bs_bs_banner
R E V I E W
Drug and Alcohol Review (January 2015), 34, 67–73 DOI: 10.1111/dar.12196
Filtration of crushed tablet suspensions has potential to reduce infection incidence in people who inject drugs HUEI NG1, RAHUL P. PATEL1, RAIMONDO BRUNO2, ROGER LATHAM3, TROY WANANDY1,4 & STUART MCLEAN1 1
Division of Pharmacy, School of Medicine, University of Tasmania, Hobart, Australia, 2Division of Psychology, School of Medicine, University of Tasmania, Hobart, Australia, 3School of Medicine, University of Tasmania, Hobart, Australia, and 4 Pharmacy Department, Royal Hobart Hospital, Hobart, Australia
Abstract Introduction and Aims. The medical complications of injecting preparations from crushed tablets can be severe, and most can be attributed to the injection of insoluble particles and micro-organisms. Previously we have shown that most of the particles can be removed by filtration, but it was not known whether bacteria could also be filtered in the presence of a high particle load. This study aims to determine the feasibility of filtration to remove bacteria from injections prepared from tablets. Design and Methods. Injections were prepared from crushed slow-release morphine tablets, in mixed bacterial suspensions of Staphylococcus aureus, Streptococcus pyogenes and Pseudomonas aeruginosa. The injection suspensions were passed through syringe filters of porosity 0.45 or 0.20 μm, or combined 0.8 then 0.2 μm, and the bacterial load was counted. Results. Bacterial concentrations in unfiltered injections were 2.5–4.3 × 106 colony forming units mL−1. Both the 0.20 and 0.45 μm filters blocked unless a prefilter (cigarette filter) was used first. The 0.2 μm filter and the combined 0.8/0.2 μm filter reduced the bacteria to the limit of detection (10 colony forming units mL−1) or below. Filtration through a 0.45 μm filter was slightly less effective. Discussion and Conclusions. Use of a 0.2 μm filter, together with other injection hygiene measures, offers the prospect of greatly reducing the medical complications of injecting crushed tablets and should be considered as a highly effective harm reduction method. It is very likely that these benefits would also apply to other illicit drug injections, although validation studies are needed. [Ng H, Patel RP, Latham R, Bruno R, Wanandy R, McLean S. Filtration of crushed tablet suspensions has potential to reduce infection incidence in people who inject drugs. Drug Alcohol Rev 2015;34:67–73] Key words: injecting drug use, filtration, micro-organism, harm reduction, pharmaceutical opioid.
Introduction Injection of illicit drug preparations carries health risks in addition to those due to the pharmacological properties of the drug itself. These additional risks include local lesions at the site of injection, cardiovascular and pulmonary complications, focal and disseminated infections, and gangrene and amputation [1–6]. Although recognised since the 1960s [2], these nonpharmacological complications remain a significant source of morbidity and mortality among people who inject illicit drug preparations.
Most of these complications are attributable to the injection of micro-organisms and insoluble particles with the drug solutions. Illicitly prepared drug injections are not sterile, and the presence of microorganisms in the equipment used, injection mixture and puncture site carry a high risk of infection [7,8]. Most non-pulmonary bacterial infections in people who inject drugs are due to Staphylococcus aureus (SA) or streptococci, often from their own colonising strains [9]. Street drugs and injection paraphernalia have been found to be a source of these and other pathogens, such as Pseudomonas aeruginosa (PAO), Candida albicans,
Huei Ng MPharmSc, Research Student, Rahul P. Patel PhD, Lecturer, Roger Latham BAgrSc (Hons), Associate Lecturer, Raimondo Bruno PhD (ClinPsych), Associate Professor,Troy Wanandy MSc, Professional Specialist Pharmacist and University Associate, Stuart McLean PhD, Emeritus Professor. Correspondence to Prof. Stuart McLean, Pharmacy, School of Medicine, University of Tasmania, Private Bag 26, Hobart, Tas. 7001, Australia. Tel: +61 3 6226 2199; Fax: +61 3 6226 2870; E-mail: [email protected] Received 14 April 2014; accepted for publication 16 July 2014. © 2014 Australasian Professional Society on Alcohol and other Drugs
