Intraoperative detection of gamma emissions using K-alpha x-ray fluorescence

Today, hand-held gamma detection probes are commonplace and are utilized widely for the intraoperative detection of various radiopharmaceuticals during cancer surgery [1]. The basic technical principles behind the functionality of these devices [1–6], as well as their many clinical applications [1], have been well characterized previously. Nevertheless, new insights into their functionality can still be gained, which in turn can lead to further improvement in the currently available technology.

The formation of gamma rays can be the result of primary gamma emissions from a gamma-emitting radionuclide or can be the direct by-product of positron–electron annihilations from a positron-emitting radionuclide (i.e., production of resultant high-energy 511 keV gamma photons from the annihilation of a positron colliding with an electron). Depending upon the radionuclide, there can be multiple gamma emissions of various energy levels, as well as varying degrees of both gamma emissions and positron emissions. Likewise, the higher the energy of a given gamma emission, the higher the probability that the gamma photons will pass directly through the detection crystal and remain undetected. Furthermore, lower energy broadband x-ray emissions can be generated from Compton scattering in biological tissues or from charge trapping in the detection crystal. In the end, the resultant broad spectrum of emission energies from a given radionuclide can make the electrical signal detected by the gamma-detection probe system very difficult to interpret [3]. Therefore, the exact specifications of any given gamma-detection probe system may need to vary greatly in order to optimize the intraoperative detection of various radiopharmaceuticals that can be utilized during cancer surgery.

Over the past decade, there has been increasing worldwide interest in the use of positron-emitting radionuclides, such as fluorine-18 (¹⁸F) and iodine-124 (¹²⁴I), for the intraoperative detection of known and occult disease during cancer surgery [1,7–33]. While the decay pattern for ¹⁸F is rather straight forward (with 97% representing positron emissions), the decay pattern for ¹²⁴I is rather complex (with 23% representing positron emissions and 77% representing primary high-energy gamma emissions, of which 61% are at 603 keV, 10% are at 723 keV and 10% are at 1691 keV) [1,34,35]. A significant portion of the clinical investigation into the use of positron-emitting radionuclides in radioguided surgery has been conducted at The Ohio State University (OH, USA) [1,7,8,17–19,22–27,30,31]. In this regard, various gamma-detection probe systems of differing configurations and designs have been utilized for detecting these positron-emitting radionuclides. However, our intrinsic understanding of exactly how these gamma-detection probe systems detect positron-emitting radionuclides remains suspect. Up to this point, it has been tacitly assumed that the detected signal from these positron-emitting radionuclides was the result of the production of high-energy 511 keV gamma photons from positron–electron annihilation events in the case of ¹⁸F and the result of both primary high-energy gamma emissions and the production of high-energy 511 keV gamma photons from positron–electron annihilation events in the case of ¹²⁴I. Yet, this assumption about the detected signal from these positron-emitting radionuclides may not be entirely correct.

In 2007, Dr Marlin O Thurston, Professor Emeritus of Electrical Engineering at The Ohio State University and a scientist who had historically played an instrumental role in the development of gamma-detection probe technology used in the field of radioguided surgery, made a very intriguing observation [Thurston M, Pers. Comm., July 27, 2007]. In a set of preliminary phantom experiments that he conducted to evaluate 511 keV energy detection efficiencies, Dr Thurston noted that one previously commercially available gamma-detection probe, having a cadmium–zinc–telluride (CZT) crystal (15 mm in diameter and 1.5 mm in thickness), 0.025 inches of lead side-shielding and 0.75 inches of rear shielding, appreciably detected 511 keV high-energy gamma emissions. This increased detection rate for 511 keV high-energy gamma emissions was remarkable in light of the fact that the CZT crystal within this particular gamma-detection probe was of a thickness that would theoretically interact with only a negligible portion of the 511 keV high-energy gamma emissions. Spectral analysis of this particular gamma-detection probe revealed significant lower energy photon emissions in the range between 60 and 100 keV. In contrast to the high-energy 511 keV photons, these lower energy photon emissions had a much higher probability of interacting with and being detected by the CZT crystal. Based upon this intriguing observation, Dr Thurston hypothesized that these lower energy photon emissions were, in fact, secondary K-alpha (K_α) emissions from the lead shielding of this particular gamma-detection probe. In other words, the 511 keV high-energy gamma emission source was providing sufficient energy to cause the lead shielding to exhibit K_α x-ray fluorescence. In this instance, the K_α emissions are typically generated when a primary photon from the source radionuclide resultantly ejects an electron in the K shell of an atom within the lead shielding of the housing of the gamma-detection probe by exceeding the binding energy of that orbital. An electron from the L shell subsequently fills the K shell vacancy, resulting in emission of a secondary photon that leads to K_α x-ray fluorescence. As a result, Dr Thurston concluded that 511 keV high-energy gamma emissions could be measured indirectly, using a relatively thin CZT crystal in combination with a metallic faceplate, in this case lead, placed between the CZT crystal and the gamma emission source. Unfortunately, prior to the complete realization of this concept for intraoperative detection of gamma emissions using K_α x-ray fluorescence, Dr Thurston passed away on 18 September 2007. Yet a legacy to this K_α x-ray fluorescence concept has been articulated recently within several related methodology patents [101,102].

