Hyperfractionated radiotherapy in head and neck cancer: a second look at the clinical data

Hyperfractionated radiotherapy in head and neck cancer: a second look at the clinical data

Radiotherapy and Oncology 46 (1998) 127–130 Hyperfractionated radiotherapy in head and neck cancer: a second look at the clinical data Michael Bauman...

44KB Sizes 0 Downloads 110 Views

Radiotherapy and Oncology 46 (1998) 127–130

Hyperfractionated radiotherapy in head and neck cancer: a second look at the clinical data Michael Baumann a ,*, Søren M. Bentzen b, K. Kian Ang c a

Clinic of Radiation Oncology, Medical Faculty Carl Gustav Carus, Technical University of Dresden, Fetscherstr. 74, D-01307 Dresden, Germany b Danish Cancer Society, Department of Experimental Clinical Oncology, Nørrebrogade 44, Building 5, DK-8000 Aarhus C, Denmark c Department of Radiation Oncology, The University of Texas, M.D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030, USA Received 1 July 1997; revised version received 4 September 1997; accepted 9 September 1997

Abstract Hyperfractionated radiation therapy (HF) is one of the most promising modifications of conventional fractionation (CF). However, two recent reviews on the efficacy of HF reached opposite conclusions: Stuschke and Thames (Int. J. Radiat. Oncol. Biol. Phys. 37: 259–267, 1997) concluded that the effectiveness of radiotherapy is consistently higher for HF than for CF, whereas Beck-Bornholdt et al. (Radiother. Oncol. 43: 1–21, 1997) concluded that evidence for a consistent therapeutic gain from HF is lacking and it cannot be ruled out that HF in head and neck tumors is detrimental. To clarify this issue, some important discrepancies between the two reviews are highlighted and a second look is taken at the clinical data published on HF in head and neck cancer. The most convincing prospective study so far is the EORTC trial 22791. This trial supports that HF allows escalation of the biologically effective dose to the tumor without a significant increase in late complications.  1998 Elsevier Science Ireland Ltd. Keywords: Hyperfractionated radiotherapy; Head and neck cancer; Clinical data

1. Introduction Hyperfractionated radiation therapy (HF) implies the use of dose fractions of less than 1.8 Gy usually administered twice a day and is one of the most interesting modifications of conventional fractionation (CF) given in five doses of 1.8–2.0 Gy per week [3,10,13]. During the last decade, several prospective clinical trials have evaluated this approach in a variety of tumor entities. Many investigators have concluded that HF may be associated with a therapeutic gain compared with CF. As a consequence, many clinical departments already consider HF as an established, perhaps superior, alternative to conventional radiation therapy. However, the issue is far from being resolved since two extensive reviews yielding conflicting conclusions were published within only 2 months in the two major international radiation oncology journals [2,9]. Stuschke and Thames [9] reported in the International Journal of Radia-

tion Oncology, Biology and Physics that their ‘overview demonstrates that the effectiveness of radiotherapy is consistently higher for HF than for CF’. In contrast, Beck-Bornholdt et al. [2] stated in Radiotherapy and Oncology that ‘… we cannot even rule out that hyperfractionation (in head and neck tumors) is detrimental, the results… seem to indicate this’. They conclude that ‘… evidence for a consistent therapeutic gain from HF is still lacking’ and that ‘the dose per fraction generally used in standard radiotherapy is already a good choice’. It is quite puzzling that two groups of authors could review the same published literature data and come up with rather opposing conclusions. In this commentary, we attempt to clarify the issue by highlighting some important discrepancies between the two reviews and providing our thought on the status of HF. The discussion focuses on head and neck cancer because most HF trials deal with this cancer site.

* Corresponding author.

0167-8140/98/$19.00  1998 Elsevier Science Ireland Ltd. All rights reserved PII S0167-8140 (97 )0 0173-4

