Estimation of radiation dose received during treatment of in-stent restenosis using ionizing radiation☆

1. Introduction 

Balloon angioplasty is a standard procedure for the treatment of coronary artery disease. Angioplasty involves passing a catheter into a blocked artery and inflating a small balloon to open a stenotic vessel to restore blood flow. In many of these procedures, a biocompatible metal stent is placed at the treated site to reinforce the arterial wall and keep the artery patent. In some cases, angioplasty promotes a wound healing response (neointimal hyperplasia), and within 6 months of angioplasty, approximately 30% of patients suffer from restenosis of the treated coronary artery [1], [2].

An artery has restenosed if the lumen of the treated arterial section narrows to less than 50% of the lumen diameter of adjacent vessel segments. Restenosis may result from a hyperplastic phenomenon such as scar tissue formation. The scar tissue is an attempt of the arterial wall to heal as a result of the microscopic damage induced by angioplasty. In response to the microscopic damage, smooth muscle cells are stimulated to divide and pass through the tear in the endothelial lining of the arterial wall and form scar tissue [2], [3].

Radiation is known to be effective in treating both malignant and benign diseases and hyperplastic conditions such as keloids, pterygium, and heterotopic ossification [4]. Presumably, radiation has the same lethal and inhibitory effects on the rapidly dividing smooth muscle cells that make scar tissue in arterial vessels [3], [4]. The radiation modality used to inhibit coronary artery restenosis is known as intravascular brachytherapy (IVB).

Brachytherapy consists of placing a source of radioactivity in or near the afflicted tissue to give a localized treatment of radiation. In the case of IVB, radioactive sources are inserted in the artery lumen via catheters or stents. IVB has been found useful in inhibiting the formation of new tissue growth in such a way that there is a reduction in smooth muscle cells and thickness of the intimal tissue layer of the artery and, thus, a reduction in scar tissue formation [1], [2], [3], [4].

IVB protocols using either gamma ray or beta particle emitting radioisotopes have been tested and FDA approved. Whether IVB protocols are administered using gamma or beta radiation sources, certain dosimetry criteria must be met. The IVB dosimetry criteria includes (i) single acute fraction of 10–30 Gy to an arterial segment 2–4 cm in length, 2–5 mm diameter, and 0.5–3.0 mm thick; (ii) high dose volume localized to the region of angioplasty, with a minimum dose to normal vessels and myocardium; and (iii) dose rates >2 Gy/min [2], [4]. In addition, the radioactive sources used must have dimensions, stiffness, and flexibility that are compatible for use with angioplasty catheters.

To meet the above criteria, the ideal radioactive source requires a high specific activity, long half-life, and uniform dose distribution over a 2- to 3-mm distance [2]. Clinical trials have been performed using iridium-192, phosphorus-32, and strontium-90/ytrium-90, among others. At present, there is no ideal radioactive source and there are advantages and disadvantages to both the gamma ray and beta particle emitting sources used in the clinical trials, almost with an equal and opposite impact [2], [4]. The gamma source (Ir-192) easily penetrates target tissues with a greater homogeneity to the target volume than the beta sources do (P-32 and Sr-90/Y-90). However, the gamma radiation treatment takes longer to deliver—20 min as opposed to 3–5 min with the beta sources. Radiation safety and shielding issues present apparent problems when dealing with the high-energy gamma source, whereas the beta emitters are easily shielded, but clinically, the dose distribution diminishes rapidly from the source [3], [4].

Aside from thrombosis, no remarkable complications have occurred in treated patients [4]. However, as suggested by many investigators, several years of follow-up will be needed before ruling out late or long-term effects of IVB, such as pericardial disease, myocardial fibrosis, aneurysms, and cancer. With IVB, the body dose absorbed by the patient treated with gamma sources is excessive when compared with vascular brachytherapy procedures using beta emitters, simply by virtue of the radiation characteristics of gamma and beta sources.

