Accreditation and quality assurance for Radiation Therapy Oncology Group: Multicenter clinical trials using Stereotactic Body Radiation Therapy in lung cancer

Abstract

Starting in 2002, the Radiation Therapy Oncology Group in North America began the process of developing multicenter prospective trials in lung cancer using Stereotactic Body Radiation Therapy (SBRT). Much of the work was based on the prospective single institution trials from Indiana University that had been presented and published. In late 2004, RTOG 0236 using SBRT for medically inoperable patients with clinical stage I non-small cell lung cancer (NSCLC) was activated for accrual. Prior to activation, representatives from the Lung, Image-Guided Therapy, Physics, and Radiobiology Committees met on regular occasions to design the multicenter study and quality assurance measures. SBRT is not a black box, and the essence of the therapy had to be distilled via guidelines. Issues related to patient selection, method of dosimetry construction, equipment requirements, motion assessments and control, site accreditation, data exchange, and follow-up policies were worked out by compromise and consensus. RTOG 0236 has nearly completed its accrual. The Lung Committee has initiated the development of several other trials, each building on the last, to investigate the therapy in central tumors, in combinations with systemic therapy, in operable patients, and in lung metastases patients. The guidelines developed for RTOG 0236 will be refined to take advantage of more modern innovations including heterogeneity corrections and intensity modulation when appropriate. The development of RTOG 0618 using SBRT in operable patients with early stage NSCLC is a testament to both the enthusiasm from already published works and prospective multicenter clinical testing using SBRT techniques.

While single institution trials are often the forum for the innovation, piloting, and refining of novel cancer therapies, promising treatments should ultimately be tested in multi-center trials. Multi-center trials are typically performed by government sponsored or investigator-independent cooperative groups specializing in quantifying outcomes fairly. Bias involved in selection of patients, the conduct of treatment, the assessment of control and toxicity, and the determination of outcome is best controlled in multi-center cooperative group trials. The multi-center testing of therapy avoids conflict of interest associated with the single institution experience. If the novel therapy is important for future implementation on a wide scale, it must be documented that physicians and staff outside the originating center can be trained to perform the therapy. It must be demonstrated that the results are consistently reproducible. In the end, the single or limited institution experience creates the novel therapy while the multi-center (and particularly the cooperative group) experience finds the proper place for the therapy in the treatment arsenal. The ultimate cooperative group trial, the large phase III randomized trial, often redefines the standard of care.

The task of carrying out specialized or intricate novel therapy in a cooperative group is daunting. The first critical aspect of creating a multi-center trial is defining the therapy. The definition is critical, as often the pilot single institution experience occurred at several centers simultaneously and independently. Each center may have a different definition of what is critical to carrying out the treatment. The essence of the therapy and its conduct may not have been fully articulated by the pioneering clinicians. There may be lack of consensus. Nonetheless, in order to design a multi-center trial, there must be a clear definition of the most critical aspects of therapy. The decisions involved in formulating the concrete definition involve consulting, clarifying, testing, considering practicality, considering reproducibility, and ultimately compromise.

A novel cancer therapy, officially called stereotactic body radiation therapy (SBRT) in the United States and going by other names including extracranial radiosurgery, stereotactic radioablation, and other combinations of these words, has become popular do to impressive initial clinical results. SBRT originated from a direct translation of rigid immobilization intracranial radiosugery. Investigators from the University of Arizona performed the first treatments with stereotactic targeting and high dose per fraction delivery for tumors in the spine [1]. Independent translations of intracranial technology were carried out in Sweden at the Karolinska Hospital for treating liver tumors [2], [3]. Simultaneously, workers in Japan focused on treating lung tumors with focused techniques [4] and motion control [5]. Subsequent mostly single institution prospective testing in Germany [6], [7], the United States [8], [9], and Japan [10], created the patient base for a series of publications on the subject. International meetings have been held to exchange ideas. SBRT therapy became more uniform in its conduct. Outcome reports have shown SBRT to be a tolerable, efficacious, and promising non-invasive therapy. In other words, SBRT was ready for multi-center cooperative group testing

What is SBRT?

