Sequential TPF chemotherapy followed by concurrent chemoradiotherapy in locally advanced head and neck cancer – a retrospective analysis of toxicity and outcomes

Abstract

Background and aims

Phase III trials have shown that the addition of a taxane to cisplatin/5FU-based induction chemotherapy (TPF) improves response rates and overall survival in unresectable stage III/IV head and neck cancer. We sought to assess the tolerability, compliance and clinical outcomes of this treatment regime.

Methods

A retrospective study of patients treated within a single centre between September 2007 and November 2010. Toxicities were graded according to CTCAE version 3.0. Survival, distant metastasis and local control rates are expressed as percentages at two years using the Kaplan–Meier method.

Results

A total of 100 patients were identified (11% stage III, 86% stage IV) and 32% of patients were admitted as an emergency after TPF. The rate of neutropenic fever was 31%, this number fell to 9% when prophylactic G-CSF was used. In addition, 89% of patients underwent radical chemoradiation. Of these, 96% completed the full radiotherapy course. However, only 64% of patients received a minimum of two cycles of concurrent platinum chemotherapy. The two-year overall survival, metastasis free survival and local control rates were 62.6%, 88.5% and 73.3%, respectively.

Conclusions

TPF chemotherapy can be delivered safely in a non-trial cohort of patients. There is, however, a significant reduction in concurrent chemotherapy dose intensity. The long-term impact of this remains unclear.

Keywords

Locally advanced head and neck cancer chemoradiotherapy induction chemotherapy

Introduction

Head and neck squamous carcinoma is the fifth most common cancer worldwide and the majority (60%) are locally advanced at presentation.¹ Radical radiotherapy is the definitive treatment option for these patients and the meta-analysis of chemotherapy in head and neck cancer (MACH-NC) demonstrated an 8% survival benefit when platinum-based chemotherapy was given concomitantly.² Recurrent or persistent loco-regional disease is the major cause for treatment failure (50–60% local recurrence within two years). Due to recent advances in the management of loco-regional disease, systemic therapies to reduce distant metastasis rates are gaining importance. De novo squamous head and neck cancers are more sensitive to systemic treatments when compared to most other solid cancers but the response is not durable. Therefore, new strategies aim to improve results by a sequential approach with induction chemotherapy in addition to local therapy. There is, however, concern that induction therapy may compromise the delivery of definitive loco-regional treatment.

The role of induction chemotherapy followed by concomitant chemoradiotherapy has been tested in three randomised trials and the results are favourable (TAX 324 and Spanish Intergroup).^3–5 However, a recent phase II trial showed that only 45% of patients receiving three cycles of induction TPF chemotherapy completed concomitant chemoradiotherapy.⁶ The role of sequential therapy is debatable, and further randomised trials are awaited before change of standard treatment. We performed a retrospective analysis of toxicities and clinical outcomes after induction TPF chemotherapy.

Materials and methods

Study design

This is a retrospective study of toxicities and clinical outcomes in patients receiving induction TPF chemotherapy followed by concurrent chemoradiotherapy for locally advanced head and neck cancer. Patients were treated within a single centre between September 2007 and November 2010.

Treatment

Local policy is to use induction TPF for patients with inoperable, bulky stage III/IVA-B disease. TPF chemotherapy consisted of docetaxel 75 mg/m² (day 1), cisplatin 75 mg/m² (day 1) and 5FU 750 mg/m² (days 1–5). Radical radiotherapy was delivered using a CT planned 3D conformal technique to a dose of 68Gy in 34 fractions (a small number of patients underwent IMRT as availability of this technique emerged). Concurrent chemotherapy consisted of two to three cycles of cisplatin (100 mg/m²) given on a three weekly basis.

Data collection

Eligible patients were identified using an electronic chemotherapy prescribing system; no exclusion criteria were applied. A customised data collection form was devised and sources of information included the medical notes and electronic laboratory, radiology and pathology results systems. Toxicities were graded according to the common toxicities criteria (CTCAE version 3.0). Local control and distant metastasis rates were obtained via electronic results systems and medical letters. Local control was defined as either the absence of local recurrence or persistent loco-regional disease after maximal therapy. Overall survival data was obtained both electronically and by contacting primary health care providers. Follow-up began from the day of the first cycle of TPF chemotherapy.

Statistical analysis

Estimates of survival rates and local control rates are expressed as percentages at two years using the Kaplan–Meier method. Kaplan–Meier curves are shown up to 45 months follow-up. All analyses were carried out in Stata version 12.0.