While K_α x-ray fluorescence is certainly not a new concept to the field of medical diagnostics [36,37], the application of this technology to the design of hand-held probes could allow for the intraoperative detection of various radionuclides exhibiting a broad range of gamma emission energies by simply detecting the resultant lower energy secondary K_α emissions that occur from a lead faceplate that is placed in the pathway of the incoming primary gamma emissions. Such a method of detection would be especially useful for higher-energy gamma-emitting radionuclides and would potentially improve the overall detection efficiency of all higher-energy gamma emissions. By applying the K_α x-ray fluorescence methodology to a hand-held probe design, it should be possible to produce a resultant intraoperative detection device of a much reduced size compared with the currently commercially available, large, bulky, high-energy gamma-detection probes.

Any primary gamma emissions with energies greater than the threshold electron binding energy for lead (i.e., 88 keV) are capable of producing secondary K_α emission from the lead faceplate of the K_α x-ray fluorescence probe design. The percent yield of K_α emissions diminishes as the excitation energy increases well beyond the threshold electron-binding energy for lead. Although this means that the detection efficiency varies with the energy of the gamma emission, the gamma events that can be detected by using the crystal–lead faceplate combination are significantly improved compared with using the crystal alone. This is particularly true at higher gamma emission energies where a substantial increase in CZT crystal thickness and applied electrical field would be necessary to detect even a small fraction of the primary gamma emissions. By translating the gamma events from a higher energy to a fixed low energy, a thin CZT crystal could be used to capture the 73- and 75-keV K_α x-ray fluorescence emitted by the lead faceplate that is placed between the CZT crystal and the gamma emission source. Moreover, any gamma emissions in excess of the 88 keV could be detected with such a device. This would include the detection of ^99mTc (140 keV primary gamma emissions), the radionuclide that is currently used exclusively for the performance of sentinel lymph node biopsy procedures in the clinical staging of all breast cancers and melanomas. Additionally, K_α x-ray fluorescence from the lead faceplate could be applied to the detection of other radionuclides that can potentially be utilized in radioguided surgery, such as ⁶⁷Ga (93 keV, 184 keV and 300 keV primary gamma emissions), ¹²³I (159 keV primary gamma emissions), ¹¹¹In (171 keV and 247 keV primary gamma emissions), ¹⁸F (511 keV gamma emissions from proton–electron annihilations) and ¹²⁴I (511 keV gamma emissions from proton–electron annihilations and 603 keV primary gamma emissions) [1,201].

As a closing point of interest, some consideration should be given to the fact that such a K_α x-ray fluorescence probe design would not allow for differentiation between primary gamma emissions and other gamma emissions of reduced energy that are the result of tissue scattering or crystal irregularities. These factors can reduce the spatial resolution of such a probe and can translate into greater difficulty in specifically localizing the gamma emitting source. However, any loss in spatial resolution due to the K_α methodology could be compensated for by applying real-time statistical analysis to the gamma-emitting events.

In conclusion, the K_α x-ray fluorescence concept represents an innovative methodology for the detection of radionuclides that have gamma emission energy greater than approximately 88 keV, making it attractive for the detection of ^99mTc, ⁶⁷Ga, ¹²³I, ¹¹¹In, ¹⁸F and ¹²⁴I, and thus making it a nearly universal radionuclide detection probe.

Financial & competing interests disclosure

Edward W Martin Jr has equity in Actis, Ltd. Vish V Subramaniam has a related patent (US Publication No. US-2009-0208417-A1). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.