Cytotoxic drug contamination in healthcare settings is a growing concern throughout Europe and other countries, including the United States. Many studies demonstrate that parenteral drug contamination may affect the health of pharmacy, medical, and nursing staff being exposed to cytotoxic drugs.1-7 Exposure to intravenous (IV) cytotoxic agents may occur during administration, reconstitution, and dispensation of IV drugs, or from contamination on the outside of drug vials.8
We know from earlier reports that air and wipe samples have shown measurable levels of exposure to cyclophosphamide among pharmacists and nurses.1 A recent study in 6 Canadian hospitals showed that levels of cyclophosphamide were over the limit of detection in wipe samples collected from frequently contacted surfaces at every stage of the medication system.9 However, a large proportion of contaminated samples in that study came from healthcare workers who are not responsible for drug reconstitution,9 such as oncologists, dietitians, and other support staff. In a French study, doxorubicin and cyclophosphamide were found on approximately 30% of the gloves of nurses who primed IV sets before patient administration.2 Some cytotoxic agents can penetrate through barriers and into the body; studies on renal excretion of antineoplastics have shown cyclophosphamide in urine samples of hospital staff handling infusions.10
The carcinogenic, mutagenic, and reprotoxic properties of cytotoxic agents have been demonstrated in several studies since the 1970s. Among nurses handling these drugs, a substantial association was observed between spontaneous abortions or stillbirths and occupational exposure to the agents in French, Finnish, and American studies.3-5 Healthcare staff exposed to alkylating agents were found to have a significantly greater risk for chromosomal aberration than the nonexposed population.6
A prospective, follow-up study of a large-scale nursing cohort in Hungary supports earlier findings7,11 and emphasizes the increasing frequency of immunodeficiency among nurses. A high-risk lifestyle (eg, smoking and other negative health behavior) increased the mutation rates associated with antineoplastics. Low-dose occupational cytostatic drug exposure was observed to be immunotoxic; occupational exposure to antineoplastic drugs increased B-cells but decreased the frequency of activated T-lymphocytes.11
Such immunologic shifts are considered to impair the immune system of exposed individuals.11 Anemia, thyroid dysfunctions, and tumors (benign and malignant) were found to be more common among nurses exposed to cytotoxic drugs than among healthy controls.7
Therefore, reducing cytotoxic contamination to minimize the risk for harmful effects is a major challenge for hospital pharmacists; of note, the contamination in healthcare settings may also transfer to the patients’ homes.12 Studies show that implementation of proper safety measures can decrease the negative impact of cytotoxic drugs on health, and that the clinical use of a closed-system drug-transfer device to handle hazardous drugs is likely one of the most important safety measures available to combat it.7,11
The US National Institute for Occupational Safety and Health defines a closed-system drug-transfer device as “a drug transfer device that mechanically prohibits the transfer of environmental contaminants into the system and the escape of hazardous drug or vapor concentrations outside the system.”13 Tevadaptor, the closed-system drug-transfer device that was used in this study, was approved by the US Food and Drug Administration in 2007,14 was reclassified as a closed system under the ONB code in 2015,15 and was cleared for marketing to prevent vapor, aerosol, or liquid drugs from being released during drug preparation and administration.15
Centralized infusion preparation units of hospital pharmacies participated in the present study. Closed-system drug-transfer devices were launched shortly before this study as a tool to minimize IV drug contamination from the work environment, and had been implemented mainly in hospital pharmacies. Three Hungarian centers had been using the Tevadaptor device before this study and, therefore, were ideal candidates for participation.
To maximize the likelihood of obtaining valid and reliable data, centralized preparation units servicing entire hospitals with standardized, good manufacturing practice–based production were primarily considered for the study, as opposed to decentralized inpatient and outpatient units, where antineoplastics are also handled by local healthcare and nursing staff. Drug dispensation is the most complicated part of IV drug handling, posing the highest occupational risks (ie, reconstitution, withdrawal, and aerosolizing); therefore, this dispensing area of the pharmacy environment was considered an appropriate challenge for successful use of the safety device.
The primary purpose of our study was to establish baseline measurements of cytotoxic drug contamination and assess the values in 13 hospital pharmacies in Hungary. To our knowledge, such data were not available in Hungary before this study. The other objective was to determine the efficacy of a closed-system drug-transfer device at reducing IV cytotoxic drug contamination.
