Assessment of chemical contamination by cancer drugs during use of the RIVA TM compounding robot: A pilot study

Abstract

Introduction

Many hospitals are now investing in robotic compounding system for the preparation of cytotoxic agents. The objective of the present study was to describe contamination by cytotoxics inside and outside the RIVA^TM robot (ARxIUM, Winnipeg, Canada).

Material & Methods

We applied a risk analysis to determine which locations inside and outside the compounding robot should be monitored. Samples were collected by swabbing with a wet swab (using 0.1 mL of sterile water) before the robots was cleaned. Ten cytotoxics compounded with the robot were screened for using LC-MS/MS. We determined the percentage contamination rates inside (CR_in) and outside (CR_out) the robot and the amounts of each contaminant (in ng/cm²). If a sample was found to be positive, a corrective action was implemented.

Results

Our risk analysis highlighted 10 locations inside the robot and 7 outside. Ten sampling campaigns (10 samples per campaign) were performed. The mean CR_in (40%) was significantly higher than the mean CR_out (2%; p < 10⁻⁴). Gemcitabine and cyclophosphamide were the main contaminants. After the implementation of corrective measures (such as daily cleaning with SDS/isopropyl alcohol), the CR_in fell from 60% to 10%.

Discussion/conclusion

The frequency of contamination was lower for robotic compounding than for manual compounding in an isolator. However, robotic compounding tended to generated larger mean amounts of contaminant; this was related to incidents such as splashing when syringes were disposed of after the compounding. The implementation of corrective actions effectively reduced the CRs. Further longer-term studies are required to confirm these results.

Keywords

Occupational exposure antineoplastic drugs cancer drugs compounding robot robotization

1. Introduction

In hospital settings, many healthcare professionals (nurses, orderlies, cleaning staff, couriers, physicians, and pharmacy staff) may be potentially exposed to antineoplastic drugs.^1–7 Indeed, each step in the cancer drug supply and use chain (reception, compounding, administration, and even the management of waste and patients’ excreta) involves a risk of exposure.^8–11

A number of guidelines and regulations provide information on the design of compounding areas, the choice of the equipment or medical devices used for compounding and administration, and the requirements for protecting and training the compounding staff. These include good manufacturing practice regulations, the European Pharmaceutical Inspection Co-operation Scheme, the U.S. Pharmacopeia's monograph 800 on “Hazardous Drugs – Handling in Healthcare”, and recommendations from professional bodies and learned societies (e.g., the American Society of Health Pharmacists, the National Institute of Occupational Safety and Health, and the International Society of Oncology Pharmacy Practitioners).^12–17 The main recommendations focus on educating staff about the risks associated with handling cytotoxics as soon as the latter arrive in the compounding unit or hospital ward.^14,16 This handling involves drug compounding and the cleaning and decontamination of compounding facilities. Many other measures have been introduced to protect and monitor staff and raise awareness.^8–10 Even though (i) in-hospital exposure to cytotoxics was evidenced long ago and (ii) procedures are available for reducing this exposure, a number of recent studies have highlighted the continuing presence of cytotoxic drug contamination in hospitals.^18–27

Intravenous cytotoxic drug compounding is typically performed in dedicated, centralized compounding units.^22,25 Manual compounding by pharmacy technicians is performed in barrier isolators or vertical laminar air-flow hoods. Using closed-system drug transfer devices for the compounding step strongly reduces contamination and limits the exposure of pharmacists and pharmacy technicians.^18,19,26 Some studies have found that worksurfaces in compounding units were regularly contaminated by manually compounded cancer drugs.²³

A recent monthly monitoring study performed in our compounding unit showed how surface contamination by several cytotoxic drugs changed over time. Although the contamination decreased significantly at some locations in the unit, it was still present regularly at others.²²

The automated compounding of intravenous cytotoxics enables units to increase the number of compounded preparations - notably in the context of dose banding.²⁸ However, most automatic systems are equipped with a needle sampling device, which is known to increase the risk of contamination to cytotoxic drugs.²⁹ A number of researchers have evaluated this contamination in automated compounding units. Schierl et al. compared cyclophosphamide contamination for manual compounding vs. robotic compounding with the APOTECA Chemo^TM robot.³⁰ Krämer et al. studied 5-fluorouracil (5-FU) and platinum salt contamination associated with the Cytocare^TM robot.³¹ Sessink et al. monitored the surface contaminations inside and outside the Cytocare^TM on two consecutive days and compared them with literature data on manual compounding.³² These three studies showed that the implementation of robotic systems was associated with a significant decrease in contamination in the compounding room.