68
H. Ng et al.
Clostridia species and Bacillus anthracis [5,7–9]. Although most infections are due to commensal organisms on the skin of the individual or their associates, or in their immediate environment [10], there are instances of infections from practically every possible source. The oral flora (e.g. Strep. Milleri, Eikenella corrodens) have been implicated in infections after crushing tablets between the teeth; Pseudomonas has been sourced to contaminated tap water, and Candida infections have followed the use of lemon juice to dissolve heroin or cocaine [7]. Viral and parasitic infections (e.g. malaria, even in Europe and the USA) have been associated with sharing blood-contaminated equipment [7]. Illicit drugs often contain insoluble additives, fillers and contaminants which can produce chronic granulomatous inflammation, angiothrombosis and pulmonary complications [2,6]. Injections prepared from crushed pharmaceutical tablets present a particular hazard as they typically contain large numbers of insoluble particles of talc and other fillers of various sizes whose deposition can cause abscesses, emboli and ischaemic necrosis [3,6,11,12]. This tissue damage and ischaemia can promote the development of infection when micro-organisms are introduced [5]. Injection of pharmaceutical tablets, mainly opioids, is a worldwide problem seen in the USA [13], Canada [14], India and other countries of South Asia [15], Europe [16,17] and Australia [18,19]. The type of filter, and whether any is used, varies considerably among people who inject tablets [20,21]. Makeshift filters (cigarette filters, cotton balls) are readily available and easy to use but are ineffective at removing all but the largest particles (>50 μm) and retain significant amounts of drug unless flushed [20,22]. Syringe (or ‘wheel’) filters are far more effective but tend to clog when the pore size is small (0.2 or 0.45 μm). This can be prevented by first using another coarser filter (even a makeshift filter), but this adds another step. This problem can be at least partly overcome by the use of filters specifically designed to replace makeshift filters (e.g. Sterifilt®, Apothicom, Paris, France). These have 10 μm pores and are less liable to block but also admit more particles and are not sterilising [17]. Morphine formulated in an insoluble matrix for prolonged effect after an oral dose (MS Contin®, Mundipharma, Cambridge, UK) is the most commonly injected pharmaceutical opioid among people who frequently inject drugs in Australia [23] and has been particularly associated with harm [24]. We have previously shown that an injection prepared from a single crushed tablet, using the methods adopted by people who inject drugs, contains tens of millions of particles with sizes ranging from less than 5 μm to © 2014 Australasian Professional Society on Alcohol and other Drugs
greater than 400 μm [20]. Passing the solution through a cigarette filter, a common practice among injectors, only removed the particles of greater than 50 μm. A 0.22 μm syringe filter was capable of removing essentially all of the tablet-derived particles, although it would block unless used after another coarser prefilter (such as a cigarette filter). In pharmaceutical practice, solutions for injection may be sterilised by passing them through a bacterialretentive membrane filter with a nominal pore size of 0.22 μm or less [25]. Therefore, effective particle filtration also offers the prospect of a microbiologically clean injection. However, pharmaceutical syringe filters are not designed to filter the heavy load of particles produced from crushed tablets, even that remaining after passage through a cigarette filter. Membrane filters are most effective when the filtered load is low. This is because in addition to sieving at their pores, membrane filters also remove particles by surface effects (impaction and adsorption) and by trapping within pores, all of which are saturable processes [26]. Heavily contaminated solutions can block these sites and can also clog the pores, thus stopping the flow through the filter which then fails to sterilise all of the preparation. It is also possible that the integrity of filters could be damaged by the particles and the excessive pressures sometimes required for filtration, leading to leakage of micro-organisms. This study aims to determine the effectiveness of membrane filters at removing bacteria from injections prepared from crushed slow-release oral morphine tablets. These tablets were selected for investigation as they are the most commonly injected, and our previous work has demonstrated them to carry a greater particulate burden than oxycodone or buprenorphine tablets [20,22,27]. Bacterial suspensions were prepared, mixed with crushed tablets to prepare the injection suspensions, then filtered by a variety of methods.The bacteria in the suspensions were enumerated before and after filtration.