In this study, we used a pretest–posttest design, which was conducted in 13 hospital pharmacies of regional oncology centers. Participating sites were selected from 30 centers across Hungary to enhance the representativeness of the sample. Each center followed the wipe-sampling protocol developed and used in the European Society for Oncology Pharmacy (ESOP) study.16
Wipe samples were obtained from 5 agreed-on surface areas within the pharmacy, using the Berner sampling kit.17 These 5 areas included the inside of the safety cabinet, the floor in front of the safety cabinet, gloves of personnel working under the hood, the door handle or knob inside the clean room, and the transportation box used to deliver ready-to-administer infusions to the oncology unit. In one of the hospitals, a sample was also obtained from the surface of the elastomeric pumps used for home therapy. Because the pump samples were obtained in only 1 hospital, results for those samples were not considered for the comparative analysis.
Where the size of the surface allowed, samples were obtained using a standard 35-cm by 35-cm area (Figure). If the chosen sampling area was smaller than recommended by the sampling kit (ie, ≤2500 cm2 overall), the actual size was recorded and forwarded to the analytical site so that all results could be standardized.
A 3-step wiping protocol was followed for all samples. This protocol required wiping in 3 different directions to best capture any residue on surfaces (Figure). To ensure use of the correct wiping technique, pharmacy staff were trained beforehand. Each sampling process was assisted and supervised by experienced study coordinators.
Two sets of samples were obtained, both at the end of the day, before decontamination. The first sample was taken at study onset (ie, baseline). The second sample (ie, postintervention) was taken after the closed-system drug-transfer device was used for 2 weeks or 1 month (ie, the study end point). The specific end point of 2 weeks or 1 month was selected by individual pharmacists, because different pharmacies had various resources available to purchase the device. The duration of the closed-system drug-transfer device application was based on the rate of selected drugs used daily at the particular site. To be able to provide contamination data for a similar volume of each cytotoxic drug at each site, participating personnel were asked not to change their actual protocols for preparing the cytotoxic drugs and decontamination before or during the study.
Samples were frozen and delivered to the Institute of Energy and Environmental Technology (IUTA) in Duisburg, Germany. IUTA was chosen as the analytical site because the ESOP study monitoring protocol was implemented in Europe,16 which proved that IUTA had developed a valid analytical method to detect surface contamination from wipe samples.
The potential presence of the following 8 compounds—5-fluorouracil (5-FU), cyclophosphamide, docetaxel, etoposide, gemcitabine, ifosfamide, methotrexate, and paclitaxel—was investigated using the wipe samples. Although the limit of detection was set by the IUTA at 0.1 ng/cm2, values below this level were estimated and were only provided as information. Essentially, results below the limit of detection indicated that traces of the drug were undetectable on the surface.
Each pharmacy used the device for 5-FU, gemcitabine, cyclophosphamide, and paclitaxel for 2 weeks or for 1 month after baseline samples had been obtained. In addition, each pharmacy selected an additional 2 to 4 compounds from the list of 8, for which the drug-transfer device also was used. Nonselected substances were reconstituted without using this device.
Participating hospital pharmacies used 4 main components of the drug-transfer system to ensure effective closure, including a vial adaptor for safe reconstitution from a vial; a syringe adaptor to withdraw and dispense drugs from vials; a spike adaptor to transfer the drug from the syringe through an access port into the infusion bag; and a luer lock adaptor attached to the elastomeric pump to help ensure a closed system for filling pumps with 5-FU.
All hospital pharmacies used safety cabinets to avoid aerosol contamination. Each study center used a special cleaning process, which included 0.1 mol/L NaOH and alcohol-based sequential purification, according to ESOP and Hungarian national guidelines. During the study period, no practices or the use of any safety measure or protective equipment were changed. The only experimental intervention was the implementation of the closed-system drug-transfer device.
Statistical tests were conducted using SPSS Statistics Software (Version 20.0; IBM, Armonk, NY). To assess for normal distribution of the study sample, the 1-sample Kolmogorov-Smirnov normality test was used. Differences between baseline and postintervention measures were determined by the Wilcoxon test for paired t-tests. The level of significance was set at 5%; 1-tailed tests were used for computation.
