Correspondence
Treatment of Head and Neck Cancer
N Engl J Med 1998; 339:1330-1331October 29, 1998
- Article
To the Editor:
Brizel et al. (June 18 issue)1 conclude that hyperfractionated irradiation with concurrent chemotherapy “is more efficacious and not more toxic than hyperfractionated irradiation alone.” At three years, the estimated rate of overall survival was 55 percent in the combined-therapy group and 34 percent in the hyperfractionation group (P=0.07). This difference is not statistically significant.
Feeding tubes were required in 44 of 56 patients (79 percent) in the combined-treatment group and 29 of 60 patients (48 percent) in the hyperfractionation group. Soft-tissue necrosis occurred in 11 patients (20 percent) in the combined-treatment group and 7 patients (12 percent) in the hyperfractionation group. Sepsis developed in 14 patients (25 percent) in the combined-treatment group (including 1 who died) and 4 patients (6.7 percent) in the hyperfractionation group. How can the authors conclude that combined treatment is “not more toxic than hyperfractionated irradiation therapy alone”?
1 ReferencesRichard A. Evans, M.D.
1011 Augusta Dr., Suite 109, Houston, TX 77057-20151
Brizel DM ,Albers M ,Fisher SR , et al. Hyperfractionated irradiation with or without concurrent chemotherapy for locally advanced head and neck cancer. N Engl J Med 1998;338:1798-1804
Full Text | Web of Science | Medline
To the Editor:
Brizel et al. conclude that “combined treatment [chemotherapy and irradiation] for advanced head and neck cancer is more efficacious and not more toxic than hyperfractionated irradiation alone.” The data presented, however, do not provide very strong support for this conclusion.
The same number of patients in each group (six) had distant metastases. The contention that metastasis might have been prevented if five patients in the combined-treatment group had received additional chemotherapy is unsupported. Likewise, the proportion of patients with a local first recurrence of disease (at the primary site) did not differ significantly between the two groups: 35 percent in the hyperfractionation group and 29 percent in the combined-treatment group (P=0.55, by a two-sided Fisher's exact test). The greatest difference in the rates of treatment failure involved recurrences in regional nodes (in 15 of the 60 patients in the hyperfractionation group vs. 0 of 56 in the combined-treatment group).
The authors acknowledge that the two groups differed with respect to the nodal stage at diagnosis. According to our calculation, the P value for this difference (in a two-by-two comparison with a two-sided Fisher's exact test) is 0.09, not 0.31. The difference in nodal stage could explain the worse survival in the hyperfractionation group, a difference of borderline significance. Since nodal dissection was not performed in all the patients, conclusions about differences in regional control and survival are confounded. Differences in overall rates of recurrence at the primary site (with or without nodal recurrence) were not reported but would be of interest in interpreting the only statistically significant finding in the study, the difference in locoregional control. In addition, knowledge of the rates of salvage surgery and ultimate preservation of organ function in patients initially enrolled for the purpose of preserving organ function (nearly half the patients) is important in comparing the groups.
There was greater toxicity in the combined-treatment group, with a higher incidence of sepsis (14 percent, as compared with 4 percent in the hyperfractionation group; P=0.15), which caused the death of one patient, and an increased need for a feeding tube (44 percent vs. 29 percent, P=0.08). The increased toxicity was seen despite a 15 percent reduction in the intensity of the radiation dose in the combined-treatment group. The evidence of improved survival and possibly improved local control appears to be no stronger than the evidence of increased toxicity with combined treatment.
Improvements in local control and survival that have been demonstrated with combined treatment for lung, gastrointestinal, and breast cancers; lymphoma; pediatric soft-tissue sarcomas; and nasopharyngeal carcinoma required larger studies with less heterogeneous patient populations. The hope is that incremental improvements in radiotherapy, expanding chemotherapeutic options, and a better understanding of how to integrate these approaches, thanks to studies such as that reported by Brizel et al., will lead to more convincing evidence of the benefit of chemotherapy for selected patients with head and neck cancer.
John Rescigno, M.D.
Daniel F. Heitjan, Ph.D.
Columbia–Presbyterian Medical Center, New York, NY 10032To the Editor:
I do not believe that Dr. Brizel and colleagues are entitled to draw the conclusions they do from the evidence presented. As is so often the case with prospective, randomized studies of patients with advanced head and neck cancer, the multiplicity of primary sites and stages makes comparisons difficult, and much larger groups of patients are required than the 122 who underwent randomization in their study, only 116 of whom could be included in the analysis. In the current nationwide collaborative trial in the United Kingdom, for example, which has strong support from London, Edinburgh, Glasgow, Manchester, and Birmingham — all the major urban centers — close to 1000 patients have been recruited, a number barely sufficient for a comparison of chemotherapy with no chemotherapy.