Results

Patient characteristics

A total of 100 patients were identified (84 men and 16 women). Of these, 94% of patients were ECOG performance status 0 or 1. Tumour sites included oropharynx (54%), larynx (13%), hypopharynx (10%) and nasopharynx (8%). The remainder (15%) were post-cricoid tumours or cancer of unknown primary presenting with metastatic cervical lymphadenopathy. Also, 11% had stage III disease and 86% stage IV; 89% of patients underwent chemoradiation after TPF; 5% were treated with surgery and 4% surgery and post-operative radiotherapy. Around 92% of patients received prophylactic ciprofloxacin on days 5 to 15, and 27% also received primary prophylaxis with G-CSF; an audit in late 2009 identified a high risk of neutropenic fever after TPF so primary G-CSF was used in all patients treated during 2010 (23 in total).

Compliance

The planned number of cycles of induction TPF ranged from 2 to 3. Chemotherapy was completed as planned and without dose modification in 66% of patients. In 11% of patients, dose modifications were made at the outset due to co-morbidities or performance status; 8% were dose reduced and 3% switched from cisplatin to carboplatin for subsequent cycles.

Toxicity

After TPF, 32% of patients were admitted as an emergency. Documented non-haematological toxicities (grade ≥3) included acute coronary syndrome (3%), gastritis (1%), peripheral neuropathy (1%), mucositis (2%), nausea/vomiting (1%), diarrhoea (1%) and nephrotoxicity (1%). Two deaths were observed: one patient had a cerebrovascular event immediately prior to cycle 2 and one died of neutropenic sepsis.

Grade ≥3 neutropenia (at any point within 4 weeks of TPF) was seen in 35% of patients (22% with G-CSF prophylaxis and 39% without). The rate of neutropenic fever was 31%; this number fell to 9% when G-CSF was used for primary prophylaxis. Grade ≥3 thrombocytopenia and anaemia occurred in 7% of patients.

Radiological response to TPF

Only 5% of patients had a complete response to induction TPF, 67% had a partial response, 18% stable disease and 5% progressive disease.

Concurrent chemo-radiation

Of the 89 patients receiving chemoradiation, 96% completed the full radiotherapy course. However, only 64% of patients received a minimum of two cycles of concurrent platinum chemotherapy. The principal reason was chemotherapy-related toxicity. In some cases, a decision was made to give radiotherapy without cisplatin because of the severity of toxicities after TPF (acute coronary syndrome, neutropenic sepsis). Alternatively, some patients had persisting toxicity after TPF (neuropathy). Some patients started concurrent chemotherapy but had only one cycle because of new toxicities (neutropenia, hearing loss and tinnitus). The only other reason was death during radiotherapy – one patient suffered a fatal bowel perforation after 11 fractions.

Survival analysis

A total of 99 patients were included in this analysis (1 patient had left the country). Mean follow-up time was 24 months; range 2–45. Of them, 47 patients had died, 28 patients had persistent or progressive loco-regional disease and 12 patients had recurred with distant metastases. The Kaplan–Meier curves for overall survival, metastasis free survival and local control are shown (see Figures 1 –3). The two-year overall survival rate was 62.6%, 95% CI (52% to 71.4%); the two-year metastasis free survival rate was 88.5%, 95% CI (78.9% to 93.9%); the two-year rate for local control was 73.3%, 95% CI (62.6% to 81.4%). Median survival was 35 months.

Figure 1.

Overall survival in patients receiving TPF.

Figure 2.

Metastasis free survival in patients receiving TPF.

Figure 3.

Local control in patients receiving TPF.

Discussion

Induction chemotherapy with TPF has emerged as a treatment option in locally advanced head neck cancer over the last few years. Several studies have indicated a survival benefit over PF chemotherapy,^5,7,8 a better chance of achieving a complete response^4,8–10 and higher laryngeal conservation rates.¹⁰ However, this benefit may come at a cost with increases in haematological toxicity and neutropenic sepsis rates. In addition, further results of phase III studies comparing induction chemotherapy with TPF followed by chemoradiotherapy versus the current gold standard therapy of chemoradiotherapy are awaited.

Head and neck patients are often a relatively unfit group with significant comorbidity and from a poor socio-economic class. There is significant concern that the benefits of TPF may be lost in a non-trial cohort of patients who are likely to be less fit. There were two aspects to this study. First, could TPF be delivered in a safe manner to an unselected non-trial group of patients? Second, would neo-adjuvant TPF affect the ability to complete the main radical treatment of chemo-radiotherapy and also longer-term outcome? Importantly, the patient demographics and disease stages in our patients are similar to those in recent phase III studies (see Table 1).

Table 1.

Patient details in this series and recent phase III studies.

