Post-Operative Management of Head and Neck Cancer

Articles by Bernier et al and Cooper et al, published together in 2004, changed the standard of care for the post-operative management of patients with squamous cell cancer of the head and neck (HNC).¹,² Prior to the publication of these two randomized Phase III trials, the standard for post-surgical, adjuvant radiation treatment for stage III and IV head and neck cancer was radiation alone. These Phase III trials demonstrated that there was a significant advantage to adding bolus cisplatin to a course of radiation therapy in the post-operative setting. Chemoradiotherapy (CRT) with cisplatin resulted in significantly improved local regional control (LRC) in both trials but was not as consistently robust in improving survival. There are subtle differences in selection criteria and outcomes in each trial and the final results, while similar, are not identical.

The rationale for the development of these trials was based on an analysis of prior studies, none of which provided a definitive answer about the value of adding chemotherapy to the post-operative radiotherapy in advanced stage patients. It is evident from an analysis of failure from natural history reviews and Phase II and Phase III trials that the vast majority of patients with HNC fail in the local or regional tumor bed, and less frequently from the development of distant metastases (DM). In addition, specific poor pathologic prognostic features were associated with a higher risk of failure in the post-operative setting. Thus, it was hypothesized that improving LRC, through radiosensitization with chemotherapy, would improve survival. Furthermore, a significant effect might be best identified in patients with a higher risk of failure. Based on these hypotheses and ongoing Phase III trials with cisplatinum based CRT in advanced disease, these two Phase IIII trials were launched, one in Europe and one in North America. Both trials utilized the same therapy: Bolus cisplatin for 3 doses given every 3 weeks during a course of daily radiotherapy of 6000 to 6600 cGy. The important endpoint was LRC in the North American trial, whereas the EORTC chose Progression free survival (PFS) and survival as the primary end points. In my opinion, LRC is not as important as survival or PFS as LRC has a limited benefit to the patient compared to survival. Given the historical problems of survival and LRC, the North American study through the Radiation Therapy Oncology Group (RTOG 95-01) chose to address LRC as a more important measure of radiation efficacy and avoid the confounding problems associated with medical co-morbidities such as co-existing heart disease and, unfortunately, enhanced toxicity, which can reduce survival in HNC patients.

The EORTC trial, reported by Bernier, is the smaller of the two trials. Patients were selected if they had undergone a clinically complete resection and they had T3 or T4 primary tumors, N2 or N3 nodal disease (except T3N0 larynx cancer with no adverse pathologic features), or they had lymphovascular invasion (LVI), perineural involvement (PNI), a positive surgical margin, or extra capsular spread (ECS). Over 160 patients were randomized. The North American trial, reported by Cooper, included clinically completely resected patients with 2 or more positive lymph nodes, a positive margin, or ECS and randomized over 450 patients.

In both the EORTC and North American trial the local regional control was significantly improved in the CRT treatment group. The differences were highly significant, but this only translated into a significant survival advantage in the EORTC trial. The improved survival results in the EORTC study may have been due in part from the higher number of laryngeal/hypopharyngeal patients in the EORTC study in contrast to the North American trial that had a higher percentage of oropharynx cancer patients, felt to be a more favorable cohort in regards to radiation response. The North American trial showed a trend in improved survival, but this did not achieve significance. In both trials, the rate of DM was the same in the CRT and the standard treatment groups. In fact, in both trials, DM accounted for the majority of failures in the experimental arm as LRC improved, reversing the expected failure pattern. The CRT arms engendered more acute toxicity in both studies as would be expected. Only about 65% of patients could tolerate all 3 doses of cisplatin.

