Phase 1 study of cetuximab in combination with 5-fluorouracil, cisplatin, and radiotherapy in patients with locally advanced anal canal carcinoma




This study sought to determine the feasibility and recommended phase 2 dose (RP2D) of the combination of cetuximab with chemoradiotherapy based on 5-fluorouracil (5-FU) and cisplatin (CP) in locally advanced anal canal carcinoma.


Cetuximab was administered on days 1, 8, 15, 29, 36, 43, and 50 (400 mg/m2 initial dose, then 250 mg/m2/week) concurrent with total dose radiation of 55 to 59 Gy, both starting on day 1. Escalating doses of 5-FU (96-hour infusion) and CP (2-hour infusion), both on days 1 and 29, were administered according to the following design: starting dose level (0) 5-FU/CP = 800/60 mg/m2/day and up to dose level (+2) 5-FU/CP = 1000/80 mg/m2/day.


Dose-limiting toxicity (DLT) events (uncontrolled diarrhea or febrile neutropenia) occurred in 3 of 14 assessable patients receiving escalated dose of 5-FU/CP, with 1 in dose level (0) and 2 in dose level (+2). The RP2D was 5-FU/CP = 800/80 mg/m2/day. Because of unexpected non-DLT treatment-related grade 3 (G3) adverse events (AEs) such as thrombosis/embolism, syncope, and infection occurring in ≥ 20% of patients, a safety expansion cohort with an additional 9 patients was investigated with the RP2D. The most frequent G3/G4 AEs evaluated in 23 patients were radiation dermatitis (12 patients), diarrhea (10 patients), thrombosis/embolism (6 patients), and infection (5 patients). The study was closed due to these severe AEs, although no G5 AEs occurred. Twenty of 21 patients (95%) achieved pathological complete response at primary tumor. With a median follow-up of 43.4 months, the 3-year locoregional control rate was 64.2%.


Cetuximab could not be integrated with chemoradiotherapy-cisplatin–based therapy due to the high toxicity rate. However, efficacy is encouraging and further investigation of an epidermal growth factor receptor–targeted agent (other than cetuximab) concurrent with chemoradiation should be pursued. Cancer 2013;119:2973—2980. © 2013 American Cancer Society.


Over the last decade, several strategies to improve chemoradiotherapy (CRT) outcomes in anal carcinoma have included the addition of induction or consolidation chemotherapy to the CRT backbone, radiotherapy (RT) dose escalation, and the integration of different chemotherapy agents into the CRT backbone. Yet, data from randomized phase 3 trials have demonstrated that none of those efforts have improved overall survival (OS) or disease-free survival (DFS) rates in this population.[1-3] Moreover, preliminary results of the Cancer Research United Kingdom Anal Cancer Trial (ACT II) showed equivalency for chemoradiation cisplatin (CP)-based therapy compared with chemoradiation mitomycin-C–based therapy in terms of DFS, with significantly less acute grade 3/grade 4 (G3/G4) hematological toxicity in favor of the CP group.[3]

An emerging strategy in oncology is to incorporate newer, biologically active, targeted therapies into established CRT schedules. One attractive target is the epidermal growth factor receptor (EGFR), a type I tyrosine kinase membrane receptor that regulates key functions in epithelial malignancies.[4] EGFR is often overexpressed and/or abnormally activated in almost all squamous cell carcinomas (SCCs), including anal carcinomas (in up to 90%).[5]

Cetuximab (Erbitux; Merck KgaA, Darmstadt, Germany) is an immunoglobulin G1 monoclonal antibody that specifically targets EGFR, competitively inhibiting ligand binding and ligand-dependent downstream signaling.[6] Phase 1 clinical trials demonstrated the safety of cetuximab combined with CP[7] and cetuximab combined with radiation therapy[8] in SCCs other than anal carcinoma. Also, in phase 3 clinical trials, adding cetuximab to first-line chemotherapy or RT improves outcomes in various types of cancers.[9-11]

Because of encouraging results of clinical trials with cetuximab in different tumors, the need for more effective therapies, particularly in locally advanced anal canal carcinoma (LAACC),[12] and the high expression of EGFR in anal carcinoma, we set out to develop a regimen to evaluate the safety and activity of cetuximab combined with 5-fluorouracil (5-FU)/CP concurrent with radiation for LAACC.