In the study by Brizel et al., the patients in the two groups were not well matched. Most authorities on head and neck cancer agree that the most important prognostic feature is the nodal status in the neck, yet in the more favorable N0–N1 category, there were 23 patients in the hyperfractionation group as compared with 31 in the combined-treatment group; in the much more adverse N2–N3 category, there were 37 patients in the hyperfractionation group as compared with 25 in the combined-treatment group. These are important discrepancies in such a small overall sample. There were also mismatches with respect to the tumor site (or sites). The base of the tongue, to take one of several examples, is a particularly adverse primary location, yet 13 patients in the hyperfractionation group had primary tumors at this site, as compared with only 8 in the combined-treatment group.
Brizel et al. claim that there are large differences between their two treatment approaches, yet the hyperfractionation group probably had a worse initial prognosis than the combined-treatment group. The precise role of concurrent chemotherapy in advanced head and neck cancer and, in particular, the quantification of a potential benefit remain unclear.
J.S. Tobias, M.D.
Middlesex Hospital, London W1N 8AA, United KingdomAuthor/Editor Response
Dr. Brizel replies:
To the Editor: Dr. Evans has obviously misinterpreted the toxicity data and made erroneous conclusions. Table 2 of our article clearly indicates that the data are percentages of patients, not absolute numbers (e.g., feeding tubes were required in 44 percent and 29 percent of the combined-treatment and hyperfractionation groups, respectively). We did acknowledge that acute toxic effects were more frequent with the combined-treatment regimen. However, acute toxic effects eventually resolve after the completion of treatment. The functional disabilities from mutilating surgery, uncontrolled locoregional tumors, or both are lifelong. We are pursuing strategies to reduce both the acute and chronic toxic effects of radiotherapy and concurrent chemotherapy. Meanwhile, we contend that the marked improvement in the long-term outcome outweighs the modest increase in acute toxicity.
We agree with Rescigno and Heitjan that the differences in disease-free survival and overall survival were not statistically significant (i.e., P>0.05). However, the 95 percent confidence intervals indicate the high probability that these differences were in fact real. Moreover, the curves for locoregional control and disease-free survival both plateaued before two years. In a larger patient population, the differences would most likely have been statistically significant. One should not automatically accept or reject data on the basis of a P value ≤0.05 or >0.05. Otherwise, one runs the risk of confusing statistical significance with clinical relevance.
It is worth reiterating that locoregional control was the primary end point of the study. Disease-free survival and overall survival were assessed to address the problem of competing risks. We agree that a higher nodal stage worsens the overall prognosis. As we pointed out, however, controlling for the imbalance between the groups in nodal stage did not influence the significance of the improvement in locoregional control provided by combined treatment. The study groups were well balanced with respect to resectability, tumor stage, tumor size, and anatomical site — all important prognostic variables.
Recent well-executed multi-institutional studies of radiotherapy and concurrent chemotherapy with larger, more homogeneous populations all report outcomes similar to ours.1-5 The challenge in the future is to optimize the integration of irradiation and chemotherapy in order to maximize efficacy and minimize toxicity.
5 ReferencesDavid M. Brizel, M.D.
Duke University Medical Center, Durham, NC 277101
Al-Sarraf M ,LeBlanc M ,Giri PG , et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized intergroup study 0099. J Clin Oncol 1998;16:1310-1317
Web of Science | Medline2
Wendt TG ,Grabenbauer GG ,Rodel CM , et al. Simultaneous radiochemotherapy versus radiotherapy alone in advanced head and neck cancer: a randomized multicenter study. J Clin Oncol 1998;16:1318-1324
Web of Science | Medline3
Calais G ,Alfonsi M ,Bardet E , et al. Randomized study comparing radiation alone (RT) versus RT with concomitant chemotherapy in stages III and IV oropharynx carcinoma: preliminary results of the 94.01 study from the French group of radiation oncology for head and neck cancer. Prog Proc Am Soc Clin Oncol 1998;17:385-3854
Bourhis J ,Eschwege F . Radiotherapy-chemotherapy combinations in head and neck squamous cell carcinoma: overview of randomized trials. Anticancer Res 1996;16:2397-2402
Web of Science | Medline5
Brizel DM ,Albers ME ,Fisher SR , et al. Hyperfractionated irradiation with or without concurrent chemotherapy for locally advanced head and neck cancer. N Engl J Med 1998;338:1798-1804
Full Text | Web of Science | Medline
