Protons or Photons for Hepatocellular Carcinoma? Let's Move Forward Together

Laura A. Dawson, IJROBP 2009;74:661

Hepatocellular carcinoma (HCC) is the third leading cause of global cancer mortality, with a 5-year survival rate of 5% and an increasing incidence in North America. Cure is possible in patients who are treated with liver transplantation, resection, or ablative therapies (e.g., radiofrequency ablation for tumors <3 cm). Unfortunately, in most patients, these local therapies are not options, and hepatic progression is inevitable. One may expect that radiation therapy would have an established role in HCC; yet radiation therapy is not a universally accepted treatment, and many radiation oncologists have never treated a patient with HCC. It is time to establish the role of radiation therapy in HCC.

Progress has been slow in HCC for many reasons. Underlying liver disease and impaired liver function are almost universally present. Any insult to the liver (including treatment) can trigger liver failure and hasten death. The whole liver tolerance to radiation therapy is low and is reduced further in patients with HCC compared with those with liver metastases . As the liver volume irradiated increases, the risk of radiation induced liver disease rises. Reactivation of hepatitis B and a general decline in liver function are also possible, and both are poorly correlated with radiation dose and volume irradiated. Furthermore, the tolerance of other normal tissues around the liver (e.g., the stomach, duodenum, bowels, and kidneys) needs to be respected. In addition to potential toxicities, organ motion in the upper abdomen makes delivery of conformal HCC radiation therapy particularly challenging.

Can tumorcidal doses of radiation be delivered safely to focal HCCs? Yes, this is possible because of technological advances that have occurred over the past few decades, including improved HCC imaging, conformal 3D planning, automated optimization and IMRT, better understanding of normal tissue complication risks, breathing motion management strategies and soft tissue targeting (e.g., image guided radiation therapy [IGRT]). There is a growing body of literature on the use of conformal radiation therapy in HCC ; however most papers are retrospective and/or single-institution studies. One exception is the French multi-center phase II study that reported sustained local control in 78% of patients with early HCC (one ≤5 cm or three ≤3 cm) treated with 66 Gy in 2 Gy per fraction using conformal radiation therapy. A decline in liver function was seen in some patients with impaired liver function (Child-Pugh class B) at baseline. Radiation therapy has also been used safely in patients with more advanced HCC (e.g., with portal vein thrombosis) who are not ideal candidates for other local or regional therapies.

The article by Fukumitsu from Tsukuba, Japan is provocative, as the outcomes are outstanding. In this prospective study of 51 patients, proton therapy was used to deliver 66 GyE in 10 fractions to patients with one to three HCCs (≤10 cm), diagnosed using standard HCC criteria. Of the patients, 20% were in Child Pugh class B. The 5-year local control and survival were 88% and 39%, respectively. In Child-Pugh class A patients with solitary tumors, 5-year survival was 46%, similar to results obtained after surgery. Liver function remained stable or improved in 84% of patients, and no radiation-induced liver disease was observed. Late rib fractures were seen in 3 patients, adding ribs tissue to the list of normal tissues to be avoided. No other serious late toxicities were observed. Patients with tumors near the gastrointestinal tract or porta hepatis were ineligible for this hypofractionated protocol, to reduce the risk of luminal gastrointestinal and biliary toxicities that may occur after high doses per fraction.

Did the use of protons have an impact on these excellent results? Possibly. Particle therapy is another example of a technologic advance that should benefit HCC patients by reducing the integral dose to the liver, facilitating treatment of larger tumors, and allowing more patients to be treated safely with high doses. However, the use of protons comes at a cost (in addition to the financial costs). Protons are more susceptible to dosimetric and geometric uncertainties because of the increased sensitivity to tissue heterogeneity and the range uncertainty. Although the HCC target tissues are relatively homogeneous, the proton beams often pass through tissues with variable heterogeneities (e.g., ribs and lung). Thus, changes in path length (e.g., because of breathing motion or change in the mean liver position from fraction to fraction) can lead to differences in the delivered doses from the planned doses. These adverse effects can be reduced with the use of breathing motion management and IGRT, both which were used in the Fukumitsu et al. study.

