In order to give compounded cytostatics longer shelf lives, better data are a prerequisite, and it would be opportune either to establish an international collaboration or to centralise the resources on a European level to find a common solution
Iben Larsson PhD
The Danish Research Unit for Hospital Pharmacy, Amgros, Copenhagen, Denmark
Despite the increased efficiency in cancer treatments, the demand for ready-to-use cytostatic preparations is constantly growing in hospital pharmacies. In 2010, an estimate was made of Danish hospital pharmacies’ future demands. Based on the past six years, there was an average annual growth of 4.5% in the total increase of delivery of cytotoxic drugs. In 2009, approximately 245,000 preparations were made, and this number is expected to reach 400,000 by 2020.
In Denmark, as in many other countries in Europe, the compounding of cytostatics takes place in hospital pharmacies at small, decentralised compounding units, without the use of advanced technology. The production is mainly patient-specific and order-based, with shelf lives of between 24 hours and 14 days. One solution to fulfil future requirements is to compound cytostatics with the largest turnover as batch production for stockholding, and apply/implement/adopt dosebanding whenever clinically applicable. However, such a set-up requires the documentation of chemical/physical stability data for a minimum of one-to-three months. The microbiology stability has to be validated locally and is not discussed in this article.
Hospital pharmacists are responsible for the quality of compounded products, and therefore they need to have documentation for the longer shelf lives of compounded cytostatics before stock-keeping can be introduced. The usual applied sources for documenting chemical/physical stability of compounded drugs are: the Summary of Product Characteristics (SPC); data provided by the industry, which are not included in the SPC; data from the literature; and the organisation’s own stability experiments. The data from the SPCs differ from those of the other sources in that they are approved by regulatory authorities. The problems in assessing the shelf lives from the four sources of stability data are examined below.
SPCs and supplier data
An SPC evaluation was conducted at www.produktresume.dk and at the European Medicines Agency website on 5 May 2011 for 13 of the most common cytostatics: 5FU, cyclophosphamide, oxaliplatin, cisplatin, doxorubicin, paclitaxel, vincristine, irinotecan, epirubicin, gemcitabine, docetaxel, carboplatin and cytarabine (150 SPCs in total). It was only possible to identify SPCs with shelf lives of 28 days for doxorubicin, epirubicin, gemcitabine and irinotecan. For many of the other cytotoxics, this is in contrast to findings of shelf lives from literature and also to the data provided by the suppliers. In a literature search of shelf lives for cyclophosphamide, fluorouracil, oxaliplatin and gemcitabine,(1) the maximum shelf lives were found to be:
19 weeks,(2) 112 days,(3) 90 days(4) and 35 days,(5) respectively.
The easiest and most cost-effective way for hospital pharmacies to extend the shelf lives of compounded cytotoxics would be if sufficient data were stated in the SPCs. In order to discuss this possibility, all companies participating in the Danish procurement of cytostatics for 2012 were invited to a discussion. The 14 suppliers were: Hospira, Sandoz, Actavis, Accord Healthcare, Teva, Pfizer, Medac, Fresenius Kabi, Sun Pharmaceuticals, A-Pharm, Eli Lily, Baxter, Sobi and Pierre Fabre. The company delegates ranged from marketing directors to vice presidents of different departments. They represented original manufacturers, generic manufacturers and parallel importers, and covered all of the 13 cytostatics.
These discussions with international representatives of the pharmaceutical companies revealed that uncertainties and non-uniformity in the regulatory processes across Europe are major obstacles to having extended stability information implemented in the SPCs, and that European authorities have different attitudes towards accepting extended shelf lives regardless of the quality of the stability data. Besides this, it is doubtful whether the industry will benefit from stating extended shelf lives in the SPCs, and there is no pressure on the industry to change the short shelf lives. There is no indication that change will come from the industry.
Because of a general need in European hospital pharmacies for extended shelf lives for compounded cytotoxics, many companies are offering supplementary stability data beyond the period stated in the SPC. We know from a small national and an international questionnaire that this is daily practice in Denmark, as well as The Netherlands, Norway, Sweden, Germany, Spain and Switzerland.
The practice of applying supplementary data as documentation for extended shelf lives is questionable because the suppliers do not take responsibility for use of these data but rather refer to the shelf lives stated in the SPCs. The quality of supplementary data varies and it is impossible for the hospital pharmacies to judge these data. For example, some companies have included degradation products in their assessments whereas others have not.
The third source that hospital pharmacies use for evidence for extended shelf lives is the published literature. The general aim for articles concerning the stability of cytotoxics is to report stability data under specified circumstances, such as certain diluents, containers, pH and concentrations, or to report a new method of testing the stability for a certain compound.
If the premise is accepted that literature is knowledge that can be used to extend the shelf lives, this leads to a number of problems. One is the high number of possible variations of test conditions (for example, diluents, concentration, pH) where the hospital pharmacies, in the absence of better evidence, expend much effort in trying to extrapolate literature data to their specific situation, very often without knowing exactly how these differences can effect the stability of the cytotoxic.
