Dose intensity comparison between weekly and 3-weekly Cisplatin delivered concurrently with radical radiotherapy for head and neck cancer: A retrospective comparison from New Cross Hospital, Wolverhampton, UK

Abstract

Aims. In this retrospective comparison, we describe the differences in dose intensity, delays and toxicity between weekly Cisplatin and 3-weekly Cisplatin given concurrently to patients with locally advanced squamous head and neck cancer (SCCHN) at New Cross Hospital, Wolverhampton. Materials and methods. Fifty-one patients received radical Cisplatin based chemoradiotherapy for stage 4a SCCHN of the head and neck between September 2000 and December 2004. Twenty-seven patients were treated with 3-weekly inpatient Cisplatin for 3 cycles (20 patients-80 mg/m²; 7 patients-100 mg/m²) concomitantly with radiotherapy (66–70 Gy/33–35 fractions). Twenty-four patients received a similar radiotherapy schedule but received weekly Cisplatin 33–40 mg/m². Results. More patients received a higher cumulative dose of at least 240 mg/m² if given weekly Cisplatin 40 mg/m² or 3-weekly Cisplatin 80 mg/m² compared with those receiving Cisplatin 3-weekly 100 mg/m² (p=0.04). Maximum cumulative dose achievable in the latter group was only 200 mg/m² and none achieved the full 3 cycles. Mean Cisplatin dose in the weekly Cisplatin 40 mg/m² regime (mean 202 mg/m²) and 3-weekly arm of 80 mg/m² (mean 203 mg/m²) was higher than that reached if given 3-weekly Cisplatin 100 mg/m² (mean 180 mg/m²) although statistically insignificant (p=0.39) due to the small number of patients. More delays (29% vs. 41%) and omission of chemotherapy (5.6% vs. 17.4%) occurred in the 3-weekly compared with the weekly regime. Toxicity, radiotherapy overall treatment time and delays were similar between the two groups. Conclusion. Delivery of 100 mg/m² Cisplatin 3-weekly with radiotherapy was less tolerated than 40 mg/m² weekly and resulted in less patients achieving cumulative dose beyond 200 mg/m², potentially lowering chemotherapy dose intensity.

Concurrent chemotherapy has become increasingly more important in combination with radiotherapy for squamous cell carcinoma of the head and neck. The Pignon meta-analysis has suggested this approach may provide an improvement in overall survival by an absolute figure of 8% when compared to radiotherapy alone (p < 0.0001) [1]. However a variety of chemotherapy agents were used in the randomized trials (single agents or combinations) analysed and the optimal choice of chemotherapy remains unclear.

Cisplatin is among the most common agents used in combination with radiotherapy as well as one of the most studied. It has radiosensitizing properties and its toxicity does not overlap radiotherapy. The MACH-NC update suggested that the magnitude of benefit was higher (p < 0.01) for platinum based chemoradiotherapy (0.75) compared concomitant therapy with other cytotoxic agents (0.86) [2].

Single agent Cisplatin has been used in various different schedules when given along with radiotherapy. Three recently published trials using high dose Cisplatin 100 mg/m² given every 3-weekly concurrently with radiotherapy have shown better local control compared with radiotherapy alone, Two of these have indicated an overall survival benefit [2–4]. However, Cisplatin at a dose of 100 mg/m² contributed to more severe mucositis. Furthermore, the incidence of late toxicity remain unclear as there has been relatively little recording and reporting of the late effects from chemoradiation.

Others have attempted to reduce the early toxicity and potential late toxicity by reducing total cumulative dose and fractionating the administration of Cisplatin. Huguenin et al. employed 2 cycles of Cisplatin which was divided into five doses of Cisplatin 20 mg/m² given daily [5]. The results of this study suggest less systemic toxicity and mucositis without compromising local control and overall survival. Data from two further other trials employing a lower cumulative Cisplatin dose or more fractionated Cisplatin dosing [6], [7] have shown comparable results. In light of these data and the observation that a significant number of patients could not receive the third Cisplatin dose of 100 mg/m² in a number of studies, it is important to question the long held view that 3 weekly Cisplatin at a 100 mg/m² should be the standard regime for combination with chemotherapy.

