Home > Newsroom > Publications> Newsroom > News Archive

Please note: This article is part of the Cancer Center's News Archive and is here for historical purposes. The information and links may no longer be up-to-date.

Michigan Oncology Journal Fall 98

Conformal Radiation Therapy for Patients with Intrahepatic Malignancies

---Cornelius J. McGinn, M.D. and Theodore S. Lawrence, M.D., Ph.D.

In 1998, an estimated 13,900 patients in the United States will be diagnosed with hepatocellular carcinoma (hepatoma) and intrahepatic cholangiocarcinoma (1). Only a small subset of these patients present with resectable disease. The median survival in most U.S. series, including those with patients who have undergone potentially curative or palliative resection, is generally in the range of six to ten months (2). Mortality results from local and regional progression of tumor. Although a variety of non-surgical treatment approaches have been investigated, none have been effective enough to be considered standard primary or adjuvant management.

Figure 1: Solid surface reconstruction derived from a helical treatment planning CT of a 61 year old male with an unresectable metastasis from colon cancer.

Displayed are the target volume (tumor with an additional margin for uncertainty), the liver, spinal cord and kidneys. Anterior and lateral beams (defined by a multileaf collimator) are shown as well, in this oblique view.

A larger number of patients, however, die from colorectal cancer metastatic to the liver. Aggressive local/regional approaches have been considered for these patients based on clinical information confirming hepatic dissemination as the only site of disease in 20 percent, and surgical series demonstrating long-term survival in patients with solitary or a limited number of liver metastases (3). Among these, only surgical resection can be considered standard, yet is only possible in a select group of patients.

Investigation of non-surgical local/regional treatment modalities for this patient population continues at medical centers across the U.S. Each institution tends to focus on a limited number of approaches from a wide range of options including hepatic arterial chemotherapy, chemoembolization, radioisotope administration, percutaneous ethanol injection, and cryosurgery. At the University of Michigan, we have focused on the use of external beam irradiation combined with hepatic arterial chemotherapy in a series of prospective clinical trials initiated in 1987. Our work represents the only systematic investigation of radiation dose escalation trials in this patient population, as will be discussed below. It is clear, however, that investigation and evaluation of all potential options (including advances in systemic therapy) and potential integration of approaches will be required to improve the outcome for these patients.

The Role of Conformal Radiation Therapy
Previous investigations into the use radiation therapy for patients with intrahepatic cancers have revealed that the delivery of greater than approximately 33 Gy to the whole liver causes an unacceptable incidence of radiation-induced liver disease (RILD, also called "radiation hepatitis") (4), and treatment with lower, tolerable doses of radiation produces only transient palliation. However, it has long been known that higher doses of radiation can be given to parts of the liver if sufficient normal liver is spared. The introduction of three-dimensional (3D) radiation treatment planning techniques has provided two crucial tools to allow the development of high-dose focal hepatic irradiation protocols. First, it permitted a reduction of the dose to the normal liver through the use of beams that can be focused on the lesion(s) identified on the treatment planning CT scan and are not constrained to the axial plane. Second, 3D treatment planning made possible the quantification of the fraction of normal liver irradiated.

Using these tools, we initiated a series of clinical trials for patients with intrahepatic malignancies combining high-dose 3D conformal radiation with concurrent hepatic arterial fluorodeoxyuridine (HA FUdR). Concurrent HA FUdR was administered because of its established role in the treatment of colorectal liver metastases (5) and for its potent radiosensitizing properties. In these initial protocols, the dose was based on the amount of normal liver excluded from the high-dose region (defined as the 50 percent isodose line). However, the relationship between volume of liver irradiated and dose that it could safely receive was not well understood. Thus, all patients were divided into only three categories (<33 percent of liver excluded, 33 to 66 percent of liver excluded, and >66 percent of liver excluded), with three corresponding doses (33 Gy, 48 Gy, and 66 Gy, respectively). Compared to whole liver irradiation (+/- HA FUdR), this approach produced a higher response rate and possibly improved survival, par ticularly for patients with primary hepatobiliary cancer (6, 7).

