Over the years, doxorubicin and gemcitabine have been among the most widely used drugs for hepatocellular carcinoma (HCC), with relative efficacy. The authors report the results of a phase 2 study of the combination of gemcitabine plus pegylated liposomal doxorubicin.
Patients with advanced HCC received combination chemotherapy with gemcitabine 1000 mg/m2 on Days 1 and 8, followed by pegylated liposomal doxorubicin 30 mg/m2 on Day 1. Treatment was repeated every 4 weeks to a maximum of 8 cycles. Primary endpoint was overall response rate, and secondary endpoints were time to disease progression (TTP), overall survival (OS), and toxicity.
Forty-one patients were enrolled and were evaluable for response, toxicity, and survival. A total of 194 cycles of treatment were administered. Three (7%) patients had a complete response, and 1 of these patients underwent liver transplantation. Seven (17%) patients had a partial response and, among these patients, 1 patient underwent surgical resection. Among the 31 patients who had initial alpha-fetoprotein levels >400 ng/mL, 20 (64.5%) had a >20% decrease after 2 cycles of treatment. The median TTP and OS were 5.8 and 22.5 months, respectively. Hematologic toxicity was the most common side effect, including neutropenia (17%) and anemia (7%).
Hepatocellular carcinoma (HCC) is the fifth most common cancer and the third leading cause of cancer mortality worldwide.1 The highest incidence rates are found in East Asia and in the United States; the incidence rates of HCC tripled between 1975 and 2005.2 The only curative treatments for HCC are surgical resection and liver transplantation; however, only about 15% of patients are eligible for these treatments, because of the coexistence of the advanced stage of the disease and cirrhosis at presentation.3 Over the years, numerous cytotoxic agents have been tested with rather disappointing results.4 Underlying cirrhosis, impaired liver function, portal hypertension, and thrombocytopenia make systemic chemotherapy difficult.
Gemcitabine is a pyrimidine antimetabolite with demonstrated activity against HCC,5 exhibiting cell phase specificity, primarily killing cells undergoing DNA synthesis. Moreover, the favorable nonhematologic toxicity spectrum and the mild and reversible hematological toxicity profile have made this agent an attractive and ideal drug for combination regimens.
Doxorubicin, a topoisomerase II inhibitor, is another of the most widely used and active agents in HCC, with a response rate between 0 and 44%,6 although in more recent studies this has not exceeded 25%7; however, its use is limited by elevated toxicity. In contrast to free doxorubicin, pegylated liposomal doxorubicin, which avoids uptake by the reticuloendothelial system, has shown longer circulation times, higher tumor selectivity, greater antitumor activity, and reduced toxicity.8 Neutropenia, stomatitis, and palmoplantar erythrodysesthesia (PPE) were found to be the major dose-limiting adverse effects of pegylated liposomal doxorubicin. Regarding its activity in advanced HCC, various studies have shown response rates of 10% to 14%,9 with a favorable toxicity profile.
Since 2008, sorafenib has become the standard first-line treatment for advanced HCC, on the basis of the results of the SHARP trial; this study showed an increased median time to progression (TTP) and overall survival (OS) of about 3 months when compared with placebo10; despite the significant survival benefit obtained, further improvement in the treatment of advanced HCC remains mandatory.
For all these reasons, in particular for the demonstrated single-agent cytotoxic action, the different mechanisms of action, the lack of cross-resistance, and nonoverlapping toxicity, a phase 2 study was undertaken to examine the efficacy and safety profiles of a combination chemotherapy with gemcitabine plus pegylated liposomal doxorubicin. In our study, gemcitabine was administered before pegylated liposomal doxorubicin, because various studies11, 12 have demonstrated that gemcitabine induces an increase in topoisomerase II expression; thus, a subsequent pegylated liposomal doxorubicin-mediated antitopoisomerase effect may cause a more significant cytotoxicity. Moreover, because in a prior phase 2 study13 of pegylated liposomal doxorubicin at a dose of 50 mg/m2 in refractory ovarian cancer significant skin toxicities and mucositis were been reported, whereas a subsequent study14 demonstrated that pegylated liposomal doxorubicin at a dose of 30 mg/m2 in advanced HCC had low systemic toxicity, it was decided to use pegylated liposomal doxorubicin at a dose of 30 mg/m2.
