External beam radiotherapy for liver tumors

Historical context and whole-liver radiation therapy

The efficacy of radiation therapy (RT) is predicated on the ability to deliver the tumoricidal RT dose while keeping the dose to the surrounding normal tissues the same. The balance between the probability of tumor control and the risk of normal tissue complications is a measure of the therapeutic ratio of the treatment. Irradiation of tumors using external beam RT (EBRT) will expose the surrounding organs to doses that may range up to the prescription dose of the tumor. Unfortunately, doses necessary to achieve tumor control usually overlap with those that can cause complications. In case of liver tumors, doses necessary to control disease are higher than doses that cause dysfunction of the liver parenchyma. Improvements in the therapeutic ratio for liver tumors have been primarily predicated on achieving greater conformality of RT thereby limiting exposure of the liver parenchyma and nearby gastrointestinal (GI) tract to tolerable limits. This is more challenging for larger liver tumors or when there is underlying liver disease.

Early clinical experience that revealed the low tolerance of normal liver parenchyma to RT involved irradiation of the entire liver or large parts of the organ. Radiation-induced liver disease (RILD) was described as a clinical syndrome characterized by hepatomegaly and ascites, often associated with liver function test abnormalities, most consistently with alkaline phosphatase elevation. ^, This clinical syndrome was associated with venoocclusive histologic changes characterized by sinusoidal congestion with fibrin deposition in central veins. In the largest series of 40 patients treated with whole-liver doses ranging from 1300 to 5100 cGy, over 40% of patients who received 3800 cGy and above developed RILD, whereas no patient receiving less than 3000 cGy did. Some of the patients with RILD recovered after medical management with diuretics and occasionally steroids; others subsequently died of liver failure. Due to concerns of toxicity arising from large volumes of liver receiving low doses, and despite the fact that a contemporaneous report of 26 hepatocellular carcinoma (HCC) cases treated with RT at Memorial Hospital from 1923 to 1959 with tumor doses ranging from 700 to 4500 cGy (average 2956 cGy) showed regression in 14 cases in patients who received greater than 2000 cGy, the enthusiasm for RT as treatment for liver cancer waned.

The next several decades saw several failed attempts to improve the therapeutic ratio of liver irradiation by varying the fractionation size and/or combining low-dose whole-liver irradiation with cytotoxic or radiopharmaceutical agents. ^, Radiation Therapy Oncology Group (RTOG) study 84-05, although conducted for liver metastasis, is notable for reconfirming that dose escalation beyond 30 Gy to the whole liver is associated with high risk of complications. This was a dose escalation study of whole-liver RT delivered in twice-daily fractions of 1.5 Gy based on the assumption that smaller fraction size with the same total dose may potentially lessen the effect on normal tissues and the twice-daily schedule would keep the overall treatment time the same, thus maintaining the treatment efficacy. None of the 122 patients receiving 27 Gy or 30 Gy to the whole liver developed liver injury; 5 of 51 patients receiving 33 Gy did. A subsequent RTOG study conducted in 194 patients with HCC showed that combining low-dose daily whole-liver irradiation to 2100 cGy with chemotherapy or using hyperfractionation to 2400 cGy in twice-daily fractions was associated with response rates of 22% and 18%, respectively. This served as the background for future studies, which hypothesized that smaller liver volumes could tolerate higher RT doses and that higher doses would be more effective.

Partial-liver irradiation: Three-dimensional conformal radiation therapy (3D-CRT)

Although it was known that normal tissue tolerance depends on the amount of tissue irradiated, it was not until the tools to quantify the relationships became available in the 1980s that partial liver volume dose escalation became feasible. ^, The rationale for pursuing partial-liver irradiation was at least in part based on the observation that liver function can be preserved after a significant part of the liver is surgically resected. A clinical protocol carried out at the University of Michigan designed to stay within a given normal tissue complication probability allowed for delivery of focal radiation ranging from 48 to 72.6 Gy in 11 patients with localized HCC and 9 patients with cholangiocarcinoma. These doses were associated with an objective response in 11 of the 11 evaluable patients compared with 1 of 6 patients who received whole-liver irradiation to 36 Gy. Following these initial steps, a series of dose escalation protocols by the same group helped to quantify the relationship between dose, irradiated volume, and RILD, and provided initial data on the dose-response relationship for liver tumors. An important step was development of a model that quantitatively described the probability of normal tissue complication for a given dose and volume irradiated. Using this model with parameters that were revised as more data were accumulated, the investigators gained a better ability to describe the risk of RILD, compare alternate radiation plans, and individualize radiation on the basis of the predicted probability of toxicity. These data also showed that patients with primary liver tumors had a greater risk of RILD than patients with metastatic disease, likely secondary to underlying cirrhosis often associated with the former.

