tinctoria, Isatis indigotica), licorice (Glycyrrhiza glabra, Glycyrrhiza uralensis), maitake (Grifola frondosa), milk thistle (Silybum marianum, Carduus marianum), European mistletoe (Viscum album, Viscum coloratum), mume fruit (Prunus mume), noni (Morinda citrifolia), pau d'arco (Tabebuia impetiginosa, Tabebuia avellanedae, Tabebuia heptaphylla), pennyroyal (Mentha pulegium, Hedeoma pulegioides), pinellia rhizome (Pinella ternata, Pinella pedatisecta), pokeweed (Phytolacca americana), rabdosia rubescens (Rabdosia rubescens), rhizoma iphigenia indica (Iphigenia indica), rhubarb (Rheum palmatum, Rheum officinale), saw palmetto (Serenoa repens), skullcap (Scutellaria baicalensis, Scutellaria lateriflora), sheep sorrel (Rumex acetosella), shiitake mushroom (Lentinula edodes), slippery elm (Ulmus rubra), sophora root (Sophora subprostata), soy (Glycine max), spreading hedyotis (Hedyotis diffusa, Hedyotis corymbosa), stillingia (Stillingia sylvatica), trichosanthes root (Trichosanthes kirilowii, Trichosanthes japonica, Trichosanthes uniflora), turmeric (Curcuma longa, Curcuma domestica)
Immune-system-related Astragalus (Astragalus membranaceus), bee pollen, bitter melon (Momordica charantia), blue-green algae (Spirulina sp., Aphanizomenon flos-aquae), cat's claw (Uncaria tomentosa), cordyceps (Cordyceps sinensis), Coriolus versicolor (Coriolus versicolor, Trametes versicolor, Polyporus versicolor, Polystictus versicolor), echinacea (Echinacea purpurea, Echinacea angustifolia, Echinacea pallida), American ginseng (Panax quinquefolius), Asian ginseng (Panax ginseng), Siberian ginseng (Eleutherococcus senticosus, Acanthopanax senticosus), glossy privet fruit (Ligustrum lucidum Ait), goldenseal (Hydrastis canadensis), maitake (Grifola frondosa), European mistletoe (Viscum album, Viscum coloratum), mume fruit (Prunus mume), noni (Morinda citrifolia), reishi mushroom (Ganoderma lucidum), rhubarb (Rheum palmatum, Rheum officinale), shiitake mushroom (Lentinula edodes)
Alopecia Fenugreek (Trigonella foenum-graecum), psoralea fruit (Psoralea corylifolia), swertiae herba (Swertia pseudochinensis, Swertia diluta, Swertia chinensis, Swertia yunnanensis, Swertia japonica)
Nausea and vomiting Ginger (Zingiber officinale), peppermint (Mentha haplocalyx, Mentha piperita, Mentha lavanduliodora, Mentha arvensis)
Peripheral neuropathy and pain Achyranthes root (Achyranthes bidentata, Achyranthes longifolia, Achyranthes aspera), aloe vera (Aloe barbadensis, Aloe capensis), capsaicin (Capsicum frutescens, Capsicum annuum), corydalis (Corydalis turtschaninovii), devil's claw (Harpagophytum procumbens), divaricate saposhnikovia root (Ledebouriella divaricata, Ledebouriella seseloids, Siler divaricatum, Saposhnikovia divaricata, Stenocoelium divaricatum), eucommia bark (Eucommia ulmoides), fourstamen stephania root (Stephania tetrandra), Asian ginseng (Panax ginseng), maté (Ilex paraguariensis), noni (Morinda citrifolia), rhizoma sinomenii acuti (Sinomenium acutum), schefflera (Schefflera arboricola, Schefflera kwangsiensis, Schefflera venulosa), St. John's wort (Hypericum perforatum), willow bark (Salix alba)
Inflammation Achyranthes root (Achyranthes bidentata, Achyranthes longifolia, Achyranthes aspera), aloe vera (Aloe barbadensis, Aloe capensis), arnica (Arnica montana), berberidaceae (Berberis soulieana, Berberis wilsonae, Berberis poirettii, Berberis vernae), black nightshade (Solanum nigrum), bupleurum (Bupleurum chinense, Bupleurum scorzoneraefolium), butcher's broom (Ruscus aculeatus), cat's claw (Uncaria tomentosa), german chamomile (Matricaria recutita L.), chaparral (Larrea tridentate, Larrea divaricata), cinnamon (Cinnamomum verum, Cinnamomum zeylanicum), comfrey (Symphytum officinale), devil's claw (Harpagophytum procumbens), divaricate saposhnikovia root (Ledebouriella divaricata, Ledebouriella seseloids, Siler divaricatum, Saposhnikovia divaricata, Stenocoelium divaricatum), eucommia bark (Eucommia ulmoides), fenugreek (Trigonella foenum-graecum), forsythia fruit (Forsythia suspensa, Syringa suspensa, Forsythia viridissima, Forsythia koreana), fourstamen stephania root (Stephania tetrandra), gotu kola (Centella asiatica, Hydrocotyle asiatica), licorice (Glycyrrhiza glabra, Glycyrrhiza uralensis), lobelia (Lobelia