Just where does the isolator fit in?

08/01/2003

Just where does the isolator fit in?

By Johan VandenbrouckE, Ph.D and Hank Rahe

Editor's Note: There's considerable debate in the compounding pharmacy segment concerning the proper design of a safe and manageable contamination control system incorporating isolators. In this month's Unfiltered, we decided to share a recent e-mail dialogue that we facilitated through our editorial offices between Dr. Johan Vandenbroucke, senior pharmacist at the University Hospital Ghent in Ghent, Belgium, and CleanRooms columnist and Editorial Advisory Board member, Hank Rahe. If you wish to continue this dialogue, please e-mail [email protected]

Where does the isolator fit in?

In the recent article by Hank Rahe, published in the Life Science section of CleanRooms magazine, the author claims that secondary containment is a must when designing a contamination control system [March 2003, pg. 14].

With secondary containment, Mr. Rahe puts barrier isolators on a first plan as a back-up when the first level of protection fails. I question that attitude, speaking from my experience as a senior hospital pharmacist responsible in the cytotoxic reconstitution center of a university hospital in Belgium, and from the many international contacts I have with colleagues working in the same field.

First of all, isolators do not prevent contamination. They retain, in a certain amount and time, the contamination that occurs during handling of hazardous drugs within the isolators. In most cases, people working in an isolator have a false sense of security and they work with open systems (spikes, needle and syringes).

Using open systems leads to the introduction of contamination; especially when admitting that contamination begins the moment a single droplet, or aerosol, or vapor of a hazardous drug comes into the environment.

Recent studies in the U.K. (Davy Collin at the European Association of Hospital Pharmacists Congress in Vienna, 2002) and in France (Favier, Rull, Bertucat, J Pharm Clin, 2001) clearly demonstrate that the containment of isolators is doubtful. In the U.K. study, negative pressure isolators were compared with traditional biological safety cabinets (BSC). In the French study, positive pressure isolators were compared to BSC. Contamination was tested via wipe-sampling in and around the isolator/BSC during both studies.

The results are remarkable. In each setting where the isolators are used, a higher contamination was found in and outside the isolator compared to the setting where people are working in BSC. Davy Collin made the following statement: People are the weak spot in the system, and isolators should be considered as a perfect engineering tool at the level of personal protection measure—but not more than that.

The European directive in protecting people at work against cancerogenic agents is also very clear. One should follow the following hierarchic order to achieve prevention:

• Level 1 is to replace the product by a less or non-toxic one. (If Level 1 is not feasible, then use Level 2.)
• Level 2 is the use of closed systems to prevent the occurrence of any form of contamination. (If Levels 1 and 2 are not achieved or possible, go to Level 3.)
• Level 3 is the use of local and general ventilation measures.

(If Levels 1, 2 and 3 are not achieved or possible, go to Level 4.)

• Level 4 is the use of personal protection measures.

My opinion is that isolators and BSCs fit in the Level 3 category, under the conditions that they are well-engineered, controlled and validated. Special attention should be given to the exhaust of the isolator/BSC. There is proof that hazardous drugs like cytotoxic agents can and will evaporate even at room temperature. HEPA or absolute filters do not retain the vapors of these drugs, so an external exhaust with booster ventilation to provide a negative pressure in the ducts is essential.

If contamination occurs in the protected working zone (isolator/BSC), it will come out together with the prepared infusion bags or syringes, and with the waste. Some isolators have the ability to seal the waste and the prepared products in plastic bags to overcome that issue. It is only a temporary solution, because the nurses need to unpack it before they can administer it to the patient. At the University of Bath in the U.K., Prof. Graham Sewell has tested the infusion bags that leave the negative pressure isolator, and found that more then 33 percent of them were contaminated—a contamination control system and policy cannot end at the doors of the preparation room.

My second remark is about the testing and validation of the systems. I agree with the author that there is a need for testing, and that testing is not an easy task to do. Differences in sampling method and protocol, in materials used, in analytical method and detection limits make it difficult to compare different studies.

On the other hand, within a single study using the same sampling and analytical method, a difference of a factor 5 to 10 can be considered as valuable and proven. An even more reliable way of testing the contamination is the analysis of urine samples from workers, because differences in sampling are excluded and it gives you direct data of the product in the body of the personnel. Until now, not a single study using urine analysis has been conducted in isolator settings.

Our hospital has performed a study using wipe and urine samples to examine the possible difference between the use of an open (needle/syringe) system in combination with BSC, and the use of a Swedish closed device (PhaSeal) in combination with BSC. The Swedish system meets the demands of Level 2 of the European directive.

The result of the urine and wipe samples clearly indicates a difference in both amount and frequency of contamination in favor of the closed system. (J Oncol Pharm Practice, Vol. 6, No.4, 2001; pp. 146-152).The results of our study have been confirmed by other studies done in Sweden, U.S. and Canada.

It is clear that the use of closed systems prevents contamination and thus reduces existing contamination levels. As a result of that, a secondary containment system is no longer a must except for the microbiological prevention of the product. A combination of a closed system and an isolator could be the perfect marriage between safety and GMP.

Johan Vandenbroucke, Ph.D
Senior Pharmacist Production
University Hospital Ghent, Ghent, Belgium

Rahe in response...

Dear Dr. Vandenbroucke:

I read with great interest your prospective on contamination control. I have written a number of chapters in books concerning contamination control and have been a proponent of what I call the hierarchy engineering controls for containment.

The hierarchy recognizes a properly designed "closed system" as the top of the pyramid of controls. The key is that it must be "properly designed" because any system must pass a "risk assessment" that looks at the consequence of failure.

Most closed systems have secondary containment features because each has a component of high risk. Single barrier systems, such as the PhaSeal, have a very high consequence of failure from a defective vial or the failure of the device itself. Failure can result in direct exposure to a complete vial of hazardous drug. Partial failure (improperly seated connection) can be even a higher risk since there is no indication of failure and no secondary containment.

Isolators offer a secondary containment feature in case of failure and have been validated to very low exposure levels. Like any tool, proper use and training are critical to achieving acceptable results. A U.S.-based barrier isolator company is coming out with secondary containment for the product from the isolator to the delivery point, and will not require direct contact with the product bag by the individual administrating the product to the patient.

Methods of sample collection and validated methods used in most hospital studies seem to be lacking. A qualified industrial hygienist who understands design of experiments and validation of recovery from surface materials has done few of the studies. But exposure in terms of eight-hour level or short-term levels seems to be non-existent.

Most studies seem to focus only on the presence of hazardous drugs without the critical link of how much per period of time. If this information was available, then it could be compared to product exposure limits to determine if the activity is truly safe.

The Life Sciences Supplement published by CleanRooms in February 2002 carried three articles concerning containment, risk assessment and contamination control using isolators, and may be interesting reading. Of interest in the barrier isolator article is information on the impact of vapors when passing through a HEPA filter. The HEPA filter seems to change the state of the gas back to solid, which follows the first law of thermodynamics concerning the states of matter, and may be a source of the contamination found with non-vented BL-2s.

—Hank Rahe