From March 2006 to February 2008, 32 patients with advanced HNC were enrolled in the study: 23 in cohort A and 9 in cohort B (). Twenty-five patients (78%) had squamous cell carcinoma. Twenty patients had newly diagnosed tumors, nine patients had recurrent tumors and three patients had a second primary cancer. Thirty-one patients were assessable for toxicity and 30 for DLT, while 2 patients were replaced (); one patient in cohort A withdrew consent due to transportation issues on the second week of treatment and was not considered assessable for toxicity; another patient in cohort A had a sudden death probably from aspiration that was not considered treatment related.
Among 23 patients in cohort A, 19 received three cycles of pemetrexed (83%), 3 patients received two cycles and 1 patient (who withdrew from study) received one cycle. In cohort B, eight patients (89%) received three cycles and one patient two cycles. A total of five patients (16%) were not able to complete the planned cycles of pemetrexed. In cohort A, 1 patient received nine weekly doses of cetuximab, 14 patients received eight doses, 4 patients seven doses, 3 patients four to six doses, and 1 patient one dose. In cohort B, eight patients received eight weekly doses of cetuximab and one patient seven weekly doses. The radiation doses delivered (median [range]) were 70 Gy (38–74 Gy) in cohort A and 66 Gy (60–70 Gy) in cohort B. All but three patients in cohort A received at least 70 Gy (six of these patients had delays in radiotherapy of 2–10 days); one patient discontinued radiotherapy at 40 Gy due to dose-limiting febrile neutropenia and grade 4 mucositis, one patient had sudden death after receiving 38 Gy, and one withdrew from the study on the second week of treatment without developing significant toxic effects. In cohort B, all but one patient, who developed neutropenic fever, received 66 Gy.
Five patients experienced a DLT (). In cohort A, one patient on each dose level of pemetrexed developed febrile neutropenia with ANC <500/μl. An additional patient at dose level −1 developed fever with ANC >500/μl, which did not meet criteria for DLT. Therefore, dose level +1 (500 mg/m2) was determined as the MTD of pemetrexed for not previously irradiated patients. In cohort B, two patients, both at pemetrexed dose level +1, experienced a DLT. One had febrile neutropenia, whereas the other presented with perforated duodenal bleeding and subsequent grade 5 sepsis with grade 4 thrombocytopenia and neutropenia. Hence, the MTD of pemetrexed for previously irradiated patients was determined at 350 mg/m2.
Hematological toxic effects are shown in . In cohort A, six patients developed grade 3 and five patients grade 4 neutropenia. In cohort B, three patients experienced grade 4 neutropenia. Serious thrombocytopenia was rare; one patient in group A and two patients in group B developed grade 3–4 thrombocytopenia.
Grade 2–4 hematologic toxic effects by dose level
Mucositis, dysphagia, rash and dermatitis were the most common non-hematologic toxic effects (). Sixteen patients developed grade 3 mucositis and one patient (at dose level 3) developed grade 4 mucositis. In cohort A, the rate of grade 3–4 mucositis was 4 of 7 (57%), 5 of 7 (71%) and 8 of 9 (89%) at dose levels −1, 0, and +1, respectively. Gastrostomy tube was placed in six patients during treatment, whereas seven patients had G-tube placed before treatment initiation. Five patients remained G-tube dependent upon last follow-up; however, only one of them in cohort B, who had G-tube before treatment, was also progression free. Nine patients developed infection with the most common being aspiration pneumonia (four patients). Thirteen patients developed grade 1 and two grade 2 hypomagnesemia that was reversible. All patients developed cetuximab-associated skin rash of any grade, and among them only four patients had grade 3 rash.
Grade 2–4 non-hematologic toxic effects by dose level
Of 21 patients assessable for response in cohort A, 12 patients (4 on each dose level) achieved a complete response and 7 a partial response (4 at dose level +1; 2 at dose level 0; 1 at dose level −1). Of seven assessable patients in cohort B, four patients achieved a complete response (three at dose level 0; one at dose level +1) and three a partial response (two at dose level 0; one at dose level +1). With a median follow-up of 36 months, the 3-year PFS was 41% in cohort A and 22% in cohort B, and the 3-year OS was 55% and 37%, in cohorts A and B, respectively. Two patients died during treatment; one death was possibly treatment related. One additional patient developed severe neck fibrosis and facial edema and died without documented disease progression 8 months after treatment completion. Among 13 patients with locally advanced stage IV SCCHN (5 with an oropharyngeal primary, 4 laryngeal, 2 hypopharyngeal, and 2 with unknown primary) who completed radiotherapy (an additional patient who did not complete radiotherapy progressed and died), 6 patients (46%) remained progression free but only 2 recurred locoregionally (i.e. 85% locoregional control) with a median follow-up of 33 months. Seven of these 13 patients had available tumor specimens that were tested for human papillomavirus (HPV) by in situ hybridization and p16 by immunohistochemistry; all but 1 patient had HPV- and p16-negative tumors (the patient with HPV-/p16-positive tumor had a base of tongue primary and developed disease progression in mediastinal lymph nodes). A patient with anaplastic thyroid cancer and a patient with medullary thyroid carcinoma remained progression free after 32 and 36 months of follow-up, respectively.
MTHFR and TS genotypes
Twenty-eight patients underwent genotyping for MTHFR and TS. The frequency of the MTHFR and TS polymorphisms are shown in . The genotype frequencies were in Hardy–Weinberg equilibrium. Polymorphisms in MTHFR and TS did not correlate with severity of treatment-related toxic effects (data not shown). Furthermore, the different variants of MTHFR and TS did not seem to affect either the response rate or the survival.