MATERIALS AND METHODS
This study was a phase 2, open-label, single-arm, single-institution trial. All patients were required to provide written informed consent before study participation. Patients were eligible for the study if they met the following criteria: pathologically proven HCC or an alpha-fetoprotein (AFP) level >400 ng/mL associated with imaging evidence on computed tomography (CT) scan or nuclear magnetic resonance (NMR) of HCC and inoperable or metastatic HCC. Inoperable disease was defined as either lesion too large for resection, regional lymph node involvement, presence of multifocal disease, or invasion of major blood vessels. Other inclusion criteria were as follows: measurable disease according to the Response Evaluation Criteria in Solid Tumors (RECIST)15 on CT or magnetic resonance imaging; no previous systemic chemotherapy regimens for HCC; age >18 years; Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 to 2; compensated Child-Pugh stage A or B cirrhosis; adequate renal function (creatinine <120 μmol/L), adequate blood cell counts (neutrophil count >1.5 × 109/L, platelet count >100 × 109/L), and adequate hepatic function (bilirubin ≤3 mg/dL, alanine aminotransferase and aspartate aminotransferase levels up to 7× the institutional normal limits). Prior allowed treatments included surgery, transarterial chemoembolization, percutaneous ethanol injection, and radiofrequency ablation. Exclusion criteria included concurrent malignancy, known central nervous system metastases, significant medical comorbidities, clinically significant cardiovascular disease, and any evidence of hepatic encephalopathy.
Treatment consisted of gemcitabine 1000 mg/m2 diluted in 250 mL normal saline infused over 30 minutes on Days 1 and 8, followed by pegylated liposomal doxorubicin 30 mg/m2 in 500 mL 5% dextrose intravenously over 1 hour on Day 1. Each cycle was given every 4 weeks.
The pretreatment evaluation included a medical history, physical examination, serum hepatitis B and C serology, blood chemistry including AFP level, and hepatic and renal function tests. Required radiologic studies included chest x-ray, cardiac echocardiogram, abdomen CT, or NMR imaging scan.
Physical examinations, full blood counts, serum bilirubin, and AFP assay were recorded before the beginning of each cycle; CT or NMR imaging scans were performed every 3 cycles. Follow-up evaluations included physical examination, blood chemistry, and CT or NMR imaging scans every 2 months. Response and progression were evaluated using the RECIST system.15 Treatment was continued until disease progression, chemotherapy delay for >3 weeks, unacceptable toxicity, patient refusal, or completion of 8 courses of therapy. Salvage chemotherapy was allowed for patients who were refractory to this treatment and was left to the discretion of the investigators.
Premedication included ranitidine 150 mg, hydrocortisone 125 mg, chlorphenamine 10 mg, and ondansetron 4 mg. Granulocyte colony-stimulating factor was not recommended for prophylaxis of neutropenia.
Toxicity Evaluation and Dose Modification
All patients who received at least 1 dose of treatment were considered evaluable for safety. Toxicity was evaluated according to the National Cancer Institute Common Toxicity Criteria version 2.
On Day 1, the patient was required to have an absolute neutrophil count ≥1500 and platelets ≥100,000 to receive treatment. The dose of gemcitabine was reduced by 50% in case of grade 1 thrombocytopenia or grade 2 neutropenia; it was discontinued if patients developed higher toxicity. Pegylated liposomal doxorubicin was reduced by 25% or discontinued in case of grade ≥2 PPE.
Treatment of common adverse effects seen with this regimen, such as anorexia, nausea, and emesis, was allowed using standard interventions.
The primary endpoint was the overall response rate. Sample size was calculated to reject a 10% response rate in favor of a target response rate of 30%, with a significance level of .05 and a power of 80% by using Simon 2-stage design.16 In the initial stage, a total of 12 assessable patients were evaluated for response. If >1 response was observed, then 20 additional patients were selected to enter the second stage to achieve a sample size of at least 32 evaluable patients.
Secondary objectives included TTP, OS, safety, and identification of the factors that influence TTP and OS with this regimen. Follow-up was calculated from the outset of treatment.
TTP was calculated from the start of therapy until tumor progression or death. OS was calculated from the start of therapy to the date of death. The Kaplan-Meier method was used to estimate survival. The analyses were performed on an intent-to-treat basis.