Methods Bacterial suspension Three bacterial strains which commonly cause infections in people who inject drugs [7,8] were chosen for study. SA (ATCC 29213), Streptococcus pyogenes (SP; ATCC 19615) and PAO (reference strain PAO1) were grown overnight on tryptone soy broth solidified with 1.5% agar (Oxoid CM0129 and LP0011, ThermoFisher Scientific, Scoresby, Vic, Australia) at 35°C in a humid atmosphere enriched with 5% CO2. The bacteria were then suspended in sterile 0.9% NaCl (saline). Concentrations of bacteria were estimated
Filtration of illicit drug injections
using standard curves of absorbance at 600 nm against bacterial concentration. The actual concentrations of bacteria were retrospectively determined by serial dilution and plate count as described below (‘Enumeration of bacteria’). The bacterial suspensions were mixed together, with the actual concentration of each bacterial species in the final suspension ranging from 1.5 to 8.6 × 106 colony-forming units (CFU) mL−1, then mixed with crushed slow-release morphine tablets as detailed next. Injections and filtration The methods of preparing injections from crushed tablets (sustained-release 60 mg morphine sulphate; MS-Contin) were developed following an interview with people who inject drugs and have been described previously, as have most of the methods of filtration [20]. In brief, crushed tablets were suspended in 3 mL of the mixed bacterial suspension. The membrane filters (also known as ‘wheel’ filters) were the type which attach to a syringe (we used 25 mm diameter filters and 5 mL syringes).The filter porosity was either 0.45 or 0.2 μm, or a combination of two filters in sequence: 0.80 μm followed by 0.20 μm (combined 0.8/0.2 filter). All filters were obtained from Acrodisc® (Pall Life Sciences, Ann Arbor, MI, USA), and details and images can be seen at the Pall Life Sciences website (see http://www.pall.com/pdfs/Laboratory/Sterile _Acrodisc_Syringe_Filters_with_Supor_Membrane_PN _102370H.pdf. Accessed 8 June 2014 and archived by WebCite® at http://www.webcitation.org/6QBwk ZNYh). In trials, it was found that the 0.45 and 0.20 μm filters became blocked before the total injection volume had been filtered (as evidenced by increased back pressure), so the extracts were first filtered through an unused filter for hand-rolled cigarettes (Ranch Slims, Stuart Alexander and Co., Sydney, Australia). This pre-filtration process was not required for the combined 0.8/0.2 μm filter [22]. After filtration, the filters were rinsed with 2 mL saline, and this volume was also added to the unfiltered mixtures. Control injections were prepared in the same way, except that the tablet was omitted. The particle counts for the filtered and unfiltered injections are detailed in our previous publication [20]. Enumeration of bacteria Bacterial suspensions were serially diluted in saline then plated on brain heart infusion broth solidified with 1.5% bacteriological agar (BHIA, Oxoid CM1135 and LP0011). Aliquots (100 μL) of each dilution were spread on two BHIA plates which were then incubated for 24 h at 35°C in a humidified atmosphere enriched with 5% CO2.The number of colonies of each bacterial
69
species was determined by macroscopic examination of the plates and comparison with typical colony morphologies. Counts from duplicate plates were averaged and are expressed as CFU mL−1. Counts between 30 and 300 CFU per plate were used for accurate quantitation and correspond to 300–3000 CFU mL−1. In filtrates where plate counts were
R E V I E W
Drug and Alcohol Review (January 2015), 34, 67–73 DOI: 10.1111/dar.12196
Filtration of crushed tablet suspensions has potential to reduce infection incidence in people who inject drugs HUEI NG1, RAHUL P. PATEL1, RAIMONDO BRUNO2, ROGER LATHAM3, TROY WANANDY1,4 & STUART MCLEAN1 1