Study	TPF details	N	Age	Sex (%)	Stage (%)
Current Series	Docetaxel 75 mg/m² D1 Cisplatin 75 mg/m² D1 5FU 750 mg/m² D1-5	100	56 (23–79)	Male (84) Female (16)	II (2) III (11) IV (86) Not documented (1)
Paccagnella et al. 2010⁹	Docetaxel 75 mg/m² D1 Cisplatin 80 mg/m² D1 5FU 800 mg/m² D1-4	50	58 (36–75)	Male (92) Female (8)	III (16) IV (84)
Pointreau et al. 2009¹⁰	Docetaxel 75 mg/m² D1 Cisplatin 75 mg/m² D1 5FU 750 mg/m² D1-5	110	57 (33–72)	Male (91.8) Female (8)	III/IV (NR) All hypopharynx/larynx
Vermorken et al. 2007⁸	Docetaxel 75 mg/m² D1 Cisplatin 75 mg/m² D1 5FU 750 mg/m² D1-5	177	53 (31–70)	Male (89.8) Female (10.2)	NR
Posner et al. 2007⁵	Docetaxel 75 mg/m² D1 Cisplatin 100 mg/m² D1 5FU 1000 mg/m² D1-4	255	55 (38–82)	Male (84) Female (16)	III (16) IV (84)
Hitt et al. 2005⁴	Paclitaxel 175 mg/m² D1 Cisplatin 100 mg/m² D1 5FU 500 mg/m² D2-6	189	56 (31–75)	Male (94) Female (6)	III (15) IV (85)

NR: not recorded.

However, 66% of patients received TPF without dose delay/modification; this is in keeping with the published trials in which between 62.7% and 75.7% of patients completed induction chemotherapy without compromising dose intensity.^4,5,8,10 Two patients had a fatal event while on TPF (one cerebrovascular accident and one neutropenic sepsis); again, this is similar to trial data in which treatment-related mortality rates ranged from 0% to 2.7%.^4,5,8–10 In contrast, we observed a 31% incidence of neutropenic fever/sepsis, which is much higher than in published series (see Table 2). After we started using primary prophylaxis with G-CSF, the neutropenic fever rate fell to 9%, which provides evidence to support its on-going use.

Table 2.

Grade 3/4 Haematological toxicity after induction TPF chemotherapy.

Reference	Neuts	Plts	Hb	Febrile neutropenia	G-CSF	Prophylactic antibiotics
Current series	35%	7%	7%	26% (9% G-CSF + 31% G-CSF-)	Primary prevention starting 2010	D5-15
Paccagnella et al.⁹	NR	NR	NR	NR	Secondary prevention	D5-15
Pointreau et al.¹⁰	31.5% (G4)	7.8%	3.7%	10.9%	Secondary prevention	D5-15
Vermorken et al.⁸	76.9%	5.2%	9.2%	5.2%	Secondary prevention	D5-15
Posner et al.⁵	83%	4%	12%	12%	No	D5-15
Hitt et al.⁴	37%	NR	NR	8%	No	No

NR: not recorded; D: day; G: grade.

In our experience, induction TPF did not affect subsequent treatment with radiotherapy (96% completed radiotherapy as planned). Of concern, however, is the fact that only 64% received at least two cycles of concurrent chemotherapy. In the recent updated meta-analysis of concurrent chemotherapy, it was noted that at least 150 mg/m² of cisplatin over the radiotherapy treatment time was required for the best outcome.² The long-term outcome of our reduced dose of concurrent cisplatin chemotherapy remains to be seen.

Our data support trial results, suggesting TPF is effective in down staging bulky disease prior to definitive treatment (see Table 3). Our clinical outcomes are also very similar to those seen within phase III studies (see Table 4). However, our data are retrospective and we cannot draw direct comparisons. Our rates on non-haematological toxicity are low, but toxicities are often not clearly documented in medical notes and these results are likely to be an underestimation. However, despite these limitations, our results do suggest that clinical outcomes similar to trial data can be achieved in a non-trial cohort of patients.

Table 3.

Radiological response to TPF.

Reference	Radiological Response to TPF
Reference	CR (%)	PR (%)	SD (%)	PD (%)
Current series	5	67	18	5
Paccagnella et al.⁹	6.5	63	19.7	10.8
Pointreau et al.¹⁰	41.8	38.2	11.8	3.2
Vermorken et al.⁸	8.5	59.3	NR	NR
Posner et al.⁵	17	55	12	7
Hitt et al.⁴	33	47	11	4

NR: not recorded; CR: complete response; PR: partial response; SD: stable disease; PD: progressive disease.

