lable at ScienceDirect

Biomaterials 32 (2011) 7711e7720

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

A pH-responsive mesoporous silica nanoparticles-based multi-drug deliverysystem for overcoming multi-drug resistance

Qianjun He a,1, Yu Gao b,1, Lingxia Zhang a, Zhiwen Zhang b, Fang Gao b, Xiufeng Ji b, Yaping Li b,*,Jianlin Shi a,*a State Key Laboratory of High Performance Ceramics and Superfine Microstructure, Shanghai Institute of Ceramics, Chinese Academy of Sciences, 1295 Ding-Xi Road,Shanghai 200050, Chinab Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zu Chong-Zhi Road, Shanghai 201203, China

a r t i c l e i n f o

Article history:Received 16 June 2011Accepted 26 June 2011Available online 4 August 2011

Keywords:Mesoporous silicaNanoparticleControlled drug releaseMulti-drug deliveryOne-pot self-assemblyMulti-drug resistance

* Corresponding authors. Tel.: þ86 21 52412714; faE-mail addresses: ypli@mail.shcnc.ac.cn (Y. Li), jlsh

1 Contributed equally to this work.

0142-9612/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.biomaterials.2011.06.066

a b s t r a c t

A type of pH-responsive nano multi-drug delivery systems (nano-MDDSs) with uniform particle size(100 � 13 nm) and excellent monodispersity was developed by in situ co-self-assembly among water-insoluble anti-cancer drug (doxorubicin, DOX), surfactant micelles (CTAB) as chemosensitiver andsilicon species forming drugs/surfactant micelles-co-loaded mesoporous silica nanoparticles (drugs@-micelles@MSNs or DOX@CTAB@MSNs) via a micelleseMSNs self-assembly mechanism. The nano-MDDSDOX@CTAB@MSNs had a highly precise pH-responsive drug release behavior both in vitro and in vivo,and exhibited high drug efficiencies against drug-resistant MCF-7/ADR cells as well as drug-sensitiveMCF-7 cells by the MSNs-mediated transmembrane delivery, the sustained drug release and the highanti-cancer and multi-drug resistance (MDR)-overcoming efficiencies. The MDR-overcoming mechanismwas proved to be a synergistic cell cycle arrest/apoptosis-inducing effect resulted from the chemo-sensitization of the surfactant CTAB. These results demonstrated a very promising nano-MDDS for thepH-responsive controlled drug release and the cancer MDR overcoming.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Cancer is one of the deadliest killers to human lives, and boththe incidence and mortality of cancers are continuously rising [1].Unfortunately, a wholesale and thorough cure for cancers remainselusive for many reasons. One of the well known reason should bestrong toxic side effects of the naked drugs or traditional drugdelivery systems (DDSs) mainly due to the drug leakage beforereaching the cancer focus. Besides, themulti-drug resistance (MDR)in cancers should be responsible for the high recurrence rate andthe final failure in cancer chemotherapy: more than 90% of malig-nant tumor patients die from the certain extent MDR [2]. Therefore,the MDR in cancer become one of the major obstacles in thechemotherapeutic treatment of many human cancers [3]. Theovercoming of the MDR in cancer is continuously a hot topic inanti-cancer researches. The current strategies to overcomethe tumor MDR generally resort to the multi-drug combined

x: þ86 21 52413122.i@sunm.shcnc.ac.cn (J. Shi).

All rights reserved.

chemosensitization, the reconstruction of primary drugs, and bio-/nano-technologies. The combined use of two or several strategies isbeing recognized to be a realistic route to the successful chemo-therapeutic treatment [4,5]. By integrating the multi-drug che-mosensitizationwith the nanotechnology, some nano drug deliverysystems (nano-DDSs) based on organic or inorganic nano-carriershave been designed to overcome the MDR and also enhance thedrug efficacies against both drug-sensitive and drug-resistantcancer cells mostly by improving drug bio-accessibility and che-mosensitivity [6�8]. The common organic nano-carriers, such asliposomes, lipid nanocapsules, nano hydrogels, polymer nano-particles and micelles, can easily solubilize and encapsulate poorlywater-soluble drugs into their hydrophobic cores, however are alsophysicochemically instable and consequently subject to unex-pected drug leakage [9].

In contrast, inorganic nano-carriers, such as mesoporous silicananoparticles (MSNs), carbon nanomaterials and gold nano-particles, are highly stable physicochemically and biochemically.Especially, MSNs possess some unique features, such as uniformand tunable particle size/pore size/morphology, high surface areaand pore volume, facile surface functionalization, etc [10�13]. Theas-synthesized MSNs before removing the structure directing

Q. He et al. / Biomaterials 32 (2011) 7711e77207712

agents, mainly surfactant micelles, which is named as micel-les@MSNs here, can be regarded as a kind of organiceinorganiccomposite nano-carriers. Such a nano-carrier would probably showan almost perfect combination of the water-insoluble drug loading/delivery, the high physicochemical and biochemical stability, theanti-leaking and the controlled drug release [14�16]. In this work,we will demonstrate that such a composite nano-carrier, micel-les@MSNs (e.g. CTAB@MSNs when the CTAB is selected as thesurfactant micelles), can indeed combine these advantages in one.

Further, the multi-drug resistance in cancer is frequentlyresulted from over low intracellular drug concentrations whenadministrated with some poorly water-soluble anti-cancer drugs,such as doxorubicin, camptothecin, colchicines, vinblastine, pacli-taxel, etc. Therefore, the use of nano-carriers for improving theaqueous solubility of poorly water-soluble drugs and enhancingintracellular drug concentration (bio-accessibility) has attractedgreat attentions [17], and here we find that the present micel-les@MSNs is a good choice of composite nano-carriers thanks to thesolubilization effect of the micelles for hydrophobic drugs. In thiswork, we successfully encapsulated a common poorly water-soluble anti-cancer drug doxorubicin (DOX) into the micel-les@MSNs to construct a nano multi-drug delivery system (nano-MDDS) drugs@micelles@MSNs by a facile one-pot co-self-assemblystrategy. Noticeably, some surfactants, such as CTAB (cetyl tri-methyl ammonium bromide), Tween 80, Triton X-100, P123, F127,Nonidet P-40, DTAB (dodecyl trimethyl ammonium bromide) andSDBS (sodium dodecyl benzene sulfonate), are all well qualified asboth anti-cancer active drugs and chemosensitizers for overcomingthe MDR in cancer [18�22]. Herein, CTAB was chosen as botha structure-directing agent for constructing a drugs@mi-celles@MSNs nano-MDDS (DOX@CTAB@MSNs) and a chemo-sensitizer for overcoming the MDR and enhancing the drugefficacies (Scheme 1), because the anti-cancer activity of CTAB hasbeen proved to be exceptionally high [19,20] and the particle sizeand dispersivity of the CTAB-templated MSNs could be facilelycontrolled [23].