There are minimal data available on patient whole-body dose as a result of IVB, and subsequently, little is known about the potential long-term radiation effects, in particular radiation carcinogenesis. The purpose of this research is to estimate, experimentally, the radiation doses received by patients undergoing coronary artery restenosis inhibition using Ir-192.

Historically, radiation has been used to treat malignancies for which, without radiation treatment mortality, is inevitable and the risk of secondary malignancies is well known and documented. Restenosis is not a malignant progression. Furthermore, the restenosis problem may necessitate repeated cardiac procedures, such as angioplasty, IVB, and bypass surgery, which not only carry an invasive risk but risk of additional radiation exposures due to clinical procedures that use fluoroscopy. Patients should be provided with a reasonable estimate of radiation dose and associated risk for an intended treatment to give informed consent.

To express radiation risk, the effective dose equivalent1 (EDE) could be provided with a comparison of the EDE as a magnitude or fraction of background radiation or occupational dose. Without a good estimate of the radiation dose to the whole body, the health risks associated with IVB procedures are not realized.

2. Materials and methods 

3. Results 

Experimental dosimetry data were obtained during actual patient treatments for in-stent restenosis. The patient treatments were performed under FDA-approved clinical investigation procedures. The results and data that follow consider the peripheral radiation dose on the surface of patients from the iridium-192 and the skin dose from fluoroscopy, as recorded by the TLDs. In addition, in-air exposure around the patients from the gamma rays was estimated from ion chamber measurements.

TLDs were placed on the skin surface to obtain a measure of the gamma dose at selected anatomic locations. TLDs were positioned on the left and right shoulders, sternal notch, forehead, and zyphoid process of 48 patient cases from January through July 2000. Of the 48 patients, 2 cases were aborted before the Ir-192 source was deployed, 2 cases were performed to assess placebo effects, and 1 case used a beta emitter, P-32. The remaining 43 cases were performed using Ir-192. The number of iridium seeds (10, 14, 17, 19, or 23) selected coincided with the length of the lesion treated. For the research, the average activity temporarily implanted was 307 mCi, with an average dwell time of 21 min. Patient gender was recorded for 41 of 48 patients, with 75% of the patients being male and 25%, female. The average patient age was 63 years and the average patient weight was 88.1 kg.

The measurements indicate that the average body dose on the skin surface from all TLDs, clinical requirements, and gamma source configurations varies from 0.95 mSv at the head to 27.06 mSv at the sternal notch. The overall dose for each sampling location for the 43 iridium patients is presented in Table 1. The wide range of measured dose for each anatomical location from all patients is attributed to the placement of the TLDs, variation in patient size, and the coronary artery treated, as well the location of the lesion within the coronary artery.

Table 1. Patient dataa Ir-192

	Location	DoseAVG (mSv)	Dose range (mSv)
	Left shoulder	8.12±4.22	2.13–23.28
	Right shoulder	5.83±2.97	1.35–12.69
	Sternal notch	27.06±24.43	6.49–130.0
	Head	0.95±0.69	0.27–3.72
	Zyphoid process	6.87±5.89	0.89–25.31

a The data presents the average±1σ and ranges from all TL elements placed at the indicated location from all patients treated with Ir-192.

In addition to obtaining surface dose measurements with TLDs, in-air radiation exposure measurements using an ion chamber were taken around the patients during irradiation. The ion chamber measurements represent exposure rates from the Ir-192 source and not scattered X-radiation. Measurements were taken from the 43 Ir-192 cases. The average patient exposure rates at one foot from the patient's chest and left and right shoulders were 4.65±2.4 R/hr, 869±353, and 754±417 mR/hr, respectively. There are substantial variations in the recorded measurements made with the ion chamber primarily due to patient size and location of measurements. The typical areas surveyed for exposure are indicated in Fig. 1.

View Large Image Download to PowerPoint

Fig. 1. Exposure rates (mr/h) vs. distance from patient with 382 mCi in the left heart. Not to scale. Victoreen 450 m.