Our pharmacy recently acquired a RIVA^TM compounding robot (ARxIUM, Winnipeg, Canada). Our unit compounds 52,000 cancer drug preparations (60% of which are in infusion bags) a year, and a 4% annual increase in activity has obliged us to compound dose-banded preparations. The implementation of this robot obliged us to modify our preparation and administration procedures over the whole supply chain.³³ To the best of our knowledge, cytotoxic drug contamination of the RIVA^TM compounding robot has not previously been evaluated.

The objectives of the present study were to assess the levels of surface contamination to cytotoxic drugs inside and outside the RIVA^TM robot and to compare them with those observed for manual compounding. Furthermore, we studied cross-contamination between drug bags prepared by the robot.

2. Materials and methods

2.1. Operation of the RIVA^TM compounding robot

The present study was conducted in a French university hospital. The RIVA^TM robot is an ISO 5 class laminar airflow station for the preparation and reconstitution of sterile injectable drugs (Figures 1 and 2). It has a single, anthropomorphic, articulated arm and contains three areas: (i) a “vestibule” for loading raw materials, (ii) a preparation area containing the robotic arm and preparation stations for both bags and syringes, and (iii) a storage area supplying the system.³⁴ The materials are decontaminated manually first by surface treatment with H₂O₂ when they enter the storage area and then manually using 70% ethanol, before the vials and bags are loaded into the robot. A Datamatrix chip is used to identify and track objects during the compounding.

The robot's first compounding step consists in a microbial decontamination of the drug vials and bag septa with high-intensity UV light. During the preparation, various objects are carried from one station to another by the articulated arm, which allows the robot to perform several actions simultaneously at the different stations. An in-process gravimetric check is made at the end of compounding. Lastly, preparations are labeled inside the area and delivered to the pharmacy technician.

The robot is cleaned daily by the pharmacy technician; with the application of a standard biocide at specific locations: “syringe manipulators up” (SMU) and ‘syringe manipulators down’ (SMD), the robot arm, and the area's floor. All the robot's inner and outer surfaces are thoroughly cleaned weekly by two technicians, using 70% ethanol.

2.2. Risk analysis of surface contamination and sampling plan

2.2.1. Sampling plan

For many years, we have monitored surfaces in locations potentially contaminated by cytotoxic drugs.²² The robot's arrival obliged us to change the number and nature of locations. We therefore sought to measure contamination (i) of the vial when loading in the area, (ii) during the withdrawal of the dose at the SMU, (iii) when injecting the dose into the bag, (iv) when storing the vial on the temporary vial storage shelf, (v) when shaking vials on the vial mixers, (vi) when placing the empty vial into the waste bin, (vii) when removing the needle, and (viii) when canceling a preparation during the dose sampling. For the contamination of the ISO 7 area (outside the robot), the steps with a risk of contamination are (i) handling of drug vials before loading, (ii) removal of the finished preparation (bag or syringe), (iii) disposal of vials after full cleaning, (iv) replacement of the waste bins, and (v) replacement of the needle bin.

This risk analysis was designed to limit the cost of surface monitoring by choosing the most relevant locations. As per our usual procedure for manual compounding, samples were collected monthly.

2.2.2. Sampling method

The study was carried out during the first 11 months after the robot had been qualified. Ten sampling campaigns were conducted. T0 corresponded to the qualification, i.e., before the robot had compounded any cytotoxics. We sought to monitor surface contamination and cross-contamination between compounded drugs.

Initially, eight drugs compounded with robotic assistance. This number has risen to 18, with 12 cytotoxics (5-FU, carboplatin, cyclophosphamide, cytarabine, doxorubicin, epirubicin, etoposide, gemcitabine, irinotecan, methotrexate, pemetrexed, and vincristine) and six monoclonal antibodies. Dacarbazine, ganciclovir, ifosfamide, and vinblastine are not currently compounded with the robot. The most frequently compounded drug was 5-FU (with an average of 138 bags per month, corresponding to 13% of the total activity), followed by gemcitabine (124 bags, 11%), irinotecan (107 bags, 10%), pemetrexed (65 bags, 5%), carboplatin (51 bags, 5%), cyclophosphamide (40 bags, 4%) and doxorubicin (33 bags, 3%).

Firstly, the robot inside and outside surfaces were sampled, according to our usual procedure. Swabs were moistened with 0.1 mL of sterile water and then applied to a 10×10 cm-square by swabbing the surface first horizontally and then vertically until the swab was dry. Thereafter, the swabs were placed in sterile 15 ml conical bottom centrifuge tubes (product reference 188271, Greiner Bio-One, Kremsmünster, Austria) and stored at −18°C until analysis. A negative control was included in each sampling campaign: a swab was placed directly into a centrifuge tube without being manipulated. Surface samples were always collected before the daily cleaning. The RIVA^TM's inside and outside surfaces were sampled at the same time. Samples were also collected from the outside surface and the injection site of prepared infusion bags.