Compared with the traditional drug post-loading route, in thisreport the simultaneous loading of multiple drugs including poorlywater-soluble drugs and surfactant chemosensitizer was achievedby an in situ drug co-loading strategy, or in another word, a facileone-pot co-self-assembly route, thanks to the hydrophobic natureof the core part of the surfactant micelles in micelles@MSNs, asillustrated by Scheme 1. Moreover, the constructed drugs@mi-celles@MSNs nano multi-drug delivery system (nano-MDDS) show

Scheme 1. The methodological comparison between the traditional MSNs-based nano-DDco-loading route.

a highly desirable precise pH-responsive drug release performance,i.e. loaded drug will not leak in normal tissues where the pH valueis usually constant at 7.2e7.5, but will release in a sustained waywithin cancer cells where the pH value is a little lower through theion exchanging interaction between Hþ/H3Oþ and positivelycharged drugs@micelles. Therefore the present nano-MDDS canhopefully resolve the difficulties in the traditional post-loading anduncontrolled release of poorly water-soluble drugs and the low pH-responsivity of the MSNs-based nano-DDSs. Furthermore, this pHvalue mediated multi-drug co-release is expected to greatly favorthe drug accumulation within cancer cells and the multi-drugcombined chemosensitization for overcoming MDR in cancer. Inaddition, in spite of the unavailable cytotoxic specificity of bothCTAB and DOX, the present nano-MDDS DOX@CTAB@MSNs isexpected not only to show a passive targeting capability via theenhanced permeability and retention (EPR) effect of abnormaltumors, but also to have a specific drug release character: thesustained release of CTAB and DOX within acidic cancer cells buthardly release in normal neutral tissue or cells.

2. Materials and methods

2.1. Preparation of DOX@CTAB@MSNs

Poorly water-soluble drug DOX (0.5 mmol, Beijing HuaFeng United TechnologyCo., Ltd., Beijing) was solubilized in 500 mL of CTAB (Sigma-Aldrich Co.) solution(10 mM) at 80 �C under intensive stirring and light-sealed environment. Afterseveral minutes, a clear solution was obtained and then NH4F (0.04 mol, SinopharmChemical Reagent Co. Ltd., Shanghai) which was used as a catalyzer for acceleratingthe hydrolyzation and condensation of silicon sources was added. Immediately,tetraethyl orthosilicate (TEOS, 0.04 mol, Sinopharm Chemical Reagent Co. Ltd.,Shanghai) was added dropwise for 20 min. After 1 h, the reaction solution in a semi-transparent colloidal state was centrifugated for 10 minwith the centrifugal force of18000 g in a high speed refrigerated centrifuge. Nanoparticles were collected andwashed with ethanol and deionized water three times in turn for completelyremoving residual reactants. Finally, products were dispersed in deionized waterand the freeze drying power was used for measurements. The control sampleCTAB@MSNs without loading drugs was synthesized via the similar method byadding no drugs in the initial reaction solution. To obtain pure MSNs carrier, sampleCTAB@MSNs was extracted several times with a mixed solution of ethanol (150 mL)and hydrochloric acid (36e38%, 2 mL) to completely remove CTAB.

2.2. Nanoparticles characterization

The UV adsorption spectra of clear reaction solution before adding TEOS andupper clear solution after 1 h reaction and 10 min centrifugation were collected ona Shimadzu UV-3101PC UV-vis absorption spectrophotometer. According to theadsorbance difference, the drug loading capacity of DOX@CTAB@MSNs was calcu-lated by the BeereLambert law. The drug loading capacity of DOX@CTAB@MSNs was

S by the drug post-loading route and the proposed nano-MDDS by the drug in situ

Q. He et al. / Biomaterials 32 (2011) 7711e7720 7713

measured by dissolving DOX@CTAB@MSNs into a clear aqueous solution of HF andthen determining the concentration of free drugs (w 2.4 mg DOX and 600 mg CTABper 1 g MSNs).

The morphology and mesostructure of nanoparticles were observed viatransmission electron microscopy (TEM). TEM micrographs were obtained ona JEMe2010 electron microscope with an accelerating voltage of 200 kV. The mes-ostructure ordering was characterized by small-angle X-ray diffraction (SA-XRD).SA-XRD data were recorded on Rigaku D/Max-2550 V diffractometer using Cu Karadiation (40 kV and 40mA) at a scanning rate of 0.4�/min over the range of 0.5e6.0�

with a step width of 0.002� . The particle size distribution data were collected bya DLS method in a Mastersizer 2000 analyzer (Malvern Instruments Ltd. UK).

2.3. Cytotoxicity against MCF-7 and MCF-7/ADR cells

MCF-7 cells were cultured in RPMI 1640 (GIBCO, New York) containing 10% fetalbovine serum (FBS, Sijiqing Biological Engineering Materials Co., Ltd., Hangzhou).MCF-7/ADR cells were cultured in RPMI 1640 containing 10% FBS and 1 mg mL�1 ofDOX. Cells were maintained at 37 �C in a humidified and 5% CO2 incubator. For allexperiments, cells were harvested by the use of 0.25% trypsin (Sigma) in D-Hank’ssolution (0.40 g KCl, 0.06 g KH2PO4, 8.00 g NaCl, 0.35 g NaHCO3, 0.048 g Na2HPO4,1000 mL H2O) and resuspended in fresh medium before plating. In vitro cytotoxicityagainst MCF-7 and MCF-7/ADR cells was assessed by the standard Cell Counting Kit-8 (CCK-8, Beyotime Institute of Biotechnology, Jiangsu) assay. The statistical evalu-ation of data was performed using a two-tailed unpaired Student’s t-test. A p-valueof less than 0.05 was considered statistically significant. Each data point is repre-sented as mean � standard deviation (SD) of eight independent experiments (n ¼ 8,n indicates the number of wells in a plate for each experimental condition). The timeand dose dependences of the cytotoxicity were investigated at different time pointsof incubation (24 h, 48 h and 72 h) at different concentrations.