inflata), nettle (Urtica dioica), pennyroyal (Mentha pulegium, Hedeoma pulegioides), pseudoginseng (Panax notoginseng), pygeum (Pygeum africanum, Prunus Africana), rehmannia root (Rehmannia glutinosa), reishi mushroom (Ganoderma lucidum), rhizoma sinomenii acuti (Sinomenium acutum), rhubarb (Rheum palmatum, Rheum officinale), sassafras (Sassafras albidum), saw palmetto (Serenoa repens), skullcap (Scutellaria baicalensis, Scutellaria lateriflora), sheep sorrel (Rumex acetosella), slippery elm (Ulmus rubra), solomon's seal (Polygonatum multiflorum, Polygonatum odoratum, Polygonatum pluriforum), tea tree oil (Melaleuca alternifolia), tree peony bark (Paeonia suffruiticosa), turmeric (Curcuma longa, Curcuma domestica), willow bark (Salix alba)
Fatigue syndrome Blue-green algae (Spirulina sp., Aphanizomenon flos-aquae), cordyceps (Cordyceps sinensis), guarana (Paullinia cupana), maté (Ilex paraguariensis), St. John's wort (Hypericum perforatum)
Non-cancer-related Acorus (Acorus gramineus), alangiaceae (Alangium chinense), atractylodes rhizome (Atractylodes lancea, Atractylodes chinensis), biond magnolia flower (Magnolia biondii, Magnolia fargessi, Magnolia dedudata, Magnolia heptepeta, Magnolia liliflora, Magnolia quinquepeta), birthwort fruit (Aristolochia contorta, Aristolochia debilis), black cohosh (Cimicifuga racemosa), borage (Borago officinalis), buffalo horns (Bubalus bubalis), cattle gallstone (Bos taurus domesticus), chasteberry (Vitex agnus castus), Chinese gentian root (Gentiana manshurica, Gentiana scabra, Gentiana triflora, Gentiana regescens), Chinese lobelia (Lobelia radicans, Lobelia chinensis), Chinese magnoliavine fruit (schisandra) (Schisandra chinensis), Chinese plantain and plantain seed (Plantago asiatica, Plantago major, Plantago depressa), chocolate vine stem (Akebia quinata, Aristolochia manshuriensis), chrysanthemum (Chrysanthemum morifolium, Chrysanthemum sinense, Chrysanthemum japonense), climbing groundsel (Senecio scandens), cocklebur fruit (Xanthium sibiricum, Xanthium strumarium), common aucklandia root (costus root) (Saussurea lappa, Saussurea costus, Aucklandia costus, Aplotaxis lappa), corn silk (Zea mays), croton seed (Croton tiglium), danshen (Salvia miltiorrhiza), dwarf lilyturf tuber (Ophiopogon japonica, Liriope spicata), earthworm (Pheretima aspergillum, Allolobophora caliginosa), endothelium corneum glgeriae galli (Mucusa ventriculi gali), ephedra (Ephedra sinica, Ephedra equisetina), feverfew (Tanacetum parthenium), figwort (Scrophularia marilandica, Scrophularia nodosa, Scrophularia ningpoensis, Scrophularia buergeriana), finger citron fruit (Citrus medica L. var. sarcodactylis), fleeceflower root (Polygonum multiflorum), folium ilicis chinensis (Ilex chinensis), gelatin, ginkgo (Ginkgo biloba), glehnia (Glehnia littoralis), green tangerine orange peel (Citrus reticulata, Citrus tangerina, Citrus sinensis, Citrus aurantium, Citrus wilsonii), gypsum mineral, hawthorn (Crataegus monogyna), herba euphorbiae humifusae (Euphorbia humifusa), horse chestnut (Aesculus hippocastanum), ilex (Ilex pubescens), kava (Piper methysticum), largehead atractylodes rhizome (Atractylodes macrocephala), lightyellow sophora root (Sophora flavescens), nux vomica seed (Strychnos nux-vomica), passionflower (Passiflora incarnate), qing hao (Artemisia annua), red clover (Trifolium pratense), rhizoma menispermi (Menispermum dauricum), safflower flower (Carthamus tinctorius), senega root (Polygala tenuifolia), shrubalthea flower (Vitex negundo L. var. cannabifolia), snakegourd fruit (Trichosanthes kirilowii, Trichosanthes uniflora, Trichosanthes braeteata, Trichosanthes multiloba, Trichosanthes truncata, Trichosanthes hylonoma, Trichosanthes japonica), song luo (Usnea diffracta, Usnea longissima), spicebush root (Lindera strychnifolia), stemona root (Stemona japonica), szechwan lovage rhizome (Ligusticum chuanxiong), tall gastrodia tuber and armillaria (Gastrodia elata), tangerine peel (Citrus reticulate), tendrilleaf fritillary bulb (Fritillaria cirrhosa, Fritillaria unibracteata, Fritillaria przewalskii, Fritillaria delavayi), valerian (Valeriana officinalis), wild indigo (Baptisia tinctoria), wolfberry tree bark (Lycium chinense), wolly grass root (Imperata cylindrica), wormwood (Artemisia absinthium)