Secondly, cross-contamination between two consecutive preparation batches was assessed by simulating the compounding of cancer drugs. Two cases were studied: the first consisted in mixing bags of antineoplastic drugs in the same production line and preparing them in series, whereas the second consisted in cleaning the robot between two consecutive series of different drugs. Five batches of 4 cytotoxics (5-FU, cyclophosphamide, irinotecan, and gemcitabine) were prepared alternately with four batches of five simulated compounded bags (the transfer of water for injection into 0.9% NaCl bags). A 1 mL sample was withdrawn from each 0.9% NaCl bag, in order to measure the potential presence of cytotoxics from the previously compounded bags. To ensure the absence of cytotoxics before compounding, the blanks consisted of samples from the 0.9% NaCl bag before it was loaded into the robot.

2.2.3. Analytical assay

Samples were assayed in our hospital's toxicology laboratory, using a previously described liquid chromatography tandem mass spectrometry (LC-MS/MS) method.^22,26,27 The compounds were extracted from the swabs using 2 mL of a methanol (gradient grade for liquid chromatography LiChrosolv^®, Merck, Darmstadt, Germany)/formic acid 0.1% (>98%, Sigma-Aldrich, Steinheim, Germany) mixture for 60 min. Fifty μL of an internal standard mix solution (5-FU¹⁵N₂, methyl-clonazepam, and beta-hydroxyethyl-theophylline in methanol) was added to the swabs just before the extraction, and the samples were centrifuged (4500 rpm/5 min). After retrieving the wipes, the solvent was evaporated under high-pressure air, and the dry residue was dissolved in 100 μL of 5 mM ammonium formate (97%, Sigma-Aldrich) in ultrapure water. The concentration was measured by LC-MS/MS (Xevo TQ-XS, Waters, Guyancourt, France). The analytes were separated over 5 min on a stationary phase Acquity UPLC^® HSS C₁₈ column (1.8 μm, 2.1 × 150 mm) using gradient elution with 5 mM ammonium formate buffer and 0.1% formic acid in acetonitrile (ULC/MS quality, Biosolve, Dieuze, France). The injection volume was 10 μL. The method was validated according to French national standards. The limit of detection (LOD) and limit of quantification (LOQ) for each drug are given in Table 1.

Table 1.
LOD and LOQ values for the assayed drugs.

LOD (ng) LOQ (ng)

5-FU 10.0 25.0

Carboplatin 10.0 25.0

Cyclophosphamide 1.0 1.0

Cytarabine 5.0 10.0

Dacarbazine 1.0 1.0

Doxorubicin 1.0 2.5

Epirubicin 1.0 2.5

Etoposide 1.0 1.0

Ganciclovir 1.0 1.0

Gemcitabine 1.0 1.0

Ifosfamide 1.0 1.0

Irinotecan 1.0 1.0

Methotrexate 1.0 1.0

Pemetrexed 5.0 10.0

Vinblastine 2.5 10.0

Vincristine 1.0 1.0

	LOD (ng)	LOQ (ng)
5-FU	10.0	25.0
Carboplatin	10.0	25.0
Cyclophosphamide	1.0	1.0
Cytarabine	5.0	10.0
Dacarbazine	1.0	1.0
Doxorubicin	1.0	2.5
Epirubicin	1.0	2.5
Etoposide	1.0	1.0
Ganciclovir	1.0	1.0
Gemcitabine	1.0	1.0
Ifosfamide	1.0	1.0
Irinotecan	1.0	1.0
Methotrexate	1.0	1.0
Pemetrexed	5.0	10.0
Vinblastine	2.5	10.0
Vincristine	1.0	1.0

2.2.4. Expression of the results and statistical analysis

Surface contamination results are quoted as the percentage contamination rate (CR, in %), calculated by dividing the number of contaminated samples (i.e., > LOD) by the total number of samples and multiplying by 100. The CRs were computed for each compounding period (P) and for the study as a whole. CR_in and CR_out correspond to samples collected respectively inside and outside the robot.

In a second step, CRs at different locations were compared in a chi-squared test or Fisher's exact test, depending on the sample size. The contamination values were expressed in ng/cm² and compared in a non-parametric Mann-Whitney test. All tests were two-tailed. The threshold for statistical significance was set to p < 0.05.

