Improving topical non-melanoma skin cancer treatment: In vitro efficacy of a novel guanosine-analog phosphonate

Actinic keratosis, a frequent carcinoma in situ of non-melanoma skin cancer (NMSC), can transform into life-threatening cutaneous squamous cell carcinoma. Current treatment is limited due to low complete clearance rates and asks for novel therapeutic concepts; the novel purine nucleotide analogue OxBu may be an option. In order to enhance skin penetration, solid lipid nanoparticles (SLN, 136-156 nm) were produced with an OxBu entrapment efficiency of 96.5 ± 0.1%. For improved preclinical evaluation, we combined tissue engineering with clinically used keratin-18 quantification. Three doses of 10(-3) mol/l OxBu, dissolved in phosphate-buffered saline as well as loaded to SLN, were effective on reconstructed NMSC. Tumour response and apoptosis induction were evaluated by an increase in caspase-cleaved fragment of keratin-18, caspase-7 activation as well as by reduced expression of matrix metallopeptidase-2 and Ki-67. OxBu efficacy was superior to equimolar 5-fluorouracil solution, and thus the drug should be subjected to the next step in preclinical evaluation.

Interaction of drug-carrier systems with targets--a study using atomic force microscopy

To learn about the interaction between drug agents and nanoparticular carrier systems, the physical analytical methods of parelectric, electron spin and fluorescence spectroscopy have proven helpful tools to yield descriptive models of such complex systems. For a deeper understanding of drug absorption from body surfaces and drug distribution into the tissues, however, the lack of knowledge about the interaction between such agents and membranes on different levels is a severe drawback. This gap can be closed by the application of atomic force microscopy at normal temperatures and under the admission of liquid surroundings. Moreover, this method allows the inspection of such system-membrane interactions in dependence on time. We studied membrane topography in liquid and gel-phase mixtures, structural changes of membranes during their destruction by aqueous peptide solutions as well as the stability of the membranes exposed to surfactants of increasing concentration and to lipid nanoparticles (solid lipid nanoparticles, nanostructured lipid carriers). For future modelling we can describe the geometry of lipid nanoparticles as well.

Cyclosporine-loaded solid lipid nanoparticles (SLN®): Drug–lipid physicochemical interactions and characterization of drug incorporation

Solid lipid nanoparticles (SLN) were produced loaded with cyclosporine A in order to develop an improved oral formulation. In this study, the particles were characterized with regard to the structure of the lipid particle matrix, being a determining factor for mode of drug incorporation and drug release. Differential scanning calorimetry (DSC) and wide-angle X-ray scattering (WAXS) measurements were employed for the analysis of the polymorphic modifications and mode of drug incorporation. Particles were produced using Imwitor 900 as lipid matrix (the suspension consisted of 10% particles, 8% Imwitor 900, 2% cyclosporine A), 2.5% Tagat S, 0.5% sodium cholate and 87% water. DSC and WAXS were used to analyse bulk lipid, bulk drug, drug incorporated in the bulk and unloaded and drug-loaded SLN dispersions. The processing of the bulk lipid into nanoparticles was accompanied by a polymorphic transformation from the beta to the alpha-modification. After production, the drug-free SLN dispersions converted back to beta-modification, while the drug-loaded SLN stayed primarily in alpha-modification. After incorporation of cyclosporine A into SLN, the peptide lost its crystalline character. Based on WAXS data, it could be concluded that cyclosporine is molecularly dispersed in between the fatty acid chains of the liquid-crystalline alpha-modification fraction of the loaded SLN.

In vitro skin absorption and drug release – A comparison of six commercial prednicarbate preparations for topical use

Reconstructed human epidermis is a useful tool for in vitro skin absorption studies of chemical compounds. If this may hold true also for topical dermatics, we investigated the glucocorticoid prednicarbate applied by two sets (innovator and generic) of cream, ointment and fatty ointment using the commercially available EpiDerm model. Moreover, stability and local tolerability of the preparations as well as drug release were studied, to estimate an influence on prednicarbate absorption and metabolism. While release ranked in the order cream<fatty ointment<ointment for both sets of preparations, prednicarbate penetration and permeation of the EpiDerm model did not. Less PC uptake observed with the generic ointment and fatty ointment appeared to be linked to impaired enzymatic ester cleavage within the tissue. Thus with drugs subject to skin metabolism, cutaneous uptake is not to be derived from drug release studies, yet has to be studied experimentally with viable skin or reconstructed human epidermis.