Complementary and alternative medicines that have multiple uses are listed in their respective categories.

OncoRx-MI can detect 2750 pairs of interactions between CRegs and CAMs, making up a total of 4408 DCI pairs. CAMs for cancer treatment and prevention (39.0%) and inflammation (16.8%) have the highest proportions for CReg–CAM interacting pairs, while those for alopecia (0.1%) and nausea and vomiting (4.1%) have the lowest (Table 3). Majority of the DCIs are pharmacokinetic in nature (79%), involving the inhibition and induction of the cytochrome P450 isozymes (CYP2C9, 2C19, 2D6, 3A4) and induction of p-glycoprotein. On the other hand, pharmacodynamic DCIs include hepatotoxicity (39%), altered corticosteroid efficacies (30%), increased risks of hypoglycemia (4%), hypertensive crisis (2%), serotonin syndrome, and bleeding (1% each).

Table 3.

Database Statistics According to Categories of Complementary and Alternative Medicines (CAMs)

Classes of CAMs	Number of CAMs in each class ^a	Number of interacting CAMs ^a	Number of chemotherapy regimen–CAM interaction pairs (%) ^b	Number of drug–CAM interaction pairs (%) ^b
Cancer treatment and prevention	66	18	1692 (39.0)	2684 (38.9)
Alopecia	3	1	5 (0.1)	6 (0.1)
Fatigue syndrome	5	5	306 (7.1)	447 (6.5)
Immune-system-related	20	7	550 (12.7)	826 (12.0)
Inflammation	39	7	730 (16.8)	1212 (17.6)
Nausea and vomiting	2	2	178 (4.1)	296 (4.3)
Peripheral neuropathy and pain	15	5	418 (9.6)	690 (10.0)
Non-cancer-related	64	3	456 (10.5)	732 (10.6)

CAMs that have multiple uses are listed in their respective categories.