MCF-7 cells were seeded in 96-well plates at a density of 104 cells per well. Afterincubation for 24 h at 37 �C in 100 mL RPMI 1640medium containing 10% FBS, culturemedium was discarded and then cells were treated with 100 mL pH 7.4 D-Hank’ssolution of drugs at different concentrations. At the end of each incubation (24 h,48 h or 72 h), 10 mL of CCK-8 solution was added into each wells. After cells wereincubated for another 4 h, the absorbance was monitored at 450 nm on a micro-plate reader (Bio-Tek ELx800). A culture medium without nanoparticles was usedas the blank control. The cytotoxicity was expressed as the percentage of the cellviability as compared with the blank control.

2.4. Evaluation of MCF-7/ADR cell death mechanisms

Annexin V-FITC apoptosis detection kit (Invitrogen, Oregon, USA) was used toquantify the apoptotic and necrotic cells by a standard FACS assay. MCF-7/ADR cellswere double stained with Annexin V and PI (propidium iodide) and subject to flowcytometry (FCM). Firstly, MCF-7/ADR cells were seeded into 6-well plates (2 � 106

cells per well). After incubation for 24 h, cells were treated with 100 mL pH 7.4 D-Hank’s solution of drugs. The same particle concentration of 100 mg mL�1 was usedfor DOX@CTAB@MSNs and CTAB@MSNs. Correspondingly, the free CTAB solution(60 mg mL�1) containing the same drug concentration with the DOX@CTAB@MSNssolution was used for comparison, but the concentration of free DOX was magnifiedten times so as to cause cell death, in order to reveal the death mechanisms and cellcycle distributions of MCF-7/ADR cells. At the end of each incubation (1 day, 2 daysor 3 days), culture medium was discarded and cells were washed twice with the D-Hank’s solution. Then the fluorescein isothiocyanate (FITC)-labeled Annexin V (1 mL,1 mg mL�1; Responsif, Erlangen, Germany) was added to each well. Cells wereincubated at room temperature on the shaker (100 rpm) for 15 min in darkness.Then the PI stock solution (5 mL, 750 mM) was added to each well. After cells wereincubated for another 5 min, cells were washed twice with the D-Hank’s solutionagain. For each experiment, drug-untreated cells were used as a blank control.Signals collected on a FACSCalibur flow cytometry (Becton Dickinson, Zürich,Switzerland) with a CellQuest software were gated to separate live/apoptotic/necrotic cells: AV�/PI� (lower left quadrant), intact live cells; AVþ/PI� (lower rightquadrant), early apoptotic cells; AVþ/PIþ (upper left quadrant), necrotic cells; AV�/PIþ (upper right quadrant), late apoptotic/secondarily necrotic cells.

2.5. Intracellular ATP level assay

MCF-7/ADR cells were seeded into 96-well plates (5000 cells per well), incu-bated for 24 h at 37 �C, and then treatedwith drugs or carrier alone. Intracellular ATPlevels were determined using the luciferin-luciferase-based ATP luminescence assaykit (Beyotime Institute of Biotechnology, Jiangsu) as instructed by the manufacturer.

2.6. Cell cycle distribution by FCM DNA analysis

MCF-7/ADR cells were seeded into 6-well plates (2 � 106 cells per well), andthen treated with drugs for 24 h. After incubation for another 24 h, adherent cellswere detached by the addition of trypsin, washed with PBS and collected byrefrigerated centrifugation. Then cells were resuspended in ice-cold PBS, and thenfixed with 70% precooled ethanol at 4 �C overnight. Ethanol-fixed cells were washed

twice with ice-cold PBS again, and then incubated with 1 mg mL�1 RNase A for20 min at 37 �C, and stained with 10 mg mL�1 PI for 30 min in the dark. The DNAcontent was measured on a FACSCalibur flow cytometry (Becton Dickinson, Zürich,Switzerland), and cell cycle distribution was determined by a ModFit software(Verity Software House, Topsham, ME).

2.7. In vitro drug release behaviors of DOX@CTAB@MSNs in buffer solutions ofdifferent pH values

Sterilized dialysis bags with dialyzer molecular-weight cut-off 10,000 Da wereused to carry out the drug release experiments. These dialysis bags were pretreatedprior to use as follows. These dialysis bags were fully immersed into 50% aqueoussolution of ethanol and boiled 1 h, and then washed with 50% ethanol, 10 mmol L�1

NaHCO3 and 1 mmol L�1 EDTA in turn. Phosphate buffered saline (PBS) of pH ¼ 7.4and acetic buffer solutions (ABS) of pH ¼ 4, 5 and 6.5 were used as the drug releasemedia to simulate normal blood/tissues and tumor environments. The sampleDOX@CTAB@MSNs (60 mg) was dispersed into 2 mL release media, and then thesolutions were put into pretreated dialysis bags. The sealed dialysis bags were putinto brown bottles and then 58 mL release media was added. These bottles wereshaken at a speed of 100 rpm at 37 �C under a light-sealed condition. At certain timeintervals, 3 mL of the release media were taken out for measuring the released drugconcentrations by the UV-vis absorption technique, and then were returned to theoriginal release media. The concentrations of released drug were calculated by theBeereLambert law according to the absorbances of the release media at a certaincharacteristic adsorption wavelength.

2.8. In vitro intracellular uptake and drug release behaviors of DOX@CTAB@MSNs inMCF-7/ADR cells

Green emitting fluorescein isothiocyanate (FITC, Sigma-Aldrich Co.) was graftedonto DOX@CTAB@MSNs (denoted by ‘DOX@CTAB@MSNs-FITC’) by a facile co-condensation approach as described in our previous report [24]. In a typical proce-dure, 104 MCF-7/ADR cells were cultured for 8 h at 37 �C in RPMI 1640 mediumsupplemented with 10% FBS in each well. Then cells were gently washed twice withD-Hank’s solution of pH 7.4, and subsequently D-Hank’s-bufferedDOX@CTAB@MSNs-FITC at a final concentration of 160 mg mL�1 added into petri dishes. After incubationfor 4 h, MCF-7/ADR cells were washed for several times with D-Hank’s solution toremove the residual nanoparticles, and then were totally lysed under an ultrasoniccell disruptor and centrifugated for 10 min with the centrifugal force of 18000 g ina high speed refrigerated centrifuge. The intracellular drug-released concentrationswere measured bymonitoring collected upper clear solutions using a TECAN infiniteF200 microplate reader (emission wavelength of 595 nm, excitation wavelength of480 nm). The intracellular concentrations of all DOX, including non-released andreleased DOX, were measured by monitoring the dissociated (not lysed) MCF-7/ADRcells by flow cytometry. Besides, the intracellular uptakes and the localizations ofDOX@CTAB@MSNs-FITC and released drug DOX were directly visualized via a LeicaTCS confocal microscope (Leica Microsystems, Germany). In the assay, all experi-ments were carried out under a light-sealed condition to avoid photo-bleaching.