3. Results and discussion

3.1. Risk analysis

Our contamination risk analysis defined 10 sampling locations inside the robot and 7 outside, including areas such as the robot gripper, the SMU, the waste area, and the needle removal station (Figures 1 and 2). The latter seven points included each step in the compounding process before (n = 2), during (n = 2), and after the preparation (n = 3). The use of robotic systems should considerably reduce staff exposure to cytotoxic drugs - mainly because the most critical operations are performed in a closed area, in the absence of human intervention.^35–37 The compounding method (whether robotic or manual) influences the risk of contamination, as does the packaging of the preparations.^30,31,37 Previous studies have highlighted at-risk locations in a robotized process. In a study of chemical contamination inside the Apoteca^TM robot, Schierl et al. used a risk analysis (as we did) to define sampling locations.³⁰ Some of the locations were similar, such as the surfaces under the dosing device where the compounding takes place, the area where the vials are temporarily stored before or after use, and the area where the final products are stored.^30,37 Sessink et al. monitored chemical contamination inside the Cytocare^TM robot by checking for cyclophosphamide at four sampling locations: the entrance grid, the carousel, the storage shelf, and the surface within the robot.³² We also chose these sampling locations for the RIVA^TM. According to our initial results, the needle waste container area appears to less relevant for monitoring because no corrective actions can be undertaken, and the consumables are thrown away after each preparation (limiting the accumulation of chemical contamination). Our present results (the frequency of positive samples and the contamination intensity) have made us more aware of our surface monitoring strategy. Sampling surfaces need to be regularly reevaluated after a few months of follow-up, in order to check that they are still relevant. The less frequent monitoring of some locations will allow others (those more at risk) to be monitored more frequently.

3.2. Analysis of surface contamination

3.2.1. Surface contamination inside and outside the compounding robot

Ten sets of 10 samples were collected, i.e., 100 samples in total. The mean time interval between two campaigns was 39.6 days (range: 22–91 days). All the samples collected at T0 and all the negative controls were contamination-free.

In all, 30 of the 100 samples were found to be contaminated (Figure 3), and eight of these were contaminated by at least two drugs. For samples collected outside the robot (7 locations), one was contaminated by two drugs. These results show that the contamination was significantly higher inside the robot than outside (CR_in= 40.0%, CR_out= 1.43%; p < 0.0001) (Figure 3). These results are encouraging when compared with those for our manual compounding in an isolator, using a closed-system drug transfer device; the contamination rate in the compounding room (around isolators) was 37.1%.²² The analysis of our manual compounding (in which surface contamination was analyzed at three different locations) showed that the contamination rate decreased over time but that an initial contamination spread through the unit.

The mostly frequently detected contaminant drugs were gemcitabine (35% of the contaminations), cyclophosphamide (17.5%), irinotecan (15%), and 5-FU (15%).²² Various drug tracers have been used in the literature: cyclophosphamide ifosfamide, and 5-FU have been studied most frequently.^18,20,24,25 Sessink et al.³² and Schierl et al.³⁰ studied cyclophosphamide contamination, and Krämer et al. studied platinum and 5-FU.³¹ Our choice of tracers was prompted by how frequently the various compounds were handled in our unit. Moreover, specific assay were available locally. Connor et al. commented on the difficulty in comparing literature data generated with different methodologies.¹ Indeed, the assay sensitivity, the swabbing protocol, and the surface area sampled are not standardized and may prevent comparisons.

As shown in Figure 3, CR_in changed over time: the maximum value (60.0%) was observed in P1 and P2, and the minimum value (10.0%) was observed in P9. Inside the robot, four surface zones were frequently contaminated: the waste area (in 7 of the 10 periods), the needle removal station (in 6), the SMU (in 9) and the surface under the SMU (in 4). The median amount of inside chemical contamination was 25.4 ng/cm² (range: 1.07–504,206 ng/cm²; Table 2), and the highest concentrations were found in the waste area. Two samples showed a non-quantifiable contamination on the gripper and the waste area in P3 and P7. Nevertheless, only a few samples were massively contaminated, such as the waste area in P1 (with 5-FU) and SMU in P7 (with pemetrexed). Most contaminations were due to several compounds, with an average of four per contaminated sample.

Table 2.

Results of surface monitoring tests inside the RIVA^TM CarboPt: carboplatin; CYC: cyclophosphamide; DOXO: doxorubicin; 5-FU: 5-fluorouracil; GEM: gemcitabine; IRI: irinotecan; PEME: pemetrexed. ND: not detected (the instances of “ND” in bold italics indicate the presence of contamination noted in the previous sampling campaign), SMD: syringe manipulators down, SMU: syringe manipulators up; < LOQ: detected but below the limit of quantification.