Percentages may not add up to 100% due to rounding of data.

User demographics and feedback on the database

As of February 2011, there were a total of 130 users of the database. Most users were from the United States and Canada (41%), followed by Europe (23%). Surprisingly, large proportions were also from Australia (16%), Asia (9%), and the United Kingdom and Ireland (5%). Other users included those from the Middle East (3%), South America (2%), and New Zealand (1%). Approximately 42% of the users were from academia or affiliated with an academic institution, and nearly half (46%) were working in hospitals, cancer/medical centers, and health service provider organizations. A small percentage of users were in private practice, from nonprofit governmental organizations or resource providers of cancer drug and integrative medicine information, and pharmacies and the pharmaceutical industry (2% each).

Approximately 64% were health care professionals, inclusive of pharmacists (28%), physicians (25%), nurses (9%), and other allied health professionals (2%). Among the first three groups of health care professionals, a large percentage specialized in oncology (36% pharmacists, 47% physicians, and 75% nurses). The proportion of naturopaths was also high (13%), edging over that of nurses. Researchers and trainees/students made up approximately one fifth of the users, with the latter including those from medicine, pharmacy, life sciences, and biomedical engineering fields.

A total of 26 of 70 users (37%) completed the feedback form from the period of July 21, 2010 to January 31, 2011. On the whole, the majority of those who completed the feedback form agreed that the database in this study was better than their current sources of oncology DCI information (Table 4), with 23 users indicating that the content on drugs, regimens, and CAMs was better than their current sources of information. The same number of users also felt that the interaction-related data provided by OncoRx-MI was better. One user indicated that the content in OncoRx-MI was “a lot worse” because he could not find interactions with a curcumin product. However, curcumin was indexed under a different name (turmeric) in the database. In his comments, he stated that he would like to see more CAMs in the database so that more DCIs could be identified, but nevertheless, he agreed that this database would be useful with a continual expanded list of interactions.

Table 4.

Ratings of OncoRx-MI Parameters Based on User Feedback

	Number of users who provided rating (%)
OncoRx-MI parameters	A lot better	A little better	Makes no difference	A little worse	A lot worse
Content on drugs, regimens and CAMs	9 (35)	14 (54)	2 (8)	0 (0)	1 (4)
Interaction-related content	10 (39)	13 (50)	2 (8)	0 (0)	1 (4)
Management plans	6 (23)	13 (50)	7 (27)	0 (0)	0 (0)
Noninteracting alternatives	3 (12)	12 (46)	11 (42)	0 (0)	0 (0)
Ease-of-use	6 (23)	9 (35)	9 (35)	2 (8)	0 (0)

CAMs, complementary and alternative medicines.

Nineteen (19) users indicated that the management plans provided by OncoRx were better than their current information sources, but only a little more than half (15 users) indicated the same for the noninteracting alternatives. The same number of users indicated that OncoRx-MI's ease-of-use was better, but 2 users also indicated that its usability was “a little worse.” According to the feedback from 1 of them, he felt confused due to the multiple selections of drug and CAM categories, and said that he preferred to have just one selection of the ACD and interacting CAM. In fact, this option is provided for users as the default search interface. A statement describing this default search interface as part of the instructions might clarify users' doubts on how to carry out a search in OncoRx-MI.

