Keywords

1 Radiotherapy of Liver Metastases

Cancer confined to the primary tumor site and to the locoregional lymph nodes is usually considered curable when treated with surgery or radiation therapy. Immunotherapy, i.e. for metastatic melanoma may lead to long lasting remission, but in most cases of widespread metastatic cancer in adult patients, the treatment options are limited to cytotoxic therapy with the intention to prolong life for incurable patients. Hellman and Weichselbaum suggested that cancer progression is a multistep process with a state of oligometastases between the stages of purely localized stage and the stage of widely metastatic disease [1]. Patients with oligometastatic cancer may potentially be cured or may obtain long lasting remission if both the primary cancer and the metastases are treated aggressively.

The European Society for Medical Oncology (ESMO) consensus guidelines 2016 for management of metastatic colorectal cancer ranks SBRT as equivalent to radiofrequency ablation (RFA) and other ablation techniques in patients with inoperable oligometastases and recommends that the appropriate ablative tool from the “toolbox” based on patients and cancer-related characteristics should be chosen on the individual patient [2]. However, most MTDs still consider radiation therapy for patients where no other local ablation therapy is possible. The efficacy in terms of improvement of survival of surgery or other local ablative therapies of oligometastases has not been proven and no randomized trial has compared the efficacy of surgery, thermal ablation and radiation therapy. A matched comparative analysis of 60 patients treated for CRC liver metastases with either RFA or robotic SBRT showed the longest disease-free survival in patients treated with SBRT. However, there was a trend for longer overall survival in the RFA group, which did not reach a statistically significant level [3]. Trials comparing different modalities like surgery or RFA and radiation succumb to poor accrual. An example is the RAS study (NCT01233544) randomizing CRC liver metastasis patients between RFA and SBRT that unfortunately was terminated because of poor accrual.

Although there is no proven benefit from local therapy of metastases, there are substantial numbers of cohort studies supporting that selected patients may have a favorable prognosis when the liver metastases are treated locally, even without receiving systemic therapy.

2 Radiotherapy with or Without Systemic Antineoplastic Therapy

Delaying the progression of the cancer and thereby allowing a delay of onset or avoidance of systemic antineoplastic therapy is often an argument for radiation therapy of metastases. There are no randomized trials evaluating the effect of systemic therapy in addition to radiation therapy for liver metastases, but there is also no evidence for omission of systemic therapy. Two published randomized trials have explored the effect of addition of systemic therapy to the surgical resection of colorectal carcinoma (CRC) liver metastases. They both found that chemotherapy improved the progression-free or disease-free survival [4, 5]. In the largest retrospective cohort study of patients treated with SBRT for metastases both pre- and post-SBRT chemotherapy was related to improved overall survival, thus supporting the effect of addition of systemic therapy [6]. Of 321 oligometastatic patients included in this analysis, 201 patients (66%) were treated for liver metastases.

A local therapy such as SBRT may potentially benefit patients who have more widely disseminated metastatic cancer that does not necessarily meet the definition of oligometastasis. By combining locally ablative therapy with systemic therapy, there is potential to prolong progression-free survival and overall survival. A study of 24 patients with non-small cell lung cancer where erlotinib was combined with consolidative SBRT of the detectable metastatic lesions. This resulted in progression free and overall survival rates of 15 and 20 months, respectively, in patients who had failed platinum-based chemotherapy with no more than six metastases [7]. EGFR mutation was not part of the routine work-up for this study and none of 13 patients in whom the analysis was done had an EGFR mutation. The outcome of this study is better than expected when compared to similar patient cohorts treated with erlotinib alone. Consolidative local ablation combined with systemic antineoplastic therapy for metastatic cancer, as well as local therapy for metastatic oligoprogression where remaining metastatic lesions are controlled by systemic therapy may be important areas for future clinical research.

The addition of chemotherapy to whole-liver radiation therapy (WLRT) has been explored in the past. Most often, fluoro-deoxyuridine or 5-fluorouracil was used to enhance the response rates. No modern chemotherapy regimens have been tested together with WLRT. Chemotherapeutic drugs did not seem to improve the objective or symptomatic responses when comparing to studies where WLRT was given alone [8,9,10,11].