In 2004 after several years of planning, the lung committee of the Radiation Therapy Oncology Group (RTOG) finalized plans to carry out a multicenter trial of SBRT in patients with medically inoperable non-small cell lung cancer (NSCLC). As this was the first cooperative group multicenter trial of its kind, the first step was to define the therapy. Previously, a working group from the American College of Radiology and American Society for Therapeutic Radiology and Oncology had formulated guidelines for the conduct of SBRT [11]. The guidelines described the following essential components collectively unique to its conduct:

1. Secure immobilization avoiding patient movement for the typical long treatment sessions;

2. Accurate repositioning of the patient from planning sessions to each of the treatment sessions;

3. Proper accounting of inherent internal organ motion including breathing motion consistently between planning and treatment;

4. Construction of dose distributions confidently covering tumor and yet falling off very rapidly to surrounding normal tissues. The dosimetry must be extremely conformal in relation to the prescription isodose line compared to the target outline but may allow very heterogeneous target dose ranges;

5. Registration of the patient's anatomy, constructed dosimetry, and treatment delivery to a 3-D coordinate system as referenced to fiducials. Fiducials are “markers” whose position can be confidently correlated both to the tumor target and the treatment delivery device. A “stereotactic” treatment is one directed by such fiducial references;

6. Biologically potent dose prescriptions using a few (i.e., 1–5) fractions of very high dose (e.g., generally a minimum of 6 Gy per fraction but often as high as 20–30 Gy per fraction).

This therapy is used to treat well demarcated visible gross disease up to 5–7 cm in dimension. It is not used for prophylactic (adjuvant) treatment as the intent is to totally disrupt clonogenicity and likely disrupt all cellular functioning of the target tissues (i.e., the definition of an ablative therapy).

Effectively, SBRT is a treatment that can ablate or totally destroy that to which it is aimed. Such a treatment, properly directed would constitute a most potent form of cancer therapy. In turn, if misdirected or used too liberally, SBRT could lead to debilitating toxicity. Whether the potent SBRT dose can truly be placed primarily within tumor using stereotactic targeting, motion control, ideal immobilization and specialized dosimetry techniques remains to be proven in all clinical circumstances. At any rate, SBRT is not similar to conventionally fractionated radiation therapy (CFRT) in its conduct, toxicity, or ability to control cancer. As such, it was early recognized by the RTOG Lung Cancer Committee that typical quality assurance (QA) and conduct guidelines for CFRT would fall short for carrying out a prospective multicenter trial with SBRT. Fortunately, within the RTOG there exist mechanisms for finding consensus in the development of guidelines for novel therapies. In turn, a committee was formed to develop the protocol using expertise from the following committees and subcontractors: Lung Committee, Image-Guided Therapy Committee, Physics Committee, Translational Science Committee, Radiobiology Committee, Radiological Physics Center, and Advanced Technology Consortium. While many aspects of the therapy and outcomes were considered, this paper will focus on the accreditation and dosimetry aspects of the protocol that came to be called RTOG 0236.

Dose selection

In RTOG 0236, the prescription dose was based on single institution data from Indiana University treating patients with medically inoperable stage I non-small cell lung cancer. That group performed a prospective phase I dose escalation trial where uniformly selected patients were treated with SBRT [8], [12]. Starting at a low dose unlikely to cause excessive toxicity (8 Gy per fraction times 3 fractions — total 24 Gy) the group increased the dose per fraction in successive cohorts. As such, the only treatment variable in this prospective trial was prescription dose. In the end, a dose of 20 Gy per fraction times 3 fractions — total 60 Gy — was found to be tolerable in all patient groups. The RTOG Lung Committee chose to use this dose for RTOG 0236 because it constituted a very potent therapy with prospective data indicating acceptable toxicity. RTOG 0236 then would take this therapy to the next step in a phase II trial.

Target selection

CFRT often treats large volumes of normal tissue to the same or close to the same dose as the gross tumor. While this may be intentional as a method to address areas at risk for microscopic tumor involvement (i.e., an adjuvant therapy), there is no place for such prophylactic treatment in SBRT. Indeed the Indiana trial that formed the basis for dose selection in RTOG 0236 treated very small volumes relative to CFRT. The Gross Tumor Volume (GTV), defined only as the solid abnormality on computed tomography, was not expanded in the traditional fashion with CFRT to form a Clinical Target Volume, CTV. Instead, by nature of the disease treated it was assumed that any microscopic tentacles coming from the gross tumor would be eradicated by the dose falloff, also called penumbra. The GTV was expanded to a Planning Treatment Volume (PTV) which accounts for set-up variability and tumor motion. In RTOG 0236, the expansion from GTV was set by protocol to be no more than 0.5 cm in the axial plane (left, right, anterior, and posterior), and 1.0 cm in the craniocaudal plane (superior and inferior). Of course, some tumors particularly close to the diaphragm would move more than these allowances with free breathing. In response, the protocol required that institutions demonstrate a feasible method of accountability for motion that would allow such small margins without missing the tumor target.