In terms of clinical relevance, 20 users indicated that OncoRx-MI was a useful tool in their practices, and that it also helped increase their knowledge on managing oncology DCIs. There were 21 users who also indicated that the database helped improve their drug-related decision making. One (1) user, who was an oncology practitioner, found this database useful when handling a patient query about DCIs. Despite several neutral ratings and some discontent regarding the amount of drugs that OncoRx-MI can identify interactions with, many of the users gave positive comments regarding the database, saying that OncoRx-MI was “much more complete for oncology regimens and CAMs.” One (1) user suggested that more practitioners should know about the database, while another mentioned that the alternative medications suggested by OncoRx-MI could help improve pharmaceutical care by providing “clear instructions on drugs to take and drugs to avoid.” The following comment given by a user summarizes how users felt regarding the usefulness of OncoRx-MI:

I have found the level of detail with regard to the interaction to surpass that in most other databases.…There are a couple [of] interactions identified in your database that are completely lacking in other databases.…The referencing is excellent.…I usually check more than a single database and I have added OncoRx to my list of routinely searched databases. —Feedback comment by a user

Discussion

As far as the authors know, there is currently no oncology-specific database that caters to the search of DCIs via individual ACDs and CRegs. OncoRx-MI, where “MI” stands for the Mobile Internet, is the first database application of its kind that allows users to search for DCIs of CRegs on Smartphones.

OncoRx-MI is different from its online parent version (called OncoRx) that was previously published on the World Wide Web.¹⁴ OncoRx-MI caters to a smaller resolution of mobile devices, and the interface has been modified to allow better usability for iPhone users. The data in OncoRx-MI is organized in longitudinal format, so that users can scroll in an up–down fashion instead of panning sideways to read the DCI information. In addition, users do not need to frequently zoom in and out of the iPhone screen to read the DCI content. This not only saves the hassle, time, and frustration on the clinicians' part during on-the-spot patient consultations, but may also reduce the risk of misinterpretation errors, since it allows them to concentrate on a particular piece of information they require to manage the DCI. Emphasis has been placed on reducing the screen resolution of OncoRx-MI to one that is optimal for mobile Web browsing without compromising in terms of drug- and interaction-related content. By leveraging on the use of 3G networks via mobile devices, the DCI content obtained by the mobile user is exactly of the same quality as that acquired using the online version.

The primary advantage of OncoRx-MI is that it enables the search for DCIs of CRegs quickly and conveniently in the clinical setting. This is practical since patients with cancer are often on combination regimens within which the individual drugs are usually standard. DCI information is usually limited and scattered among the various available sources, which include primary literature, databases,^23,24 textbooks on CAMs,^20,25 and conventional drug information manuals.^15,17,18 As such, to conduct a comprehensive search of DCIs with a particular combination CReg, clinicians need to plow through a large number of resources and make sense of the huge amount of data themselves. Furthermore, due to the time limitation during patient consultations, clinicians may be prone to errors of omission during an interaction search, which can pose a risk of patients suffering from serious DCI effects. OncoRx-MI is able to hasten the search process because it is a centralized database that has information collated from a variety of sources and includes DCIs identified based on similar metabolic pathways and class effects. By allowing searches based on acronyms of CRegs rather than by individual ACDs, OncoRx-MI not only reduces the time and effort required to search for DCIs of a particular combination regimen, but also reduces the probability of drug-related problems.

Furthermore, OncoRx-MI provides users with conflicting information regarding DCIs that are substantiated with references, as well as overall management plans for the interacting CRegs, so that clinicians can have a holistic understanding of the interaction and better manage the patient based on clinical judgment. In addition, OncoRx-MI draws on the popularity of mobile handheld devices and the widespread availability of 3G and Wi-Fi networks to deliver information regarding DCIs to clinicians “on-the-spot” during patient consultations, as and when required. This can be helpful to clinicians working in inpatient settings due to the existence of certain environmental barriers in hospitals, such as connection or firewall issues that impede intranet access, and/or old, slow, or limited number of computers in wards or treatment areas.³⁷

