Results of a phase III multi-centre randomised controlled trial of intensity modulated (IMRT) vs conventional radiotherapy (RT) in head and neck cancer (PARSPORT: ISRCTN48243537; CRUK/03/005)
Chris Nutting
Royal Marsden Hospital, London, UK
Background
Xerostomia
is the commonest late toxicity of RT to the head and neck. IMRT dose
distributions reduce the dose delivered to parotid gland. PARSPORT investigated
the role of IMRT in reducing xerostomia in patients with head and neck cancer.
Method
The
PARSPORT trial compared two radiotherapy delivery methods in the treatment of
patients with pharyngeal tumours (T1-4, N0-3, M0). Patients received 65Gy in
30 fractions over 6 weeks delivered using either CT planned parallel opposed
lateral fields or parotid-sparing IMRT. Stratification was by site of tumour and
centre. The primary endpoint was incidence of LENT-SOMA ≥G2 xerostomia
one year after treatment. Secondary endpoints included acute toxicities (CTCAE
v3) and other late RTOG and LENT-SOMA radiation toxicities. Saliva collection
was attempted from both parotid glands. Proportions
of patients with ≥G2 toxicity were compared using exact tests. For
secondary endpoints a significance level of 1% was used.
Results
94
patients (47 RT; 47 IMRT) were randomised between 2003 and 2007 from six UK centres. 80 patients had oropharyngeal tumours and 14 hypopharyngeal. Radiotherapy was
given as primary treatment in 71 patients and post-operatively in 23. 22
patients had AJCC stage I/II disease. Median follow-up was 31.9 months (IQR:
23.6 – 38.8). Twelve month LENT-SOMA ≥G2 xerostomia scores were observed
in 74% (25/34) of RT and 39% (15/38) of IMRT patients (p=0.004). Corresponding
values at 18 months were 71% (15/21) and 29% (9/31) (p=0.003). On the RTOG
scale, 12 month ≥G2 xerostomia was reported in 64% (21/33) RT vs 41%
(15/37) IMRT patients (p=0.05). The 18 month incidence was 81% 17/21 RT vs 20%
(6/30) IMRT (p<0.001). Acute radiotherapy related ≥G2 fatigue was
more prevalent in the IMRT group (76% vs 41% p=0.001). No differences in acute
mucositis or pain scores were seen. At 12 months, no statistically significant
differences were seen in other late toxicities. Sialometry measurements
confirm the LENT-SOMA outcome with 18% (3 months), 33% (6 months) and 24% (12
months) of patients having contralateral measurable saliva in the IMRT arm
compared with 0% at each time point in the conventional RT arm (p=0.008 at 12
months). No differences were observed between overall survival and
locoregional control rates.
Conclusion
Sparing
the salivary glands through use of IMRT preserves salivary flow and
significantly reduces the incidence of xerostomia in patients with pharyngeal
tumours.