Division of Pharmacy, School of Medicine, University of Tasmania, Hobart, Australia, 2Division of Psychology, School of Medicine, University of Tasmania, Hobart, Australia, 3School of Medicine, University of Tasmania, Hobart, Australia, and 4 Pharmacy Department, Royal Hobart Hospital, Hobart, Australia
Abstract Introduction and Aims. The medical complications of injecting preparations from crushed tablets can be severe, and most can be attributed to the injection of insoluble particles and micro-organisms. Previously we have shown that most of the particles can be removed by filtration, but it was not known whether bacteria could also be filtered in the presence of a high particle load. This study aims to determine the feasibility of filtration to remove bacteria from injections prepared from tablets. Design and Methods. Injections were prepared from crushed slow-release morphine tablets, in mixed bacterial suspensions of Staphylococcus aureus, Streptococcus pyogenes and Pseudomonas aeruginosa. The injection suspensions were passed through syringe filters of porosity 0.45 or 0.20 μm, or combined 0.8 then 0.2 μm, and the bacterial load was counted. Results. Bacterial concentrations in unfiltered injections were 2.5–4.3 × 106 colony forming units mL−1. Both the 0.20 and 0.45 μm filters blocked unless a prefilter (cigarette filter) was used first. The 0.2 μm filter and the combined 0.8/0.2 μm filter reduced the bacteria to the limit of detection (10 colony forming units mL−1) or below. Filtration through a 0.45 μm filter was slightly less effective. Discussion and Conclusions. Use of a 0.2 μm filter, together with other injection hygiene measures, offers the prospect of greatly reducing the medical complications of injecting crushed tablets and should be considered as a highly effective harm reduction method. It is very likely that these benefits would also apply to other illicit drug injections, although validation studies are needed. [Ng H, Patel RP, Latham R, Bruno R, Wanandy R, McLean S. Filtration of crushed tablet suspensions has potential to reduce infection incidence in people who inject drugs. Drug Alcohol Rev 2015;34:67–73] Key words: injecting drug use, filtration, micro-organism, harm reduction, pharmaceutical opioid.
Introduction Injection of illicit drug preparations carries health risks in addition to those due to the pharmacological properties of the drug itself. These additional risks include local lesions at the site of injection, cardiovascular and pulmonary complications, focal and disseminated infections, and gangrene and amputation [1–6]. Although recognised since the 1960s [2], these nonpharmacological complications remain a significant source of morbidity and mortality among people who inject illicit drug preparations.
Most of these complications are attributable to the injection of micro-organisms and insoluble particles with the drug solutions. Illicitly prepared drug injections are not sterile, and the presence of microorganisms in the equipment used, injection mixture and puncture site carry a high risk of infection [7,8]. Most non-pulmonary bacterial infections in people who inject drugs are due to Staphylococcus aureus (SA) or streptococci, often from their own colonising strains [9]. Street drugs and injection paraphernalia have been found to be a source of these and other pathogens, such as Pseudomonas aeruginosa (PAO), Candida albicans,
Huei Ng MPharmSc, Research Student, Rahul P. Patel PhD, Lecturer, Roger Latham BAgrSc (Hons), Associate Lecturer, Raimondo Bruno PhD (ClinPsych), Associate Professor,Troy Wanandy MSc, Professional Specialist Pharmacist and University Associate, Stuart McLean PhD, Emeritus Professor. Correspondence to Prof. Stuart McLean, Pharmacy, School of Medicine, University of Tasmania, Private Bag 26, Hobart, Tas. 7001, Australia. Tel: +61 3 6226 2199; Fax: +61 3 6226 2870; E-mail: [email protected] Received 14 April 2014; accepted for publication 16 July 2014. © 2014 Australasian Professional Society on Alcohol and other Drugs