		Amount of contaminant (ng/cm²)
Area	Drug	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10
Waste area (around the door opening area)	5-FU	26,000	ND	ND	ND	ND	ND	< LOQ	ND	ND	ND
	CYC	20.4	367.1	ND	ND	2.0	ND	ND	ND	ND	3.6
	GEM	ND	9.4	1929.8	ND	ND	ND	ND	1.2	ND	ND
	IRI	6000	ND	4.6	ND	ND	ND	ND	ND	ND	ND
Robot gripper	5-FU	ND	ND	ND	< LOQ	ND	ND	ND	ND	ND	ND
Needle removal area	DOXO	ND	83.5	ND	ND	ND	ND	ND	ND	ND	7.08
	IRI	ND	707.8	ND	1.2	ND	ND	ND	ND	ND	ND
	CarboPt	ND	ND	64.3	ND	ND	ND	ND	ND	ND	ND
	5-FU	ND	ND	ND	41.2	25.4	ND	ND	ND	ND	ND
	GEM	ND	ND	ND	ND	8.6	ND	14.2	ND	ND	ND
SMU	GEM	43.1	71.8	52.3	ND	ND	ND	ND	1.1	3.7	ND
	IRI	ND	ND	ND	ND	5.5	ND	ND	ND	ND	ND
	PEME	ND	ND	ND	ND	ND	55.4	504,206	ND	ND	39.39
Under the SMU	GEM	3	235.2	136.3	ND	ND	ND	ND	ND	ND	ND
	CYC	ND	ND	ND	ND	ND	3.7	ND	ND	ND	ND
	IRI	ND	ND	ND	ND	ND	3.7	ND	ND	ND	ND
	PEME	ND	ND	ND	ND	ND	69.7	ND	ND	ND	ND
SMD	5-FU	ND	ND	ND	ND	ND	ND	36.4	ND	ND	ND
	CYC	ND	ND	ND	ND	ND	ND	ND	ND	ND	1.07
	GEM	ND	ND	ND	ND	ND	ND	ND	ND	ND	3.74
Exit area preparation	CYC	1.7	ND	ND	ND	ND	ND	ND	ND	ND	ND

In the literature on contamination, corrective action is never discussed. In contrast, we implemented several prospectively action to reduce contamination. Firstly, a protocol involving a home-made admixture of a surfactant and an alcoholic solvent (i.e., Sodium Dodecyl Sulphate – SDS/isopropanol) was used to decontaminate the most critical areas daily after the production: the waste area (around the door opening), the SMU (the vial gripper and syringe gripper), and the SMD.^20,27 After the high CR_in was observed in P1, the decontamination protocol was implemented for the waste area. After P2, this protocol was extended to the SMU, the area under the SMU, and the SMD; after this change, CR_in at this zones fell from 60% to 10%. A similar effect of decontamination has been previously demonstrated for manual compounding process.²² Secondly, we reorganized the robotic process. Initially, the two bins contained empty vials and consumables, such as bags and syringes. The process for cancer drug vials and waste inside the robot was separated: vials were loaded on one side, and bags and other consumables were loaded on the other side.

Only one outside contamination (in P2, on the syringe waste bin) was observed. It involved both cyclophosphamide and gemcitabine. Given that this frequency of contamination was very low (at least compared to the inside), no corrective measures have been implemented.

Our analysis showed that the main chemical contaminations during the robotic preparation process were related to the generation of droplets. This type of contamination can result from an inappropriate handling parameter or an incident when the syringe is disposed of. We envisaged a change in the robot arm's movement during syringe disposal when a preparation procedure failed. Lastly, for preparation failures, it might be valuable to close syringes with Luer-lock caps before they are disposed of into the waste containers. During our qualification of the RIVA^TM's performance, we performed contamination tests with an invisible tracer (quinine): no contamination was observed. For each new drug handled inside the robot, we determine the optimal settings and seek to minimize the pressure inside the vial. Moreover, the monitoring results can prompt us to reevaluate our parameters. Lastly, we started to wipe the vials before loading the robot, in order to limit input contamination. In fact, it is well known that the external surface of industrial drug vials is often contaminated.^38–40 In the present study, we analyzed chemical contamination inside the RIVA^TM robot. However, several different materials (e.g., various plastics, and stainless steel) are present inside the isolator. It would be interesting to determine whether the extent of chemical contamination depending on the material and the efficiency of the cleaning protocol. It should be noted that we studied a RIVA^TM robot equipped with a needle sampling system. Indeed, most of today's robotic preparation systems use needle sampling; only one (the Equashield pro^TM) uses a closed system inside the robot. Several studies of chemical contamination during manual preparation have evidenced a reduction in chemical contamination when closed systems are used.^26,27 Unfortunately, data on chemical contamination in the Equashield pro^TM do not appear to have been published.

