- © 2005 by American Society of Clinical Oncology
Concurrent Hyperfractionated Radiotherapy and Low-Dose Daily Carboplatin/Paclitaxel in Patients With Early-Stage (I/II) Non–Small-Cell Lung Cancer: Long-Term Results of a Phase II Study
- Address reprint requests to Branislav Jeremić, MD, PhD, International Atomic Energy Agency, Division of Human Health, Applied Radiation Biology and Radiotherapy, Wagramer Strasse 5, A-1400 Vienna, Austria; e-mail: b.jeremic{at}iaea.org
Abstract
Purpose Feasibility and activity of concurrent hyperfractionated radiotherapy (Hfx RT) and low-dose, daily carboplatin and paclitaxel were investigated in patients with early-stage (I/II) non–small-cell lung cancer in a phase II study.
Patients and Methods Fifty-six patients started their treatment on day 1 with 30 mg/m2 of paclitaxel. Hfx RT using 1.3 Gy bid to a total dose of 67.6 Gy and concurrent low-dose daily carboplatin 25 mg/m2 and paclitaxel 10 mg/m2, both given Mondays through Fridays during the RT course, started from the second day.
Results There were 29 complete responses (52%) and 15 partial responses (27%), and 12 patients (21%), experienced stable disease. The median survival time was 35 months, and 3- and 5-year survival rates were 50% and 36%, respectively. The median time to local progression has not been achieved, but 3- and 5-year local progression-free survival rates were 56% and 54%, respectively. The median time to distant metastasis has not been achieved, but 3- and 5- year distant metastasis-free survival rates were 61% and 61%, respectively. The median and 5-year cause-specific survivals were 39 months and 43%, respectively. Acute high-grade (> 3) toxicity was hematologic (22%), esophageal (7%), or bronchopulmonary (7%). No grade 5 toxicity was observed. Late high-grade toxicity was rarely observed (total, 10%).
Conclusion Hfx RT and concurrent low-dose daily carboplatin/paclitaxel was feasible with low toxicity and effective in patients with stage I/II non–small-cell lung cancer. It should continue to be investigated for this disease.
INTRODUCTION
Early-stage non–small-cell lung cancer (NSCLC) corresponds to stage I (T1N0 and T2N0) and II (T1N1 and T2N1) disease.1 Since several years ago, it also includes T3N0 tumors in stage II because of a similar outcome of these patients to those having T1-2N1 disease.2 Although surgery remains the treatment of choice for the majority of these patients,1-3 some patients never undergo surgery because of a severe comorbidity, age, or refusal. Such patients have traditionally been treated with radiotherapy (RT). Numerous reports from the last several decades indicated effectiveness of RT for this disease,4-24 although a number of questions remained unanswered, such as those regarding “optimal” total dose and fractionation, necessity of elective nodal coverage, or establishment of prognostic factors,25 mostly because of the lack of prospective studies.
Chemotherapy (CHT) did not play an important role in the primary treatment of these patients. Occasional reports26 included a small percentage of patients treated with combined RT and CHT, because CHT has not been considered a necessary adjunct to RT in localized disease, and characteristics of patients frequently did not allow its administration. However, with more surgical candidates refusing surgery in recent years, and with an observation that fit elderly patients can tolerate RT/CHT treatment similar to their younger counterparts,27,28 we designed a phase II study trying to build on a premise that low-dose daily CHT may enhance RT effects on the locoregional level. We have also taken into account the fact that our previous studies in stage I and II NSCLC with hyperfractionated (Hfx) RT given alone ended up with excellent results29,30 and the fact that low-dose CHT including carboplatin (and etoposide) was effective and well tolerated in patients with stage III NSCLC.27,31-33 Also, carboplatin was administered in between the two daily fractions, aimed at enhancing local tumor control by being present in tumor cells at the time that cell-repair processes (after the first daily fraction) are taking place and to be present when the second daily fraction is taking place. Paclitaxel was added to carboplatin (instead of etoposide) because of mounting evidence that it may act as a radioenhancer and is effective in NSCLC, either alone or in combination platinating agents.34-36 It was also shown that carboplatin had no perceived effect on the pharmacokinetics of paclitaxel.37 Paclitaxel was administered initially 1 day before the first RT fraction to enable cell-cycle synchronization, thought to be its major mode of action, although reoxygenation with more fractionated drug administration and induced apoptosis were likely to occur.36 Low daily doses were administered before the first daily RT fraction to continue with cell synchrony. Low daily doses should have provided low plasma levels of the drug for prolonged exposure times, which is thought to be more important than peak serum concentrations.38,39
PATIENTS AND METHODS
To be eligible for this study, adult (≥ 18 years) patients had to have histologically or cytologically confirmed NSCLC classified as stage I or II by the International System,1 a Karnofsky performance score (KPS) of ≥ 70%, weight loss of less than 5%, and no previous therapy. Exclusion criteria included postoperative thoracic recurrence and a history of any prior or concurrent cancer (except that of skin), unless the patient had shown no evidence of disease for more than 5 years or malignant pleural effusion.