Table 4.

Overall survival, distant metastases and loco-regional failure rates.

Reference	Locoregional failure	Distant Metastasis Rate	Overall survival
Current series	26.7% at two years	11.5% at two years	62.6% at two years
Paccagnella et al.⁹	NR	NR	61% at two years
Pointreau et al.¹⁰	33.3% after 36 months F/U	10.9% after 36 months F/U	75% at two years
Vermorken et al.⁸	48.6% after 32.5 months F/U	12.9% after 32.5 months F/U	43% at two years
Posner et al.⁵	30% after 42 months F/U	14% after 42 months F/U	67% at two years
Hitt et al.⁴	NR	NR	66.5% at two years

NR: not recorded; F/U: follow up.

Within our centre, we have defined strict criteria for eligibility for induction chemotherapy with TPF (locally advanced, bulky stage III/IV disease requiring down-staging, ECOG performance status ≤1, <65 years of age, no significant co-morbidity, <10% weight loss). This is a potentially highly toxic chemotherapy regime and requires expert clinical knowledge and close monitoring of patients. Currently, we cannot say that induction TPF chemotherapy provides a survival benefit over standard chemoradiotherapy and as such the standard of care remains chemoradiotherapy with TPF being reserved for a selected group of patients requiring downstaging of very large tumours. We await the publication of several phase III studies comparing TPF and subsequent chemoradiotherapy against the gold standard of chemoradiotherapy with great interest.^11,12

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Seiwert

Cohen

EEW

. State-of-the-art management of locally advanced head and neck cancer. Br J Cancer 2005; 92: 1341–1348.

Pignon

le Maitre

Maillard

. Meta-analysis of chemotherapy in head and neck cancer (MACH-NC): An update on 93 randomised trials and 17,346 patients. Radiother Oncol 2009; 92: 4–14.

Hitt R, Grau JJ, Lopez-Pousa A, et al. Final results of a randomized phase III trial comparing induction chemotherapy with cisplatin/5-FU or docetaxel/cisplatin/5-FU follow by chemoradiotherapy (CRT) versus CRT alone as first-line treatment of unresectable locally advanced head and neck cancer (LAHNC). J Clin Oncol 2009; 27: abstr 6009.

Hitt

Lopez-Pousa

Martinez-Trufero

. Phase III study comparing cisplatin plus fluorouracil to paclitaxel, cisplatin, and fluorouracil induction chemotherapy followed by chemoradiotherapy in locally advanced head and neck cancer [Erratum appears in J Clin Oncol 2006; 24: 1015]. J Clin Oncol 2005; 23: 8636–8645.

Posner

Hershock

Blajman

. Cisplatin and fluorouracil alone or with docetaxel in head and neck cancer. N Engl J Med 2007; 357: 1705–1715.

Lefebvre JL, Pointreau Y, Rolland F, et al. Sequential chemoradiotherapy (SCRT) for larynx preservation (LP): Preliminary results of the randomized phase II TREMPLIN study. J Clin Oncol 2009; 27: abstr 6010.

Lorch

Goloubeva

Haddad

. Induction chemotherapy with cisplatin and fluorouracil alone or in combination with docetaxel in locally advanced squamous cell cancer of the head and neck: long-term results of the TAX 324 randomised phase 3 trial. Lancet Oncol 2011; 12: 153–159.

Vermorken

Remenar

van Herpen

. Cisplatin, fluorouracil, and docetaxel in unresectable head and neck cancer. N Engl J Med 2007; 357: 1695–1704.

Paccagnella

Ghi

Loreggian

. Concomitant chemoradiotherapy versus induction docetaxel, cisplatin and 5 fluorouracil (TPF) followed by concomitant chemoradiotherapy in locally advanced head and neck cancer: a phase II randomized study. Ann Oncol 2010; 21: 1515–1522.

10.

Pointreau

Garaud

Chapet

. Randomized trial of induction chemotherapy with cisplatin and 5-fluorouracil with or without docetaxel for larynx preservation. J Natl Cancer Inst 2009; 101: 498–506.

11.

Haddad R, Tishler R, Adkins D, et al. The PARADIGM trial: A phase III study comparing sequential therapy (ST) to concurrent chemoradiotherapy (CRT) in locally advanced head and neck cancer (LAHNC): Preliminary toxicity report. J Clin Oncol 2010; 28: abstr 5563.

12.

Docetaxel based chemotherapy plus or munis induction chemotherapy to decrease events in head and neck cancer (DeCIDE) [database on the Internet]. Available at http://clinicaltrials.gov/show/NCT00117572 [cited 21st January 2012].