Study Design and Statistics

This was an open-label, dose escalation, and single-center phase 1 study. The primary objectives were to determine the recommended phase 2 dose (RP2D) and safety of 5-FU and CP in combination with cetuximab and concurrent RT in LAACC. Clinical activity was also assessed. The study and its subsequent amendments were approved by institutional review boards (IRBs) and National Ethics Review Boards. Informed consent was obtained from all patients. Descriptive statistics on patient characteristics, analysis of toxicities, and outcomes were performed for all patients.

Eligibility Criteria

Patients with pathologically confirmed SCC of the anal canal were eligible if they were 18 to 70 years old, had a Karnofsky performance score of at least 70%, had category T3 to T4 tumors or were node-positive (defined by the AJCC Cancer Staging Manual[13]) and adequate renal, hepatic, and bone marrow function. For patients presenting with palpable inguinal nodes felt to be clinically positive and/or with inguinal nodes on imagery, inguinal node status was assessed by needle aspiration cytology or biopsy confirmation. Enlarged inguinal, perirectal, or pelvic nodes on computed tomography (CT) scan, ≥ 1.0 cm were considered to be clinically positive. Exclusion criteria included T1 or M1 tumor, severe comorbidities (uncontrolled infection, psychiatric disorders, cardiovascular disease, collagenosis, and known human immunodeficiency virus infection), or major malignancy (unless successfully treated and disease-free for at least 5 years). Pregnancy or women with childbearing potential who lacked effective contraception were also excluded.

Study Assessment and Follow-Up

Patient demographics and medical history were evaluated before treatment. Adverse events (AEs), physical examination, digital rectal examination, vital signs, performance status, hematology, clinical chemistry, and creatinine clearance were measured before treatment and weekly during treatment. Toxicities were assessed using Common Terminology Criteria for Adverse Events, version 3.0.

Imaging staging studies (chest, abdominal, and pelvic CT and pelvic magnetic resonance), rectosigmoidoscopy, electrocardiography, and echocardiography were performed before treatment and 8 weeks after treatment. Tumor specimens from pre- and posttreatment biopsies were collected for molecular analysis according to the Brazilian National Tumor Bank guidelines. Assessment of complete pathological response (pCR; defined asabsence of carcinomatosus cells in the specimens analyzed) on primary tumor (pT0), and clinical response (CR) according to Response Evaluation Criteria in Solid Tumors (RECIST), took place at the eighth week after treatment. Afterward, follow-up visits were performed every 3 months for 2 years, and annually thereafter. Locoregional control (LRC) was defined as the time from day 1 of treatment to either disease recurrence in the anus and/or inguinal or pelvic lymph node, or death from any cause other than distant metastasis.

Study Treatment

Standard cetuximab dose combined with escalated doses of 5-FU/CP concurrent with moderate dose of RT were investigated (Fig. 1).

Figure 1.

Study treatment schema is shown for cetuximab added to chemoradiation based on 5-fluorouracil (5-FU)/cisplatin for locally advanced anal canal carcinoma.


Intravenous administration of 5-FU (96-hour infusion) and CP (2-hour infusion) were administered intravenously on days 1 and 29 preceding cetuximab infusion. Four dose-level escalations were planned (Table 1) in consecutive cohorts of 3 patients, with expansion to 6 patients if 1 of the 3 initial patients experienced a dose-limiting toxicity (DLT). If 2 or more of 6 patients exhibited DLT at the “starting dose” level (0), then dose level (−1) would be the recommended dose to be tested. Up to 6 patients would be treated at this dose level to ensure information on the safety profile at that dose is complete. If 2 patients presented DLT at dose level (−1), the study should be stopped. The maximum tolerated dose (MTD) was defined at the dose level (+1) or dose level (+2), if 2 of 3 or 2 of 6 patients developed DLT. The immediate dose level below the MTD was considered the RP2D. Doses were escalated to the next cohort after 12 weeks (8 weeks of treatment plus 4 weeks of toxicity observation) if there was no evidence of limiting toxicity after this period. No intrapatient dose escalation was allowed.