Independent prognosticators for TTP and OS were studied by forward stepwise approach in Cox regression analyses. Variables with a P value <.25 in univariate analysis were included in the multivariate regression. P values were based on 2-sided testing, and statistical analyses were performed with SPSS 15 statistical software (SPSS Inc, Chicago, Ill). Differences with a P≤.05 were considered significant.
Between November 2003 and May 2008, 41 patients with advanced-stage HCC were enrolled in the study. Patient and disease characteristics are shown in Table 1. There were 33 (80.5%) men and 8 (19.5%) women. The median age was 63.2 years (range, 44-78.1 years); the median ECOG PS was 1 (range, 0-2); 19 (46.3%) patients and 22 (53.7%) patients, respectively, were in Child-Pugh stage A and B; metastatic disease was present in 46.3% of patients; and 31 (75.6%) patients had an elevated AFP level (>400 ng/mL) at the time of entry. Hepatitis B surface antigen was positive in 12.2% of patients, and 48.8% of patients were hepatitis C virus seropositive. Nineteen (46.3%) patients had stage II or III disease, according to the Okuda system. Twenty-two (53.6%) patients had received previous treatment, including surgical resection and/or locoregional therapy (transarterial chemoembolization, radiofrequency thermal ablation, and percutaneous ethanol injection). Regarding second-line treatment, this was left to the discretion of the investigators; most patients were treated with oral capecitabine or intravenous 5-fluorouracil.
Overall, 194 cycles of treatment were administered; the median number of courses of pegylated liposomal doxorubicin plus gemcitabine was 4 (range, 1-8). At the time of analysis, 20 patients were still involved in the study. Median follow-up time was 8.2 months (range, 1.7-52.8 months). In the 20 patients who were still alive, the median follow-up time was 22.5 months (range, 2.3-52.8 months).
The primary study endpoint was objective response (Table 2). All patients were evaluable for response. Three (7.3%) patients had a complete response (CR) and, in particular, complete pathologic remission by chemotherapy alone was documented in 1 patient in the specimen who underwent resection. Moreover, 1 of these patients underwent liver transplantation after the response to chemotherapy; to date he is still alive (52.8 months from the start of systemic treatment).
Table 2. Response to Treatment
Complete response (CR)
Partial response (PR)
Overall response rate (CR+PR)
Seven (17.1%) patients had a partial response (PR) and, among these patients, 1 patient underwent surgical resection; in this case the TTP was 18.9 months.
Of 31 patients who had an elevated AFP level (>400 ng/mL) at baseline, a decrease in AFP of >20% after 2 cycles of treatment was observed in 20 (64.5%) patients.
The overall response rate (CR + PR) was 24.4%. The disease control rate (CR + PR + stable disease [SD]) was 58.5%. However, 17 (41.5%) patients had disease progression.
The median TTP of the entire population was 5.8 months (95% confidence interval [CI], 2.6-8.9 months) (Fig. 1) and the median OS was 22.5 months (95% CI, 4.6-40.3 months) (Fig. 2).
All patients were assessable for toxicity. Adverse events related to the study drugs are listed in Table 3. The combination of gemcitabine plus pegylated liposomal doxorubicin was generally well tolerated, and no treatment-related death occurred; moreover, there was no significant difference in tolerance to chemotherapy between patients with Child-Pugh stage A and B. Thirteen grade 3-4 adverse events were observed. Hematologic toxicity was the most common severe toxicity and consisted of neutropenia (17%), anemia (7%), and thrombocytopenia (2%). Nonhematologic grade 3-4 toxicity consisted of nausea/vomiting in (2%) 1 patient and stomatitis in 1 (2%) patient. Only 3 (7%) grade 1-2 hand-foot syndromes were observed. Liver toxicity was only grade 1-2, and the majority (61%) of patients had viral hepatitis at baseline.
Table 3. Toxicity Profile
HFS indicates hand-foot syndrome.
The results of univariate and multivariate analyses, carried out to identify factors predicting TTP and OS, are given in Table 4 and Table 5.
ECOG PS (P = .04) and Okuda staging (P<.001) were found to be independent predictors for TTP on multivariate analysis.
Multivariate analysis indicates that OS was significantly longer in patients with good PS (P <.001), earlier disease stage (P = .002) and lower AFP (<400 ng/mL) levels (P = .02).