Table 1 shows the average concentration levels of the different drugs used in the 5 pretest (S1) and 5 posttest (S2) samples for each study center. Three centers (ie, centers 7, 9, and 12) had been using the closed-system drug-transfer device before the study. The prestudy use period ranged from 3 to 8 months. Center 9 did not use docetaxel or paclitaxel, because of the specialization of the hospital.
Table 2 provides comparative descriptive statistics studies (ie, baseline and postintervention contamination values) for the 10 pharmacies where the device had not been used before the study, and for the 3 pharmacies that had begun using it before the study.
Table 3 displays comparative statistics for the contamination levels of the 8 cytotoxic drugs detected by wipe samples, including differences in mean values between baseline and postintervention. The data show reductions in contamination postintervention as a result of the use of the device.
Table 4 presents 5-FU data for samples obtained from the pharmacy of the National Institute of Oncology in Budapest, Hungary, where the device was used to fill elastomeric pumps after baseline samples had been taken. Testing these samples was considered important to determine whether contamination was apparent on the surface of these pumps, and to assess the risk for transmission to patients’ homes.
Overall, the findings showed that the level of contamination was higher at baseline, and was reduced significantly for some substances with the use of the closed-system drug-transfer device, but not below the lower limit of detection after 2 weeks of monitoring.
Our findings demonstrated higher levels of cytotoxic contamination at baseline, which were reduced postintervention. As shown in Table 2, pharmacies that did not use the drug-transfer device before the study had significantly higher levels of contamination at baseline than pharmacies that had used the device beforehand.
Some peak contamination values were high. For example, baseline values of 5-FU (2167 ng/cm2), gemcitabine (525 ng/cm2), and ifosfamide (2083 ng/cm2) were 21,000, 5000, and 20,000 times above the detection limit established for our study (0.1 ng/cm2), respectively. Although these values are unusually high, such results are not unique. Rioufol and colleagues detected large levels of contamination on nurses’ gloves, with the maximum amount of 32,600 ng of cyclophosphamide found on a pair of gloves.2
Our data confirm the prevalence of environmental contamination in the 13 pharmacies studied, and emphasize the importance of observing appropriate safety measures and protocols for handling cytotoxic drugs. As noted before, pharmacy staff and nurses can be a source of contamination in healthcare settings by being in the first line of exposure. Pharmacists and pharmacy assistants need to be aware of this potential risk.
Our study shows that despite the use of best practices, contamination may be found at varying degrees on different surfaces within the pharmacy environment. If pharmacy staff are not sufficiently cautious when working with elastomeric pumps, they may leave cytotoxic residue on the surface of the pumps and thus risk contaminating their patients and their homes.
However, as shown in this study, applying a closed-system drug-transfer device was successful in reducing surface contamination of pumps to levels almost below the limit of detection set for this study. These results support the conclusion of Hon and colleagues that contamination at the drug-preparation stage continues through subsequent stages of medication delivery.9
The device used in this study proved successful in reducing contamination for 6 of the 8 substances investigated. Although results for those substances were statistically significant, the contamination level was below the lower limit of detection for some substances. This may, in part, be the result of the short period of use of the device. Although use of a closed-system drug-transfer device for ≥2 weeks may help reduce contamination, this period may not be long enough to determine the full efficacy of the device. A longitudinal study to observe how these devices fare long-term is recommended.
That contamination levels for all substances were not reduced below the limit of detection may be explained by the short time of the device use and the wide variety of data because of the substantial physical and chemical differences of the studied compounds, as well as the significant differences in daily infusion production (range, 50-500 preparations daily) among the centers.
This study has several limitations. Some of the conditions of our study were not standardized among the participating institutions. In addition, substantial differences were found in infusion production capacity, dosing of selected compounds, and human resources among the study centers. However, this study was not designed to be a standardized, controlled, laboratory experiment, but rather an exploration of actual pharmacy practice.
Furthermore, the results varied among hospitals, because of the different pharmacy methods used in the participating institutions. Nevertheless, contamination was evident in all participating pharmacies, and the device used in this study demonstrated a favorable trend in all participating institutions. However, the short period of using the device in this study—2 weeks or 1 month—was not sufficient to determine its long-term effects.
In addition, we did not collect data from oncology units to establish their levels of contamination.
Finally, our study focused on the hospital pharmacy environment and did not address the safety concerns of pharmacy staff. However, medical and nursing staff who come into contact with hazardous drugs must also be protected during administration.