Polymorphic behaviour of Compritol888 ATO as bulk lipid and as SLN and NLC

CompritolR888 ATO (glycerol behenate) is widely used as a pharmaceutical excipient in the field of solid dosage forms due to its lubricating properties. It is an amphiphilic material with a high melting point (approximately 70 degrees C) and, therefore, it can also be used to prepare aqueous colloidal dispersions. The aim of this paper is to study the suitability of CompritolR888 ATO for the production of solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC) for the entrapment of a lipophilic model drug. This study assesses the crystalline structure of the bulk lipid, as well as the changes that occur in its crystal lattice with the addition of 'impurities', such as oil (alpha-tocopherol) and drug (ketoconazole), using DSC and X-ray diffraction analysis before and after thermal stress. Aqueous SLN and NLC dispersions were produced using an appropriate surfactant/co-surfactant system and their physicochemical stability was assessed by PCS, LD, DSC and by WAXS. It was found that the crystalline lattice of CompritolR888 ATO is composed of very small amounts of the unstable alpha polymorphic form characteristic of triacylglycerols, which disappears after thermal stress of bulk lipid. Mixing oils and drug molecules which are soluble in this lipid decreased its lattice organization and, thus, was revealed to be suitable for production of lipid nanoparticles containing ketoconazole. However, particle growth could not be avoided during shelf life.

Interaction of drug molecules with carrier systems as studied by parelectric spectroscopy and electron spin resonance

According to recent investigations of nanoparticular carrier systems the mode of drug-particle interaction appears to influence drug penetration into the skin. For a more detailed insight into the molecular structure of drug loaded particles the two independent analytical methods, namely the parelectric spectroscopy (PS) and the electron spin resonance (ESR) have been applied to 4,5,5,-trimethyl-1-yloxy-3-imidazoline-2-spiro-3'-(5'()-cholestane) as a model drug. Spectra have been analyzed in dependence on the concentration of the spin label. Changes in the concentration-dependent dipole mobility and dipole density given by PS and the concentration-dependent rotational correlation time (ESR) which are a measure of the vicinity of carrier and/or the surfactant and guest molecule were studied with cholestane-labeled solid lipid nanoparticles (SLN), nanoparticular lipid carriers (NLC) and nanoemulsions (NE). The spin probes were attached to the SLN surface which consists of two distinct sub-compartments: the rim and the flat surface of the disk-like shapes. The shape could be observed by freeze-fraction electron microscopy. Spin probes, however, were incorporated into the carrier matrix in the cases of NLC and NE. Results of PS are verified by ESR which allows a more detailed insight. Taking the results together a detailed new model of 'drug'-particle interaction could be established.

Oral bioavailability of cyclosporine: Solid lipid nanoparticles (SLN®) versus drug nanocrystals

For the development of an optimized oral formulation for cyclosporine A, 2% of this drug has been formulated in solid lipid nanoparticles (SLN, mean size 157 nm) and as nanocrystals (mean size 962 nm). The encapsulation rate of SLN was found to be 96.1%. Nanocrystals are composed of 100% of drug. For the assessment of the pharmacokinetic parameters the developed formulations have been administered via oral route to three young pigs. Comparison studies with a commercial Sandimmun Neoral/Optoral used as reference have been performed. The blood profiles observed after oral administration of the commercial microemulsion Sandimmun revealed a fast absorption of drug leading to the observation of a plasma peak above 1,000 ng/ml within the first 2 h. For drug nanocrystals most of the blood concentrations were in the range between 30 and 70 ng/ml over a period of 14 h. These values were very low, showing huge differences between the measuring time points and between the tested animals. On the contrary, administration of cyclosporine-loaded SLN led to a mean plasma profile with almost similarly low variations in comparison to the reference microemulsion, however with no initial blood peak as observed with the Sandimmun Neoral/Optoral. Comparing the area under the curves (AUC) obtained with the tested animals it could be stated that the SLN formulation avoids side effects by lacking blood concentrations higher than 1,000 ng/ml. In this study it has been proved that using SLN as a drug carrier for oral administration of cyclosporine A a low variation in bioavailability of the drug and simultaneously avoiding the plasma peak typical of the first Sandimmun formulation can be achieved.