Our study had a number of limitations. Firstly, the time interval between samplings was irregular and ranged from 22 to 91 days, rather than monthly as initially planned. This irregularity was due to logistic issues and the lack of availability of the LC-MS/MS system. Secondly, we sought to limit the cost of surface monitoring by using a risk analysis to select the most relevant sampling locations; it might have been useful to analyze additional locations. Lastly, the number and type of compounding operations varies from one sampling day to another; this might have been a source of bias.

3.2.2. External surface contamination of infusion bags

Twenty-five bags were analyzed for 5-FU, irinotecan, gemcitabine, doxorubicin, and cyclophosphamide. Traces of one or more of these drugs were found in all tested samples. The median contamination values at P1 and P2, respectively, in ng/cm²) were 597 and 457 for 5-FU, 32 and 685 for irinotecan, 1236 and 873 for gemcitabine, 860 and 787 for cyclophosphamide, and 30.7 and 17.4 for doxorubicin. The latter value was computed based on four values as one result was detected but not quantified (i.e., trace). The only significant difference in contamination between P1 and P2 was observed for irinotecan (Table 3).

Table 3.
Contamination of the outside surface of bags.

Drug Dose (mg) Contamination P1 (ng/cm²) Contamination P2 (ng/cm²) p-value

5-Fluorouracil 650 597 ND 0.295

375 776 747

365 2378 ND

800 235 457

800 209 341

Irinotecan 180 1.0 685 0.008

156 52 852

240 32 231

240 182 633

240 3.4 1169

Gemcitabine 2640 415 711 0.421

1148 1236 873

2724 2005 1195

2160 1191 1299

2160 1809 783

Cyclophosphamide 1054 1659 152 0.548

1000 413 1354

976 1027 342

1431 860 787

844 760 983

Doxorubicin 49.8 71.8 81.5 0.310

84 15.7 28.4

50 30.7 Traces*

68 44.4 6.3

84 10.3 4.4

Drug	Dose (mg)	Contamination P1 (ng/cm²)	Contamination P2 (ng/cm²)	p-value
5-Fluorouracil	650	597	ND	0.295
	375	776	747
	365	2378	ND
	800	235	457
	800	209	341
Irinotecan	180	1.0	685	0.008
	156	52	852
	240	32	231
	240	182	633
	240	3.4	1169
Gemcitabine	2640	415	711	0.421
	1148	1236	873
	2724	2005	1195
	2160	1191	1299
	2160	1809	783
Cyclophosphamide	1054	1659	152	0.548
	1000	413	1354
	976	1027	342
	1431	860	787
	844	760	983
Doxorubicin	49.8	71.8	81.5	0.310
	84	15.7	28.4
	50	30.7	Traces*
	68	44.4	6.3
	84	10.3	4.4

* Trace contamination was defined as an amount between the LOD and the LOQ.

3.3 Cross-contamination study

We analyzed 20 bags of real cytotoxic preparations and 20 bags of simulated preparations. No traces of cytotoxics were detected in any of the simulated preparations. Unfortunately, we could not find any comparable literature data on cross-contamination by a chemotherapy compounding robot.

4. Conclusion

Our results showed that contamination to cytotoxic drugs was significantly lower outside the RIVA^TM robot than inside. The implementation of corrective actions (notably cleaning with specific decontamination solutions) enabled us to reduce the CR over time. Further studies are required to evaluate the influence of regular decontamination on the persistence of contamination in the robot and on bag surfaces. Our results also showed that serial cross-contamination between batches is unlikely. Further studies are required to compare robot compounding and manual compounding with regard to precision and/or to compare several compounding robots.

Footnotes

Acknowledgment

The authors thank David Fraser, PhD (Biotech Communication SARL, Ploudalmézeau, France) for editorial assistance.

Author contributions

Conceptualization, M.B. and M.V.; methodology, M.B., N.S. and M.V.; formal analysis, M.B., N.B., J.C., M.P. and M.V.; investigation, M.B., N.S. and M.V.; resources, M.V. and N.B.; data curation, M.B., M.V.; writing—original draft preparation, M.B., N.S. and M.V.; writing—review and editing, M.P., N.B., D.A. and P.O.; supervision, D.A., N.S. and P.O.; project administration: N.S., M.V., P.O. All authors have read and agreed to the published version of the manuscript.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Michèle Vasseur

Nicolas Simon

References

Connor

Lawson

Polovich

, et al. Reproductive health risks associated with occupational exposures to antineoplastic drugs in health care settings: a review of the evidence. J Occup Environ Med 2014; 56: 901–910.