Limitations and future work

The main limitation of OncoRx-MI is that the interaction search is limited to a CReg–CAM pair at any one time. Searches including two or more CAMs with CRegs will require separate search entries. It is intended that future updates of OncoRx-MI will allow users to search for multiple CReg–CAM interactions, as well as to include DCI information with other natural products and dietary supplements, such as vitamins and minerals. Besides the iPhone, the OncoRx-MI interface has not been tested on other Smartphones. As such, differences in the layout of the interface and slight discrepancies in the screen resolution may occur when OncoRx-MI is assessed using other mobile handheld devices. Ongoing improvements to the compatibility of OncoRx-MI with other mobile devices are being made so that a wider target audience of health care professionals can be reached. Improvements in usability of the interface will also be considered in future versions. OncoRx-MI is nonetheless still useful in its current phase because it is able to detect a considerable number of interactions between CAMs and the CRegs used in clinical practice.

Conclusions

OncoRx-MI is the first mobile Web-based application of its kind targeted to detect DCIs for CRegs. It is intended to complement other currently available mobile drug interaction resources for detection of DCIs. This application is useful to health care practitioners in the clinical oncology setting, or to those who deal with patients with cancer in their daily practice. It is also of assistance to clinicians who are new to oncology practice, and other clinical specialists who may not be familiar with chemotherapy. It is hoped that OncoRx-MI will increase the awareness of DCIs among health care professionals working in the clinical oncology setting, and can ultimately improve the pharmaceutical care of patients with cancer.

Footnotes

Acknowledgments

This article is written within a project funded by the start-up grant awarded by the Faculty of Science, National University of Singapore.

Disclosure Statement

The authors have no conflicts of interest that are directly relevant to the content of this study.

References

Boon

, Olatunde

, Zick

. Trends in complementary/alternative medicine use by breast cancer survivors: Comparing survey data from 1998 and 2005. BMC Women's Health, 2007; 7:4.

Ernst

, Cassileth

. The prevalence of complementary/alternative medicine in cancer: A systematic review. Cancer, 1998; 83:777–782.

Shih

, Chiang

, Chan

. Complementary and alternative medicine (CAM) usage in Singaporean adult cancer patients. Ann Oncol, 2009; 20:752–757.

Mansky

, Wallerstedt

. Complementary medicine in palliative care and cancer symptom management. Cancer Journal, 2006; 12:425–431.

Tascilar

, de Jong

, Verweij

, Mathijssen

RHJ

. Complementary and alternative medicine during cancer treatment: Beyond innocence. Oncologist, 2006; 11:732–741.

Beijnen

, Schellens

JHM

. Drug interactions in oncology. Lancet Oncol, 2004; 5:489–496.

Meijerman

, Beijnen

, Schellens

JHM

. Herb-drug interactions in oncology: focus on mechanisms of induction. Oncologist, 2006; 11:742–752.

Epocrates Inc. Epocrates Rx. www.epocrates.com/products/rx/. 2011 January 10.

Lexi-Comp Inc. Lexi-Comp^® On-Hand. http://webstore.lexi.com/Store/ON-HAND. 2011 January 10.

10.

Clauson

, Polen

, Marsh

. Clinical decision support tools: Performance of personal digital assistant versus online drug information databases. Pharmacotherapy, 2007; 27:1651–1658.

11.

Clauson

, Polen

, Peak

et al. Clinical decision support tools: Personal digital assistant versus online dietary supplement databases. Ann Pharmacother, 2008; 42:1592–1599.

12.

Terry

. Medical apps for Smartphones. Telemed e-Health, 2010; 16:17–22.

13.

Manufacturing, and Logistics IT. Doctor+iPhone=better healthcare. www.logisticsit.com/absolutenm/templates/article-mobile.aspx?articleid=3701&zoneid=8. 2011 February 3.

14.

Yap

KY-L

, Kuo

, Lee

JJJ

et al. An onco-informatics database for anticancer drug interactions with complementary and alternative medicines used in cancer treatment and supportive care: An overview of the OncoRx project. Support Care Cancer, 2010; 18:883–891.