Individualized Dose Determination
The data from these initial studies suggested that local control and survival of patients with intrahepatic cancers might be increased if we could escalate the dose of radiation further. Yet if we tried to escalate by dose alone, considering only the volume of liver excluded from the 50 percent isodose line and broad volume groupings (or "bins", <33 percent vs. 33 to 66 percent vs. >66 percent excluded from the high dose region), patients would be subjected to a wide range of risk. For example, those in the low end of the 33 to 66 percent bin would get the same dose as those in the high end of the bin, yet would have a higher risk of toxicity (less normal liver spared). Conversely, those in the high end of the bin would be exposed to a low risk of toxicity, and perhaps not receive as high a dose as they could have tolerated.

We therefore analyzed the entire dose distribution (all isodose lines, rather than just the 50 percent isodose line) across the normal liver for patients with and without RILD in these earlier protocols, and developed parameters for a normal tissue complication probability (NTCP) model that described this clinical experience. We then designed a protocol in which each patient received an individualized, maximum possible dose while being subjected to an estimated fixed probability of RILD. The estimated risk is calculated from the entire dose distribution (once the 3D plan for that patient has been created) using the NTCP model. This contrasts with standard phase I trial design, which delivers a target dose without regard to the volume of normal tissue, and with our previous protocol, which based the dose on the fraction of normal liver receiving < 50 percent of the isocenter dose.

Figure 2: Pre-treatment CT scan for the patient shown in Figure 1

In the first level of this trial, the dose prescribed was that which subjected each patient to a 10 percent risk of RILD. Our hypotheses were that this approach would allow safe delivery of a higher dose than we would have prescribed on our previous protocol, and that the model would predict the observed complication probability. In the first 21 patients who were evaluable for liver toxicity, the mean dose delivered was 56.6 Gy (range 40.5 to 81 Gy) compared to a mean of 46.0 Gy (range 33 to 66 Gy) if they had been treated according to the prior study (p<0.01) (8). One of these 21 patients developed RILD, for a complication rate of 4.8 percent. These findings supported our initial hypotheses, but also suggested that the NTCP model may have overestimated risk. As a result, we are continuing to accrue patients in the second level of the trial, where each will receive an individualized, maximum possible dose while being subjected to an estimated 20 percent risk RILD.

Response to treatment and duration of response have not been formally analyzed as yet. However, in agreement with our previous approach, patients have experienced encouraging outcomes (Fig 1-2).

Summary
Our ongoing investigations into the management of patients with intrahepatic malignancies suggest that advances in 3D conformal radiation therapy techniques, along with mathematical modeling of complication probabilities (made possible by 3D techniques) can be used prospectively to safely deliver far higher doses of radiation than with prior approaches. Additional follow-up and accrual will be required to determine if these higher doses produce further improvements in response and survival.

return to top

References
1. Landis SH, Murray T, Bolden S, et al. Cancer Statistics, 1998. CA Cancer J Clin. 48:6, 1998.
2. Stuart KE, Anand AJ, Jenkins RL. Hepatocellular carcinoma in the United States: prognostic features, treatment outcome, and survival. Cancer. 77:2217, 1996.
3. Fong Y, Cohen AM, Fortner JG, et al. Liver resection for colorectal metastases. J Clin Oncol. 15:938, 1997.
4. Lawrence TS, Robertson JM, Anscher MS, et al. Hepatic toxicity resulting from cancer treatment. Int J Radiat Oncol Biol Phys. 31:1237, 1995.
5. Harmantas A, Rotstein LE, Langer B. Regional versus systemic chemotherapy in the treatment of colorectal carcinoma metastatic to the liver. Is there a survival difference? Meta-analysis of the published literature. Cancer. 78:1639, 1996.
6. Robertson JM, Lawrence TS, Andrews JC, et al. Long term results of hepatic artery fluorodeoxyuridine and conformal radiation therapy for primary hepatobiliary cancers. Int J Radiat Oncol Biol Phys. 37:325, 1997.
7. Robertson JM, Lawrence TS, Walker S, et al. The treatment of colorectal liver metastases with conformal radiation therapy and regional chemotherapy. Int J Radiat Oncol Biol Phys. 32:445, 1995.
8. McGinn C, Ten Haken R, Ensminger W, et al. The treatment of intrahepatic cancers with radiation doses based on a normal tissue complication probability model. J Clin Oncol. 16:2246, 1998.

Return to top