2.9. In vivo drug release behaviors of DOX@CTAB@MSNs in tumors

The tumor-bearing mousemodel was induced by injecting MCF-7/ADR cells intothe right infra-axillary dermis of nude mice. MCF-7 tumor-bearing mice wereintravenously injected with DOX@CTAB@MSNs at a dose of 5 mg g�1 through the tailvein. The administrated mice were maintained daily under a 12 h light/dark cycle atthe Animal Care Facility, and fresh distilled water and food for all animals wereavailable ad libitum. After mainline administration for 2 days, mice were sacrificedvia the cervical dislocation. The tissues such as liver, spleen, lung and tumor werediscretized, rinsed with sterile physiological saline, and then blotted dry with filterpaper. All the discretized tissues were cut into slices of 5 mm in thick followingpolyoxymethylene fixing, paraffine embedding and actin staining (Actin-TrackerGreen Kit, Beyotime, Jiangsu) for directly observing the localization and drug releaseof DOX@CTAB@MSNs on a Leica TCS confocal microscope. The in vivo animalexperiments were carried out under the guideline approved by the InternationalAnimal Care and Use Committee (IACUC) of Shanghai Institute of Materia Medica,Chinese Academy of Sciences.

3. Results and discussion

3.1. Synthesis and characterization of DOX@CTAB@MSNs

An important feature in the construction of the present nanomulti-drug delivery system (nano-MDDS) DOX@CTAB@MSNs is thenear neutral reaction conditions. As the co-loaded drug DOX willeasily hydrolyze under either basic or acidic conditions, therefore,we used NH4F as a catalyzer for accelerating the hydrolyzation and

Q. He et al. / Biomaterials 32 (2011) 7711e77207714

condensation of silicon sources, consequently improving the dis-persivity and uniforming the particle size of nano-sized MSNsunder a near neutral reaction condition. From transmission elec-tron microscopy (TEM) images, it can be found that bothCTAB@MSNs and DOX@CTAB@MSNs have fairly uniform particlesizes (112 � 14 nm and 101 �12 nm, respectively) and almost idealmonodispersity (Fig.1a and b) in accordancewith the dynamic lightscattering results (Fig. 1c and d), and show a partially orderedmesoporous structure in conformity with small-angle X-raydiffraction (SA-XRD) measurements (Fig. 1e). Furthermore, a reac-tion temperature of 80 �C was found to be suitable to accelerate thesolubilization of DOX without significant damage to DOX under thepresent near neutral reaction condition.

3.2. In vitro cytotoxicity of the nano-MDDS against MCF-7 andMCF-7/ADR cells

Next, we evaluated the cytotoxicity of the synthesized nano-MDDS DOX@CTAB@MSNs against both MCF-7 and MCF-7/ADRcells (human breast cancer cells). As for the drug-sensitive MCF-7cells, the cytotoxicity of free CTAB and DOX has apparent drugconcentration and treatment time dependences (Fig. 2a, b and c),and both free CTAB and free DOX show remarkable cytotoxicity atvery low concentrations, and free CTAB is evenmore toxic than freeDOX at the same concentrations, as previously identified as aneffective anti-cancer agent [20]. However, this cytotoxicity differ-ence between free CTAB and free DOX become less at the increaseddrug treatment time. After drug treatment for 3 days, the IC50 (50%inhibiting concentration) values of free CTAB and DOX are about

Fig. 1. TEM images of CTAB@MSNs (a) and DOX@CTAB@MSNs (b), and particle diameter disand SA-XRD patterns of CTAB@MSNs and DOX@CTAB@MSNs (e).

1.8 mg mL�1 and 2.9 mg mL�1, respectively. By comparison with freedrugs, the cytotoxicity of the constructed nano-MDDS DOX@C-TAB@MSNs also show strong drug concentration and treatmenttime dependences (Fig. 2d e and f). Importantly, the nano-MDDSDOX@CTAB@MSNs has higher cytotoxicity against MCF-7 cellsthan CTAB@MSNs at the same particle concentrations and for thesame treatment time, in spite of lower cytotoxicity of free DOX thanfree CTAB as above mentioned. This suggests that the nano-MDDSDOX@CTAB@MSNs can be well uptaken by MCF-7 cells, and haveremarkably improved the intracellular accessibility of the poorlywater-soluble drugs. After drug treatments for 3 days, the IC50values of CTAB@MSNs and DOX@CTAB@MSNs are about18.6 mg mL�1 and 17.3 mg mL�1, respectively, based on particleconcentrations. Such low IC50 values of the nano-MDDSs indicatetheir high drug efficiencies in vitro against MCF-7 cells, and alsoimply the expected low using amounts of drug and carrier. Inaddition, the cytotoxicity between free drugs and the nano-MDDSat the same drug concentrations might be incommensurabledirectly because the drugs were released into cells from the nano-MDDS in a highly sustained way as under-mentioned and theintracellular uptaking fashions of free drug molecules and thenano-MDDS were potentially different.

As for drug-resistant MCF-7/ADR cells, both free CTAB and freeDOX do not bring visible cytotoxicity in wide ranges of drugconcentration and treatment time (Fig. 3a b and c), indicating therepresentative MDR characteristics of used MCF-7/ADR cells. It isworth noting that comparedwith free drugs, the synthesized nano-MDDS DOX@CTAB@MSNs and CTAB@MSNs exhibit remarkablecytotoxicity and their cytotoxicities are dependent on the drug

tributions of CTAB@MSNs (c) and DOX@CTAB@MSNs (d) measured by the DLS method,

Fig. 2. Cytotoxicity of free drugs (aec) and the nano-MDDS (def) against MCF-7 cells after treated for 1 day (a, d), 2 days (b, e) and 3 days (c, f).