For an IVB procedure, the total fluoroscopy time to include pre- and postgamma irradiation varies significantly from patient to patient. During the actual source dwell time, fluoroscopy is used to check the positioning of the catheter post source insertion and source positioning while dwelling. At most, fluoroscopy time in this portion of the procedure is 2 min, while total fluoroscopy time, on average, may be 18 min.

The placebo cases represent a simulated brachytherapy procedure (for the purposes of clinical validation). The placebo data for this research were especially important because an assessment of the fluoroscopy dose could be made. Fluoroscopy times for the two placebo cases were each less than 2 min. Thus, the TLD measured exposure for the placebo cases represents the exposure received from fluoroscopy only. There were also two cases ultimately aborted before the source was inserted. The TL elements had been placed on the patients upon cue; however, just prior to source insertion, complications developed and the treatment with Ir-192 was aborted. Nonetheless, these aborted cases provide an estimation of the potential dose from fluoroscopy.

For a beta/gamma source comparison, experimental dose data were obtained from one beta case. As expected, the dose distribution at even very short distances from the beta source is significantly lower than the gamma dose. For beta cases, the available activity is 600 mCi or less, with dwell times of only a few minutes. The exposure rate around the patient is minimal. Staff remains in the cardiac catheterization laboratory during treatment with the beta sources.

Table 2 presents the measured fluoroscopy exposure (placebo and aborted cases) and the exposure from the beta case. For beta sources, the dose varied from 0.11 mSv at the head to 0.49 mSv at the sternal notch. The fluoroscopy contribution to the body dose (15-minexposure time) was 0.10 mSv to the head and 2.57 mSv to the sternal notch.

Table 2. Placebo, fluoro, and beta doses (doses in mSv)

	Location	Placebo patient 1	Placebo patient 2	15-min fluoro (aborted case)	3-min fluoro (aborted case)	Beta patient (p-32)
	Left shoulder	0.14	0.20	1.95	0.26	0.25
	Right shoulder	0.16	0.09	0.96	0.65	0.48
	Sternal notch	0.16	0.26	2.57	0.51	0.50
	Head	0.03	0.08	0.10	0.05	0.11
	Zyphoid	0.08	0.11	0.39	0.19	0.24
	Average	0.12	0.15	1.20	0.33	0.32

4. Discussion 

The dose distribution around sealed sources not only depends on distance and source size but on absorption in tissues and scattering by tissues as well. At distances less than 10 cm from the source, the decrease in dose may be described by the inverse square law. At these distances, tissue absorption decreases the dose and scattering increases the dose, resulting in an equal and opposite effect with no net influence on the dose distribution. At distances greater than 10 cm, the dose distribution is affected by absorption and scatter, which are no longer equal and opposite, and deviations from the inverse square law may be as great as 20%, depending on the energy of the source [8].

For the research, TL elements were placed at distances of 10 cm and greater from the source. The research did not intend to determine dose at distances proximal to the treated lesions, but rather, doses distally in an effort to estimate overall body dose on the skin surface. The investigator used the methods describe in Ref. [8] and depth dose charts provided by Best Industries for the iridium sources to calculate the dose at specified distances. The calculated dose is based on a prescribed dose of 14 Gy at 2.0 mm. The following formula from Ref. [8] was used to calculate the dose rate from an assumed point source:

(1)

The commonly used T(r) values are valid for a range of 1−10 cm. Ref. [8] establishes T(r) values for Co-60, Ir-192, and Cs-137 for distances of 10−60 cm.

The calculated doses presented in Table 3 were computed from Eq. (1) for the patient procedures using 10, 14, and 17 seeds of Ir-192. The air kerma strength conversion factor used was 4.03 U mCi −1 (AAPM TG-43). Doses are reported in mGy. The average dose for all seed configurations at 10 cm is 48.20 mGy and 0.08 mGy at 60 cm.