These trials together tell us that those patients who have had definitive surgery for HNC and have poor prognostic features would benefit from CRT as opposed to radiotherapy alone. At the least, LRC would be improved. Patient selection is critical to success. Although a combined analysis of the data suggested that only positive surgical margins and ECS were important determinants of the benefited treatment population, the broader EORTC criteria may still apply since the analysis of the North American trial was retrospective. It is possible that LVI and PNI, in the absence of other adverse features, identify a population who would have significant benefit from CRT and there is, from the EORTC trial, an indication to treat those patients. Given that 60% of patients treated with CRT failed in the North American trial and 50% in the EORTC trial, there is considerable room for improvement. Weekly therapies, with less toxicity, may improve drug delivery and LRC and are being tested. In addition, new agents such as cetuximab, or alternative sensitizing drugs, such as TAXanes, are also being tested in Phase II trials. These enhancements should improve LRC and may impact on DM. Phase III trials will eventually answer these questions.

Locally Advanced HNC

The Phase III trial of Bonner et al. comparing radiotherapy with or without cetuximab for locally advanced HNC is the first Phase III trial to identify a biologic treatment that significantly improved survival in patients with locally advanced HNC.³ This trial established cetuximab as an attractive, alternative option for the curative treatment of patients with locally advanced HNC. The primary endpoint of this trial was local regional control. Although not stated, survival was a secondary end point in the trial. Cetuximab was given as an initial loading dose of 400 mg and then weekly (250 mg), with radiotherapy, based on the early Phase I and Phase II trials in HNC. In this trial radiotherapy was given according to the institutional preference since there were several different radiotherapy regimens in use. Given the immense difficulty in getting consensus on radiation therapy scheduling, this compromise was necessary for timely accrual, and this concept has been adopted by other ongoing Phase III trials. Quality was reviewed centrally to assure adequate coverage and dosing. This trial was designed specifically to compare cetuximab-sensitized radiotherapy to radiotherapy alone and not to compare cetuximab sensitized-radiotherapy to cisplatinum-based chemoradiotherapy, which had become the standard of care during the analysis of this trial.

Over 400 patients were randomized and at the planned analysis there was a significant improvement in both the LRC and survival in the cetuximab arm. LRC was reduced 32%, and the median time to progression or death was extended from 14 to 24 months. This was highly significant. At 3 years LRC was 47% vs 34%, a significant difference. There was a 26% reduction in the risk of death in the cetuximab group and at 3 years survival was significantly better at 55% vs. 45%. Cetuximab had no effect on DM, which were the same in both groups. Importantly, cetuximab did not increase the rate of mucositis; however, serious hypersensitivity reactions occurred in 14% and skin rash was common and severe in 17%. In clinical practice both of these are important clinical issues. Rash can be extremely symptomatic and dose limiting.

Cetuximab was approved for curative CRT of locally advanced HNC on the basis of this trial, but the data remain incomplete in terms of establishing the best use of the agent. There has been no comparison of cetuximab to the accepted standard of cisplatinum based CRT. Several cisplatinum based trials were completed which demonstrated very significantly improved survival in the cisplatinum based CRT group compared to radiotherapy alone, and generally, those trials included more patients with very advanced disease. Furthermore, survival was relatively greater with the cisplatinum based CRT then that observed in this single cetuximab trial.

Given the highly multiple positive studies with cisplatin and the relative differences in enhanced survival, we generally recommend cisplatinum based CRT for our patients. Bolus cisplatinum is relatively toxic and is associated with poor compliance. We and others will recommend cetuximab to patients who we feel cannot tolerate cisplatinum and are candidates for primary radiotherapy. There are several studies ongoing to add cetuximab to cisplatinum or docetaxel or combine it with docetaxel plus cisplatinum and 5-fluoruracil (TPF) and carboplatinum/paclitaxel regimens. The RTOG is currently accruing to a randomized Phase III trial comparing accelerated chemo-radiation using bolus cisplatinum versus a similar regimen in addition to weekly cetuximab (RTOG 0522). Cetuximab may find its most valued place as a part of a multi-drug regimen.

