2150 The Effect of Radiation on an Ambulatory Chemotherapy Infusion Pump Mario D. Lacerna, M.D.1 Michael B. Sharpe, Ph.D.2 John M. Robertson, M.D.2 1 Department of Radiation Oncology, Albany Medical College, Albany, New York. 2 Department of Radiation Oncology, William Beaumont Hospital, Royal Oak, Michigan. BACKGROUND. Ambulatory infusion pumps are used to deliver concurrent chemotherapy with pelvic radiation therapy for patients with rectal carcinoma. The pump is worn around the waist and may be exposed to direct as well as scattered radiation, possibly leading to a complete malfunction, requiring a new pump, and/or changes in the pump timing, with clinically significant reductions in chemotherapy administration. METHODS. Two new ambulatory chemotherapy pumps were irradiated using a 6-megavolt linear accelerator. The first pump received gradually increasing doses to determine whether a complete malfunction were possible and the approximate dose. The second pump was irradiated with a single large dose of 20 Gray (Gy) followed by smaller doses of 2 Gy to characterize the dose better. After each dose of radiation was given to both pumps, an internal self-diagnostic test and an independent assessment of the pump timing were performed. RESULTS. The first pump malfunctioned completely at a cumulative dose of 38.6 Gy after receiving an individual dose of 20 Gy. The second pump tolerated the single dose of 20 Gy without difficulty and completely malfunctioned at doses of 40 – 42 Gy. The second pump exhibited a reduction in pump timing by 25% at a cumulative dose of 40 Gy, which resolved spontaneously by approximately 2 hours. CONCLUSIONS. Even if removed from the direct radiation beam, an individual pump could accumulate enough radiation for complete failure during the treatment of fewer than 20 patients. Prior to a complete malfunction, the flow rate of chemotherapy may decrease by 25% for a number of hours without detection. Additional work will be necessary to define further the nature of the reduction in pump timing observed. Cancer 1999;86:2150 –3. © 1999 American Cancer Society. KEYWORDS: ambulatory chemotherapy, infusion pumps, rectal carcinoma, radiation therapy. C Presented at the 39th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, Orlando, Florida, October 19 –23, 1997. Address for reprints: John M. Robertson, M.D., Department of Radiation Oncology, William Beaumont Hospital, 3601 W. Thirteen Mile Road, Royal Oak, MI 48072. Received March 17, 1999; revision received July 6, 1999; accepted July 6, 1999. © 1999 American Cancer Society ontinuous infusional 5-fluorouracil has been shown to be superior to bolus administration when combined with concurrent pelvic radiation therapy for patients with resected rectal carcinoma.1 This is accomplished using an ambulatory pump worn around the waist, where it may be exposed to direct radiation as well as scattered radiation from the pelvic fields (Fig. 1). Even if placed 3–5 cm outside of the pelvic field, the pump would receive approximately 5% of the dose of radiation due to scatter. Because a single ambulatory infusion pump could be used for a large number of patients, even small amounts of scattered radiation due to the treatment of an individual patient could result in a considerable cumulative radiation exposure during the functional life of the pump. Although the manufacturers of ambulatory infusion pumps advise against direct irradiation, to our Effect of Radiation on Chemotherapy Pump/Lacerna et al. 2151 Unlike cardiac pacemakers, complete failure of an ambulatory chemotherapy infusion pump would have little clinical consequence because the damage is obvious and the pump could be replaced easily. However, undetected changes in the pump timing could result in meaningful clinical consequences because the flow rate of chemotherapy would be altered. To determine whether this was possible, we obtained two new ambulatory chemotherapy infusion pumps and exposed them to megavoltage radiation on a linear accelerator. After each dose of radiation, the pumps were tested with both an internal self-diagnostic test and an independent test of the timing mechanism to determine whether the flow rate of chemotherapy could change without detection by the internal selfdiagnostic test. MATERIALS AND METHODS FIGURE 1. Port film of the right lateral field used to irradiate a patient with rectal carcinoma. An ambulatory chemotherapy infusion pump (arrow) was placed outside of the irradiated area but within 5 cm of the superior border. In this position the pump would be expected to receive approximately 5% of the radiation dose. knowledge there has been no published information regarding the effect of radiation. In contrast, the result of irradiating cardiac pacemakers has been well described.2–5 Radiation doses of only 2 Gray (Gy) have been found to cause minor malfunctions, which do not pose a risk to the patient.2 Temporary changes may occur only while the radiation beam is on or persist after the beam is off but reverse with reprogramming of the pacemaker. The dose for complete failure or a permanent change in the pulse rate is quite variable, with some units failing at only 10 Gy whereas others tolerate . 