Cisplatin 40 mg/m² has been used successfully in concurrent chemoradiation for squamous cell carcinoma of the cervix with relatively low acute toxicity. The largest head and neck study delivering weekly Cisplatin 40 mg/m² randomized 350 patients with nasopharyngeal carcinoma to concurrent chemoradiation or radiotherapy alone. Concurrent chemoradiation was found to be well tolerated despite the large radiotherapy fields employed in the treatment of nasopharyngeal carcinoma with an acceptable 4.6% of these patients having grade 4 mucositis and 12.6% grade 3 leukopenia [8].

Since January 2002 weekly cisplatin as been the standard cytotoxic schedule for concomitant administration with radiotherapy for squamous cell carcinoma of the head and neck (excluding nasopharyngeal carcinoma) in our department. Prior this, 3-weekly cisplatin had been the standard schedule. Delivering cisplatin weekly as an outpatient avoided the problem of inadequate inpatient beds and the resultant delay in the administration of chemotherapy.

The aim of our audit was to review the difference in dose intensity, delays and toxicity between weekly Cisplatin and 3-weekly Cisplatin given concurrently to patients with locally advanced SCC of Head and Neck (SCCHN) at our centre.

Materials and methods

Data collection

All patients treated for SCCHN stage 4a/b with radical concurrent chemoradiation at the Deansley Centre, New Cross Hospital between September 2000 and December 2004 were identified and case notes reviewed retrospectively.

Data concerning disease site, TNM stage, radiotherapy dose/ fractionation, and chemotherapy details were collected and the incidence of delays in therapy, acute toxicity, dose reduction and treatment omissions of both chemotherapy and radiotherapy were also recorded. Information on toxicity of therapy is readily available as acute toxicity is routinely recorded and graded according to Common Toxicity Criteria (CTC) Version 2.0 during treatment by clinical nurses and radiographers as part of the department's protocol.

Patient details

Fifty-one patients received chemoradiation for SCCHN (41 males and 10 females). The most common site of primary disease was the oropharynx (28 patients) followed by hypopharynx (8), nasopharynx (7), and larynx (5 including 2 stomal recurrences). One patient was treated for extensive lymph node disease from an unknown primary and one patient was treated for locoregional recurrence in the neck. Staging was established clinically from clinical examination and nasoendoscopy as well as radiologically from CT/MRI scans. The distribution of TNM staging and disease site for these patients are shown in Table I.

Table I.  Patient and treatment details.

	Weekly Cisplatin (n = 24)	3 weekly Cisplatin (n = 27)
T Staging
T1	6 (25%)	3 (11%)
T2	2 (8%)	4 (15%)
T3	4 (17%)	2 (7%)
T4	10 (41%)	16 (59%)
Tx (recurrent or UKP)	2 (8%)	2 (7%)
N Staging
N0	0	3
N1	1	0
N2	16	15
N3	6	8
Nx	0	1
Site
Oropharynx	14	14
Hypopharynx	4	4
Larynx	4	1
Nasopharynx	0	7
Unknown primary	0	1
Neck recurrence	0	1
Chemotherapy
Concurrent Cisplatin Dose	20 (40 mg/m^2)	7 (100 mg/m^2)
	4 (33 mg/m^2)	20 (80 mg/m^2)
Radiotherapy Dose Fractionation
70Gy/35#	0	2
66GY/30#	23	23
60Gy/30#	1 (stomal recurrence)	0
Unable to complete	0	2

Chemotherapy

Twenty-four patients were treated as outpatients with weekly Cisplatin for 6 cycles during radiotherapy. Weekly 40 mg/m² and 33 mg/m² were commenced in 20 and 4 patients respectively. When the department first moved to delivering weekly Cisplatin, 33 mg/m² was used in preference to 40 mg/m² for reasons of caution. IV hydration with 1 litre 0.9% saline over 2 hours was given pre and post chemotherapy.

Twenty-seven patients were treated on a 3-weekly inpatient chemotherapy schedule on day 1, 22 and 43 of radiotherapy, of which Cisplatin 100 mg/m² and 80 mg/m² were planned for seven and 21 patients respectively. Patients were hydrated pre and post Cisplatin with 2 litres of 0.9% Saline plus 100 ml of 20% Mannitol given pre-chemotherapy. 5HT₃ antagonists plus dexamethasone were given as anti-emetic prophylaxis. The intended maximum total Cisplatin dose to be delivered is 240 mg/m² in the weekly group and 3-weekly 80 mg/m² group; and 300 mg/m² in the 3-weekly 100 mg/m² group.