68
H. Ng et al.
Clostridia species and Bacillus anthracis [5,7–9]. Although most infections are due to commensal organisms on the skin of the individual or their associates, or in their immediate environment [10], there are instances of infections from practically every possible source. The oral flora (e.g. Strep. Milleri, Eikenella corrodens) have been implicated in infections after crushing tablets between the teeth; Pseudomonas has been sourced to contaminated tap water, and Candida infections have followed the use of lemon juice to dissolve heroin or cocaine [7]. Viral and parasitic infections (e.g. malaria, even in Europe and the USA) have been associated with sharing blood-contaminated equipment [7]. Illicit drugs often contain insoluble additives, fillers and contaminants which can produce chronic granulomatous inflammation, angiothrombosis and pulmonary complications [2,6]. Injections prepared from crushed pharmaceutical tablets present a particular hazard as they typically contain large numbers of insoluble particles of talc and other fillers of various sizes whose deposition can cause abscesses, emboli and ischaemic necrosis [3,6,11,12]. This tissue damage and ischaemia can promote the development of infection when micro-organisms are introduced [5]. Injection of pharmaceutical tablets, mainly opioids, is a worldwide problem seen in the USA [13], Canada [14], India and other countries of South Asia [15], Europe [16,17] and Australia [18,19]. The type of filter, and whether any is used, varies considerably among people who inject tablets [20,21]. Makeshift filters (cigarette filters, cotton balls) are readily available and easy to use but are ineffective at removing all but the largest particles (>50 μm) and retain significant amounts of drug unless flushed [20,22]. Syringe (or ‘wheel’) filters are far more effective but tend to clog when the pore size is small (0.2 or 0.45 μm). This can be prevented by first using another coarser filter (even a makeshift filter), but this adds another step. This problem can be at least partly overcome by the use of filters specifically designed to replace makeshift filters (e.g. Sterifilt®, Apothicom, Paris, France). These have 10 μm pores and are less liable to block but also admit more particles and are not sterilising [17]. Morphine formulated in an insoluble matrix for prolonged effect after an oral dose (MS Contin®, Mundipharma, Cambridge, UK) is the most commonly injected pharmaceutical opioid among people who frequently inject drugs in Australia [23] and has been particularly associated with harm [24]. We have previously shown that an injection prepared from a single crushed tablet, using the methods adopted by people who inject drugs, contains tens of millions of particles with sizes ranging from less than 5 μm to © 2014 Australasian Professional Society on Alcohol and other Drugs
greater than 400 μm [20]. Passing the solution through a cigarette filter, a common practice among injectors, only removed the particles of greater than 50 μm. A 0.22 μm syringe filter was capable of removing essentially all of the tablet-derived particles, although it would block unless used after another coarser prefilter (such as a cigarette filter). In pharmaceutical practice, solutions for injection may be sterilised by passing them through a bacterialretentive membrane filter with a nominal pore size of 0.22 μm or less [25]. Therefore, effective particle filtration also offers the prospect of a microbiologically clean injection. However, pharmaceutical syringe filters are not designed to filter the heavy load of particles produced from crushed tablets, even that remaining after passage through a cigarette filter. Membrane filters are most effective when the filtered load is low. This is because in addition to sieving at their pores, membrane filters also remove particles by surface effects (impaction and adsorption) and by trapping within pores, all of which are saturable processes [26]. Heavily contaminated solutions can block these sites and can also clog the pores, thus stopping the flow through the filter which then fails to sterilise all of the preparation. It is also possible that the integrity of filters could be damaged by the particles and the excessive pressures sometimes required for filtration, leading to leakage of micro-organisms. This study aims to determine the effectiveness of membrane filters at removing bacteria from injections prepared from crushed slow-release oral morphine tablets. These tablets were selected for investigation as they are the most commonly injected, and our previous work has demonstrated them to carry a greater particulate burden than oxycodone or buprenorphine tablets [20,22,27]. Bacterial suspensions were prepared, mixed with crushed tablets to prepare the injection suspensions, then filtered by a variety of methods.The bacteria in the suspensions were enumerated before and after filtration.