3 Pattern of Practice in Radiation Therapy for Liver Metastases

With steadily improving systemic therapies, we expect to see an increasing number of patients with oligometastases to the liver [12]. A substantial proportion may be eligible for radiation therapy. A large survey on the use of SBRT targeting individual radiation oncologists had 1007 responses [13]. It showed a marked increase in the use of SBRT for metastases. Sixty-one percent used SBRT for treatment of patients with 1–3 metastases; 75% for treatment of metastases to the liver. In a survey targeting centers with an active program on radiation therapy for metastases, 69 of 80 responders (86%) treated metastases to the liver with radiation therapy [14]. Fifty had an active liver SBRT program. Fewer used WLRT for palliation of liver metastasis related symptoms. Thirty-two responders (40%) treated more than five patients with WLRT per year.

4 3-D Conformal X-Ray Therapy

The Ann Arbor group treated 22 patients with CRC liver metastases with conventional 3-D conformal radiation therapy (CRT) and concomitant intra-arterial hepatic chemotherapy (fluoro-deoxyuridine). With total doses of 48–73 Gy in 1.5–1.65 Gy given twice a day, the response rate was 50%. However, only 25% of the patients were without hepatic progression within 1 year [15]. In a risk-adapted NTCP-based dose-escalation study from the Ann Arbor on CRT of primary and secondary liver cancer using a median total dose of 60.75 Gy (range, 40–90 Gy), 1.5 Gy twice daily and concomitant fluoro-deoxyuridine, the response rate for CRC patients was 60% [16, 17]. The patients included in this study had a large tumor burden with median tumor size of 10 × 10 × 8 cm. The median survival time for these patients was 17 months and radiation dose above 60.75 Gy was associated with favorable survival for the entire study group.

In general, response rates after CRT are high, but local control rates are lower than what is seen after the high-dose-per-fraction and high total dose radiation therapy associated with SBRT. However, CRT studies are not directly comparable to SBRT studies, as patient selection is different, generally including larger metastases in CRT cohorts. CRT, with the possibility of integration of IGRT and IMRT, is most useful in selected cases where SBRT is not possible, e.g., when the target is close to the bowel.

5 Stereotactic Body Radiation Therapy

SBRT was developed on the principles of stereotactic brain radiosurgery, and initial attempts were performed using a fixed body frame for patient immobilization and with image guidance primarily based on simple mega-voltage portal imaging to deliver 1–6 large fractions of radiation to a malignant tumor outside the brain. The first paper on clinical results of SBRT was published by a research group from the Karolinska Institute, Stockholm in 1995 [18] and since then the technique has evolved dramatically and it is now one of the important cornerstones in modern radiation oncology.

The early publication from Stockholm reported a local control rate of treated tumors that was much higher than expected, but a large number of publications confirm the high probability of local control after hypofractionated radiotherapy with high biological equivalent doses. Table 18.1 gives the results of the prospective and the largest retrospective cohort studies in SBRT for liver metastases [6, 19,20,21,22,23,24,25,26,27,28,29,30,31]. Most reports of local control rates are in the range 80–100%. A few studies report lower local control rates, i.e., the study by de Vin with a local control of 33%. This study included patients with a broad range of metastasis sites with liver metastases accounting for only 25% of the cases.

Table 18.1 Results from prospective and large retrospective studies of SBRT for liver metastases
Full size table

A number of studies have reported a dose-response effect with higher local control probability with the use of high biological equivalent doses [6, 21, 27, 28]. In a pooled analysis from three North-American centers, the actuarial local control rates were 86 and 42% at 1-year with doses of above and below BED10Gy of 75 Gy, respectively [27]. In the study from Denmark, the overall 2 years local recurrence rate was 13% and the rate was considerably reduced with BED10Gy over 100 Gy (hazard ratio 0.34) [6].