Q. He et al. / Biomaterials 32 (2011) 7711e7720 7715

concentration and treatment time again (Fig. 3d e and f). Thissuggests that the nano-MDDS DOX@CTAB@MSNs have significantlyimproved the intracellular drug accessibility into MCF-7/ADR cellswhich would be proved further by monitoring the intracellulardrug concentrations both qualitatively and quantitatively, as under-mentioned. Moreover, CTAB has successfully played a role of che-mosensitizer which further enhances the sensitivity of MCF-7/ADRcells to the poorly water-soluble drug DOX, similar to the case ofMCF-7 cells as mentioned above. After drug treatments for 3 days,the IC50 values of CTAB@MSNs and DOX@CTAB@MSNs are about78.5 mg mL�1 and 65.2 mg mL�1, respectively, for MCF-7/ADR cellsbased on the concentrations of the nano-MDDS. In addition, thecarrier MSNs itself did not bring visible cytotoxicity against bothMCF-7 cells and MCF-7/ADR cells in wide ranges of particleconcentration and treatment time (Figs. 2 and 3). To sum up, thepresent nano-MDDS DOX@CTAB@MSNs remarkably enhanced theMCF-7/ADR intracellular accessibility to poorly water-solubledrugs, and had much higher drug efficiencies in vitro against bothdrug-sensitive MCF-7/ADR cells and drug-resistant MCF-7 cells, ascompared with free drugs.

3.3. Death mechanisms of MDR cancer cells

Further, the death mechanisms of MCF-7/ADR cells treated withthe nano multi-drug delivery system (nano-MDDS) DOX@C-TAB@MSNs and free drugs for different time periodswere evaluatedby flow cytometry (FCM) and fluorescence-activated cell sorting(FACS) protocols (Fig. 4). The carrierMSNs showvery little influence

on the apoptosis and necrosis of MCF-7/ADR cells. However, freeCTAB at a high enough concentration could simultaneously inducedistinct apoptosis and necrosis of MCF-7/ADR cells, and causepreferentially the necrosis in one day and then the major apoptosisand subsequent secondary necrosis in two or three days (Fig. 4).Different from free CTAB, free DOX could only induce the apoptosisof MCF-7/ADR cells slightly although the concentration of free DOXwas magnified ten times as compared with DOX encapsulatedwithin DOX@CTAB@MSNs, which is dependent on treatment time.Comparedwith free CTAB, CTAB@MSNswould induce the apoptosisand subsequent secondary necrosis rather than the necrosis, whichcould be associatedwith the low intracellular concentration of CTABowing to their sustained intracellular release behaviors. Comparedwith CTAB@MSNs, DOX@CTAB@MSNs remarkably accelerated theapoptosis of MCF-7/ADR cells at the same particle concentrations,which is also dependent on treatment time. Therefore, the nano-MDDS DOX@CTAB@MSNs could induce MCF-7/ADR cell apoptosisby a synergistic apoptosis-accelerating effect between DOX andchemosensitizer CTAB, which should be attributed to the sustainedintracellular release of DOX even at a very small DOX-releasedpercentage as under-mentioned, and the synergistic chemo-sensitization effect between DOX and CTAB as a chemosensitizer (asconcluded from Fig. 3).

In addition, we further made a preliminary evaluationon theeffect of drug and carrier on the intracellular ATP (adenosinetriphosphate) level from the energy point of view because CTAB hasbeen identified as amitochondrial inhibitor for anti-cancer [20]. It iswell known that P-glycoprotein (P-gp) protein as a drug effluxpump

Fig. 4. Evaluation of the death mechanisms of MCF-7/ADR cells treated with free drugs and nano-MDDSs for different time periods (1 day, 2 days and 3 days) by flow cytometry(FCM) and fluorescence-activated cell sorting (FACS) protocols.

Fig. 3. Cytotoxicity of free drugs (aec) and the nano-MDDS (def) against MCF-7/ADR cells after treated for 1 day (a, d), 2 days (b, e) and 3 days (c, f).

Q. He et al. / Biomaterials 32 (2011) 7711e77207716

Fig. 5. Intracellular ATP levels in MCF-7/ADR cells treated with the carrier, free drugsand the nano-MDDS for different time periods.

Q. He et al. / Biomaterials 32 (2011) 7711e7720 7717

is over-expressed on themembrane ofMCF-7/ADR cells, resulting inthe MDR phenotype, and the drug efflux by P-gp is dependent oncell energy [25,26]. The intracellular ATP (adenosine triphosphate)level in MCF-7/ADR cells treated with the carrier MSNs couldrecover after incubation for 2 days (Fig. 5), though there is a transi-tory reduction in 1 day as reported by Tao [27]. However, free CTABat a high concentration could cause the sharp and unrecoverabledecrease of the intracellular ATP level. Compared with free CTAB,CTAB@MSNs could advance the slow but ever-increasing inhibitionto the intracellular ATP level, owing to the sustained CTAB release.FreeDOX can only result in a slight but time-dependent reduction inthe intracellular ATP level, although the concentration of free DOXwas magnified ten times as compared with DOX encapsulatedwithin DOX@CTAB@MSNs. However, the nano-MDDS DOX@C-TAB@MSNs could induce a more intensive reduction in the intra-cellular ATP level than free DOX and CTAB@MSNs (Fig. 5), which

Fig. 6. Cell cycle distribution histograms of MCF-7/ADR cells treated with free drugs and thecolumn graphs represent the death cells, G1 phase (DNA pre-synthetic gap phase), S phasrespectively.

indicates that the nano-MDDS DOX@CTAB@MSNs could disturb/impair the mitochondrial functions of MCF-7/ADR cells by anenhanced ATP-inhibiting effect, leading to the much reduced drugefflux capability of P-gp. Therefore it is believed that there wouldexist a synergistic effect in killing MDR cancer cells between thesurfactant and the intracellularly released drug: surfactant CTABplays an important role in inhibiting mitochondrial functions/ATPcontent [20], and consequently prevents the drug efflux by P-gp andfavors the released DOX to diffuse into the nuclei without distur-bance, or with much diminished disturbance, from the drug effluxeffect of P-gp. Thus the intracellularly released drugs could beaccumulated continuously within the MDR cells by this synergisticmechanism and consequently induced the apoptosis of MCF-7/ADRcells, as under mentioned.