128

M. Baumann et al. / Radiotherapy and Oncology 46 (1998) 127–130

2. Definition of hyperfractionation The study inclusion criteria and the literature search method were clearly presented by Stuschke and Thames [9] but not by Beck-Bornholdt et al. [2]. Scrutiny of the papers reveals that the difference in the definition of HF and hence selection of trials for review is a major cause of disparity. Both groups of authors consider that HF represents a regimen characterized by administration of a higher number of smaller fractions (i.e. less than 1.8 Gy per fraction) relative to CF [13]. Not agreed upon, however, is whether the total radiation dose should be part of the definition of HF and whether regimens with planned treatment breaks should be considered. Consequently, Stuschke and Thames [9] analyzed only trials in which the total radiation doses of the HF arm were greater than, or at least equal to, those in the CF arms, whereas Beck-Bornholdt et al. [2] assessed all trials addressing a higher number of smaller fractions regardless of the total dose. Therefore, the latter group included RTOG trial 79-13 [8] which prescribed a lower dose in the HF arm. EORTC trial 22811 [11] that tested a particular split-course accelerated and hyperfractionated regime was analyzed by Beck-Bornholdt et al. [2] but not by Stuschke and Thames [9]. The radiobiologic peculiarity of the regimen tested in EORTC 22811 will be addressed later.

3. Tumor control versus therapeutic ratio Readers of any review should have a clear idea of the question asked and of the methods used to address the ques-

tion [12]. Stuschke and Thames [9] used the approach of a meta-analysis on pooled results of several studies and focused on tumor-related outcome. They did not attempt to obtain an estimate on late normal tissue effects, which were discussed in just 15 lines. This is reasonable in so far as the results from various studies are difficult to compare quantitatively, especially as some of the papers do not provide sufficient data on late normal tissue reactions. Therefore, the conclusion of Stuschke and Thames [9] that ‘… the effectiveness of radiotherapy is consistently higher for HF than for CF’ pertains only to tumor control. In contrast, the review by Beck-Bornholdt et al. [2] clearly addressed the therapeutic gain from HF, that is whether the benefit/cost (tumor control versus late sideeffects) is superior to that of conventional fractionation. Moreover, this group of authors raised the question of whether the observed outcome of HF is better than what one may achieve from the simple escalation of the CF dose. In other words, the authors questioned whether the clinical data support the biological rationale of HF, namely the use of small dose fractions to exploit the dissociation in fractionation sensitivity between tumors and late-responding normal tissues. The latter is, indeed, a crucial clinicalradiobiology question! In reference to tumor control, BeckBornholdt et al. [2] were in agreement with Stuschke and Thames [9] that there is a trend in favor of HF. However, they concluded that HF was not proven to be superior to CF because of an alleged increase in late normal tissue injury. So this judgment stands or falls with the late complication data of the three studies summarized in Table 4 of their review paper. Therefore, it is worthwhile to take a closer look at the data of these studies (Table 1).

Table 1 Randomized trials providing quantitative information on late complications after hyperfractionated versus conventional radiation therapy in head and neck cancer Trial

Total doses/dose per fraction (Gy)

Conclusion of Beck-Bornholdt et al. [2]

Comment

Our conclusion

RTOG 79-13 (Marcial et al., 1987)

CF 70/1.8–2.0, HF 60/1.2

Severe late effects after HF significantly increased

Incorrect re-calculation of the published data

No significant difference in late effects after HF versus CF

EORTC 22791 (Horiot et al., 1992)

CF 70/1.8–2.0, HF 80.5/1.15

Severe late effects after HF not significantly increased. Late grade 2 reactions are not relevant. Since trial was too small, no reliable conclusion concerning grade 3 late reactions can be made

It is statistically and radiobiologically reasonable to consider grade 2 plus 3 reactions. The argument on the number of patients can not be used as evidence against HF

HF allows escalation of biological effective dose to the tumor (i.e. local control) without appreciably increased late reactions

EORTC 22811 (Van den Bogaert et al., 1995)

CF 70/2.0 or 75/1.7, HF 67.4/1.6 or 70.4/1.6

Severe late effects after HF significantly increased

The treatment schedule is no typical HF, but accelerated and hyperfractionated split-course. Consequential late effects and incomplete repair of sublethal damage may be of concern

No significant difference after CF and HF with regard to xerostomia and fibrosis. Increased late edema after HF. Paper should not be included in an analysis of late effects after HF

CF, conventional radiotherapy; HF, hyperfractionation. The conclusions drawn in the review by Beck-Bornholdt et al. [2] are summarized and commented on.