Cisplatin and 5-fluorouracil induction (PF) chemotherapy was delivered prior to chemoradiotherapy in 18 patients (10 and 8 patients in the weekly and 3-weekly chemotherapy group). The majority (83%) received 2 cycles of neoadjuvant chemotherapy Table II

Table II.  Cisplatin dose intensity.

	Weekly Cisplatin 40 mg/m^2	3-weekly Cisplatin 80 mg/m^2	3-weekly Cisplatin 100 mg/m^2
	n = 20	n = 20	n = 7	P-value
Neoadjuvant Cisplatin/ 5FU	10 (50%)	5 (25%)	3 (43%)	0.11
Total Cisplatin delivered	4034 mg/m^2	4069 mg/m^2	1260 mg/m^2
Dose reduction	4 (17%)	2 (10%)	2 (29%)
Mean Cisplatin dose	202 mg/m^2	203 mg/^2	180 mg/m2^2	0.39
Cumulative dose achieved ≥200 mg/m^2	13 (65%)	12 (60%)	4 (57%)	0.91
Cumulative dose achieved ≥240 mg/m^2	9 (45%)	11 (55%)	0	0.04

Radiotherapy

The standard dose and fractionation was 66Gy in 33 daily fractions over 45 days in all patients except in five patients. In two patients, macroscopic lymphadenopathy was boosted to a total dose of 70 Gy in 35 fractions. One patient received re-irradiation to a dose of 60 Gy in 30 fractions for a stomal recurrence. Two patients were unable to complete radiotherapy. Patients were planned either conventionally on the simulator or CT planned, using a perspex beam directed shell. Treatment was prescribed to the ICRU 50 reference point.

All patients were reviewed weekly while on treatment by the medical team and weekly by clinical nurse specialist following completion of therapy until acute reactions had began to settle. Patients had subsequent follow-up at the joint Oncology-ENT outpatient clinic.

Statistical analysis

Overall treatment time to complete radiotherapy for the weekly and 3 weekly chemotherapy regime was compared using the Mann-Whitney U Test. The χ² test was used to compare the number of patients who experienced delays in radiotherapy completion. Only patients who received 33 fractions of radiotherapy were analysed in order to make an accurate comparison. The difference in Cisplatin dose intensity for patients receiving weekly 40 mg/m², 3 weekly 80 mg/m² and 3 weekly 100 mg/m² was analysed using Fisher's Exact Test of proportions. Chi Square analysis was employed to test the proportions of acute toxicity differences, median radiotherapy overall treatment time and delays in radiotherapy between the weekly and 3-weekly chemotherapy regime.

Results

More patients were able to receive a higher cumulative dose of at least 240 mg/m² if given weekly Cisplatin 40 mg/m² or 3-weekly Cisplatin 80 mg/m² compared with those receiving Cisplatin 3-weekly 100 mg/m² (p = 0.04). Maximum cumulative dose achievable in the latter group was only 200 mg/m² and none of these patients completed the full 3 cycles.

When analysing dose intensity, patients receiving weekly Cisplatin 40 mg/m², 3-weekly Cisplatin 80 mg/m²and 3-weekly 100 mg/m² received a total mean dose of 202 mg/m², 203 mg/m² and 180 mg/m² respectively throughout their chemoradiation. Although the differences were not statistically different (p = 0.39), this was probably due to the small number of patients in this series. The dose intensity between patients who received 2 cycles of neoadjuvant chemotherapy PF chemotherapy prior to commencing chemoradiation and those who did not appeared broadly similar.

In terms of chemotherapy dose compliance, none of the 7 patients who commenced concomitant 3-weekly Cisplatin 100 mg/m² completed the full 3 cycles of treatment. For patients treated with 3-weekly lower dose of Cisplatin (80 mg/m²), 13/20 patients (65%) managed to complete 3 cycles of chemotherapy. Patients who received weekly chemotherapy managed to complete six, five and four cycles in 10 (41.7%), 9 (37.5%) and 5 (20.8%) patients, respectively. Delays in chemotherapy was experienced by 11/27 patients (41%) compared with 7/24 patients (29%) in the weekly and 3-weekly regime respectively. Eleven of sixty three (17.4%) chemotherapy treatments on the 3-weekly regimes were not given on the scheduled day of treatment compared with 7/124 (5.6%) chemotherapy treatments on the weekly regime. Dose reductions required due to side effects of therapy were necessary in 4/24 patients (17%) receiving weekly Cisplatin, 2/7 patients (29%) in those receiving Cisplatin 100 mg/m² 3-weekly and 2/20 patients (10%) receiving Cisplatin 80 mg/m² 3-weekly.