The pretreatment evaluation included medical history, physical examination, complete blood count, biochemical screening tests, chest radiography, bronchoscopy, and computed tomography (CT) of the thorax and upper abdomen. Brain CT scanning and pulmonary-function tests were performed whenever possible.
The treatment design is shown in Figure 1. On day 1 (always on a Monday), patients received 30 mg/m2 of paclitaxel during a 1-hour infusion. Premedication was not systematically administered but was prepared for the case of any adverse effect and included dexamethasone, ranitidine, and diphenhydramine. Concurrent RT/CHT started from day 2. Hfx RT with 1.3 Gy bid to a total dose of 67.6 Gy with concurrent intravenous CHT consisting of carboplatin 25 mg/m2 and paclitaxel 10 mg/m2 on each RT day. The daily paclitaxel dose was administered by using a 1-hour infusion scheduled to start at least 2 hours before the first daily fraction, thus leaving a 1-hour gap between the end of paclitaxel infusion and the first daily RT fraction. Carboplatin was administered as a short (30-minute) infusion administered 3 to 4 hours after the first daily fraction (ie, 1.5 to 2 hours before the second one). This study was performed after approval was obtained from the institutional ethics committee and informed consent was obtained from all patients.
RT was administered with 6- to 10-MV photons by using linear accelerators. The typical target volume included the primary tumor and ipsilateral hilum with a 2-cm margin (stage I), whereas in cases of stage II patients, it also included ipsilateral mediastinum from the suprasternal notch to a level 6 cm below the carina (upper and middle-lobe lesions) or to the diaphragm (lower-lobe lesions). The initial target volume was treated with a minimum dose of 49.4 Gy, after which the RT field was reduced to include all detectable tumor (total dose, 67.6 Gy). Doses were specified at mid-depth at central axis for parallel-opposed fields or at the intersection of central axes for other techniques. The maximum dose of 20.8 Gy was used for the contralateral lung, 44.2 Gy for the entire heart, 52 Gy for the spinal cord, and 57.2 Gy for the esophagus. Two daily fractions of 1.3 Gy were administered five times per week with an interfraction interval of 4.5 to 6 hours. An interfraction interval was assigned nonrandomly to each patient. Under no circumstances was it allowed to change from the shorter interval to the longer interval or vice versa.
Patients were first examined at the end of Hfx RT, then every month for 6 months after completion of Hfx RT, then every 2 months for 2 years thereafter, and finally every 4 months thereafter. A medical history for the intervening period was obtained, and physical examination, complete blood count, biochemical tests, and chest radiography were performed at each visit. Symptomatic biopsies were not performed routinely but were recommended if signs of tumor persistence or recurrence were present. Restaging at the time of any progression was performed by using CT scans of the thorax, brain, and abdomen; bone scan; abdominal ultrasound images; and radiographs of the bones, whenever necessary, in addition to the previously noted procedures.
Response criteria included a complete response, defined as the disappearance for at least 4 weeks of all measurable or assessable disease and the absence of new lesions. For measurable disease, partial response (PR) was defined as a 4-week reduction of greater than 50% of the sum of the products of the cross-sectional diameters of all measurable lesions, together with the absence of new lesions. For assessable lesions, PR was defined as a decreased tumor size observed for at least 8 weeks. Stable disease was defined as a reduction of less than 50% or an increase of less than 25% in the sum of the products of the cross-sectional diameters of all measurable lesions and no clear pattern of either regression or progression of disease for at least 8 weeks. Progression of disease was defined as an increase of greater than 25% in the sum of the products of the cross-sectional diameters of measured lesions together with an increase in assessable disease or the appearance of new lesions.