M. Baumann et al. / Radiotherapy and Oncology 46 (1998) 127–130

Marcial et al. [8] published the results of the RTOG HF trial in 1987. Beck-Bornholdt et al. [2] re-analyzed the data and argued that this study showed a significant increase in late sequelae in contradiction to the opinion of the RTOG investigators. This argument, however, arose from misinterpretation of the published data. Marcial et al. [8] reported the percentage rather than the number of patients developing severe late complications and assessed complications in the complete responders (i.e. a little more than half of the patients) rather than in all randomized patients as assumed by Beck-Bornholdt et al. [2]. Analysis after correction of these two parameters reproduced the original conclusion of Marcial et al. [8]. It would have been difficult to comprehend how a reduction in both fraction size and total dose would increase the late complication rate even after accounting for a moderate degree of incomplete repair of sublethal injury between fractions (see below). Horiot et al. [7] reported the data of EORTC trial 22791 and concluded that there was no difference in the actuarial incidence of combined grade 2 and 3 late side-effects between the HF and CF arms [7]. Beck-Bornholdt et al. [2] argued that only grade 3 reactions should be considered dose-limiting because of the severity of the malignant disease and because more than 50% of the patients show combined grade 2 and 3 late side-effects, which in their opinion gives rise to doubt about the relevance of these complications. Although this argument may have some clinical merit (e.g. radiation oncologists may choose to ignore mild to moderate toxicity) it is not well-founded from a radiobiological point of view. First, escalation of the biological effective dose will increase the whole spectrum of late normal tissue toxicity instead of grade 3 only. Second, a change in the complication rate from around the 50% level is detectable with a higher statistical power because it is situated at the steepest part of the dose-response curve. Another disputable argument is that Beck-Bornholdt et al. [2] gave prominence to the crude incidence of grade 3 reactions which was 12 and 7% after HF and CF, respectively. They point out that it will take about 1000 evaluable patients to resolve this difference with a given statistical power and a 10% significance level. This calculation is correct for determining the necessary sample size before starting a trial that is aimed to determine a 5% or greater difference of late effects. However, it is statistically not valid to use the non-significant difference of 5% to calculate a posteriori the necessary sample size because, by definition, non-significance means that the ‘real’ difference may as well be 0% or even assume negative values. Thus, the fact that the EORTC study with 356 patients could not resolve a difference of 5% in grade 3 late effects cannot be taken as circumstantial evidence that this difference is real. A radiobiological approach is more helpful for interpretation of the results of EORTC trial 22791 with regard to late normal tissue reactions. If we assume that late normal tissue injury can not be spared by reducing the dose per fraction from 2.0 to 1.15 Gy, then the 80.5 Gy delivered in the HF

129

arm of this trial would correspond to a 15% escalation of biological dose compared to CF. From our current knowledge on the steepness of dose-response curves we can estimate that such a dose escalation would raise the incidence of late reactions by at least 30%, say, from 35 to 65% [4]. From the number of patients followed up, it can be determined that the power of the EORTC study to detect such an increase was 99.8%. Thus, EORTC trial 22791 leaves little doubt that late normal tissue injury can be spared by hyperfractionation. The inclusion and interpretation of EORTC trial 22811 reported by Van den Bogaert et al. [11] in the review of Beck-Bornholdt et al. [2] are quite contentious. Van den Bogaert et al. [11] provided actuarial estimates for grade 2 plus 3 reactions for xerostomia, fibrosis and persistent edema and concluded that there was no significant difference between the HF and CF arms with respect to xerostomia and fibrosis, whereas persistent edema was significantly more frequent in the HF arm. However, Beck-Bornholdt et al. [2] chose to emphasize the crude incidence of severe grade 3 reactions and presented this end-point in their summary table. In addition, the regime tested in EORTC 22811 is a pseudo-altered fractionation schedule by prevailing criteria. It delivers two courses of radiation with 3–4 weeks gap; both courses consist of three fractions of 1.6 Gy with 4 h intervals to 48 Gy in 10 days and 19.2–24 Gy in 4–5 days, respectively. Since the total dose and the overall duration, including the break, are in the range of those of CF, it cannot be considered a pure hyperfractionation nor an accelerated fractionation regimen. Moreover, some of the complications could represent consequential late reactions (i.e. secondary to very severe acute reactions) or result from incomplete repair of sublethal damage due to the use of 4 h intervals. Experimental and clinical studies show that it takes more than 6 h for cellular repair to approach completion in a number of tissues of concern in radiotherapy of head and neck cancer [1,5,6]. Delivery of three fractions a day with 4 h intervals compounds this problem. Therefore, we support the approach of Stuschke and Thames [9] of excluding this series in the review. All in all, despite a number of very relevant methodological concerns raised by Beck-Bornholdt et al. [2], we do not agree that the data of the clinical studies published so far show an appreciable increased late toxicity in the HF arms.