Radiotherapy overall treatment time (OTT) ranged between 10–57 days in the 3-weekly regime and 43–54 days in the weekly regime Table III. Two patients unfortunately died during the treatment in the 3-weekly regime and failed to complete the intended radiotherapy schedule. The median OTT, excluding the two patients who died and those who received 30 or 35 fractions, was 46 days in the 3-weekly regime compared with 45 days in the weekly regime (p = 0.26).

Table III.  Radiotherapy overall treatment time.

	*Weekly Cisplatin (n = 23)	*3 weekly Cisplatin (n = 23)	P-value
Range	43–54 days	45–57 days
Median	45 days	46 days	0.26
Delay	10 (42%)	12 (44%)	0.77

*Excluded patients unable to complete radiotherapy due to death (n = 2), receiving 30 or 35 fractions.

When comparing haematological toxicity, haemoglobin dropped by a mean of 3.1 g/dl after chemoradiation in patients receiving the 3-weekly chemotherapy (range of lowest haemoglobin recorded 9.7–12.8 g/dl) and a mean of 3.3 g/dl for patients receiving weekly chemotherapy (range of lowest haemoglobin recorded 10.4–13.7 g/dl). There was no significant difference between the two groups with respect to grade 3 neutropenia (Table IV). However, a proportionally higher number of patients having grade 3 neutropenia on the 3-weekly regime experienced neutropenic fever (83% vs. 40%) compared to the weekly regime but this was insignificant. A similar frequency of use of prophylactic antibiotics in both groups was recorded. Renal toxicity was similar and there was no grade 3/4 renal toxicity. There was no significant difference in the gastro-intestinal or neurological toxicity experienced and no grade 4 toxicity reported.

Table IV.  Treatment toxicity (no. of patients).

	Weekly Cisplatin (n = 24)	3 weekly Cisplatin (n = 27)	P-value
Haematological
Anaemia (Grade 1/2)	20	20	0.64
Neutropenia (Grade 3)	5	6	0.83
Neutropenic fever	2	5	0.33
Renal
Grade 1/2	5	5	0.89
Gastrointestinal
Vomitting (Grade 3)	2	6	0.33
Constipation (Grade 3)	1	1	0.33
NG/PEG tube
• during treatment	22/24 (92%)	27/27 (100%)	0.43
• at 3 months	17/21 (81%)	18/24 (75%)	0.91
• at 12 months	3/19 (16%)	3/16 (19%)	0.83
Neurological
Neuropathy (Grade 1)	3	1	0.52
Skin
Radiation Dermatitis (Grade 3)	13/23 (26%)	6/23 (56%)	0.07

Assisted feeding in the form of either PEG or NG tube was used in all patients in the 3-weekly regime and 22/24 patients in the weekly regime. At 3 months, 18/24 (75%) evaluable patients in the 3-weekly regime were recorded to be still dependent on assisted feeding compared with 17/21 (81%) evaluable patients in the weekly arm. At 12 months, only 3/16 patients (18%) in the 3-weekly regime still had a PEG tube compared with 3/19 (16%) patients in the weekly regime. These patients who still had a PEG tube at 12 months were mostly managing liquids or a soft diet and were using the PEG to supplement oral intake. Moist desquamation of the skin was seen in 13/23 (56%) and 6/23 (26%) evaluable patients in the 3-weekly and weekly regime respectively.

Six weeks post-chemoradiation, 25/27 patients in the 3-weekly regime was assessable for response compared with all 24 patients in the weekly regime. In the 3-weekly regime, 20 patients had a complete response, three patients had a partial response and two patients had progressive disease. Two patients in the 3-weekly group died from treatment related toxicity; pneumonia complicating grade 2 neutropenia in both cases. In the weekly regime, 21 patients had a complete response, three patients had a partial response and none had progressive disease.

Local control and overall survival are 63% and 52% respectively in the 3-weekly Cisplatin group with a median follow up of 49 months, and 79% and 71% respectively in the weekly Cisplatin group with a median follow up of 26 months. One patient in the weekly regime died within 30 days of completion of treatment secondary to uncontrolled haemorrhage from the primary tumour. Relapse within the radiotherapy field was observed in 11 patients in the 3-weekly group, with 4 patients relapsing distantly. In the weekly group, local relapse was observed in four patients with a further four patients relapsing distantly.