Table 1. Dose Escalation Based on Dose-Limiting Toxicity
Dose LevelNo. of Patients AccruedCetuximab First Dose (mg/m2)Cetuximab Weekly (mg/m2)5-FU (mg/m2)Cisplatin (mg/m2)Daily RT Dose (Gy)
  1. Abbreviations: 5-FU, 5-fluorouracil; RT, radiotherapy.


DLTs were defined as: neutrophil count < 0.5 × 109/L for more than 5 days, febrile neutropenia, platelets < 25 × 109/L; diarrhea ≥ G3 for more than 5 days despite optimal loperamide use (uncontrolled diarrhea); other effects of > G3 thought to be treatment-related except skin rash ≥ G3 outside in-field radiation, G4 in-field RT of skin; and patients received dose intensity < 70% for each drug (cetuximab, 5-FU, and CP).


All patients received a dose of 45 Gy in 25 fractions of 1.8 Gy over 5 weeks to the tumor with margin and risk lymph nodes, as iliac and inguinal, using photon energy of 6 mV and 2 opposite fields. Fourteen patients were positioned in dorsal decubit, and the posterior field was smaller to protect femoral heads in cases in which it was possible, after a planning CT scan proved that an anterior field was able to cover all inguinal nodes. Three-dimensional conformal RT was used in 9 patients, with delineation of target and organs at risk, based on the Radiation Therapy Oncology Group atlas. Delineation of lymph nodes included common iliac, internal and external iliac, and inguinal chair, and margin around the tumor also received RT. These patients received field-in-field treatment.

After 45 Gy, all patients received an additional boost of 10 to 14 Gy in 2-Gy fractions (total dose of 55-59 Gy in 30-32 fractions over 6.0-6.5 weeks), to the original tumor volume, with 1-cm margin and previously node-positive. Four fields (expanded for AP-PA) were used in patients at the same decubit of the first phase or a direct photon or electron perineal field, with patient in lithotomy position, in cases that tumor was preferentially of margin. For node-positive patients, an anterior electron beam was performed to the inguinal affected region, using the most appropriate energy. In patients who received conformal treatment, 3 fields were used. There were no planned modifications of RT dose for toxicity, but treatment breaks were permitted.


Starting on day 1 of RT, patients received weekly intravenous cetuximab (400 mg/m2 during 120-minute infusion on day 1 and 250 mg/m2 during 60-minute infusion on days 8, 15, 22, 29, 36, 42, and 50) at least 2 hours before radiation.

For dose modifications, if neutrophil counts were < 1000/mm3 and/or platelets were < 50,000/mm3, 5-FU/CP dose was reduced by 50%. If serum creatinine value was between 1.5 and 2.0 mg/mL, with G2 peripheral neuropathy and G2 ototoxicity, CP dose was decreased by 50%. The next course of CP was not administered if these toxicities reached G3.

The dose of 5-FU was decreased by 50% if G3 diarrhea or stomatitis occurred. Options for cetuximab G3 skin toxicity included a 1-week delay, with dose reductions in increments of 50 mg/m2 (minimum cetuximab dose of 150 mg/m2), or a 2-week delay without a dose reduction. A patient was excluded from the study in case of cetuximab G4 skin toxicity.


Patient Characteristics

The study reached its MTD and RP2D goals after the accrual of 14 patients between August 2006 and November 2008. However, exacerbation of non-DLT G3 events such as diarrhea and infections and occurrence of G3 unexpected deep venous thrombosis and G3syncope AEs in ≥ 20% of patients led the investigators and IRB to recommend the use of RP2D in a safety expansion cohort of an additional 9 patients before proceeding to the phase 2 trial. The accrual was restarted on November 2009 until November 2011.