Another problem is that the method of analysis used is validated before the results can be trusted. Even though the current ICH guidelines are not meant for/do not cover production of unlicensed drugs, the ICH guideline Q2 (R1): Validating of Analytical Procedures: text and methodology can be applied for the purpose of validating the methods used. The typical validation characteristics listed and defined in this document are: accuracy, repeatability, intermediate precision, specificity, detection limit, quantitation limit, linearity and range.
The validating characteristics often stated in the papers are: accuracy, intra-days variation, referring to repeatability in the ICH guideline, and inter-days variation referring to intermediate precision. The number of concentrations together with number of replicates and relative standard deviation (% RSD) gives the linearity. The detection limit and the quantitation limit are also indirectly given by the standard deviation.
It is rare that all of these validation characteristics are stated in the literature. It would be very difficult to deem one of these characteristics as being more important than the others because they, in common, indicate the reliability of the data. For example, when recovery is not stated, it is unknown whether the reported degradation is due to a recovery below 100%, or if the number of concentrations is not reported, it is uncertain on how many concentrations the calibration curve is constructed, maybe giving a high coefficient of determination (R2) but at a fragile basis.
That does not necessarily mean that the papers found are of poor quality, because papers in a scientific context only have to prove a new concept and be sufficiently detailed that the concept can be recreated at another laboratory.
The problem arises when the hospital pharmacies, in the absence of more robust data, are using the literature as documentation for shelf lives beyond the values stated in the SPCs. In these cases, the reported validating characteristics become extremely important, because they are the only parameters that the results of stability are relying on. There are no opportunities to draw own conclusions or to spot any trend in the experimental data. Using literature in this context makes papers with very sparse information about the characteristics (or lacking information) worthless.
At a more general level, it is questionable whether literature can constitute documentation for extended shelf lives of cytostatics. The pharmaceutical manufacturing operates under very controlled conditions, for example, by being audited by the authorities and following GMP rules. Normally this is not the case when data for a paper are produced, so a gap between GMP documentation, which is the normal for drug documentation, and data from the literature exists.
The lack of firm procedures and control with the compounding of unlicensed products at the hospital pharmacies might be explained by the gradual evolution from extemporaneous reconstitution, where the cytotoxics had a short delivery time from the hospital pharmacies until administration to a specific patient to the current situation, where stock-keeping of ready-to-use cytotoxics is desired.
It might seem that the safest documentation is provided when hospital pharmacies conduct their own stability tests. Here it can be assured that the test condition is in accordance with the conditions at the hospital pharmacy, and that the results are not biased by other interpretations. However, the licensed cytotoxic product is not necessarily a constant entity. Changes may happen at the manufacturer in, for example, production method, active ingredients or excipients, and such changes might influence the shelf lives of the compounded cytotoxics. The authorities, but not the hospital pharmacies, will not receive information when such a change has been made. In the worst scenario, this means that the stability data obtained by the hospital pharmacies could be worthless, maybe even before the test has been finished, owing to unknown changes in the composition or manufacture of the registered product.
When hospital pharmacists assess and assign shelf lives of compounded cytostatics, they combine and evaluate data identified in the SPC, in literature and handbooks, and from information supplied by the drug manufacturer/supplier. This is common practice throughout a number of European countries; however, none of the evaluated data/sources are complete, even if combined, because they do not complement each other. Own stability tests require frequent repetition in order to ensure that the results are updated, to which most of the European hospital pharmacies are not financially leveraged. All the data sources have their pitfalls and drawbacks regarding assessment of extended shelf lives. Therefore, the quality of the data provided is often questionable, and none of these data sources, neither alone nor in combination, can constitute the documentation needed for the hospital pharmacies to store compounded cytostatics for one-to-three months.
The absence of comprehensive information makes it difficult for the hospital pharmacists to take responsibility for the stability of compounded cytostatics on an informed basis. In the future, it is generally not an option for European hospital pharmacies to adhere to the short shelf lives stated in the SPCs, because increasing demand for compounded cytotoxics requires automated technology, improved logistics and stockholding of the compounds.
In order to give compounded cytostatics longer shelf lives, better data are a prerequisite. As this seems to be an international problem, it would be feasible to establish international collaboration or to centralise the resources on a European level in order to find a common solution.
- Generally, the chemical/physical shelf lives of ready-to-use cytotoxics given in the Summary of Product Characteristics are short.
- Nothing indicates that the industry will change this.
- The literature is not meant to be used as documentation for extended chemical/physical shelf lives in the range of one-to-three months.
- Because all hospital pharmacies have issues with extended shelf lives on ready-to-use cytotoxics, it would be appropriate to establish an international collaboration.
- Larsson I, Kart T. Evaluation of sources to document extended shelf lives of compounded cytostatics. J Oncol Pharm Pract;12 Dec 2012 [epub ahead of print].
- Kirk B et al. Chemical stability of cyclophosphamide injection. Br J Parenter Ther 1984;5:90-7.
- Quebbeman EJ et al. Stability of fluorouracil in plastic containers used for continuous infusion at home. Am J Hosp Pharm 1984;41(6):1153-6.
- Junker A et al. Stability of oxaliplatin solution. Ann Pharmacother 2009;43(2):390-1.
- Xu Q et al. Physical and chemical stability of gemcitabine hydrochloride solutions. J Am Pharm Assoc (Wash) 1999;39(4):509-13.