Turci

Minoia

Sottani

, et al. Occupational exposure to antineoplastic drugs in seven Italian hospitals: the effect of quality assurance and adherence to guidelines. J Oncol Pharm Pract 2011; 17: 320–332.

Selevan

Lindbohm

Hornung

, et al. A study of occupational exposure to antineoplastic drugs and fetal loss in nurses. N Engl J Med 1985; 313: 1173–1178.

El-Ebiary

Abuelfadl

Sarhan

. Evaluation of genotoxicity induced by exposure to antineoplastic drugs in lymphocytes of oncology nurses and pharmacists. J Appl Toxicol 2013; 33: 196–201.

Sessink

Bos

. Drugs hazardous to healthcare workers. Evaluation of methods for monitoring occupational exposure to cytostatic drugs. Drug Saf 1999; 20: 347–359.

Villa

Geshkovska

Bellagamba

, et al. Factors associated with internal contamination of nurses by antineoplastic drugs based on biomonitoring data from a previous study. Int J Hyg Environ Health 2023; 254: 114264.

Ndaw

Remy

. Occupational exposure to antineoplastic drugs in twelve French health care setting: biological monitoring and surface contamination. Int J Environ Res Public Health 2023; 20: 4952.

Connor

DeBord

Pretty

, et al. Evaluation of antineoplastic drug exposure of health care workers at three university-based US cancer centers. J Occup Environ Med 2010; 52: 1019–1027.

Hedmer

Tinnerberg

Axmon

, et al. Environmental and biological monitoring of antineoplastic drugs in four workplaces in a Swedish hospital. Int Arch Occup Environ Health 2008; 81: 899–911.

10.

Yoshida

Koda

Nishida

, et al. Association between occupational exposure levels of antineoplastic drugs and work environment in five hospitals in Japan. J Oncol Pharm Pract 2011; 17: 29–38.

11.

Chu

Hon

Danyluk

, et al. Pilot assessment of the antineoplastic drug contamination levels in British Columbian hospitals pre- and post-cleaning. J Oncol Pharm Pract 2012; 18: 46–51.

12.

Agence Nationale de Sécurité des Médicaments et des produits de santé. Bonnes Pratiques de Préparation, https://ansm.sante.fr/documents/reference/bonnes-pratiques-de-preparation (Edition 2023, accessed 26 October 2023).

13.

Pharmaceutical inspection convention pharmaceutical inspection co-operation scheme. PIC/S guide to good practices for the preparation of medicinal products in healthcare establishments – PE 010-4, https://picscheme.org/en/publications (2014, accessed 26 October 2023).

14.

Power

Coyne

. ASHP guidelines on handling hazardous drugs. Am J Health-Syst Pharm 2018; 75: 1996–2031.

15.

National Institute of Occupational Safety and Health. NIOSH list of antineoplastic and other hazardous drugs in healthcare settings, https://www.cdc.gov/niosh/docs/2016-161/pdfs/2016-161.pdf (2016).

16.

International Society of Oncology Pharmacy Practice. ISOPP standards for the safe handling of cytotoxics. J Oncol Pharm Pract 2022; 28: 1–126.

17.

US Pharmacopeia. USP General chapter <800> Hazardous drugs – handling in healthcare settings. www.usp.org (accessed 26 October 2023).

18.

Guillemette

Langlois

Voisine

, et al. Impact and appreciation of two methods aiming at reducing hazardous drug environmental contamination: the centralization of the priming of IV tubing in the pharmacy and use of a closed-system transfer device. J Oncol Pharm Pract 2014; 20: 426–432.

19.

Sessink

PJM

Trahan

Coyne

. Reduction in surface contamination with cyclophosphamide in 30 US hospital pharmacies following implementation of a closed-system drug transfer device. Hosp Pharm 2013; 48: 204–212.

20.

Anastasi

Rudaz

Queruau Lamerie

, et al. Efficacy of two cleaning solutions for the decontamination of 10 antineoplastic agents in the biosafety cabinets of a hospital pharmacy. Ann Occup Hyg 2015; 59: 895–908.

21.

Delafoy

Roussy

Hudon

, et al. Canadian Monitoring program of the surface contamination with 11 antineoplastic drugs in 122 centers. J Oncol Pharm Pract 2023; 29: 338–347.

22.

Saint-Lorant

Vasseur

Allorge

, et al. Four-year follow-up of surface contamination by antineoplastic drugs in a compounding unit. Occup Environ Med 2023; 80: 146–153.