Table 3. Calculated dosea (in mGy) at distances from an Ir-192 IVB procedure

	Distance from the source	10 Seeds	14 Seeds	17 Seeds	DoseAVG(n=31)
	10 cm	35.75	49.11	61.70	48.20±10.37
	15 cm	14.43	19.01	23.86	18.65±3.94
	20 cm	6.65	8.60	10.95	8.55±1.84
	25 m	3.31	4.37	5.45	4.27±0.90
	30 cm	1.75	2.31	2.90	2.26±0.48
	35 m	0.95	1.24	1.57	1.20±0.26
	40 cm	0.53	0.70	0.87	0.68±0.15
	45 cm	0.30	0.40	0.49	0.39±0.08
	50 cm	0.17	0.22	0.28	0.22±0.05
	55 cm	0.10	0.13	0.17	0.13±0.03
	60 cm	0.06	0.08	0.10	0.08±0.02

a Dose estimates for distances 10–60 cm were computed for patients who were treated with 10, 14, and 17 seeds. The computed doses for each distance were then averaged. A dose of 14 Gy is assumed to be delivered at 2 mm.

TL elements were placed on the sternal notch, left and right shoulders, zyphoid process, and forehead. In Table 4, the placement location of the patient TL elements and the average dose measured from patient data for each location are noted. The listed distances from the heart in Table 4 are presumed and represent a best estimate by the author.

Table 4. Measured and calculated dose at distance

	Anatomical location	Distance from the heart (cm)	Measured dose (mSv)	Calculated dose (mGy)	Ratio (measured to calculated)
	Left shoulder	20	8.12	8.55	0.950
	Right shoulder	25	5.83	4.27	1.365
	Sternal notch	10	27.06	48.20	0.561
	Head	35	0.95	1.20	0.792
	Zyphoid process	15	6.87	18.65	0.368

Ratios for measured and calculated doses are presented in Table 4. The measured doses for the shoulders and head show adequate agreement with the calculated doses for the corresponding distances—an agreement within ±37%. Although the zyphoid and sternal notch deviated by as much as 44% and 63%, respectively, the calculated doses for these sampling distances fall within the observed range for the measured skin surface doses of these locations, as presented in Table 1.

From the estimated doses by calculation, the overall average (data for 10–60 cm) is 7.59 mGy. The overall average dose, as measured on the skin surface by TLDs considering all anatomical locations, clinical requirements, and gamma source configurations, is 9.73 mSv. Ref. [3] reports that the effective dose using Monte Carlo code EGS4 from 20 Gy at 2 mm is 14.0 mSv. Theoretically, the effective dose from 10 Gy at 2 mm is 0.70 mSv. Interpolating the effective dose for 14 Gy at 2 mm yields and estimated effective dose of 9.80 mSv. Therefore, dose estimates, whether calculated or measured, present an overall body dose of less than 10 mSv.

4.1. Comparison of tissue doses from cardiac fluoroscopic procedures 

Biological effects for which the probability of the effect increases with dose without a threshold are known as stochastic effects. In radiation protection, carcinogenesis is the stochastic effect of concern. Restenosis is not a malignant disease. However, to treat restenosis, additional procedures including repeated angioplasty and bypass surgery may be necessary. These procedures carry a significant invasive risk and give rise to additional radiation exposure from fluoroscopy. Justification of using radiation to prevent restenosis as with IVB warrants a comparison of intracoronary procedures using radiation and dose received from fluoroscopic X-rays during angioplasty.

Interventional X-ray procedures demand image quality, long fluoroscopy times, and many images to adequately visualize and access the vascular system. Therefore, patient doses in interventional radiology have the potential to be high and deterministic effects, such as skin burns, are a distinct possibility. Patient doses from fluoroscopy are typically reported as the entrance skin exposure (ESE). ESE is used to evaluate the risk for deterministic effects as well as an effective dose to quantify the probability of stochastic effects. For therapeutic cardiac procedures, such as angioplasty, the skin dose may range from 1.5 to 3.0 Gy and the effective dose, 5.0 to 13.0 mSv [3], [9], [10]. From the Handbook of Selected Tissue Doses for Fluoroscopic and Cineangiographic Examination of the Coronary Arteries, the FDA makes the estimates presented in Table 5.