3.4. Cell cycle distribution

We next assessed the effects of the nano multi-drug deliverysystem (nano-MDDS) DOX@CTAB@MSNs on cell cycle progressionand cell death by the analysis of DNA content using flow cytometry.After treatment with the carrier MSNs for 3 days, the cell cycledistribution of MCF-7/ADR cells remained constant regardless ofincubation time, as compared with the blank control, which indi-cated that the carrier MSNs itself did not affect the cell cycle (Fig. 6).It is distinct that free CTAB at a high concentration directly causedthe cell death, however CTAB@MSNs induced a time-dependentarrest of MCF-7/ADR cells in the G1 phase probably owing to thetransmembrane delivery and sustained intracellular release func-tions of CTAB@MSNs. Different from CTAB, free DOX could causea time-dependent arrest of MCF-7/ADR cells in the G2 phase untillthe concentration of DOX was magnified ten times as comparedwith DOX encapsulated within DOX@CTAB@MSNs (Fig. 6). Themost important is that the nano-MDDS DOX@CTAB@MSNs co-loaded with CTAB and DOX could significantly induce the arrestof MCF-7/ADR cells in both G1 and G2 phases and the cell death bya definite and specific synergistic effect between CTAB and DOX.This synergistic cytotoxic and cell cycle blocking effect of

nano-MDDS for 3 days by flow cytometry evaluation. Four prominent peaks in the laste (DNA synthesis phase) and G2 phase (DNA post-synthetic phase) from left to right,

Fig. 7. In vitro pH-responsive drug release behaviors of the nano-MDDS DOX@C-TAB@MSNs in the release media of different pH values which were used to simulatethe alkalescent conditions in normal tissues and blood (pH w 7.4) and the acidicconditions in tumor (pH ¼ 4e6.8). The nano-MDDS DOX@CTAB@MSNs hardly releasedDOX in the pH ¼ 7.4 release medium, but responsively released DOX in pH ¼ 4e6.8acidic media.

Q. He et al. / Biomaterials 32 (2011) 7711e77207718

DOX@CTAB@MSNs was further enhanced with the increase oftreatment time, owing to the sustained intracellular drug release asunder mentioned (Fig. 7).

3.5. In vitro pH-responsive drug release behaviors

In addition to the enhancement of drug efficiency, the inhibitionof toxic side effects of drugs, especially poorly water-soluble drugs,is also vitally important and significant to anti-cancer and MDRovercoming. One of the widely accepted routes is to endow nanodrug delivery systems (nano-DDSs) with the pH-responsive drugrelease character [28�31]. The perfect case is that drugs do not or

Fig. 8. The quantitative measurement using a fluorescence microplate reader (a) and the qrelease behaviors of the nano-MDDS DOX@CTAB@MSNs-FITC in MCF/ADR cells for representSupporting Information for more detailed drug release processes). Red, green and yellowDOX@CTAB@MSNs-FITC, represently, and purple from the superposition of blue and red su

hardly release in normal tissues and blood (pH w 7.4), but canresponsively release in tumor tissues, or evenwithin cancer cells, toselectively kill cancer cells (pH ¼ 4e6.8). Though some nano-DDSshave been designed to release drugs under in vitro simulated acidicconditions [29�31], however, to effectively suppress drug releaseas slowly as possible in pH ¼ 7.4 normal physiological conditions isstill a great challenge, especially under in vivo conditions. Fortu-nately, we discover that the designed nano-MDDS drugs@C-TAB@MSNs has such a desired precise pH-responsive drug releasecharacter both in vitro and in vivo.

From Fig. 7, it is clear that a very small amount of DOX is releasedfrom DOX@CTAB@MSNs in a very slow fashion under pH ¼ 7.4 PBSsimulating normal physiological conditions, and only less than 2%of DOXwas released after immersion for as long as 14 days which isindeed an extraordinarily low drug-released amount in such a longrelease time period as compared with previous reports [29�31].When the pH values of release media decreased from 6.5 to 4 forsimulating cancer conditions, both the DOX release rates and theDOX-released concentrations became remarkably higher. Afterimmersion for 14 days, the DOX-released percentage can reachabout 26.6% in pH ¼ 4 release medium. Therefore, the pH respon-sivity of in vitro drug release of DOX@CTAB@MSNs is believed to behigh enough for the intracellular and in vivo evaluation. Though thelow release percentage, the intracellular drug release amounts ofDOX@CTAB@MSNswere high enough to kill MCF-7 andMCF-7/ADRcells efficiently by the specific synergistic effect between CTAB andDOX, as concluded above. This means that the nano-MDDSDOX@CTAB@MSNs could work in a relative long time of morethan 14 days and in a highly pH-responsive and sustained releaseway. Further, we think that the pH-responsive drug release mech-anism for the nano-MDDS drugs@micelles@MSNs (DOX@C-TAB@MSNs) should be highly related to the electrostatic self-assembly mechanism for the synthesis of MSNs. The drugs@mi-celles@MSNs is formed by the electrostatic attraction betweendrugs-encapsulated micelles (drugs@micelles) and MSNs [32],therefore positively charged drugs@micelles within drugs@mi-celles@MSNs are exchangeable with protons in acidic releasemedia, but will be almost inexchangeable in slightly basic pH ¼ 7.4release media. As a result, surfactant and drugs can be co-released

ualitative observation by the confocal microscopic imaging (bee) of intracellular drugative treatment time periods: (b) 12 h, (c) 24 h, (d) 48 h, and (d) 72 h (see Fig. S2 in therepresent the released drug DOX, the FITC-labeled carrier MSNs and the nano-MDDSggests the entry of the released DOX into the nuclei of the MCF-7/ADR cells.

Q. He et al. / Biomaterials 32 (2011) 7711e7720 7719

together under the simulated cancer conditions in a precisely andhighly pH-responsive way.