4. Conclusion Plausible explanations for why two overviews on hyperfractionated radiotherapy analyzing the same published material came to opposing conclusions are differences in the definition of HF, in the inclusion criteria for published series for review, in the interpretation of the published results and, particularly, in the question addressed. EORTC trial 22791 by Horiot et al. [7] is so far the cleanest prospective study testing the therapeutic advantage

130

M. Baumann et al. / Radiotherapy and Oncology 46 (1998) 127–130

of hyperfractionated radiotherapy. This trial supports the idea that hyperfractionation allows the escalation of the biological effective dose to tumors (i.e. local control) without a simultaneous increase in late complications. Nevertheless, the review of Beck-Bornholdt et al. [2] points out some important issues to be addressed in future research. Hyperfractionation arms are included in ongoing clinical trials on head and neck cancer sponsored by the RTOG (no. 90-03) and EORTC (no. 22962). The results of these studies will hopefully contribute to more insight into the role of hyperfractionated radiation therapy.

References [1] Ang, K.K., Jiang, G.L., Guttenberger, R., Thames, H.D., Stephens, L.C., Smith, C.D. and Feng, Y. Impact of spinal cord repair kinetics on the practice of altered fractionation schedules. Radiother. Oncol. 25: 287–294, 1992. [2] Beck-Bornholdt, H.P., Dubben, H.H., Liertz-Petersen, C. and Willers, H. Hyperfractionation: where do we stand? Radiother. Oncol. 43: 1–21, 1997. [3] Bentzen, S.M. Quantitative clinical radiobiology. Acta Oncol. 32: 259–275, 1993. [4] Bentzen, S.M. Radiobiological considerations in the design of clinical trials. Radiother. Oncol. 32: 1–11, 1994. [5] Bentzen, S.M., Ruifrok, A.C.C. and Thames, H.D. Repair capacity and kinetics for human mucosa and epithelial tumors in the head and neck: clinical data on the effect of changing the time interval between multiple fractions per day in radiotherapy. Radiother. Oncol. 38: 89–101, 1996.

[6] Hopewell, J.W. and van den Aardweg, G.J. Studies of dose-fractionation on early and late responses in pig skin: a reappraisal of the importance of the overall treatment time and its effects on radiosensitization and incomplete repair. Int. J. Radiat. Oncol. Biol. Phys. 21: 1441–1450, 1991. [7] Horiot, J.C., Le Fur, R., N’Guyen, T., Chenal, C., Schraub, S., Alfonsi, S., Gardani, G., Van den Bogaert, W., Danczak, S., Bolla, M., Van Glabbeke, M. and De Pauw, M. Hyperfractionation versus conventional fractionation in oropharyngeal carcinoma: final analysis of a randomized trial of the EORTC cooperative group of radiotherapy. Radiother. Oncol. 25: 231–241, 1992. [8] Marcial, V.A., Pajak, T.F., Chang, C.C., Tupchong, L. and Stetz, J. Hyperfractionated photon radiation therapy in the treatment of advanced squamous cell carcinoma of the oral cavity, pharynx, larynx, and sinuses, using radiation therapy as the only planned modality: (preliminary report) by the Radiation Therapy Oncology Group (RTOG). Int. J. Radiat. Oncol. Biol. Phys. 13: 41–47, 1987. [9] Stuschke, M. and Thames, H.D. Hyperfractionated radiotherapy of human tumors: overview of the randomized clinical trials. Int. J. Radiat. Oncol. Biol. Phys. 37: 259–267, 1997. [10] Thames, H.D., Peters, L.J., Withers, H.R. and Fletcher, G.H. Accelerated fractionation vs. hyperfractionation: rationales for several treatments per day. Int. J. Radiat. Oncol. Biol. Phys. 9: 127–138, 1983. [11] Van den Bogaert, W., van der Schueren, E., Horiot, J.-C., De Vilhena, M., Schraub, S., Svoboda, V., Arcangeli, G., De Pauw, M. and Van Glabbeke, M. The EORTC randomized trial on three fractions per day and misonidazole (trial no. 22811) in advanced head and neck cancer: long term results and side effects. Radiother. Oncol. 35: 91–99, 1995. [12] Weed, D.L. Methodological guidelines for review papers. J. Natl. Cancer Inst. 89: 6–7, 1997. [13] Withers, H.R., Peters, L.J., Thames, H.D. and Fletcher, G.H. Hyperfractionation. Int. J. Radiat. Oncol. Biol. Phys. 8: 1807–1809, 1982.