Patient characteristics are summarized in Table 2. All patients (n = 23) were included for toxicity analyses, and 21 were evaluated for clinical and pathological responses. One patient was found to have an ineligible histologic analysis (adenocarcinoma) on posttreatment pathological evaluation, and another patient was considered stage IV after comparative analysis of CT scans before and after treatment

Table 2. Patient Characteristics (N = 23)
CharacteristicNo. of Patients%
Age, y  
Disease stage at diagnosis 

Dose Escalation and DLTs

Seven patients were treated at the dose level 0 (5-FU/CP = 800/60 mg/m2). One of the initial 3 patients in this first cohort experienced a DLT (G4 diarrhea and febrile neutropenia). Four additional patients were included in this cohort, and none among them experienced DLT. The addition of 1 extra patient was necessary due to consent withdrawn from 1 patient for cetuximab administration during boost radiation. In addition to DLT, this dose was associated with G3 diarrhea in another 3 patients and mandatory use of loperamide, ciprofloxacin therapy, and outpatient hydration, to minimize the severity of the diarrhea, was established through a protocol amendment. In the next dose level (+1), no additional DLTs occurred. At level (+2), 2 of the 4 patients developed DLTs (uncontrolled diarrhea and febrile neutropenia). Therefore, the dose of level (+1) was selected as the RP2D.

Toxicities and Treatment Delivered

Toxicities of G3 or G4 for all 23 patients (14 initial patients plus 9 patients from the expansion cohort) enrolled in this phase 1 study are listed in Table 3. Common G3/G4 toxicities occurring in ≥ 2 patients at least once were: in-field radiation dermatitis (52%), diarrhea (43.5%), thrombosis/embolism (26%), infection (21.4%), neutropenia (17.4%), dehydration and metabolic abnormalities (17.4%), and decreased performance status (17.4%). The thrombosis/embolism AEs were the only G4 toxicities reported by more than 2 patients (13%). No treatment-related deaths occurred.

Table 3. Incidence of CTCAE Grade 3 and 4 On-Treatment Adverse Events During Phase 1
 5-Fluorouracil/Cisplatin Dose 
 800/60 mg/m2 (n = 7)800/80 mg/m2 (n = 12)1000/80 mg/m2 (n = 4)All Patients (N = 23)
Adverse EventsN(%)N(%)N(%)N(%)N(%)N(%)N(%)N(%)
  1. Abbreviations: G3, grade 3; G4, grade 4; CTCAE, Common Terminology Criteria for Adverse Events version 3.0.

Radiation dermatitis5(71)0(0)4(33)1(8)2(50)0(0)11(48)1(4)
Infection/febrile neutropenia0(0)1(14)0(0)1(8)1(25)0(0)1(4)2(8.7)
Infection with normal or grade1 or 2 neutrophils1(14)0(0)0(0)0(0)1(25)0(0)2(8.7)0(0)
Performance status0(0)1(14)1(8)1(8)0(0)1(25)1(4)3(13)
Anal/rectal pain0(28)1(14)2(16)0(0)0(0)0(0)2(8.7)1(4)

In-field radiation dermatitis is a typical skin toxicity within the radiation portal with concurrent chemoradiation. Only one G4 (4.3%) in-field radiation dermatitis was reported, and 11 patients (47.8%) developed G3 toxicity. However, some acute G3/G4 AEs were exacerbated beyond those expected with definitive concurrent chemoradiation such as diarrhea in 10 patients (43%), thrombosis/embolism in 6 patients (26%), and infection 5 patients (21%). Hospitalization was necessary for recovery of these G3/G4 AEs in all patients with infection and half of patients with diarrhea and thrombosis/embolism. Three additional patients required hospitalization, including 1 with pancreatitis related to cetuximab, 1 to perform colostomy due to vaginal fistulae, and another to improve performance status and analgesia. The last dose of 5-FU was suspended in 1 patient (due to G4 radiodermatitis) and was decreased by 50% in 2 patients (due to G3 diarrhea); all 3 of these patients were treated with the RP2D.

All patients started weekly cetuximab concurrent with CRT. The median number of cetuximab doses administered was 7 (range, 2-8 doses). Five patients received fewer than 6 doses of cetuximab. In 2 patients, cetuximab was discontinued due to DLTs (G4 diarrhea and G3 diarrhea plus G3 infection) after the fifth dose of treatment. In 2 patients, cetuximab was discontinued after the third dose due to G3 hypersensibility reaction and G3 pancreatitis. In another patient, 4 doses of cetuximab were omitted due to G4 thrombosis/embolism. The average dose intensity for cetuximab, 5-FU, and CP were 85%, 95%, and 99%, respectively.