3.6. In vitro intracellular uptake and drug release behaviors

Further, we investigated the intracellular DOX release process ofDOX@CTAB@MSNs within MCF-7/ADR cells. Firstly, we quantita-tivelymeasured the intracellular DOX concentrations inMCF-7/ADRcells at several fixed time points up to 72 h using a fluorescencemicroplate reader. It can be clearly found that the intracellular DOXconcentration inMCF-7/ADR cells increases with the increase of thetreatment time on the whole, though could transitorily decrease atcertain time points such as at 6 h incubation (Fig. 8a). This suggeststhat the nano-MDDS DOX@CTAB@MSNs have successfully over-come the drug efflux of P-gp on MCF-7/ADR cells, and favored thecontinual intracellular accumulation of released DOX molecules,which could be attributed to both the sustained and pH-responsivedrug intracellular release from endocytosized nano-MDDS, asconcluded above in the simulated cancer conditions (Fig. 7), and thedurative transmembrane transport of intercellular DOX@C-TAB@MSNs from the outside, as can be concluded by the three-daycontinual increase of intracellular DOX concentration including

Fig. 9. In vivo localization and drug release behaviors of the nano-MDDS DOX@CTAB@MSNsDOX, respectively, and therefore purple from the superposition of blue and red suggests theis especially visible after mainline administration for 2 days. The white circles indicate the zefficiency of the nano-MDDS DOX@CTAB@MSNs against the MCF-7/ADR cells-induced tum

both non-released and released DOX (Fig. S1). In addition, thequalitative monitoring by the confocal microscopic imaging alsoreveals that the nano-MDDS DOX@CTAB@MSNs can be uptaken byMCF-7/ADR cells and gradually release DOX into cells (Fig. 8b andFig. S2). These quantitative and qualitative results demonstrate thereasons why the nano-MDDS DOX@CTAB@MSNs can overcome theMDR of MCF-7/ADR cells in vitro.

3.7. In vivo localization and drug release behaviors in tumors

Further, we investigated the in vivo localization and DOX releasebehavior of the nano-MDDS DOX@CTAB@MSNs in the MCF-7/ADRcells-induced tumor and several normal tissues. It can be foundthat large numbers of the nano-MDDS DOX@CTAB@MSNs hadaccumulated in tumor after mainline administration for 1 h (Fig. 9),most probably due to the EPR effect of abnormal tumors. The mostimportant is that DOXwas visibly released into the nuclei of MCF-7/ADR tumor cells after mainline administration for 2 days as shownin Fig. 9, which is very similar to the in vitro intracellular case ofMCF-7/ADR cells as mentioned above. Moreover, the density oftumor cells whose nuclei were intruded with released DOX fromDOX@CTAB@MSNs seems significantly lower than that of other

in the MCF-7/ADR cells-induced tumor. Green, blue and red represent actin, nuclei andentry of the released DOX into the nuclei in the MCF-7/ADR cells-induced tumor, whichones of both low cell density and high drug content, implying the probable in vivo drugor.

Q. He et al. / Biomaterials 32 (2011) 7711e77207720

tumor cells (as directed bywhite circles in Fig. 9), probably implyingthe efficient in vivo MDR-overcoming action of the nano-MDDSDOX@CTAB@MSNs against the MCF-7/ADR cells-induced tumor.The accurate in vivo MDR-overcoming efficiency of DOX@C-TAB@MSNs is under investigation. In addition, DOX@CTAB@MSNswere also found distributed in three normal tissues liver, spleen andlung, however, very interestingly, these DOX@CTAB@MSNs did notrelease DOX into the nuclei of the normal cells in 2 days aftermainline administration (Fig. S3). This suggests that DOX@C-TAB@MSNs could release DOX into tumor cells, but virtually not intonormal cells in 2 days. We infer that this result can be probablyattributed to two causes: one is that DOX@CTAB@MSNs could bemore endocytosed by tumor cells owing to their distinctively higheractivity (tumor cells proliferate much faster by uptaking moreexternal nutrients) than normal tissues; and the second, moreconfidently to claim, is that DOX could be released faster in themoreacidic condition of tumor cells as compared within normal cells[33,34]. Therefore, such a nano-MDDS shows a favorable in vivo pH-responsive drug release behavior, which is consistent with thein vitro simulated cases as mentioned above.

4. Conclusions

In summary, a kind of nano-MDDS drugs@micelles@MSNs hasbeen constructed by the facile co-loading of poorly water-solubleanti-cancer drug and surfactant chemosensitizer into MSNs viaa one-pot co-self-assembly strategy among the drugs, surfactantmicelles and silicon sources. The nano-MDDS DOX@CTAB@MSNsdemonstrated a highly pH-responsive controlled drug releasebehavior, in favor of precise intracellular drug release, both underin vitro simulated conditions and under in vivo conditions with thetumor-bearingmousemodel induced by injectingMCF-7/ADR cells.Meanwhile, the nano-MDDS DOX@CTAB@MSNs also exhibited highdrug efficiencies against both drug-sensitive MCF-7 cells and drug-resistant MCF-7/ADR cells by a synergistic cell cycle arrest/apoptosis-inducing effect between poorly water-soluble anti-cancer drug and surfactant chemosensitizer.

Acknowledgements

We greatly acknowledge financial supports from the NSFC(Grant Nos. 50823007, 50972154 and 51072212), the National BasicResearch Programof China (2011CB707905 and 2010CB934000), theScience and Technology Commission of Shanghai (10430712800),CASKJCX Project (KJCX2-YW-210) and the Science Foundation forYouth Scholar of State Key Laboratory of High Performance Ceramicsand Superfine Microstructures (Grant No. SKL201001).

Appendix. Supplementary information

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.biomaterials.2011.06.066.

References

[1] Jemal A, Center MM, DeSantis C, Ward EM. Global patterns of cancer incidenceand mortality rates and trends. Cancer Epidem Biomar 2010;19:1893e907.

[2] Lage H. An overview of cancer multidrug resistance: a still unsolved problem.Cell Mol Life Sci 2008;65:3145e67.

[3] Ullah MF. Cancer multidrug resistance (MDR): a major impediment to effec-tive chemotherapy. Asian Pac J Cancer Prev 2008;9:1e6.

[4] Page R, Takimoto C. Cancer management: a multidisciplinary approach:medical, surgical, and radiation oncology. In: Pazdur R, editor. Principles ofchemotherapy. 8. New York: PRR; 2004. p. 21e38.

[5] Shabbits JA, Hu YP, Mayer LD. Tumor chemosensitization strategies based onapoptosis manipulations. Mol Cancer Ther 2003;2:805e13.

[6] van Vlerken LE, Duan ZF, Seiden MV, Amiji MM. Modulation of intracellularceramide using polymeric nanoparticles to overcome multidrug resistance incancer. Cancer Res 2007;67:4843e50.