Among the 31 patients with elevated serum AFP (>400 ng/mL) before treatment, a decrease in AFP of >20% after 2 cycles of chemotherapy was an independent favorable predictor for TTP (P<.001) and OS (P = .02) (Table 6).
Table 6. Decrease in AFP of >20% After 2 Cycles of Chemotherapy Associated With TTP and OS, Results by Univariate and Multivariate Analyses
AFP indicates alpha-fetoprotein; TTP, time to disease progression; OS, overall survival; HR, hazard ratio; CI, confidence interval.
HCC has been considered to be a chemotherapy-resistant tumor, and until recently there were no effective treatments that significantly increased survival in prospective randomized trials.17 The median survival of such patients was from 6 to 8 months.
Only with the advent of sorafenib was a significant increase in survival obtained; a randomized, prospective, double-blind, placebo-controlled phase 3 trial (SHARP) in patients with advanced HCC and Child-Pugh A demonstrated a significant survival benefit for sorafenib.10 In this study, median OS and median TTP were prolonged by nearly 3 months in the sorafenib group compared with the placebo group; median OS was 10.7 months versus 7.9 months (P<.001), and median TTP was 5.5 months versus 2.8 months (P<.001). Although sorafenib was shown to be effective, no patient had a CR, only 2% had a PR, and 71% had SD according to RECIST criteria. Recently, despite its effect on survival and TTP, in a draft guidance the National Institute for Health and Clinical Excellence suggested that sorafenib would not be a cost-effective first-line drug to treat advanced HCC.18 Also, the drug is indicated for treatment only in patients with preserved liver function (Child-Pugh A), at least in Italy.
Before the evidence of efficacy of sorafenib become available, conventional doxorubicin was among the most used and active agents in advanced HCC, with a response rate of approximately 25%.7 In a randomized study,6 free doxorubicin has been shown to be superior to palliative treatment of HCC, although a subsequent meta-analysis showed no efficacy of the drug in terms of survival.17 Conversely, the clinical usefulness of doxorubicin is limited by unacceptable toxicity, such as cardiotoxicity and myelosuppression, which may preclude adequate dosing.
To decrease its toxicity, new formulations have been created, including pegylated liposomal doxorubicin. This drug has shown efficacy in advanced HCC, with a response rate of 10% to 17% and a median OS of 6.5 months; in particular, median OS was 12 months in patients responding to treatment. Regarding safety, it showed a well-tolerated profile even in the presence of impaired liver function; hematologic toxicity and PPE were the most frequently reported side effects.19, 20
Another frequently used drug was gemcitabine, which demonstrated activity against human hepatoma cells in vitro21 and as a single agent or in combination chemotherapy regimens for advanced HCC; phase 2 trials have reported a response rate of about 18%,5 with favorable toxicity (finding grade 3-4 hematological toxicity in only about 10% of patients).
In light of these observations, the present study aimed at improving the efficacy of systemic chemotherapy with a favorable toxicity by combining 2 relatively active drugs in HCC with different mechanisms of cytotoxic action and potential synergistic activity, lack of cross-resistance, and nonoverlapping toxicity.
In this study, we analyzed the association of pegylated liposomal doxorubicin and gemcitabine; in particular, all patients received gemcitabine, a pyrimidine nucleoside analogue, followed by the administration of pegylated liposomal doxorubicin, a type 2 antitopoisomerase. Various studies11, 12 demonstrated that gemcitabine induces expression of all topoisomerase enzymes, and the subsequent administration of pegylated liposomal doxorubicin stabilizes the covalently bound topoisomerase-DNA complex,22 resulting in massive DNA damage and cell death.
With an overall response rate of 24%, a disease control rate of 58.5%, a progression-free survival of 5.8 months, and a median survival of 22.5 months, this combination regimen compares favorably with other systemic therapies for advanced HCC, including sorafenib and other targeted agents (Table 7).
Table 7. Summary of Important Prospective Clinical Trials
Predominant Etiology (%)
ORR, % (95% CI)
Median PFS/TTP (95% CI)
Median OS (95% CI)
ORR indicates overall response rate; CI, confidence interval; PFS, progression-free survival; TTP, time to disease progression; OS, overall survival; HBV, hepatitis B virus; NA, not available; PLD, pegylated liposomal doxorubicin; PIAF, cisplatin/doxorubicin/5-fluorouracil/α-interferon; GEMOX, gemcitabine/oxaliplatin; HCV, hepatitis C virus.