Motion control

No specific motion control methodology, technology, or application was required by the protocol. To do so would have penalized or hindered innovation into newer methods. Furthermore, the simple purchase by a center of a technology used by others to appropriately control motion was not enough to allow accreditation. Instead, each center was required to demonstrate that they could implement the technology in their own patients in a manner that would allow the small margins dictated in RTOG 0236. It was required that centers applying for accreditation demonstrate that they can treat patients with tumors similar to those enrolled in RTOG 0236 (e.g., in the lung or liver) with acceptable motion accounting or control. The accreditation process required the institutions to submit data to the study committee physics Co-Investigator (Dr. James Galvin) who would assess whether it was likely that RTOG 0236 criteria could be achieved.

Most accredited institutions have employed abdominal compression as a method to dampen the amplitude of inherent tumor motion. Others have used respiratory gating methods and breath-hold techniques coupled to the output of the linear accelerator. Still other methods of tumor motion control would be considered by the committee, such as tumor tracking or chasing. Generally, the immobilization and the motion control methods are combined in the submission of data relating to inter- and intra-fraction motion.

Dosimetry

Dosimetry is the purposeful construction of dose (energy deposition per unit mass) within a patient as a rendered simulation for subsequent therapy. For SBRT, this dose must be compact and minimize exposure to eloquent normal tissues. This qualitative description of appropriate dosimetry would not suffice, however, for guidelines in a clinical trial. Instead, the study committee attempted to characterize aspects of dosimetry from plans that were effective and well tolerated. It is straightforward to convey that the high prescription dose affords tumor control and that dose outside the target potentially results in toxicity. In the end, the committee decided to create dosimetry planning goals that correlate to three separate dose levels. Both the magnitude of dose and the spatial location of that dose were scrutinized.

High dose constraints

The prescription dose was defined at the margin of the PTV. Two criteria were required: 1. 95% of the PTV should receive the prescription dose (60 Gy), and 2. 99% of the PTV should receive 90% of the prescription dose (54 Gy). These criteria insure that the target is fully treated as intended to allow the best chance of tumor control. The prescription itself was defined to an isodose line. The normalization point for this isodose line was required to be at the center of mass of the PTV. It was not a requirement to normalize to the isocenter of the beam arrangement as is common with CFRT. This followed from the fact that some SBRT platforms do not use a conventional isocenter within the PTV (e.g., Tomotherapy and Peacock).

Dose from SBRT treatment is delivered to normal tissues by energy release via attenuation and scatter. More so than scatter dose, primary attenuation dose can be controlled by blocking and modulation. Block margins should be as small as possible to avoid delivering the primary fluence that would deliver this attenuation dose. Margins are recommended to be of the order of the PTV projection itself or only modestly larger. It is acceptable, in some circumstances, to make the margins even smaller than the target itself [13], [14]. However, with such small margins relative to the size of the target, target dose homogeneity is neither necessary nor possible. The protocol discourages large margins and to reinforce this it requires that the prescription line must range from 60–90%. Therefore, the plan will inherently have a “hot spot” (dose greater than the prescription) ranging from 67–100 Gy.

The location of this “hot spot” is critical as it could either aid in tumor control or cause tremendous toxicity. Thus, a constraint insists that this “hot spot” be located inside the PTV target by requiring that any dose greater than 105% of the prescription dose occur within the PTV. This constraint, together with others described below, will eliminate that possibility of using fewer than 7–10 beams to treat the target.

The final constraint related to high dose for RTOG 0236 requires that the conformality index be less than 1.2. The conformality index is the ratio of the volume of the prescription isodose to the volume of the PTV itself. Dose deposited outside the PTV increases this ratio. Requiring a conformality ratio of 1.2 is strict. Achievement of this criterion requires that multiple beams with large angles separating each beam be used. Sites are encouraged to use 10 or more non-opposing, non-coplanar beams to meet this criterion or multiple arcs with at least 180 cumulative degrees rotation.

