In our randomised phase III study, HRQOL was assessed in patients with unresectable locoregionally advanced SCCHN after treatment with induction TPF or PF followed by RT.
The global HRQOL improved during induction CT in both treatment arms. As expected per protocol, no difference in global HRQOL was seen between arms during the treatment period. Interestingly, at 6 months after the end of the RT the global QoL remained higher than at baseline only in the TPF arm. In the PF arm, the global HRQOL returned to baseline scores, as usually seen in head and neck HRQOL studies (Abdel-Wahab et al, 2005
; Curran et al, 2007
). This resulted in a difference between the arms at 6 months after RT of 9.5 points, very close to a clinically meaningful improvement. Unfortunately, the compliance was too low to draw definite conclusions. In the pros, the compliance was similar in both arms. The imbalance between the number of patients still on study at this time point was mainly because of a higher rate of treatment discontinuation in the PF arm owing to toxicity and a higher rate of deaths owing to progressive disease, pointing towards a possible bias in disfavour of the experimental arm. In the cons, the analysis technique used relies on the assumption of data missing at random (MAR), which can always be criticised.
This trend towards an improvement in global HRQOL in the TPF arm occurred parallel to an increase in OS, higher response rate, and less severe toxicity, owing to a lower dose of cisplatin and 5-FU in the TPF arm than in the PF arm (Vermorken et al, 2007
). Why HRQOL after 6 months was better in the TPF arm than in the PF arm is not completely clear, but can probably be explained, in part, by fewer recurrences of the tumour. Another explanation can be the lower dose of cisplatin and 5-FU used in TPF compared with PF. Long-term toxicity of cisplatin leads to polyneuropathy and ototoxicity, which can influence the global QoL. Patients' overall HRQOL usually results from both treatment effects/side-effects and factors linked to the disease evolution, which are often indistinguishable.
A few investigators have assessed the longitudinal changes of HRQOL in patients with SCCHN during treatment. The general picture is a deterioration during the first 3 months after the start of treatment, followed by a slow recovery (de Graeff et al, 1999
; Bairati et al, 2005
; Fang et al, 2005
). Locoregionally advanced disease patients included in the randomised trial of cetuximab with RT vs
RT alone performed better in the combined arm (Bonner et al, 2006
; Curran et al, 2007
) and, although there was a gain in OS, no differences in HRQOL were observed. This study is the first reporting HRQOL during induction CT followed by RT, showing an improvement during the first weeks after start of neo-adjuvant CT. However, we did not measure the QoL during or in the last week of the RT. Thus, we can only speculate on the QoL during the RT in the TPF and PF arm. On the one hand, it could have been better in the TPF arm, because the trend in a better QoL, which was seen after the CT before the start of Rt, continued to improve, or on the other hand, it could have been worse in the TPF arm, because docetaxel can act as a radiosensitiser (Nabell and Spencer, 2003
Swallowing dysfunction and aspiration are seen in a high proportion of patients with SCCHN after combined chemoradiation (Bentzen and Trotti, 2007
). Therefore, swallowing and coughing, although not always related to aspiration, were selected as primary domains for this analysis. A trend to a higher reduction in swallowing and coughing problems was seen in the TPF arm compared with the PF arm, but the extent of the reduction was limited. In addition less loss of appetite was observed in the TPF arm, whereas less weight loss and more weight gain were observed in the TPF arm at the end of cycle 4. Eating problems may result from both the primary location of the head and neck cancer and treatment-induced adverse effects, such as pain in the mouth, problems with dentition, decreased saliva, and problems swallowing. Hence, weight loss is reported to affect 35–50% of patients with SCCHN, and is known to increase morbidity and mortality (van Bokhorst-de van der Schuer et al, 1999
). Thus, the improvement of swallowing combined with less eating problems observed in the TPF arm is not only beneficial for HRQOL but probably causes less morbidity and mortality in the follow-up.
Our randomised controlled trial (RCT) had several limitations. Despite being a robust, well-designed, and monitored RCT, HRQOL compliance became very limited over time, making only analyses of short-term HRQOL data possible and not allowing to draw definite conclusions. However, this is not unexpected, as collecting data in head and neck trials can be difficult, and indeed, the lack of RCTs with HRQOL results in the literature may support this hypothesis. In addition, at the start of this study, not all translations of the EORTC Head and Neck module were available, hence reducing the amount of information available from the module. At last, even if, as per protocol, very precise timing for the HRQOL assessment was described, time windows need to be defined to perform the analysis and assign HRQOL data to the different time points and allow for some delays. A 3-week delay was allowed for the assessments ‘At the end of cycle 2' and ‘At the end of cycle 4', which may have caused a slight underestimation of the treatment effect.
Nevertheless, there are positive points. This was a RCT with a good sample size; a similar compliance in both arms; the use of a robust methodology under missing data of the MAR type; and no indication of a source of bias in the investigation of the missingness mechanism.
At this moment, the standard treatment of locally advanced SCCHN consists of concurrent chemoradiation. Concurrent chemoradiation was not incorporated as treatment in the EORTC 24971/TAX 323. However, at this moment, several studies are running with TPF as induction CT followed by concurrent chemoradiotherapy or RT combined with an inhibitor of the epidermal growth factor receptor (EGFR). Because these treatments cause more toxicity than RT without concurrent combination, an improvement of the QoL and swallowing after TPF would be of real value.
The field of treatment of the locally advanced SCCHN is moving quickly at this moment. The main goal of these developments is to administer a less toxic regimen to patients while keeping the same chance for cure. The use of intensity modulated RT (IMRT) and the use of targeted therapies, such as EGFR inhibitors, will lead to less toxicity and hopefully a better QoL for those patients (Feng et al, 2010
). In addition, human papillomavirus-positive patients do have a better prognosis, both after CT and RT (Fakhry and Gillison, 2006
). In future, these patients may be treated with a less-toxic regimen than the nowadays used concomitant chemoradiotherapy. The exact role in future for induction CT, that is, TPF, in this moving field is not yet clear. However, our observation of an improvement of global QoL during induction CT is important, and has to be investigated in future trials with induction CT followed by concurrent chemoradiation using IMRT, or followed by concurrent EGFR inhibition with RT.
In summary, in unresectable SCCHN patients, TPF compared with PF as induction CT before RT seemed to improve global HRQOL and swallowing in parallel with a significantly improved OS and less severe induced toxicity. These analogous improvements of a longer life with a better HRQOL in some areas can be seen as the ultimate goal of treatment of cancer patients and opens the door for further studies to determine the exact place of TPF as induction CT for the treatment of locally advanced SCCHN.