Methods Bacterial suspension Three bacterial strains which commonly cause infections in people who inject drugs [7,8] were chosen for study. SA (ATCC 29213), Streptococcus pyogenes (SP; ATCC 19615) and PAO (reference strain PAO1) were grown overnight on tryptone soy broth solidified with 1.5% agar (Oxoid CM0129 and LP0011, ThermoFisher Scientific, Scoresby, Vic, Australia) at 35°C in a humid atmosphere enriched with 5% CO2. The bacteria were then suspended in sterile 0.9% NaCl (saline). Concentrations of bacteria were estimated
Filtration of illicit drug injections
using standard curves of absorbance at 600 nm against bacterial concentration. The actual concentrations of bacteria were retrospectively determined by serial dilution and plate count as described below (‘Enumeration of bacteria’). The bacterial suspensions were mixed together, with the actual concentration of each bacterial species in the final suspension ranging from 1.5 to 8.6 × 106 colony-forming units (CFU) mL−1, then mixed with crushed slow-release morphine tablets as detailed next. Injections and filtration The methods of preparing injections from crushed tablets (sustained-release 60 mg morphine sulphate; MS-Contin) were developed following an interview with people who inject drugs and have been described previously, as have most of the methods of filtration [20]. In brief, crushed tablets were suspended in 3 mL of the mixed bacterial suspension. The membrane filters (also known as ‘wheel’ filters) were the type which attach to a syringe (we used 25 mm diameter filters and 5 mL syringes).The filter porosity was either 0.45 or 0.2 μm, or a combination of two filters in sequence: 0.80 μm followed by 0.20 μm (combined 0.8/0.2 filter). All filters were obtained from Acrodisc® (Pall Life Sciences, Ann Arbor, MI, USA), and details and images can be seen at the Pall Life Sciences website (see http://www.pall.com/pdfs/Laboratory/Sterile _Acrodisc_Syringe_Filters_with_Supor_Membrane_PN _102370H.pdf. Accessed 8 June 2014 and archived by WebCite® at http://www.webcitation.org/6QBwk ZNYh). In trials, it was found that the 0.45 and 0.20 μm filters became blocked before the total injection volume had been filtered (as evidenced by increased back pressure), so the extracts were first filtered through an unused filter for hand-rolled cigarettes (Ranch Slims, Stuart Alexander and Co., Sydney, Australia). This pre-filtration process was not required for the combined 0.8/0.2 μm filter [22]. After filtration, the filters were rinsed with 2 mL saline, and this volume was also added to the unfiltered mixtures. Control injections were prepared in the same way, except that the tablet was omitted. The particle counts for the filtered and unfiltered injections are detailed in our previous publication [20]. Enumeration of bacteria Bacterial suspensions were serially diluted in saline then plated on brain heart infusion broth solidified with 1.5% bacteriological agar (BHIA, Oxoid CM1135 and LP0011). Aliquots (100 μL) of each dilution were spread on two BHIA plates which were then incubated for 24 h at 35°C in a humidified atmosphere enriched with 5% CO2.The number of colonies of each bacterial
69
species was determined by macroscopic examination of the plates and comparison with typical colony morphologies. Counts from duplicate plates were averaged and are expressed as CFU mL−1. Counts between 30 and 300 CFU per plate were used for accurate quantitation and correspond to 300–3000 CFU mL−1. In filtrates where plate counts were