Twenty-two patients completed the full preplanned 55- to 59-Gy course of RT (one patient received 45 Gy due to G3/G4 neutropenia toxicity). The median duration of total RT was 64 days. Only 2 patients finished RT at prescribed protocol dose in the preplanned time, and RT interruptions were often necessary due to treatment-related toxicities, and occurred in approximately 65% of the patients. RT interruptions took place after the second cycle of chemotherapy and the majority of patients (73%) had received at least 34 Gy of pelvic irradiation.

Events Leading to Early Study Closure

This study was closed before moving to phase 2 due to toxicity concerns, in particular, 6 thrombosis/embolism events attributed to treatment. Early in the study, 3 patients developed jugular and subclavian G3 venous thrombosis, where peripheral inserted central catheters (PICCs) were placed. Although none of them suffered embolic events, they had their catheter removed because of subsequent G3 infection or febrile neutropenia.

The investigators, sponsor, and the IRB carefully reviewed these 3 G3 AEs and considered their potential relationship to the investigational drug and the possible toxicity as a result of contribution of other events (diarrhea, infection, hospitalization, and PICCs). The decision was to continue study accrual with replacement of the PICC for totally implanted central venous access device and particular vigilance for further significant toxicity.

Ultimately, during the expansion cohort, a series of 3 nonfatal but significant G4 thrombosis/embolism AEs occurred in short succession, prompting study closure as a result of patient safety concerns. Patients 24 and 26 developed left cervical and superior and left limb edema followed by pulmonary embolism after receiving the second and sixth dose of cetuximab, respectively. Hospitalization was necessary for anticoagulation and line removal, and both patients had recovered uneventfully. Finally, patient 29, a 56-year-old woman, developed thoracic pain and dyspnea 12 days after hospital discharge for treatment of G4 febrile neutropenia and G4 diarrhea. Imaging angiography/CT scan revealed thrombosis sequelae due to significant luminal reduction of superior vena cava.

Efficacy Assessment

Of the 21 patients assessable for clinical response, 20 had major responses (9 complete responses [CRs] and 11 partial responses [PRs]), yielding a response rate of 95% (95% confidence interval [CI] = 78%-99%).

Twenty of 21 patients (95%) achieved pCR at primary tumor (pT0). Tumor was not locally controlled in 6 cases (2 CRs, 3 PRs, and 1 stable disease). Among these 6 patients, 4 have been continuously disease-free from their anal cancer after undergoing abdominal perineal resection, with follow-up times ranging from 13.8 to 41.3 months; 2 patients died, both with locoregional disease and distant metastasis at 18 months and 42 months. With a median follow-up of 43.4 months, the 3-year LRC was 64.2%; 95% CI = 57.15%-70.40%.


Here, we report that the integration of cetuximab to CRT-cisplatin–based treatment in LAACC was not feasible in our study due to the high rate of toxicity. Analysis of toxicity profile of all 23 patients accrued in this study demonstrated that G3/G4 rates of diarrhea and infection seem to be twice as frequent than those rates experienced in other anal cancer trials (21%-31%).[1, 14] In general, treatment-related toxicities in this trial were associated with interruption of RT in 15 patients (65%), hospitalization in 12 (52%), and omission of more than 2 cetuximab doses in 5 individuals (17.4%). The contribution of a synergistic effect of cetuximab on these toxicities is possible. In line with that, a meta-analysis of randomized trials comparing cetuximab-based therapy versus non–cetuximab-based therapy demonstrated an increase of G3/G4 infection and diarrhea rates upon use of cetuximab.[15]

More recently, the French multicenter phase 2 ACCORD16 trial that explored CP-based CRT combined with cetuximab was stopped after accruing 16 patients with LAACC due to a high rate of toxicity and an insufficient efficacy.[16] Severe toxicity rates such as diarrhea (6 patients), fatigue (6 patients), neutropenia (5 patients), hyponatremia or hypokalemia (4 patients) were very similar to our findings, except for a high rate of lymphopenia (8 patients) found in the ACCORD16 trial. Conversely, a phase 2 study—the Eastern Cooperative Oncology Group (ECOG) E3205—had recruited 28 patients to receive 2 cycles of neoadjuvant CP/5-FU alone prior to cetuximab/CP/5-FU/RT. Expedited reporting was required for type I (any G5, or G4 cardiac) or type II (G4 RT skin, diarrhea) adverse events, with prespecified rates of > 5% or > 20%, respectively, defined as unacceptable. Preliminary safety and efficacy data appear encouraging with adverse type I = 4% and type II = 4%, and a locoregional failure rate at 2 years of 13%.[17]