Our study demonstrated high levels of surface contamination in various pharmacy settings in Hungary. The closed-system drug-transfer device used in our study was efficacious in reducing contamination risk for select IV cytotoxic drugs, and minimized the risk for transferring contamination outside of the pharmacy when used to fill elastomeric pumps. Our findings suggest that using such a device is a safe and effective way to minimize exposure to cytotoxic drugs for pharmacy and other healthcare staff. We recommend that these devices be used regularly in pharmacy settings to avoid harm that may be caused by working with cytotoxic drugs.
It may also be worthwhile to replicate this study to extend it to hospital administration areas and oncology care units, under more standardized conditions to assess contamination in those areas.
This study was partially sponsored by Teva Hungary. The study was managed by Lucidpharma Kft, Hungary.
Author Disclosure Statement
Dr Juhász, Dr Szücs, and Dr Süle received partial support for this study from Teva.
- McDevitt JJ, Lees PS, McDiarmid MA. Exposure of hospital pharmacists and nurses to antineoplastic agents. J Occup Med. 1993;35:57-60.
- Rioufol C, Ranchon F, Schwiertz V, et al. Administration of anticancer drugs: exposure in hospital nurses. Clin Ther. 2014;36:401-407.
- Valanis B, Vollmer WM, Steele P. Occupational exposure to antineoplastic agents: self-reported miscarriages and stillbirths among nurses and pharmacists. J Occup Environ Med. 1999;41:632-638.
- Selevan SG, Lindbohm ML, Hornung RW, Hemminki K. A study of occupational exposure to antineoplastic drugs and fetal loss in nurses. N Engl J Med. 1985;313:1173-1178.
- Stücker I, Caillard JF, Collin R, et al. Risk of spontaneous abortion among nurses handling antineoplastic drugs. Scand J Work Environ Health. 1990;16:102-107.
- McDiarmid MA, Oliver MS, Roth TS, et al. Chromosome 5 and 7 abnormalities in oncology personnel handling anticancer drugs. J Occup Environ Med. 2010;52:1028-1034.
- Tompa A, Jakab M, Biró A, et al. Chemical safety and health conditions among Hungarian hospital nurses. Ann N Y Acad Sci. 2006;1076:635-648.
- Mason HJ, Morton J, Garfitt SJ, et al. Cytotoxic drug contamination on the outside of vials delivered to a hospital pharmacy. Ann Occup Hyg. 2003;47:681-685.
- Hon CY, Teschke K, Chu W, et al. Antineoplastic drug contamination of surfaces throughout the hospital medication system in Canadian hospitals. J Occup Environ Hyg. 2013;10:374-383.
- Evelo CT, Bos RP, Peters JG, Henderson PT. Urinary cyclophosphamide assay as a method for biological monitoring of occupational exposure to cyclophosphamide. Int Arch Occup Environ Health. 1986;58:151-155.
- Biró A, Fodor Z, Major J, Tompa A. Immunotoxicity monitoring of hospital staff occupationally exposed to cytostatic drugs. Pathol Oncol Res. 2011;17:301-308.
- Yuki M, Sekine S, Takase K, et al. Exposure of family members to antineoplastic drugs via excreta of treated cancer patients. J Oncol Pharm Pract. 2013;19:208-217.
- Nygren O, Olofsson E, Johannson L. NIOSH definition of closed-system drug-transfer devices. Ann Occup Hyg. 2009;53:549.
- US Food and Drug Administration. Tevadaptor closed drug reconstitution and transfer system [letter]. September 24, 2007. www.accessdata.fda.gov/cdrh_docs/pdf7/k071741.pdf. Accessed April 26, 2018.
- US Food and Drug Administration. Tevadaptor closed drug reconstitution and transfer system [letter]. January 23, 2015. www.accessdata.fda.gov/cdrh_docs/pdf14/K141448.pdf. Accessed April 19, 2018.
- European Society of Oncology Pharmacy Project. Monitoring of surfaces contamination with antineoplastic drugs in preparation and administration areas. www.esop.li/downloads/ESOP_project.pdf. Accessed August 10, 2014.
- Berner. Cytotoxic wipe sampling set PharmaMonitor. www.berner-safety.de/cytotoxic_wipe_sampling_set_pharmamonitor_en_436.html. Accessed April 26, 2018.