Lipid nanoparticles for skin penetration enhancement—correlation to drug localization within the particle matrix as determined by fluorescence and parelectric spectroscopy

With topical treatment of skin diseases, the requirement of a high and reproducible drug uptake often still is not met. Moreover, drug targeting to specific skin strata may improve the use of agents which are prone to cause local unwanted effects. Recent investigations have indicated that improved uptake and skin targeting may become feasible by means of nanoparticular systems such as solid lipid nanoparticles (SLN), nanostructured lipid carriers (NLC) and nanoemulsions (NE). Here we describe techniques to characterize drug loading to carrier systems and skin penetration profiles by using the lipophilic dye nile red as a model agent. Since the mode of drug association with the particle matrix may strongly influence the efficiency of skin targeting, parelectric spectroscopy (PS) was used to differentiate between matrix incorporation and attachment to the particle surface and fluorescence spectroscopy (FS) to solve dye distribution within NLC particles. Nile red was incorporated into the lipid matrix or the covering tensed shell, respectively, of SLN and NLC with all the lipids studied (Compritol, Precirol, oleic acid, Miglyol). In NLC, the dye was enriched in the liquid phase. Next, nile red concentrations were followed by image analysis of vertical sections of pigskin treated with dye-loaded nanoparticular dispersions and an oil-in-water cream for 4 and 8 h in vitro. Following the SLN dispersions, dye penetration increased about fourfold over the uptake obtained following the cream. NLC turned out less potent (<threefold increase) and penetration appeared even reduced when applying a NE. In contrast to previous studies with glucocorticoids attached to the surface of SLN, a targeting effect was not detected here. Therefore, drug targeting appears to be more strictly related to the mode of interaction of drug and particle than penetration enhancement.

Glucocorticoid entrapment into lipid carriers--characterisation by parelectric spectroscopy and influence on dermal uptake

Topical glucocorticoids such as betamethasone 17-valerate (BMV) and prednicarbate (PC) are an important therapeutic option in atopic eczema. To reduce the risk of dermal atrophy, we aimed at BMV incorporation into solid lipid nanoparticles (SLN) for epidermal targeting using various lipids and emulsifiers corresponding to previous work on PC. Cutaneous absorption into excised human skin was compared to the one with a cream. While Compritol-based particles increased BMV uptake about fourfold we failed, however, to obtain epidermal targeting. To obtain insight into the location of active substance relative to the carrier, we used the recently optimised method of parelectric spectroscopy (PS). In fact, we were able to study electric dipole movements in the broad field of a frequency span from 0.1 to 100 MHz demonstrating that glucocorticoids are attached to the particle surface but are not incorporated into the lipid matrix. With BMV, the loading capacity of the particle surface lies clearly below the usual concentration of 0.1% which is not the case with PC. An adequate association of drug and carrier is essential for epidermal targeting. Parelectric spectroscopy provides insight into the interaction between drug and lipidic carrier.

RU 58841-myristate--prodrug development for topical treatment of acne and androgenetic alopecia

Acne and androgenetic alopecia are linked to androgen effects and therefore should improve following topical application of antiandrogens. We present a new antiandrogen prodrug, RU 58841-myristate (RUM) for topical therapy. Almost devoid of affinity to the androgen receptor, as derived from investigations in the mouse fibroblast cell line 29 +/GR +, RUM is rapidly metabolised to the potent antiandrogen RU 58841 by cultured human foreskin fibroblasts and keratinocytes, male occipital scalp skin dermal papilla cells, and by cells of the sebaceous gland cell line SZ95. In order to improve a specific targeting of the hair follicle, RUM was loaded on solid lipid nanoparticles (SLN), which are already known to support dermal targeting effects. Physically stable RUM loaded SLN were produced by hot homogenization. Penetration/permeation studies carried out using the Franz diffusion cell proved only negligible permeation of reconstructed epidermis and excised porcine skin within 6 h, implying a more topical action of the drug. Targeting to the hair follicle using SLN was visualised by fluorescence microscopy, following the application of Nile Red labelled SLN to human scalp skin. Transmission electron microscopy (TEM) allowed to detect intact silver labelled SLN in porcine hair follicles of preparations applied to the skin for 24 h. RUM loaded SLN should be considered for topical antiandrogen therapy of acne and androgenetic alopecia.