In the past, the attractive combination of free doxorubicin plus gemcitabine was analyzed by Yang et al.23 Unfortunately, they showed no CR, and PR was observed in 11.8% of patients; the median OS was 4.6 months, and the median TTP was 2.5 months. Likely, the lower clinical efficacy of this combination may be because of the lower therapeutic index of free doxorubicin than pegylated liposomal doxorubicin. They found frequent grade 3-4 hematological toxicity, including neutropenia in 51.4% of patients, anemia in 45.7% of patients, and thrombocytopenia in 25.7% of patients. Also, Eastern patients recruited in these studies have worse clinical features in general than those recruited in the Western world.24
In our study, we had 3 complete responses, and in 1 patient the complete remission was documented by histological examination. To our knowledge, only a few cases of complete response after systemic chemotherapy in advanced HCC have been described in the literature25, 26; 1 case was reported using the gemcitabine-oxaliplatin combination,26 and 4 patients had a complete pathological remission with a cisplatin/doxorubicin/5-fluorouracil/α-interferon regimen25; conversely, in this last chemotherapy regimen the toxicities were considerable; in fact, there were also 2 treatment-related deaths.
Moreover, our treatment was able to convert some inoperable lesions into operable lesions, increasing their overall survival; 1 patient with initially inoperable disease underwent liver transplantation and was alive and free from recurrence at 52.8 months from treatment.
Furthermore, the TTP obtained in our study was similar to that shown by sorafenib, despite the finding that in the sorafenib trial only 5% of patients had Child-Pugh B; indeed, the TTP with gemcitabine-pegylated liposomal doxorubicin regimen was 5.5 months, and 53.7% had Child-Pugh B. Thus, our regimen may be an effective treatment for patients not eligible for sorafenib treatment because of altered liver function.
In the past, the combination of gemcitabine and pegylated liposomal doxorubicin for advanced HCC was also analyzed by Poh et al.27 They showed a poor clinical efficacy of the treatment; no patient achieved a CR or PR, with 2 (22%) patients with SD; the median OS was 2.8 months. In this study, the schedule was identical to that used in our study, although the sequence of administration of the drugs is not reported; moreover, they evaluated only 9 patients with advanced HCC, among whom there were 4 (44%) patients already treated with prior systemic chemotherapy (pegylated liposomal doxorubicin and capecitabine) and 8 (89%) patients who had prior locoregional treatment. Furthermore, 8 (88%) patients had extrahepatic metastases. Other differences from our study were that all patients were from Asia and predominant etiology of HCC was chronic hepatitis B virus infection (90%). Thus, it is not possible to make a comparison with our study, which used this combination in chemonaive Western patients, only 46.3% of whom had metastases, 53.6% of whom had prior locoregional treatment, and in whom predominant etiology was hepatitis C virus (48.8%).
We also conducted an exploratory analysis of predictive factors for TTP and OS. Independent predictors for improved TTP and OS were good ECOG PS and earlier stage disease, whereas lower serum AFP was an independent predictor only for improved OS. Moreover, our treatment was effective independently of liver function, demonstrating a low hepatotoxicity of the chemotherapeutic combination.
Interestingly, a decrease in AFP of >20% after 2 cycles of chemotherapy was an independent favorable predictor for TTP and OS. Only 1 prior study showed AFP response to be a predictive factor for survival in patients undergoing systemic chemotherapy28; they analyzed 117 patients with elevated serum AFP (>20 ng/mL), showing AFP responders had better survival than nonresponders. Similar data were originally reported for tamoxifen treatment in HCC.29
Nevertheless, the toxicity of this combination was acceptable; the most severe adverse effects were neutropenia (17%) and anemia (7%), whereas the expected PPE appeared only in 7% of patients.
In conclusion, although our trial was a small phase 2 study, the combination of gemcitabine plus pegylated liposomal doxorubicin proved effective and safe in patients with advanced HCC, even in patients with impaired liver function. In particular, patients with good ECOG PS, early stage disease, and low serum AFP could derive the greatest clinical benefit. In particular, this combination may be useful in chemonaive patients not eligible for sorafenib treatment because of altered liver function. However, further randomized studies are needed, and it could be useful to evaluate this combination alone or in association with the newer targeted agents such as sorafenib, erlotinib, and bevacizumab.