In addition, a phase 1 study of cetuximab with pelvic CRT in cervical cancer also found excessive toxicities.[18] Taken together with the anal carcinoma data, there are now up to 2 clinical trials for which preliminary results show that cetuximab combined with pelvic CRT in SCC generates limiting toxicity.

In our study, an unexpected occurrence of 6 significant G3/G4 thrombosis/embolism events led to early study closure. The possible mechanism that links cetuximab and venous thromboembolism (VTE) events may be related to its antiangiogenic effect.[19, 20] Recently, a meta-analysis of 11 randomized clinical trials (7611 patients) was conducted to better evaluate the potential risk for VTE associated with anti-EGFR agents. The incidence of VTE for the cetuximab arm was 5.7% compared with 3.9% for the control arm, a difference that was statistically significant (P = .010).[21] Although, cetuximab ultimately retains antiangiogenic properties, inhibition of angiogenesis is probably not the only mechanism at work; even regarding a specific vascular endothelial growth factor inhibitor such as bevacizumab, there is uncertainty of its role in VTE development.[22]

Cisplatin is an endothelium-damaging agent and its direct role in thromboembolic toxicity has been suggested by a large retrospective analysis of 932 patients treated with CP-based chemotherapy for any type of malignancy; 169 patients (18%) developed thromboembolic events with this type of chemotherapy.[23]

It is remarkable that all patients (n = 6) in our phase 1 trial developed symptomatic deep venous thromboses in the upper extremities where a central venous catheter was placed, which in 3 cases was complicated by pulmonary embolism. In addition, hospitalizations were necessary in half of patients for treatment of other severe toxicities previous to thrombosis events, and 2 of them were due to febrile neutropenia, which is known factor that increased VTE incidence even in the low-risk subgroup.[24] We had a constellation of risk factor for VTE involved in these patients that could explain the high incidence. However, no clear conclusion may be drawn on possible additive or synergistic effects between cetuximab and CRT on the high incidence of thrombosis/embolism. It should be highlighted, though, that no thrombosis/embolism was seen in the ACCORD16 trial, which is at odds with our observations.

Despite the DLT observed in this study, the preliminary LRC rates are encouraging in LAACC. With a median follow-up time of 43.8 months, the 3-year LRC rate was 64.2%, which is superior to the historical experience at our institution for a predominantly LAACC population receiving CRT-CP–based treatment.[14] LRC rates also compare favorably with data reported in the major randomized trial supporting concurrent CRT.[3]

Some may argue that our median duration of radiation treatment of 64 days was too long, and it would be expected to show a reduction in the local control rate, which contradicts the results. One possible explanation might arise from additive effects of the weekly intravenous schedule of cetuximab in combination with daily RT, which may mean an enhanced cell destruction over all 30 to 35 fractions due to cetuximab clearance. These effects may explain the pCR rate of 95% on primary tumor (pT0) and predict for locoregional outcome.

To pursue further development of phase 1 and 2 clinical trials with addition of anti-EGFR agents to CRT in LAACC, is advisable to categorize the patients according to VTE risk models and well-defined early stopping rules.[25] Also, cetuximab's synergistic effect on diarrhea and infection/febrile neutropenia proves to be a problem of overlapping toxicity with pelvic chemoradiation.[16, 18] These acute toxicities may be reduced by the use of other EGFR-targeted agents[26] and mainly with the utilization of intensity-modulated RT.[27]

At present, cetuximab should not be integrated with 5-FU/CP concurrent with RT in cases of LAACC due to the high rate of toxicity, particularly an unacceptable VTE incidence rate of 26%. However, efficacy is encouraging, and further investigation of an EGFR-targeted agent (other than cetuximab) concurrent with chemoradiation should be pursued.


The study drug (Erbitux) was supplied by Merck KGaA, Darmstadt, Germany.


The authors made no disclosure.