[7] Barraud L, Merle P, Soma E, Lefrançois L, Guerret S, Chevallier M, et al. Increaseof doxorubicin sensitivity by doxorubicin-loading into nanoparticles forhepatocellular carcinoma cells in vitro and in vivo. J Hepatol 2005;42:736e43.

[8] Hu CMJ, Zhang LF. Therapeutic nanoparticles to combat cancer drug resis-tance. Curr Drug Metab 2009;10:836e41.

[9] Heurtault B, Saulnier P, Pech B, Proust JE, Benoit JP. Physico-chemical stabilityof colloidal lipid particles. Biomaterials 2003;24:4283e300.

[10] Yang MH, Li H, Javadi A, Gong SQ. Multifunctional mesoporous silica nano-particles as labels for the preparation of ultrasensitive electrochemicalimmunosensors. Biomaterials 2010;31:3281e6.

[11] Gan Q, Lu XY, Yuan Y, Qian JC, Zhou HJ, Lu X, et al. A magnetic, reversible pH-responsive nanogated ensemble based on Fe3O4 nanoparticles-capped mes-oporous silica. Biomaterials 2011;32:1932e42.

[12] He QJ, Zhang JM, Chen F, Guo LM, Zhu ZY, Shi JL. An anti-ROS/hepatic fibrosisdrug delivery system based on salvianolic acid B loaded mesoporous silicananoparticles. Biomaterials 2010;31:7785e96.

[13] Zhu M, Wang HX, Liu JY, He HL, Hua XG, He QJ, et al. A mesoporous silicananoparticulate/b-TCP/BG composite drug delivery system for osteoarticulartuberculosis therapy. Biomaterials 2011;32:1986e95.

[14] Horcajada P, Chalati T, Serre C, Gillet B, Sebrie C, Baati T, et al. Porous metal-organic-framework nanoscale carriers as a potential platform for drugdelivery and imaging. Nat Mater 2010;9:172e8.

[15] Corma A, Díaz U, Arrica M, Fernández E, Ortega Í. OrganiceInorganic nano-spheres with responsive molecular gates for drug storage and release. AngewChem Int Edit 2009;48:6247e50.

[16] Liu JW, Stace-Naughton A, Jiang XM, Brinker CJ. Porous nanoparticle sup-ported lipid bilayers (protocells) as delivery vehicles. J Am Chem Soc 2009;131:1354e5.

[17] Fahr A, Liu X. Drug delivery strategies for poorly water-soluble drugs. ExpertOpin Drug Del 2007;4:403e16.

[18] Riehm H, Biedler JL. Potentiation of drug effect by Tween 80 in Chinese-Hamster cells resistant to actinomycin-D and daunomycin. Cancer Res1972;32:1195e200.

[19] Loe DW, Sharom FJ. Interaction of multidrug-resistant Chinese-Hamster ovarycells with amphiphiles. Brit J Cancer 1993;68:342e51.

[20] Ito E, Yip KW, Katz D, Fonseca SB, Hedley DW, Chow S, et al. Potential use ofcetrimonium bromide as an apoptosis-promoting anticancer agent for headand neck cancer. Mol Pharmacol 2009;76:969e83.

[21] He QJ, Shi JL, Chen F, Zhu M, Zhang LX. An anticancer drug delivery systembased on surfactant-templated mesoporous silica nanoparticles. Biomaterials2010;31:3335e46.

[22] Zhang W, Shi Y, Chen YZ, Ye J, Sha XY, Fang XL. Multifunctional pluronic P123/F127 mixed polymeric micelles loaded with paclitaxel for the treatment ofmultidrug resistant tumors. Biomaterials 2011;32:2894e906.

[23] He QJ, Cui XZ, Cui FM, Guo LM, Shi JL. Size-controlled synthesis of mono-dispersed mesoporous silica nano-spheres under a neutral condition. Micro-por Mesopor Mater 2009;117:609e16.

[24] He QJ, Shi JL, Cui XZ, Zhao JJ, Chen Y, Zhou J. Rhodamine B-co-condensedsubsphaeroidal SBA-15 nanoparticles: facile co-condensation synthesis andexcellent fluorescence features. J Mater Chem 2009;19:3395e403.

[25] Germann UA. P-glycoprotein - a mediator of multidrug resistance in tumourcells. Eur J Cancer 1996;32A:927e44.

[26] Mealey KL, Barhoumi R, Burghardt RC, Safe S, Kochevar DT. Doxycyclineinduces expression of P glycoprotein in MCF-7 breast carcinoma cells. Anti-microb Agents Chemother 2002;46:755e61.

[27] Tao Z, Morrow MP, Asefa T, Sharma KK, Duncan C, Anan A, et al. Mesoporoussilica nanoparticles inhibit cellular respiration. Nano Lett 2008;8:1517e26.

[28] Lee CH, Cheng SH, Huang IP, Souris JS, Yang CS, Mou CY, et al. Intracellular pH-responsive mesoporous silica nanoparticles for the controlled release ofanticancer chemotherapeutics. Angew Chem Int Edit 2010;49:8214e9.

[29] Stuart MAC, Stuart C, Huck WTS, Genzer J, Müller M, Ober C, et al. Emergingapplications of stimuli-responsive polymer materials. Nat Mater 2010;9:101e13.

[30] Meng HA, Xue M, Xia T, Zhao YL, Tamanoi F, Stoddart JF, et al. Autonomousin vitro anticancer drug release from mesoporous silica nanoparticles by pH-sensitive nanovalves. J Am Chem Soc 2010;132:12690e7.

[31] Zhu YF, Shi JL, Shen WH, Dong XP, Feng JW, Ruan ML, et al. Stimuli-responsivecontrolled drug release from a hollow mesoporous silica sphere/poly-electrolyte multilayer core-shell structure. Angew Chem Int Edit 2005;44:5083e7.

[32] Huo QS, Margolese DI, Ciesla U, Feng PY, Gier TE, Sieger P, et al. Generalizedsynthesis of periodic surfactant inorganic composite-materials. Nature 1994;368:317e21.

[33] Griffiths JR. Are cancer-cells acidic. Br J Cancer 1991;64:425e7.[34] Wike-Hooley JL, Haveman J, Reinhold HS. The relevance of tumour pH to the

treatment of malignant disease. Radiother Oncol 1984;2:343e66.