70 Gy without effect.3 Although the precise mechanism of permanent pacemaker malfunction is unknown, the fundamental alteration in the semiconductor is a build-up of trapped positive charge in the oxide layer, which is directly proportional to the total dose of ionizing radiation.5 Thus, the total cumulative dose of radiation delivered to the circuitry is the most important factor in semiconductor failure, without regard to the time period during which the radiation was delivered.5 Two new SIMS Deltec model 5400 ambulatory infusion pumps (SIMS Deltec, Inc., St. Paul, MN) were exposed to radiation delivered by a Philips SL20 linear accelerator (Elekta Oncology Systems, Crawley, UK) using 6-megavolt photons at a distance of 100 cm. The first pump was irradiated with increasing doses of radiation to determine whether a complete malfunction was possible and establish an approximate dose range. The second pump received a large initial dose to the lower end of the range determined with the first pump followed by a series of smaller doses until complete malfunction occurred. After each dose of radiation, the pump was inspected for changes in operation, including inspection of the digital display and confirmation of delivery action. The internal self-diagnostic test was engaged after each exposure by removal and reinsertion of the battery powering the device. Comparing the programmed delivery rate with the internal clock of a separate computer assessed the timing ability of the pump. The pumps operate using a programmable circuit to trigger a series of pistons that deliver a fixed amount of drug per triggered event. The central piston of the pump was attached to a microswitch connected to the serial port of a computer. The computer recorded the number of triggered events and the average time between events. The dose rate for testing was 3 cc/hour, resulting in an event occurring once each minute. The timing ability of the first pump was monitored for 5 hours after each radiation dose. The second pump was monitored for 5 hours after the first dose and for 100 minutes for each subsequent dose. RESULTS Inspection of both pumps after the final dose revealed obvious evidence of complete malfunction. Both units 2152 CANCER November 15, 1999 / Volume 86 / Number 10 TABLE 1 Effect of Radiation on Timing Ability of the First Pump Dose given (Gy) Cumulative dose (Gy) No. of events 0.1 0.5 1 2 5 10 20 0.1 0.6 1.6 3.6 8.6 18.6 38.6 299 299 299 300 299 299 0 Time between events (secs) 59.3 59.3 59.3 59.5 59.4 59.3 DISCUSSION Gy: Gray. TABLE 2 Effect of Radiation on the Timing Ability of the Second Pump Dose given (Gy) Cumulative dose (Gy) No. of events 20 2 2 2 2 2 2 2 2 2 2 0 2 20 22 24 26 28 30 32 34 36 38 40a 40 42 299 99 99 99 99 99 99 99 99 99 74 99 0 imately 1 minute apart. The second pump showed a marked change in the timing ability, with nearly 25% fewer events at a dose of 40 Gy without evidence of a malfunction by both the external appearance or the internal self-diagnostic test. Approximately 2 hours after that dose of radiation was given, the pump was retested without further irradiation and exhibited no abnormalities. The next dose of 2 Gy caused an obvious complete malfunction at a cumulative total of 42 Gy. Time between events (secs) 59.4 59.4 59.4 59.4 59.3 59.3 59.3 59.0 59.3 59.3 79.3 59.3 Gy: Gray. a Cumulative dose at which the timing ability changed. flashed their displays, sounded their audible alarms, and were without response from the keypad. Battery removal and reinsertion would not activate the selfdiagnostic test or change the pump status. Prior to that dose, neither pump exhibited irregularities on inspection or errors after the self-diagnostic test. The amount of radiation given with each dose, the cumulative dose, the number of triggered events, and the average time between events are shown in Table 1 for the first pump tested and in Table 2 for the second pump tested. The