Are there other explanations for these excellent outcomes? The majority of tumors treated on this study were less than 5 cm (with a median maximal tumor diameter of 2.8 cm), suggesting that some patients may have been suitable for other local therapies, which could have led to similar outcomes. Approximately two thirds of all patients treated with proton therapy during the same time period were not treated in this study, emphasizing that patients were well selected. In all, 65% of patients received prior (predominantly hepatic arterial–based) therapy, which may have contributed to the excellent outcomes. The latter point is not a criticism, as there is rationale to combine local and regional/systemic therapies in HCC because hepatic cancer recurrences and/or new primary cancers commonly occur. Finally, HCC is a highly heterogeneous cancer. The great variability in etiology of liver disease, liver function, and performance status affects survival and limits interpretation of single-institution series. Nonetheless, a 5-year survival rate of approximately 40% after radiation therapy is evidence of a strong treatment effect, regardless of potential selection bias. Some individuals may argue that these results are too good to be true, but these results are consistent with prior proton and carbon ion experience For example, in a Phase I/II study of 50 to 80 GyE in 15 fractions delivered to HCC patients using carbon ions, was associated with 5-year local control and survival rates were 81% and 25% respectively. We should not ignore these results.

Could the same outcomes be obtained with quality high-dose conformal photon irradiation in similar HCC patients? There is no reason to think that this should not be possible in selected patients. The most appropriate patients for conformal photon therapy are those with Child-Pugh A liver function who can be irradiated with enough sparing of the liver (e.g., mean liver dose <28 Gy in 2-Gy per fraction). Patients with smaller tumors (<6 cm) and tumors at the dome of the diaphragm are those most likely to be able to be treated with biologically potent photon therapy safely. Patients with Child-Pugh B class liver function or large (>8 cm) central tumors, associated with higher liver doses after conformal photon therapy, have an increased risk of liver toxicity; these are the patients in whom protons should have the greatest benefits. Patients with diffuse multi-focal HCC and those with Child-Pugh C liver function are not well suited for high-dose photon or proton therapy.

In HCC, multi-center, randomized studies of the clinical application of conformal radiation therapy are a priority, rather than studies that compare specific technologies. Although radiation therapy is an accepted HCC treatment in some specialized centers, randomized trials will be required for widespread acceptance, for both early and advanced HCC. Randomized trials will control for heterogeneity in prognostic factors and will allow us to learn about how outcomes from specialized centers translate to the broader community. The studies need not be technology specific, as long as safety can be ensured through quality assurance (QA) strategies. There is a strong need for collaboration with the HCC multi-disciplinary team, educational workshops, normal tissue dosimetric guidelines, and real-time radiation plan review, as many radiation oncologists have limited experience in HCC. Protons and photons should both be included in such studies. Motion management and IGRT should be strongly encouraged to increase the likelihood that the delivered doses match the planned doses, so that dose-outcomes relationships can be better understood. The relative benefits of a specific technical advance (e.g., protons, IGRT) are challenging to prove but could be built in as secondary study endpoints.

There are several hurdles to overcome before randomized trials of radiation therapy can be conducted in HCC. The technical issues described above have been adopted with substantial variability, and they are not widely available in developing countries where HCC is endemic. Radiation therapy has been used in early disease with curative intent (as in the Fukumitsu, in locally advanced disease to improve survival and quality of life, and in the palliative setting to improve symptoms. Each of these settings is important to study. There are many competing local and regional HCC therapies, with wide variability in specific techniques, which make studies of combination therapy more challenging. Recently, Sorafenib, an oral agent targeting kinases involved in tumor growth and angiogenesis, was shown to improve median survival in advanced HCC from 7.9 months to 10.7 months compared with supportive care (hazard ratio, 0.69; p < 0.001). Although the time to radiologic progression was improved, there was no significant difference in the time to symptomatic progression (4.1 months vs. 4.9 months, p = 0.77). This study of 602 patients from 21 countries demonstrates that randomized trials are feasible in HCC. Given the radiation-sensitizing properties of Sorafenib and other biologic targeted agents, the combination of radiation therapy and targeted agents should be studied. Hopefully these combinations will lead to more dramatic improvements in HCC outcomes.

In summary, the well-conducted prospective study by Fukumitsu is a substantial contribution to the growing literature on HCC radiation therapy. This study demonstrates that cure of early-stage HCC is possible with high-quality radiation therapy. Several questions remain, including what the ideal radiation dose and fractionation are, and whether the excellent outcomes seen with protons can be obtained with conformal photons. It is time that we work together to develop multi-institutional and randomized trials to increase the level of evidence demonstrating radiation therapy as an effective treatment in the armamentarium against HCC.