Drug targeting by solid lipid nanoparticles for dermal use

Long term topical glucocorticoid treatment can induce skin atrophy by the inhibition of fibroblasts. We, therefore, looked for the newly developed drug carriers that may contribute to a reduction of this risk by an epidermal targeting. Prednicarbate (PC, 0.25%) was incorporated into solid lipid nanoparticles of various compositions. Conventional PC cream of 0.25% and ointment served for reference. Local tolerability as well as drug penetration and metabolism were studied in excised human skin and reconstructed epidermis. With the latter drug recovery from the acceptor medium was about 2% of the applied amount following PC cream and ointment but 6.65% following nanoparticle dispersion. Most interestingly, PC incorporation into nanoparticles appeared to induce a localizing effect in the epidermal layer which was pronounced at 6 h and declined later. Dilution of the PC-loaded nanoparticle preparation with cream (1:9) did not reduce the targeting effect while adding drug-free nanoparticles to PC cream did not induce PC targeting. Therefore, the targeting effect is closely related to the PC-nanoparticles and not a result of either the specific lipid or PC adsorbance to the surface of the formerly drug free nanoparticles. Lipid nanoparticle-induced epidermal targeting may increase the benefit/risk ratio of topical therapy.

Solid lipid nanoparticles: production, characterization and applications

Solid lipid nanoparticles (SLN) have attracted increasing attention during recent years. This paper presents an overview about the selection of the ingredients, different ways of SLN production and SLN applications. Aspects of SLN stability and possibilities of SLN stabilization by lyophilization and spray drying are discussed. Special attention is paid to the relation between drug incorporation and the complexity of SLN dispersions, which includes the presence of alternative colloidal structures (liposomes, micelles, drug nanosuspensions, mixed micelles, liquid crystals) and the physical state of the lipid (supercooled melts, different lipid modifications). Appropriate analytical methods are needed for the characterization of SLN. The use of several analytical techniques is a necessity. Alternative structures and dynamic phenomena on the molecular level have to be considered. Aspects of SLN administration and the in vivo fate of the carrier are discussed.

[Solubility and dissolution rate of digoxin from Digitalis lanata drug extracts]

The influence on solubility and dissolution rate was investigated for digoxin as a model drug with a very low solubility in water. The investigations were carried out with different fractions of extracts from leaves of Digitalis lanata. These fractions differ in the composition of concomitant compounds. The solubility of digoxin from the extract fractions is increased up to 42 times, with considerable differences between the fractions. The solubility depends on the weight of the extract fraction; a limit of solubility exists. Even after separation of the solved extract components the solubility of digoxin in the residues is larger than that of the pure digoxin. The dissolution rate of digoxin of "Vorgereinigter Gesamtglykosidextrakt (VE)" and the glycosid fraction G 1 is influenced significantly, whereas digoxin in the glycosid fraction G 4 has such a degree of purity that the solubility properties are not influenced by the small amount of concomitant compounds. After 10 min already 50.4% of the digoxin in the extract fraction G 1 are dissolved, while only 21.7% of the pure digoxin are dissolved in that interval. The extract fractions exhibit different wettability properties, so that the increased dissolution rate could be attributed to improved wettability of the extract fractions. Physical mixtures of crystal-line digoxin and compounds of the extracts of the almost digoxin free fraction G 2 did not exert an influence on the dissolution behavior. Different batches of the extract fractions showed different solubility in spite of comparable digoxin content.

Solid lipid nanoparticles as drug carriers for topical glucocorticoids

Recent investigations both in vitro and in human subjects proved the benefit/risk ratio of prednicarbate (PC) to exceed those of halogenated topical glucocorticoids about 2-fold. To obtain a further highly desired increase by drug targeting to viable epidermis, PC was incorporated into solid lipid nanoparticles (SLN). Keratinocyte and fibroblast monolayer cultures, reconstructed epidermis and excised human skin served to evaluate SLN toxicity and PC absorption. Well-tolerated preparations (e.g. cellular viability 94.5% following 18 h incubation of reconstructed epidermis) were obtained. PC penetration into human skin increased by 30% as compared to PC cream, permeation of reconstructed epidermis increased even 3-fold. The present study shows the great potential of SLN to improve drug absorption by the skin.

Solid lipid nanoparticles (SLN) for controlled drug delivery. II. Drug incorporation and physicochemical characterization

Solid lipid nanoparticles (SLN) are a colloidal carrier system for controlled drug delivery. The lipophilic model drugs tetracaine and etomidate were incorporated to study the maximum drug loading, entrapment efficacy, effect of drug incorporation on SLN size, zeta potential (charge) and long-term physical stability. Drug loads of up to 10% could be achieved whilst simultaneously maintaining a physically stable nanoparticle dispersion. Incorporation of drugs showed no or little effect on particle size and zeta potential compared to drug-free SLN. The optimized production parameters previously established for drug-free SLN dispersions can therefore be transferred to drug-loaded systems to facilitate product development.