first pump exhibited complete malfunction after a cumulative dose of 38.6 Gy with no detectable difference in pump timing. Because the cumulative dose prior to failure was 18.6 Gy, the second pump was irradiated in 1 setting to 20 Gy and monitored for 5 hours. After this, the pump was irradiated at 2 Gy per dose and monitored for 100 minutes, giving an average of 99 triggered events approx- The tested ambulatory chemotherapy infusion pumps failed to operate at a cumulative radiation exposure of 38 – 42 Gy. Although the experience with the first pump suggested a single dose of 20 Gy could have been related to the failure, the findings with the second pump showed that the cumulative dose most likely was more important. Although the first pump completely and obviously malfunctioned after exposure to 20 Gy, the second pump underwent testing for 5 hours after exposure to 20 Gy and was without any evidence of a malfunction. In addition, the second pump tolerated an additional nine exposures without incident. This observation is in agreement with the literature regarding pacemakers showing that the cumulative exposure is the most important factor in semiconductor damage.5 A pump placed outside of the pelvic radiation field but within 3–5 cm of the field will receive approximately 5% of the total dose. Thus, if a patient received 1.8 Gy per day to a total of 50.4 Gy, the pump would receive a dose of 0.09 Gy per day to a total of 2.5 Gy and a complete malfunction would be expected to occur after 16 patients had used the pump. Moving the pump to the level of the patient’s chest would subject it to approximately 0.5% of the total dose and is a reasonable recommendation to reduce the risk of complete malfunction. Disconnecting the pump entirely would eliminate the risk of radiation damage but most likely is impractical for routine purposes. Discovery that the pump was within the radiation field does not appear to require exchanging the device for a new one because in the current study a dose as high as 20 Gy was given in a single dose without incident and multiple fractions of 2 Gy were given without complete malfunction or timing abnormalities. In that instance, the best recommendation most likely would be to move the pump as far as possible from the field and perform a visual verification of the timing mechanism after each radiation dose. The observation of timing abnormalities in the second device is troubling because the malfunction was undetected by the internal self-diagnostic test. If Effect of Radiation on Chemotherapy Pump/Lacerna et al. timing abnormalities occurred as the pump was irradiated from 38 Gy to 42 Gy, and a pump may receive 0.09 Gy per day during radiation treatments, then a total of . 40 treatments are possible during this time period. Thus, a single patient could receive their entire 28 –30 fractions of radiation while the pump is administering an undetectable low flow rate of chemotherapy, potentially lessening their likelihood of obtaining a benefit from adjuvant combined modality treatment. However, the current study also found that the observed timing abnormality was temporary and at the most lasted for 2 hours. This effect may have been due to transient abnormalities within the semiconductor material, similar to the changes that have been speculated to cause reversible pacemaker malfunctions.2,3 In this regard, changing the flow rate for only 2 hours after each radiation fraction may have a negligible effect on clinical outcome. Further research regarding the effects of small doses of radiation between the total cumulative dose of 38 – 42 Gy will be 2153 necessary to determine whether the transient timing change could become permanent and remain undetected prior to a complete malfunction. REFERENCES 1. 2. 3. 4. 5. O’Connell MJ, Martenson JA, Wieand HS, Krook JE, Macdonald JS, Haller DG, et al. Improving adjuvant therapy for rectal cancer by combining protracted-infusion fluorouracil with radiation therapy after curative surgery. N Engl J Med 1994;331:502–7. Last A. Radiotherapy in patients with cardiac pacemakers. Br J Radiol 1998;71:4 –10. Souliman SK, Christie J. Pacemaker failure induced by radiotherapy. PACE 1994;17:270 –3. Marbach JR, Sontag MR, Van Dyk J, Wolbarst AB. Management of radiation oncology patients with implanted cardiac pacemakers: report of AAPM Task Group No. 34. Med Phys 1994;21:85–90. Hardage ML, Marbach JR, Winsor DW. The pacemaker patient in the therapeutic and diagnostic device environment. In: Barold SS, editor. Modern cardiac pacing. New York: Futura, Inc., 1985:857–73.