Discussion

Proportionally more patients receiving Cisplatin 80 mg/m² given every 3 weeks and weekly Cisplatin 40 mg/m² were able to achieve a higher total cumulative Cisplatin dose compared with the standard 3-weekly Cisplatin 100 mg/m² in our series of patients with stage 4 locally advanced disease. When delays occur in the administration of chemotherapy, this often resulted in omission of the final cycle of chemotherapy. This omission affected the 3-weekly regime more compared with the weekly regime and was due to lack of inpatient beds in many cases. Haematological, intestinal, neurological and renal toxicity appears to be comparable in both groups of chemotherapy regime.

There was comparable number of patients dependent on assisted feeding with either PEG or NG tube in both groups at 3 and 12 months. There was a relatively high proportion of patients still dependant upon assisted feeding at 12 months in both groups but was comparable to reported studies on chemoradiation [4]. Moist desquamation was however observed to be more than double the number of patients in the 3-weekly regime compared with the weekly regime (56% vs. 26%) although it was non-significant in view of the small number of patients.

There was widespread adoption of the INT 0099 regimen delivering Cisplatin 100 mg/m² on day 1, 22 and 43 with radical radiotherapy after the impressive results compared with radiotherapy alone when treating nasopharyngeal carcinoma. However, only 63% of patients in the chemoradiation arm completed all three cycles of concurrent chemotherapy and only 55% received all three cycles of adjuvant chemotherapy as planned [9].

The poor compliance with 3 cycles of high dose Cisplatin was echoed by a recent series from the Princess Margaret Hospital, Toronto. In this retrospective study of 75 patients, 42.7% received all three cycles of concurrent INT 0099 regimen chemotherapy, and only 33.3% received the intended dose without a cumulative delay of >7 days through all three cycles [10].

There have been reports of chemotherapy intensity correlating with outcomes in prospective trials for breast cancer, ovarian cancer and adult acute myelocytic leukaemia (AML) [11]. Dose intensity of cisplatin is an important factor in the optimal treatment of ovarian and testicular carcinoma [12]. Although no randomized study has been performed in head and neck cancer to demonstrate the importance of dose intensity, given that an important impact on survival has been demonstrated by synchronous chemoradiation [1], it would suggest that maintaining dose intensity during synchronous chemotherapy will be important when treating patients.

Use of neoadjuvant chemotherapy with 2 cycles of Cisplatin and 5-Fluorouracil (PF) does not seem to affect the tolerance of concomitant cisplatin based chemotherapy in our series. As neoadjuvant chemotherapy with PF has been demonstrated to show a modest 5% improvement in overall survival from the MACH-NC meta-analysis, this can be important in contributing to the overall treatment outcome [13]. Recently, two publications have reported the improved tolerability and survival benefit of docetaxel, cisplatin, 5-fluorouracil (TPF) neoadjuvant chemotherapy compared with standard PF regime for locally advanced HNSCC [14], [15]. Posner et al. reported that there were no significant differences in the concurrent chemotherapy dose delivered to each patient group but the concurrent chemotherapy regime used was weekly carboplatin which some clinicians may consider as non-standard and therefore difficult to compare [14]. Therefore, results of a 3-arm study where patients are randomised between TPF, PF or no neoadjuvant chemotherapy followed by concurrent Cisplatin 3-weekly 100 mg/m² are awaited [16]. It will be of interest to reveal if these more intensive neoadjuvant chemotherapy regimes have any effect on synchronous chemotherapy dose intensity. Without convincing evidence that neoadjuvant chemotherapy contribute more towards improving overall survival compared to synchronous chemotherapy, it will be important that synchronous chemotherapy dose intensity is not compromised.

Delivering weekly Cisplatin at a dose of 40 mg/m² in the outpatient department appears to have similar toxicity and efficacy to a 3-weekly concurrent chemoradiation regime but is less subject to delays in treatment and reductions in dose intensity due to administrative failures. This regime should be tested formally in a phase III study against the accepted standard of 3-weekly concurrent Cisplatin. Delivery of 100 mg/m² Cisplatin 3-weekly concurrently with radiotherapy was not as well tolerated in this group of patients resulting in an overall loss of dose intensity when compared to other groups. The inpatient delivery of this treatment also meant delivery of therapy was subject to administrative failures due to lack of inpatient beds. However, administering Cisplatin 40 mg/m² weekly as an outpatient requires additional capacity in the chemotherapy day case unit.