15.

British Medical Association. Pharmaceutical Society of Great Britain. British National Formulary. London, UK: BMJ Publishing Group Ltd & RPS Publishing, 2007.

16.

BC Cancer Agency. BCCA cancer drug manual. www.bccancer.bc.ca/HPI/DrugDatabase/default.htm. 2011 February 3.

17.

Lacy

, Armstrong

, Goldman

, Lance

. Drug Information Handbook with International Trade Names Index. Hudson, OH: Lexi-comp Inc., 2007–2008.

18.

Solimando

DAJ

. Drug Information Handbook for Oncology. Hudson, OH: Lexi-comp, 2007.

19.

Klein

, Chang

, Cho

et al. Integrating genotype and phenotype information: An overview of the PharmGKB project. Pharmacogenomics J, 2001; 1:167–170.

20.

Chang

H-M

, But

PP-H

, Yao

S-C

et al. Pharmacology and applications of Chinese materia medica. Singapore: World Scientific Publishing Co., 2001.

21.

Aetna InteliHealth Inc. Index of herbal medicines and supplements. www.intelihealth.com/IH/ihtIH/WSIHW000/8513/31402.html. 2011 January 10.

22.

e-MS Inc. TCM Assistant. www.tcmassistant.com/index.html. 2011 January 10.

23.

Gerstner

Jr . Memorial Sloan-Kettering Cancer Center—cancer information (about herbs, botanicals and other products) www.mskcc.org/mskcc/html/11570.cfm. 2011 January 10.

24.

Thomson Healthcare. Micromedex^® Healthcare Series (version 5.1) [intranet database] www.thomsonhc.com/hcs/librarian/PFPUI/fV1mNyY303TXxR. 2011 February 3.

25.

Jellin

, Gregory

. Natural Medicines Comprehensive Database. Stockton, CA: Therapeutic Research Faculty, 2007; 1–2304.

26.

Drugsite Trust. Drugs.com™Drug Interactions Checker. www.drugs.com/drug_interactions.html. 2011 February 3.

27.

Gold Standard Inc. Caremark Drug Interactions. http://cpref.goldstandard.com/inter.asp?r=8084. 2011 February 3.

28.

Enhanced Medical Decisions Inc. DoublecheckMD.com ^™ www.doublecheckmd.com. 2011 February 3.

29.

Apple Inc. iPhone 3G(S) Technical Specifications. www.apple.com/iphone/iphone-3gs/high-technology.html. 2010 October 25.

30.

New Dream Network LLC. DreamHost^™ web hosting. www.dreamhost.com. 2011 February 3.

31.

Velasquez

, Cabanillas

, Salvador

et al.

Effective salvage therapy for lymphoma with cisplatin in combination with high-dose Ara-C and dexamethasone (DHAP)

Blood, 1988; 71:117–122.

32.

Quinkler

, Stewart

. Hypertension and the cortisol-cortisone shuttle. J Clin Endocrinol Metab, 2003; 88:2384–2392.

33.

Yoshida

, Takayama

. Licorice-induced hypokalemia as a treatable cause of dropped head syndrome. Clin Neurol Neurosurg, 2003; 105:286–287.

34.

Walker

, Edwards

. Licorice-induced hypertension and syndromes of apparent mineralocorticoid excess. Endocrinol Metab Clin North Am, 1994; 23:359–377.

35.

Sundaram

, Swaminathan

. Total body potassium depletion and severe myopathy due to chronic liquorice ingestion. Postgrad Med J, 1981; 57:48–49.

36.

Lee

, Park

, Lim

et al. Effects of the licorice extract against tumor growth and cisplatin-induced toxicity in a mouse xenograft model of colon cancer. Biol Pharm Bull, 2007; 30:2191–2195.

37.

Tan

, Stark

, Lowinger

et al. Information sources used by New South Wales cancer clinicians: A qualitative study. Intern Med J, 2006; 36:711–717.