23.

Bussières

Tanguay

Touzin

, et al. Environmental contamination with hazardous drugs in Quebec hospitals. Can J Hosp Pharm 2012; 65: 428–435.

24.

Acampora

Castiglia

Miraglia

, et al. A case study: surface contamination of cyclophosphamide due to working practices and cleaning procedures in two Italian hospitals. Ann Occup Hyg 2005; 49: 611–618.

25.

Crauste-Manciet

Sessink

PJM

Ferrari

, et al. Environmental contamination with cytotoxic drugs in healthcare using positive air pressure isolators. Ann Occup Hyg 2005; 49: 619–628.

26.

Simon

Vasseur

Pinturaud

, et al. Effectiveness of a closed-system transfer device in reducing surface contamination in a new antineoplastic drug-compounding unit: a prospective, controlled, parallel study. PLoS One 2016; 11: e0159052.

27.

Vasseur

Simon

Picher

, et al. A decontamination process adding a tensioactive agent and isopropanol to a closed-system drug transfer device for better control of isolator contamination. A prospective, parallel study. PloS One 2018; 13: e0201335.

28.

Seger

Churchill

Keohane

, et al. Impact of robotic antineoplastic preparation on safety, workflow, and costs. J Oncol Pract 2012; 8: 344–349.

29.

Spivey

Connor

. Determining sources of workplace contamination with antineoplastic drugs and comparing conventional IV drug preparation with a closed system. Hosp Pharm 2003; 38: 135–139.

30.

Schierl

Masini

Groeneveld

, et al. Environmental contamination by cyclophosphamide preparation: comparison of conventional manual production in biological safety cabinet and robot-assisted production by APOTECAchemo. J Oncol Pharm Pract 2016; 22: 37–45.

31.

Krämer

Federici

Schierl

. Environmental and product contamination during the preparation of antineoplastic drugs with robotic systems. Pharm Technol Hosp Pharm 2018; 3: 153–164.

32.

Sessink

Leclercq

Wouters

, et al. Environmental contamination, product contamination and workers exposure using a robotic system for antineoplastic drug preparation. J Oncol Pharm Pract 2015; 21: 118–127.

33.

Caron

Vasseur

Courtin

, et al. Evaluation of cancer drug infusion devices prior to the implementation of a compounding robot. J Oncol Pharm Pract 2024; 30: 251–256.

34.

Allersma

Odou

Gareb

. Equipment. In: Le Brun P, Crauste-Manciet S, Krämer I, Smith J and Woerdenbag H (eds) Practical pharmaceutics. Cham: Springer.

35.

Carrez

Martin

Verrey

, et al. Qualification and performance evaluation of an automated system for compounding injectable cytotoxic drugs. Pharm Technol Hosp Pharm 2018; 3: 165–175.

36.

Deljehier

Bouguéon

Heloury

, et al. Simulation program of a cytotoxic compounding robot for monoclonal antibodies and anti-infectious sterile drug preparation. J Oncol Pharm Pract 2019; 25: 1873–1890.

37.

Gandré

Krämer

. Cytotoxic surface contamination in a robotic system compared with manual preparation. Eur J Hosp Pharm 2012; 19: 151. doi:https://doi.org/10.1136/ejhpharm-2012-000074.175

38.

Power

Sessink

PJM

Gesy

, et al. Hazardous drug residue on exterior vial surfaces: evaluation of a commercial manufacturing process. Hosp Pharm 2014; 49: 355–362.

39.

Schierl

Herwig

Pfaller

, et al. Surface contamination of antineoplastic drug vials: comparison of unprotected and protected vials. Am J Health-Syst Pharm 2010; 67: 428–429.

40.

Fleury-Souverain

Nussbaumer

Mattiuzzo

, et al. Determination of the external contamination and cross-contamination by cytotoxic drugs on the surfaces of vials available on the Swiss market. J Oncol Pharm Pract 2014; 20: 100–111.

Assessment of chemical contamination by cancer drugs during use of the RIVA TM compounding robot: A pilot study

Abstract

Introduction

Material & Methods

Results

Discussion/conclusion

Keywords

1. Introduction

2. Materials and methods

2.1. Operation of the RIVATM compounding robot

2.2.1. Sampling plan

2.2.2. Sampling method

2.2.3. Analytical assay

3. Results and discussion

3.1. Risk analysis

3.2. Analysis of surface contamination

3.2.1. Surface contamination inside and outside the compounding robot

4. Conclusion

Footnotes

Acknowledgment

Author contributions

Declaration of conflicting interests

Funding

ORCID iDs

References

2.1. Operation of the RIVA^TM compounding robot