Unresectable Disease

Two mature Phase III studies comparing TPF to cisplatinum plus 5-fluorouracil (PF) alone were presented at the American Society of Clinical Oncology (ASCO) meeting in 2006.⁴,⁵ The trial by Remenar et al, for the EORTC, randomized over 350 patients with unresectable disease to TPF versus PF followed by radiotherapy, while that of Posner et al, the TAX 324 trial, randomized over 500 patients but included post induction CRT with carboplatinum in both arms. The more mature European trial established TPF as a standard of care for patients with unresectable disease in Europe and North America. The primary end point of both trials was overall survival. PFS, toxicity, and quality of life were secondary end points.

The treatment plan used in the Remenar trials was based on an earlier Phase III study report with 10-year follow up by Zorat in which PF chemotherapy improved survival compared to radiotherapy in patients with unresectable but localized HNC. The EORTC felt the data for induction chemotherapy was best for unresectable cancer and limited their study to the unresectable population. At the time the trial was initiated, CRT was not considered a standard of care and was not included in the design. CRT was included in the design of the TAX 324 trial because of the high rate of local regional failure in Phase II TPF trials and important early data that CRT resulted in better LRC compared to radiotherapy alone in advanced disease. Thus, although both trials include TPF, they employ different post-induction strategies to treat the local regional area. The TAX 324 used the ore aggressive sequential approach that incorporates induction chemotherapy and CRT. In addition to the strategic differences in radiotherapy, the EORTC TPF regimen utilized 4 cycles of TPF with a cisplatinum dose of 75 mg/M2, compared to TAX 324 TPF which is 3 cycles with a cisplatinum dose of 100 mg/m2. In addition to the differences in cisplatinum dose intensity, the Europeans gave a 5 day infusion of 5-FU at 750 mg/m2/day (3750 mg/m2 total dose) compared to 4 days at 1000 mg/m2/day (4 gm/N2 total) in the TAX 324 trial. These differences my or may not be significant.

The Remenar trial demonstrated a reduction in the risk of death of 27% in the TPF group, with less toxicity and better compliance than PF. PFS and QOL were also significantly improved. The Posner trial demonstrated a 30% reduction in the risk of death, better dose intensity as a result of less global toxicity in the TPF arms. Survival in the TAX 324 study at 3 years was 56% versus 42% bin the PF arm, while the EORTC had 37% versus 24% 3 year survival for TPF and PF respectively.

TPF induction therapy appears to confer a substantial improvement in survival and QOL in patients with locally advanced HNC over a regimen incorporating only PF. The TAX 324 trial included a broader population of patients, some of whom had locally advanced but resectable disease. Despite this heterogeneous group, survival was dramatically improved. TAX 324 utilized a weekly, less intensive carboplatinum regimen for radiation sensitization that is more tolerable then cisplatin. Less neurotoxicity, nausea, vomiting, and renal toxicity would be expected with a low dose weekly regimen of this agent, and compliance during radiotherapy was excellent, with no differences between PF and TPF during the radiotherapy arms.

There are currently several Phase III trials in Europe and North America which are comparing TPF plus CRT to CRT alone and we await the results of these studies to provide level I evidence that in fact, induction regimens provide a reduction in distant metastasis and significantly improve survival. Many researchers in Europe have adopted the sequential TAX 324 approach of TPF with CRT as the comparator to CRT alone because of the results of the TAX 324 trial. Thus, TPF followed by radiotherapy or CRT represents a reasonable alternative to CRT alone. Importantly, compelling Phase III data are not available to support either CRT or TPF as an exclusive standard.

Update of 91-11 Intergroup Trial

The 5-year update of the 91-11 intergroup trial was presented by Arlene Forastiere at ASCO in 2006.⁶ The original study report was published in the New England Journal of Medicine in 2003 and described the results of a three arm trial comparing radiotherapy to CRT or induction chemotherapy with PF for intermediate stage larynx cancer. The original report changed the standard of care for larynx preservation therapy to CRT with bolus cisplatinum from induction chemotherapy with PF. The original primary end point of the trial was laryngectomy free survival (LFS) with secondary endpoints of survival, larynx preservation (LP), disease free survival (DFS), distant metastases (DM) and local regional control (LRC). Because this trial was aimed at finding a therapy that best preserved larynx function, a non-significant difference in survival was less important than the LFS and LP end points, assuming a non-detrimental effect on survival in the chemotherapy arms. It is notable that there was slight reduction in survival in the VA larynx trial in the PF induction chemotherapy arm compared to surgery alone. This update represents a radical reassessment of the original report and represents one of the only comparisons of PF induction chemotherapy to CRT in a randomized trial.