It has been claimed that liver metastases may be radio-resistant compared to metastases at other sites. The control rate of metastases is often lower in liver compared to lung, but it is unclear whether this relates to differences in radiosensitivity or to differences in radiation techniques. Lower control rates for CRC metastases were found is some studies [27, 32], but not in others [6, 21]. In the Danish study, the local relapse probability for metastases was higher in the liver compared to other sites, possibly explained by differences in radiosensitivity or by poorer imaging for target contouring of metastases to the liver compared to other sites [6].

Currently, there is no randomized phase III data to support the efficacy of SBRT for liver metastases, although a British randomized phase III trial (CORE) is expected to start recruiting patents with breast cancer oligometastases to conventional (systemic) therapy versus the same therapy plus SBRT for the detectable metastases (NCT02759783). The survival of patients in nonrandomized SBRT-studies depends in part on how well they are selected. However, studies of large cohorts of patients with metastatic cancer treated with SBRT have reported favorable survival rates even in negatively selected patients who were not eligible for surgery or radiofrequency ablation (Table 18.1).

Milano et al. found that patients with metastatic breast cancer to the liver and other sites of the body had a favorable survival compared to patients with metastatic cancer of other origin [33]. Fode et al. found a median overall survival of 6.1 year for breast cancer patients [6]. Due to the low number of patients, the survival of breast cancer patients in the Fode study was not statistically different from other patients’ survival. Katz et al. did not find the histological type influenced survival of the patients [20]. It seems obvious that some cancer types have a more indolent and less aggressive clinical appearance and for some cancer types there are several lines of systemic antineoplastic therapies that may affect the survival. Lack of differences in survival between tumor types may be a result of the patient selection.

Because of the limited knowledge on prognostic factors related to SBRT of patients with liver metastases, there is no consensus on the criteria to select patients. There is therefore considerable heterogeneity in patient and cancer characteristics among the published studies. Only few studies have patients numbers that are sufficient for assessment of prognostic factors. Cumulative GTV smaller than 3 cm was related to long overall survival in a study of patients with lung and liver metastases [28]. A large retrospective cohort study of patients primarily with liver metastases, but also some with lung and other metastatic sites, found that WHO performance status 0–1, solitary metastasis, size of largest metastasis under 3 cm, metachronous metastasis and pre-SBRT chemotherapy were related to favorable survival in a multivariate analysis [6].

SBRT of liver tumors is generally tolerated well. However, the esophagus, the stomach, the duodenum, and the large bowel should be considered in the selection of patients and in the treatment planning process because of their limited tolerance to radiation and the risk of severe adverse effects when they are exposed to large radiation doses. The liver tolerates large doses to relatively large volumes as long as a sufficient volume of liver is spared. Gastritis, gastric- or intestinal ulceration, chronic skin reaction, rib fracture, and hepatic failure seldom occur as late effects after SBRT for liver metastases [6].

There is growing use of SBRT for treatment of liver metastases. The results of prospective phase I/II trials and retrospective cohort studies are encouraging, but we are still missing high level evidence to prove its efficacy.

6 Particle Therapy with Protons and Carbon Ions

Due to their physical properties, protons and heavier ions have considerable potential in radiation therapy for primary and secondary liver tumors. The sparing is most prominent in volumes of normal tissue receiving the low-to-intermediate radiation dose. Because of the relatively low radiation tolerance of the liver, liver cancer patients are obvious candidates for particle therapy, especially if issues related to interplay effects and immature image guidance can be managed. A treatment planning study compared intensity modulated proton therapy (IMPT) and photons based intensity modulated radiation therapy (IMRT) in stereotactic body radiation therapy (SBRT) for liver tumors [34]. The study used a dose escalation risk-adapted prescription policy where the highest possible tumor dose was applied, provided that dose volume constraints to organs at risk were met. In 10 patients tested, there was sparing of normal liver tissue with protons compared to photons. With the highest dose level, the median V15Gy was reduced by 32% with use of IMPT. Nine of 10 cases could be treated at the highest dose level using IMPT whereas only two cases met this constraint at the highest dose level and six at the lowest dose level with use of IMRT. Other treatment planning studies report similar positive findings [35].