Table 5. Tissue doses from fluoroscopic and cineangiographic exams

	Tissue	Tissue dose (mGy)
	Entrance skin exposure	870
	Thyroid	1.3
	Active bone marrow	32
	Esophagus	88
	Lung	150
	Pancreas	36

Adapted from Appendix F, Table F-1 of Ref. [11].

From the experiment, the average total procedure fluoroscopy time was 18 min. The maximum fluoroscopic output permitted by federal regulations is 10 R/min. The output posted on the X-ray machine used for cardiac visualization was 3 R/min. Therefore, the average ESE for the patient cases from this experiment is 54 R. For doses greater than 100 R, the patient is at risk for a radiation skin burn [11].

The anatomical location for the body tissues in Table 5 correlates with the dosimeter locations used for patient TL placement in the experiment. (Such as, the left and right shoulders/clavicles are areas of active bone marrow. The thyroid and esophagus are in close proximity to the sternal notch, as well as the lung and pancreas to the zyphoid.) There is some correlation with the fluoroscopy tissue dose reported by the FDA and the upper range of the measured dose data, such as the lung at 150 mGy and the sternal notch at 130 mSv. The calculated and measured dose data and Monte Carlo estimates from Ref. [3] indicate that the brachytherapy procedure yields an effective dose (9.80 mSv) in the range observed for the interventional procedure (angioplasty) using fluoroscopy (5−13 mSv).

The annual occupational limit on radiation exposure is based on the concept of EDE. The EDE concept employs risk coefficients to estimate cancer risks from radiation exposure. The annual limit of EDE for occupational workers is 0.05 Sv (5 rem), and organ doses are limited to 0.5 Sv (50 rem). Therefore, the occupational limit could be interpreted to indicate that a radiation worker who receives an EDE of less than 0.05 Sv has no greater chance of suffering from stochastic effects than do workers who are not exposed to ionizing radiation. Because radiotherapy patients are not radiation workers, they are excluded from the radiation dose limits. However, the concept of EDE is applicable when assessing excess risk to patients who are undergoing clinical procedures involving radiation.

From the experiment, the overall average dose by physical measurement was 9.73 mSv. The estimated dose for all dosimeter locations by calculation was 7.59 mGy, and the EDE presumed by Monte Carlo methods is 9.80 mSv. The additional theoretical fatal cancer risk from IVB can be taken as the measured or calculated doses times 0.4×10−4/Sv. Thus, the total theoretical risk is on the order of 20.04%. Upon comparison with limits for occupational exposure (0.05 Sv) and exposure from background (3.60 mSv), the effective dose from IVB poses an acceptable risk.

However, for the patient population at hand, additional dose accumulates from prior and subsequent procedures involving radiation and therefore increases the effective dose and increases the cancer risk in an additive fashion.

5. Conclusion 

The results suggest that skin surface exposures from gamma sources used in IVB pose acceptable risks considering the medical benefits of the procedures. High energy gamma emitters, however, do present radiation protection challenges, and these are important from a health physics perspective and are important, considering that staff must be able to closely observe and communicate with patients undergoing IVB.

In clinical applications, the effective dose concept more or less provides a mechanism of dose expression for comparing various treatments that use radiation as a predictor of future disease or outcome. The latency period for many cancers is about 20 years [12]. The increase of secondary cancers arising from IVB can reasonably be limited by the natural life expectancy of the patient population receiving IVB.

Medical exposure, in essence, is subject to dose limitation in that unnecessary exposure should be avoided and necessary exposures should be justified in terms of benefits that would not otherwise have been received. The dose data presented in this research may serve to improve the risk–benefit discussion between the clinician and patient, minimize misperceptions, and help assure patient/research subject protection when obtaining informed consent.