The CRT arm used the current North American Standard of bolus cisplatinum which has proven to improve survival in advanced disease and as an adjuvant therapy. The induction arm was PF chemotherapy using the same plan as in the original VA larynx trial and the radiation therapy was given once daily. Over 300 patients were entered. In the original analysis there was a significant advantage to CRT for LFS and LP compared to radiotherapy alone. The re-analysis with a minimum of 5 years follow up continues to show that laryngeal preservation, dead or alive with a larynx, is still significantly greater in the CRT arm than the other 2 arms. However, now, with the mature follow up, both CRT and PF result in an equal and significant increase in LFS, alive with an intact larynx, over radiotherapy alone. Both chemotherapy arms reduced DM equally well, although not significantly compared to radiotherapy. CRT is still significantly better for LRC, but surprisingly PF and CRT had equal DFS. Most importantly with the LFS result, PF was slightly better, although not significantly at 5 years, for survival. The overall survival at 5 years was between 50% and 60%. These data suggest that there were unexplained deaths, possibly from toxicity, in the CRT group that reduced survival, or that the PF arm resulted in a better outcome in several other areas, such as second primaries.

While the equivalence of PF and CRT for LFS and the slight advantage to PF for survival will not change CRT as the standard of care for intermediate stage disease, the results can be interpreted to mean that PF induction chemotherapy is at least equivalent to CRT in this setting. The recent presentations of three papers including a larynx preservation trial of TPF versus PF demonstrated in three different settings with over 1000 patients that TPF is superior to PF and less toxic. By implication TPF prior to chemo-radiation might contribute further to improved local control and survival outcomes in patients with locally advanced HNC. It will be interesting to see how this plays out in light of recent data emerging in regards to the rising incidence of HPV related oropharyngeal cancers. HPV cancers tend to respond well to radiation and may obviate the need for more aggressive regimens. In contrast, there remains considerable room for improvement for patients with advanced laryngeal and hypopharyngeal cancers and direct comparisons are currently underway although these trials are neither reported nor mature.

It is important that our prejudices in many clinical settings, even when justified by good scientific data, must still be tested where appropriate and that there remains an open question as to whether TPF therapy or CRT is the best treatment for locally advanced HNC.

References

¹ Bernier J, Domenge C, Ozsahin M, Matuszewska K, et al. EORTC Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med 2004;350:1945-1952.

² Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med 2004;350:1937-1944.

³ Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 2006;354:567-578.

⁴ Posner M, Hershock D, Le Lann L, et al. A Phase III multicenter trial of docetaxel cisplatin and 5-fluorouracil versus cisplatin plus 5-fluorouracil induction chemotherapy followed by chemoradiotherapy in patients with locally advanced squamous cell carcinoma of the head and neck. Proceedings of the American Society of Clinical Oncology: Special Scientific Session. Atlanta, Georgia. 2006.

⁵ Remenar E, Van Herpen C, Germa Lluch J, et al. A randomized Phase III multicenter trial of neoadjuvant docetaxel plus cisplatin and 5-fluorouracil versus neoadjuvant PF in patients with locally advanced, unresectable squamous cell carcinoma of the head and neck. Final analysis of EORTC 24971. Proceedings of the American Society of Clinical Oncology. Atlanta, Georgia. 2006. Abstract #5516.

⁶ Forastiere A, Maor M, Weber R, et al. Long term results of Intergroup RTOG 91-11: A Phase III trial to preserve the larynx - Induction cisplatin/5-FU and radiation therapy versus concurrent cisplatin and radiation therapy versus radiation therapy. Proceedings of the American Society of Clinical Oncology. Atlanta, Georgia. 2006. Abstract # 5517.