There are a number of published studies on proton and carbon-ion therapy of hepatocellular carcinoma (see Chap. 14). The particle therapy experience in treatment of metastases is limited to few case-reports and it does not allow for conclusions on the efficacy in treatment of oligometastases. Fortunately, a dose escalation study giving 3 × 12–20 Gy by use of passive scattering protons is actively recruiting patients with liver metastases at Loma Linda University (NCT01697371).

7 Radioembolization

Radioembolization or selective internal radiation therapy (SIRT) involves intra-hepatic administration of microspheres containing Yttrium-90 (90Y). The primary aims of radioembolization are to delay cancer progression and to improve survival. Studies of radioembolization without systemic antineoplastic therapy showed response rates of 35–75% in patients with CRC liver metastases (Table 18.2). When combined with systemic chemotherapy, the overall response rates increased to 76–91% [36, 37]. Two small randomized trials compared the combination of radioembolization and chemotherapy to chemotherapy alone in patients with CRC liver metastases [38, 39]. They both found improved response rates and progression-free survival and one study found improved survival in the SIRT arm. The large SIRFLOX study randomized 530 patients with mCRC liver-predominant disease referred for first-line chemotherapy between FOLFOX (±bevacizumab) chemotherapy and SIRT versus FOLFOX alone [40]. In this well-powered study the two arms had similar progression-free survival, but SIRT considerably delayed the progression in the liver by almost 8 months. Results of a similar multicenter study, the randomized FOXFIRE study, are also awaited [41].

Table 18.2 Results from randomized and large cohort studies of radioembolization (SIRT) for liver metastases
Full size table

In cohort studies of neuroendocrine tumors, radioembolization resulted in favorable response rates of 63–90% [42, 43]. Both of these studies report favorable median survival for this patient group.

Relatively common adverse effects are flu-like symptoms, moderate pain, gastro-intestinal symptoms, and neutropenia, whereas complications such as radiation induced liver disease (RILD), radiation cholangitis, vascular injury, infection/liver abscess, radiation pneumonitis, and gastric perforation occur less frequently. Grade III–IV toxicities in terms of intestinal ulceration, neutropenia, and transient hepatotoxicity were observed in a phase I study of combined radioembolization and FOLFOX chemotherapy in patients with CRC liver metastases [44].

8 Interstitial Brachytherapy for Liver Metastases

CT-guided interstitial brachytherapy with insertion of needles and use of afterloading technique with Iridium-192 has been used in highly specialized centers as a salvage of patients with liver tumors who are unsuitable for RFA and SBRT. Catheters are inserted to achieve target doses of 20 Gy or higher. There are reports of cohort studies describing successful control of even very large liver tumors over 10 cm in diameter and with doses exceeding 20 Gy the local control rates most often exceeds 70–90% [45].

The 2-year local control rate was 81% in a retrospective cohort study of 80 patients with 179 inoperable colorectal liver metastases treated with interstitial brachytherapy [46]. Local progression was significantly related to the size of the metastasis with control rates of 87 and 59% for lesions smaller than or larger than 40 mm, respectively. The survival rate at 3 years after treatment for liver metastases was 41%. In a phase II study of 41 patients with metastatic breast cancer, the local control was 87% and overall survival 60% 2 years after interstitial brachytherapy for liver metastases [47].

Metastases in the hilum of the liver has been considered a relative contraindication for brachytherapy, but a recent publication describes the successful ablation of metastases with a distance of less than 5 mm to the common bile duct or hepatic bifurcation with only 4 of 34 (12%) lesions recurring after CT-guided brachytherapy [48].

A matched pair-analysis of patients (18 patients with 36 liver metastases) who had received interstitial laser therapy or brachytherapy for metastases smaller than 50 mm in diameter concluded that brachytherapy was associated to a higher local control probability compared to laser therapy [49].