Moderate dose constraints

In an idealized plan, there would be no dose delivered outside the target. As this is impossible and some dose is “spilled” in the process of delivering dose to the target, it is critical that this extra-target dose be minimized. The dose gradient falls, hopefully rapidly, from the high target dose to a safe low dose. The moderate dose ranges between these extremes must be kept within as small a volume as possible and also respect normal tissue limits. Indeed, the most difficult aspect of SBRT treatment planning is to idealize this moderate dose distribution. It is often the difference between a good outcome and severe toxicity.

As with the target dose conformality ratio, a similar ratio is defined for the moderate dose. The ratio of the 30 Gy volume (50% of prescription) to the ratio of the PTV is defined as the R50. Acceptable values as a function of PTV volume were tabulated from idealized plans and are shown in Table I. Table I, however, is actually for RTOG 0618 where plans are to be generated using appropriate tissue heterogeneity algorithms. When kept appropriately low, this ratio defines a plan which minimizes unnecessary spillage of dose outside the target.

Table I.  Dose Constraints for RTOG 0618

	Ratio of Prescription Isodose Volume to the PTV		Ratio of 50% Prescription Isodose Volume to the PTV, R50%		Maximum Dose 2 cm from PTV in Any Direction, D2cm (Gy)		Percent of Lung Receiving 20 Gy Total or More, V20 (%)
	Deviation		Deviation		Deviation		Deviation
Maximum PTV Dimension (cm)	None	Minor	None	Minor	None	Minor	None	Minor	PTV Volume (cm^3)
2.0	<1.2	1.2–1.4	<4.1	4.1–4.3	<29.5	29.5–31.5	<10	10–15	1.8
2.5	<1.2	1.2–1.4	<4.1	4.1–4.3	<29.5	29.5–31.5	<10	10–15	3.8
3.0	<1.2	1.2–1.4	<4.1	4.1–4.3	<29.5	29.5–31.5	<10	10–15	7.4
3.5	<1.2	1.2–1.4	<4.1	4.1–4.3	<29.5	29.5–31.5	<10	10–15	13.2
4.0	<1.2	1.2–1.4	<4.0	4.0–4.2	<31.9	31.9–33.9	<10	10–15	21.9
4.5	<1.2	1.2–1.4	<3.9	3.9–4.1	<34.3	34.3–36.3	<10	10–15	33.8
5.0	<1.2	1.2–1.4	<3.8	3.8–4.0	<36.9	36.9–38.9	<10	10–15	49.6
5.5	<1.2	1.2–1.4	<3.7	3.7–3.9	<39.3	39.3–41.3	<10	10–15	69.9
6.0	<1.2	1.2–1.4	<3.5	3.5–3.7	<41.7	41.7–43.7	<10	10–15	95.1
6.5	<1.2	1.2–1.4	<3.3	3.3–3.5	<44.1	44.1–46.1	<10	10–15	125.8
7.0	<1.2	1.2–1.4	<3.1	3.1–3.3	<46.5	46.5–48.5	<10	10–15	162.6

Note 1: For values of PTV dimension or volume not specified linear interpolation between table entries is required.

Note 2: Protocol deviations greater than listed here as “minor” will be classified as “major” for protocol compliance.

Next, the moderate dose should be compact around the target. This criteria follows from the fact that with current understanding, it is not clear anatomically which specific tissues within a large organ like the lung can tolerate SBRT more than others. While specifically named normal tissue constraints (like the spinal cord tolerance) must be respected, in the absence of this information dose spillage should be kept as compact around the target as possible nearly depicting isotropic dose falloff [15]. RTOG 0236 attains this objective by requiring that at a radial distance of 2 cm in all directions from the edge of the PTV be kept below an idealized maximum dose limit. These limits are again based on idealized plans and tabulated as a function of PTV as shown on Table I.

Low dose constraints

There are no dose constraints in RTOG 0236 for low dose within the patient. Philosophically, this follows from the “radiosurgery” point of view that it is better to decrease high and moderate dose volume within the patient, even at the expense of increasing low dose volume, in order to decrease toxicity. This approach is reasonable so long as toxicity follows a sigmoid dose response curve. The low dose would be below the threshold for injury even when delivered to large volumes of normal tissue. In reality, the low dose volume may put the patient at risk for negative effects including second malignancies. Still, in comparison to the threat of the malignancy at hand, these risks seem acceptable.