[The importance of concomitant compounds in plant extracts]

Plant extracts are multi-composed mixtures that can be subdivided into main active substances and concomitant compounds. Concomitant compounds are called co-effectors because they can change the physicochemical properties of the main active substances and therefore influence the biopharmaceutical parameters, e.g. solubility and bioavailability. The composition and the properties of plant extracts depend on a multitude of different factors such as quality and degree of reduction of the plant, and the process of extraction and drying. Concomitant compounds affect the physical and chemical stability of plant extracts. The hygroscopicity also depends on the composition of the concomitant compounds. It is an important property for the further processing of the plant extract to solid dosage form, e.g. tablets. Solubility and dissolution rate can be influenced by the concomitant compounds. Both properties are important for the absorption of the active substances though in many cases the responsible substances and mechanisms are still unknown. Solubilisation by surface-active agents, formation of soluble drug-concomitant compound-associates and solid dispersion, in which the drug is finely dispersed or X-ray amorphous, are often responsible for the increased solubility. Improved wettability in the presence of the concomitant compounds results in better dissolution. Supersaturated solutions of the active substances are stabilized for months by inhibition of crystallization. Numerous examples are known for influencing the absorption by natural concomitant compounds. The absorption is enhanced by the increased amount of dissolved drug or by the alteration of membrane permeability. Adsorption or complex formation with the concomitant compounds decrease the absorption rate, in some cases also the extent of absorption.

Solid lipid nanoparticles (SLN) for controlled drug delivery--drug release and release mechanism

Solid lipid nanoparticles (SLN) are particulate systems for parenteral drug administration with mean particle diameters ranging from 50 up to 1000 nm. The model drugs tetracaine, etomidate and prednisolone were incorporated (1, 5 and 10%) to study the drug load, effect of drug incorporation on the structure of the lipid matrix and the release profiles and mechanism. SLN were produced by high pressure homogenization of aqueous surfactant solutions containing the drug-loaded lipids in the melted or in the solid state (500/1500 bar, 3/10 cycles). In case of tetracaine and etomidate, high drug loadings up to 10% could be achieved when using Compritol 888 ATO and Dynasan 112 as matrix material. The melting behavior of the drug loaded particles revealed that little or no interactions between drug and lipid occurred. A burst drug release (100% release < 1 min) was observed with tetracaine and etomidate SLN, which was attributed to the large surface area of the nanoparticles and drug enrichment in the outer shell of the particles. In contrast, prednisolone loaded SLN showed a distinctly prolonged release over a monitored period of 5 weeks. Depending on the chemical nature of the lipid matrix, 83.8 and 37.1% drug were released (cholesterol and compritol, respectively). These results demonstrate the principle suitability of SLN as a prolonged release formulation for lipophilic drugs.

Cytotoxicity of solid lipid nanoparticles as a function of the lipid matrix and the surfactant

PURPOSE

Assessment of the in vitro cytotoxicity of solid lipid nanoparticles (SLNs) as a function of lipid matrix (Dynasan 114, Compritol ATO 888), and stabilizing surfactant (poloxamers, Tween 80, soya lecithin, and sodium dodecyl sulphate). Comparison with other colloidal carriers should determine their potential use in the clinic.

METHODS

SLNs were produced by high pressure homogenisation. Cytotoxicity was assessed by measuring the viability of HL60 cells and human granulocytes after incubation with SLNs. Particle internalisation was quantified by chemiluminescence measurements.

RESULTS

The nature of the lipid had no effect on viability; distinct differences were found for the surfactants. Binding to the SLN surface reduced markedly the cytotoxic effect of the surfactants, e.g., up to a factor of 65 for poloxamer 184. The permanent HL60 cell line-differentiated from cells with granulocyte characteristics by retinoic acid treatment-yielded results identical to freshly isolated human granulocytes. In general, the SLNs showed a lower cytotoxicity compared to polyalkylcyanoacrylate and polylactic/glycolic acid (PLA/ GA) nanoparticles.

CONCLUSIONS

Because the results are identical when using human granulocytes, differentiated HL60 cells can be used as an easily accessible in vitro test system for i.v. injectable SLN formulations. The SLNs appear suitable as a drug carrier system for potential intravenous use due to their very low cytotoxicity in vitro.