With the purpose of improving local control in the liver, interesting approaches include the combination of brachytherapy with either hepatic artery infusion of chemotherapy or laser induced thermotherapy. The combination of regional hepatic artery chemotherapy and brachytherapy in CRC liver metastases tested in a phase II study resulted in local control and overall survival rates that are similar to those observed after brachytherapy alone and the sequence between the two treatments did not influence the results [50]. The combination of laser therapy and brachytherapy in a phase I–II study resulted in a local control rate comparable to the rate observed after brachytherapy alone. However, the two groups of patients who received or did not receive laser therapy were not identical [51].

Major complications are rare, but liver failure, ulceration of the stomach and duodenum, bleeding in the liver or thoracic wall have been observed after 5% of the procedures and pleural effusion and risible pain and nausea may relatively frequently occur after brachytherapy of the liver [47].

Brachytherapy is a highly efficient ablation therapy for liver metastases. Local control rates are in the range of 70–90%, but severe complications may occur and brachytherapy should be limited to experienced centers.

9 Low-Dose Whole-Liver Radiation Therapy

Low-dose whole-liver radiation therapy (WLRT) may be used for palliation for patients with incurable primary or secondary liver cancer with the primary focus on reducing pain and other cancer-related symptoms (Fig. 18.1). Pain relief ranges from 55 to 80% in studies using WLRT alone [8, 52,53,54,55,56]. The relatively low radiation tolerance of the liver is the main reason for the relatively sparse use of WLRT. However, WLRT seems to be safe with less than 5% risk of radiation induced liver disease (RILD) as long as total ED2Gy to the whole liver is kept below 33 Gy and in most of cases RILD resolves within 1–2 months [57]. Early on, the focus of WLRT was on improvement of disease control and overall survival with intensifying radiation dose and combining with systemic therapy. However, most patients treated with WLRT are end-stage cancer patients with short survival expectation and no other effective treatment options available. The relevant endpoints to report nowadays are therefore symptom relief and quality of life.

Fig. 18.1
figure 1

Multiple breast cancer metastasis to the liver (a) and a treatment plan for 2 × 5 Gy whole-liver radiation therapy for palliation (b)

Full size image

The Radiation Therapy Oncology Group (RTOG) pilot study used dose-fractionation schedules ranging from 21 to 30 Gy in 7–19 fractions to treat 109 patients [53]. Response in terms of partial or complete symptom control was observed for abdominal pain, nausea and vomiting, ascites, anorexia, abdominal distension, jaundice, and night sweats/fever, with complete response rates for individual symptoms up to 34% and improvement in performance status in 25% of the patients. Leibel et al. found an 80% response rate for pain (complete in 54%) and improved performance status in 28% [8]. Pain relief occurred quickly and had a median duration of 13 weeks.

Systemic or regional chemotherapy with fluoro-deoxyuridine or 5-fluorouracil used in addition to WLRT did not improve the objective or symptomatic responses when tested in randomized clinical trials [8,9,10,11].

Most experience on WLRT is based on normo- or hyperfractionation, but a prospective Australian study used a hypofractionated palliative schedule of 2 × 5 Gy in two consecutive days in a cohort of 28 patients [52]. At 2 months after treatment, 53–66% of individual symptoms present at base-line improved. Pain relief was observed in 65% at 2 weeks and 53% of the patients surviving at 10 weeks still recorded a pain response. A newly published Canadian phase II study using a single fraction WLRT schedule of 8 Gy to the tumor-bearing part (majority) of the liver or the whole liver was performed on 41 patients of whom 20 had metastases and 21 had primary liver cancer [57]. The patients were also examined with the EORTC QLQ-C30 and FACT-hep quality of life and symptom scores. On average, 48% experienced improvement in cancer-related symptoms at 1-month post treatment. Pain was the most frequent base-line symptom and it improved in 26% of the patients, insomnia improved in 50% and nausea and vomiting in 26%. In 26% of the patients, overall global health also improved.

With an increasing number of systemic therapies available for patients with metastatic cancer, the need for palliative WLRT is declining. However, we consider WLRT as an option for patients suffering with symptoms from liver metastases and it is considerably underused. Clinical studies have shown considerable palliative effect after low-dose palliative WLRT, with pain relief in 26–80% of the patients and improvement of quality of life in a proportion of the patients.