Normal tissue constraints

The dose and volume limits for CFRT are fairly well known. Most are based on experience, particularly bad experience, where more than acceptable numbers of patients suffered toxicity beyond a certain limit. This information was conveyed to the radiation oncology community at large and tolerable limits became defined. For SBRT, this same “trial and error” approach is problematic. First, when designing trials using dose fractionation scheme never previously utilized, there is no literature or case reports to draw upon for determining limits. Second, the cost of misapplying a constraint is potentially much more dramatic for SBRT given its ablative character. Furthermore, the nature of side effects themselves may be different for SBRT than CFRT, even in the same organ. For example, with CFRT the most prevalent and concerning side effect for lung treatment is radiation pneumonitis. In the Indiana trials of SBRT, pneumonitis was rarely observed [8]. Instead, a bronchial-type toxicity was seen not previously characterized for radiation therapy to the lung [8], [16]. Protocols have to be designed to react to unexpected toxicity so as to properly collect and analyze the effects toward patient outcome.

In designing dose constraints for the three fraction regimen of SBRT used in RTOG 0236, best information was collected. For poorly tolerant serially functioning structures like the spinal cord, there is some established sense of dose tolerance. In the end, generally the committee accepted a biologically effective dose (BED) conversion from CFRT limits with some additional conservative reduction. Straight BED conversions from multiple fraction CFRT levels to few fraction treatments with SBRT may be problematic [17]. Indeed, the power series fit of the cell survival curves after irradiation in normal tissues using the linear-quadratic model will likely over predict the potency of therapy for large dose per fraction regimens. To overcome this issue, the RTOG committee further reduced tolerance values by 5–10% beyond BED conversions for critical normal structures like the esophagus.

A structure particularly prone to toxicity from ablative SBRT fractionation is the skin. One of the newest RTOG protocol being developed, RTOG 0618 using SBRT for operable stage I NSCLC, has a constraint for skin of 18–24 Gy total in 3 fractions. This is in response to toxicities seen in a number of institutions with severe burns along the entry path of beams. The constraint is reasonably easy to meet so long as many beams (i.e., ≥10) are used to target the tumor. When the tumor is chest wall based in a thin person, this constraint may require the dosimetrist to use some higher energy beams to facilitate skin sparing. Table II shows the normal tissue constraints for RTOG 0618.

Table II.  Normal Tissue Constraints for RTOG 0618

Organ	Volume	Dose (cGy)
Spinal Cord	Any point	18 Gy (6 Gy per fraction)
Esophagus	Any point	27 Gy (9 Gy per fraction)
Ipsilateral Brachial Plexus	Any point	24 Gy (8 Gy per fraction)
Heart/Pericardium	Any point	30 Gy (10 Gy per fraction)
Trachea and Ipsilateral Bronchus	Any point	30 Gy (10 Gy per fraction)
Whole Lung (Right & Left)	(See table in Section 6.4.2)	(See table in Section 6.4.2)
Skin	Any point	24 Gy (8 Gy per fraction)

Tissue heterogeneity considerations

There are two primary mechanisms for dose deposition in tissues as a result of photon irradiation: attenuation and scatter. Proper dose calculating algorithms after simulation should properly account for both of these factors. The majority of dose from radiation therapy is delivered by attenuation. The process involves degrading the primary fluence of the beam by direct interactions with tissue occurs along the trajectory of the beam and is depth and tissue density dependant. Scattering events that occur, redirect photons, and subsequently deposit energy (dose) elsewhere is important but has less impact on delivered dose. The dependence of dose delivery on geometry and density can become burdensome to model and may have little impact for many situations common in radiotherapy. As such, for the majority of the history of radiation oncology, tissue heterogeneity or density changes were not accounted for in the dose calculation. Instead the body was assumed to be totally water (unit) density.