Atomic force microscopy studies of solid lipid nanoparticles

PURPOSE

Solid Lipid Nanoparticles (SLN) are an alternative carrier system for the controlled delivery of drugs. In most cases prednisolone loaded SLN show a biphasic release behaviour. The initial phase is characterised by a fast drug release, which is followed by a sustained drug release over several weeks.

METHODS

The particles are produced by high pressure homogenisation of a lipid (e.g. compritol, cholesterol) dispersed in an aqueous surfactant solution. In this study atomic force microscopy was used to image the original unaltered shape and surface properties of the particles. The crystallinity of the nanoparticles was investigated by differential scanning calorimetry.

RESULTS

The AFM investigations revealed the disc like shape of the particles. From differential scanning calorimetry data it can be concluded that the particle core is in the crystalline state. Additionally it was proven that the particles are surrounded by a soft layer.

CONCLUSIONS

Thus it is conceivable that the fast initial drug release during in vitro dissolution tests takes place by drug release of the outer noncrystalline layers of the particles. The following sustained drug release can be assigned to the predisolone release of the inner crystalline particle layers.

Phagocytic uptake and cytotoxicity of solid lipid nanoparticles (SLN) sterically stabilized with poloxamine 908 and poloxamer 407

Solid lipid nanoparticles (SLN) as alternative intravenous colloidal drug carriers were produced by high pressure homogenisation of melted lipids (glycerolbehenate, cetylpalmitate). Their surface was modified by using hydrophilic poloxamine 908 and poloxamer 407 blockcopolymers in order to reduce the phagocytic uptake by the reticuloendothelial system (RES) after i. v. injection. The phagocytosis reducing effect of the polymers was investigated in vitro in cultures of human granulocytes, uptake was quantified by chemiluminescence. Modification of the SLN with poloxamine 908 and poloxamer 407 reduced the phagocytic uptake to appr. 8-15% compared to the phagocytosis of hydrophobic polystyrene particles. The modified SLN proved more efficient in avoiding phagocytic uptake than polystyrene particles surface-modified with these blockcopolymers (48% and 38%, respectively). Viability determinations revealed the SLN to be 10 fold less cytotoxic than polylactide nanoparticles and 100 fold less than butylcyanoacrylate particles.

Penetration of hydrocortisone into excised human skin under the influence of cyclodextrins

Hydrocortisone (HC) penetration into excised human skin from 1% (w/w) HC containing formulations with white petrolatum, an o/w cream and an aqueous polyacrylate gel is described. The experiments were performed with pure HC and the inclusion compounds with beta-cyclodextrin and hydroxypropyl-beta-cyclodextrin. The smallest HC amounts in the skin were found with the petrolatum preparations. No differences between the incorporation of free HC and the inclusion compounds was found. The highest HC amounts in the skin are found with the o/w cream and hydrogel formulation containing nonincluded HC. Incorporation of both inclusion compounds into these vehicles diminishes the HC concentration in the upper skin layers but not in the dermis. It is assumed that the preferred penetration route for the easily soluble inclusion compounds is a transappendageal diffusion rather than a transdermal one. A comparison with release results by using an ointment liberation model gives a good correlation with the penetration results in the dermis but not in the other skin layers.

Leukemia following chemotherapy for ovarian cancer

An international collaborative group of cancer registries and hospitals identified 114 cases of leukemia following ovarian cancer. We investigated the possible etiologic role of chemotherapy, radiotherapy, and other factors, using a case-control study design, with three controls matched to each case of leukemia. Chemotherapy alone was associated with a relative risk of 12 (95 percent confidence interval, 4.4 to 32), as compared with surgery alone, and patients treated with both chemotherapy and radiotherapy had a relative risk of 10 (95 percent confidence interval, 3.4 to 28). Radiotherapy alone did not produce a significant increase in risk as compared with surgery alone. The risk of leukemia was greatest four or five years after chemotherapy began, and the risk was elevated for at least eight years after the cessation of chemotherapy. The drugs cyclophosphamide, chlorambucil, melphalan, thiotepa, and treosulfan were independently associated with significantly increased risks of leukemia, as was the combination of doxorubicin hydrochloride and cisplatin. Chlorambucil and melphalan were the most leukemogenic drugs, followed by thiotepa; cyclophosphamide and treosulfan were the weakest leukemogens, and the effect per gram was substantially lower at high doses than at lower doses. The extent to which the relative risks of leukemia are offset by differences in chemotherapeutic effectiveness is not known.