Two consecutive prospective trials conducted between 1996 and 2003 relied on the strategy of maximizing tumor dose using the prospectively validated predictive model of toxicity related to dose and volume of liver treated. In doing so they were able to establish objective volumetric dose parameters or constraints that minimize the risk of RILD and laid the foundation for further study of RT in the treatment of inoperable liver tumors. Between April 1996 and April 2003, 81 patients with primary liver cancers along with 47 patients with colorectal liver metastases were treated with hepatic arterial floxuridine and focal liver RT to 40 to 90 Gy (median, 60.75 Gy) in 1.5-Gy twice-daily fractions with a planned 2-week break. Median survival rates in HCC and cholangiocarcinoma, slightly inferior to colorectal metastatic cancer, were 15.2 and 13.3 months, respectively, with a trend for improved survival with higher doses. Importantly, the total dose was the only significant factor associated with survival. Similar results were noted in a prospective phase II trial from France that included Child-Pugh (CP) A/B cirrhotic patients with small-size HCC (1 nodule ≤5 cm, or 2 nodules ≤3 cm) unsuitable for other curative treatments. Twenty-seven patients were treated with a dosing strategy that conformed to the approach put forth by the University of Michigan, with all except for one patient receiving 66 Gy in 2 Gy/fraction using three-dimensional conformal radiation therapy (3D-CRT). Objective response was 92% with only transient grade 4 liver toxicity noted exclusively in CP B patients. As new technologies emerged, they were investigated in the treatment of primary liver tumors in an effort to achieve even better tumor control.

Intensity-modulated radiation therapy and stereotactic body radiation therapy

Intensity-modulated radiation therapy (IMRT) is an advanced radiation technology that further improves on the conformality of dose delivery to the target and allows for greater control of doses that spill into normal organs. It involves a computational optimization process that selects RT beam directions and modulates the intensity of each beam to gain the desired target coverage while minimizing the dose to the normal organs based on the tumor target and normal structures outlined by the physician on the simulation computed tomography (CT) or, more recently, magnetic resonance imaging (MRI). IMRT allows for greater control of dose distribution, which makes individual radiation plans more customizable to the patient’s individual anatomy. Due to its superior dosimetry, IMRT has significant advantages over 3D-CRT whenever there is a radiosensitive organ such as the stomach or duodenum near the tumor. Stereotactic body radiotherapy (SBRT) usually incorporates IMRT along with image guidance and organ motion management to deliver larger doses per treatment with increased precision. The mechanism of cell killing for these higher doses of RT is predicted to be more multifaceted than that of standard fractionation and may result in a more ablative effect. However, not all the doses delivered with the SBRT technique are high enough to have an ablative effect. Attention should be paid when interpreting the results of SBRT studies whether ablative or nonablative doses are used.

For image guidance, radiopaque fiducial markers can be percutaneously introduced into or at the periphery of the liver tumor to aid with tumor alignment, but soft tissue image guidance is critical to assess the dose delivery to organs in proximity to liver tumors, which move from day to day, such as the stomach and duodenum. The imaging components of radiation treatment delivery systems have evolved from two-dimensional radiographs to three-dimensional CT or MRI that allows for soft tissue visualization. The most significant recent innovation is real-time adaptive planning, which enables changing the radiation dose distribution every day to address changes in internal organ position. This capability is very valuable for left lobe tumors due to the variable relative position of the stomach relative to the liver from day to day.

In the past, respiratory motion and day-to-day organ motion were largely ignored. Larger volumes were treated with lower doses with limited benefit. Motion management techniques were critical innovations that enabled the delivery of ablative doses near radiosensitive organs such as the stomach. Diaphragmatic motion during breathing is associated with significant liver displacement and deformation. The average displacement during quiet breathing in the supine position has been reported by a number of investigators to range between 8 and 25 mm. There are multiple strategies to control or compensate for respiratory motion. The most common and most straightforward technique to control breathing motion is inspiration or expiration breath-hold. Respiratory gating and tracking during treatment are ways to account for breathing motion. Gating allows for the beam to only turn on at the prespecified portion of the breathing cycle, typically end expiration when the diaphragm is still and relaxed. Tracking is an emerging option that allows for the treatment beam to follow the internal motion. Compensating for the range of tumor motion during breathing with or without abdominal compression to minimize diaphragmatic excursion is sufficient for many liver tumors that are not near the stomach, colon, or duodenum. Respiratory gating is available in a number of standard radiation delivery systems. More versatile systems also allow for it to be combined with soft tissue image guidance to address day-to-day organ motion.