At the time RTOG 0236 was developed, the RTOG Physics Committee did not have official recommendations about tissue heterogeneity correction for protocols. With concern that some centers would use inferior algorithms incorporated into their planning systems, it was decided not to allow tissue correction at all for RTOG 0236. Instead, monitor units for each beam would be calculated for a plan assuming all unit density with the prescription delivered to the edge of the PTV (not isocenter). In order to better quantity effects, all sites where asked to send a separate plan to the quality assurance center where they used the same monitor units per beam but calculate dose with their system's tissue heterogeneity algorithm activated. When RTOG 0236 has completed accrual, a report will be issued about the effect of these issues on dose calculations. Currently, the RTOG Physics Committee has recommendations about proper tissue heterogeneity accounting in a planning system. RTOG 0618 will require such planning and dose constraints to tissues have been modified accordingly for the new protocol.

Phantom irradiation

As part of the accreditation process for RTOG SBRT protocols, all centers are required to irradiate a phantom supplied by the Radiological Physics Center (RPC). Models of this phantom can simulate a variety of tumor geometry and motion patterns as would be related to breathing. Dosimeters are strategically placed inside the phantoms at critical points where dose would be scrutinized for protocol quality assurance in actually treated patients. In order to pass this phantom irradiation, the center must scan, plan and treat with a specified dose prescribed to the edge of the defined target (also incorporated into the phantom). Criteria are established to define acceptable observed minus expected measurements. In this way, an institution proves it can target, generate treatment plans with compact high dose distributions, and accurately deliver therapy even to moving targets.

Dry run

The last step of the accreditation process for RTOG 0236 involved central committee scrutiny of the first enrolled case from a center. This process is made possible by the digital plan submission requirement through the Image-Guided Therapy Center (ITC) which is part of the Advanced Technology Consortium (ATC). The first patient is consented, enrolled, simulated and a plan is generated by the enrolling site. The plan including all images and the dose-volume matrix is forwarded to the ITC digitally and reconstructed on the ITC computers. The protocol committee chairs can then review the plan using a web-based tool. If the plan appropriately meets constraints including appropriate contouring and dose-volume parameters to targets and normal tissues, the enrolling center is notified and may start treatment. If there are major deviations, the patient is not allowed to be treated until the matter is resolved. Suggestions are given to the treating site regarding how to avoid deviations. The initial dry run review is completed using the ITC resources typically in less than 48 hours.

Application to subsequent protocols

Rigorous treatment planning and quality assurance constraints for RTOG 0236 are being applied to subsequently developing protocols. Not all the constraints are appropriate for all sites of disease; however, the general approach has been successfully modified. In the RTOG Lung Committee, four other SBRT protocols are being developed including the following: 1. RTOG 0618 for operable patients with peripherally located early stage NSCLC (Robert Timmerman, Principal Investigator, P.I.); 2. RTOG 0624 for medically inoperable patients with T2 or T3 tumors from NSCLC (Brian Kavanagh, P.I.); 3. a trial for medically inoperable patients with centrally located early stage NSCLC (Andrea Bezjak, P.I.); and 4. a trial for patients with limited pulmonary metastases (Volker Stieber, P.I.). In each of these protocols, the dose volume constraints, accreditation, and quality assurance measures of RTOG 0236 will be applied.

Conclusions

The Radiation Therapy Oncology Group (RTOG) in North America has developed a working process for carrying out multicenter trials for Stereotactic Body Radiation Therapy (SBRT) in lung sites. This process includes a stringent but workable plan for scrutiny of site qualifications and quality assurance. Site qualifications are reviewed via an accreditation process, a phantom irradiation review, and via digital submission of plans prior to enrolling the first patient on protocol. Dosimetry constraints have been devised not just to ensure target coverage, but also to require rapid falloff of dose to surrounding normal tissue and respect organ tolerance limits. The collective efforts of many individuals and committee actions has facilitated the first cooperative group SBRT trial, RTOG 0236 for medically inoperable patients with peripherally situated lung cancer, to enroll patients rapidly despite rigorous quality assurance. The quality assurance template created for RTOG 0236 is now being applied to a number of other developing trials such that cooperative group multicenter prospective data will be available in as short a period of time as possible for allowing physicians to make treatment decisions for their patients.

This work was presented as an oral presentation at the 3^rd Acta Oncologica Symposium “Stereotactic Body Radiotherapy,” June 15, 2006, Copenhagen, Denmark. This work was supported by the following grants from the United States National Institutes of Health: Radiation Therapy Oncology Group (RTOG) cooperative group, R21 Quick Trials 5R21CA097721-02, and the Advanced Technology Consortium U24 CA 8164.