The RT-induced effects on normal tissue were assessed as either acute or late phenomena, according to the criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organization for the Research and Treatment of Cancer.40 End points in this study included overall survival and relapse-free survival as well as toxicity. Survival and relapse-free survival rates were calculated from the first day of treatment by the Kaplan-Meier method, and the differences were evaluated by the log-rank method. The interaction of each prognostic factor and their effect on overall survival were analyzed by using the Cox proportional-hazards model. In calculating recurrence-free survival rates, patients who developed either type of failure were considered at risk for the other end point and censored at the time of last evaluation. Statistical analyses were carried out by one of us (B.M.) using the computer program SPSS (SPSS Inc, Chicago, IL). Statistical tests were based on a two-sided significance level.
RESULTS
Between January 1996 and October 1998, 56 patients were treated with this approach. All patients completed treatment as planned. Patient characteristics are listed in Table 1. Complete response was observed in 29 patients (52%) and PR in 15 patients (27%), thus making the objective response rate 79%, whereas 12 patients (21%) had stable disease. The median survival time (MST) was 35 months and 1- to 5-year survival rates were 82%, 62%, 50%, 36%, and 36%, respectively (Fig 1). When cause-specific survival (CSS) was used as an end point, the median time was 39 months, and 1- to 5-year CSS rates were 82%, 68%, 55%, 43%, and 43%, respectively (Fig 1). The median time to local progression has not been achieved, and 1- to 5-year local progression-free survival rates were 82%, 69%, 56%, 54%, and 54%, respectively (Fig 2). The median time to distant metastasis has not been achieved, and 1- to 5-year distant metastasis-free survival rates were 84%, 74%, 61%, 61%, and 61%, respectively (Fig 2). At the time of analysis, 31 (55%) patients had died of disease, 20 (36%) patients were alive with no evidence of the disease, and 5 (9%) patients had died of intercurrent diseases while having no evidence of the disease otherwise. Analysis of the pattern of failure showed that local progression alone was observed in 11 patients (20%), distant metastasis alone was observed in 7 patients (12%), and failure occurred both locally and distantly in 13 patients (23%).
On survival analysis of potential prognostic factors influencing survival (Table 2), females did significantly better than males (P = .0137), whereas age was not found to influence survival (P = .24). Patients with a higher KPS did better than those with a lower KPS (P = .0028), as did those with peripheral tumors when compared with those having a central tumor location (P = .01). Although patients in stage I did better than those in stage II, the difference was not significant (P = .68). The same was found when stage grouping was broken per exact T and N substaging (P = .47). Finally, histology significantly influenced survival (P = .042), with patients having squamous cell carcinoma (SCC) histology doing better than other histological variants, although the difference versus NSCLC not otherwise specified and the difference between the three non-SCC histologies was not significant (data not shown). The reason for not undergoing surgery significantly influenced survival (P < .0001). In particular, patients refusing surgery did significantly better than those with coexisting morbidity (P < .0001) but not better than those not operated on at an advanced age, although there was a strong trend favoring those who refused surgery (P = .094). Elderly patients also had better survival than patients with coexisting morbidity (P = .04).
In a Cox univariate model (Table 2), sex was also found to influence survival, whereas age did not. KPS again favored patients with a higher KPS, but stage did not influence survival. Neither TNM staging nor histology influenced survival in this analysis. Finally, the reason for not undergoing surgery influenced survival. Therefore, sex, KPS, and the reason for not undergoing surgery were entered into a multivariate model. It was shown that both KPS and the reason for not undergoing surgery independently predicted survival outcome, whereas sex was of borderline significance (Table 2).
Acute high-grade (≥ 3) toxicities observed during this study are listed in Table 3. Hematologic (22%), esophageal (7%), and bronchopulmonary (7%) toxicities were observed. Because of acute toxic effects, irradiation was interrupted in 4 (7%) patients for 8 to 11 days (median and mean, 9.5 days). Dose modifications were not made. Late high-grade (≥ 3) toxicity was infrequent (Table 4). No grade 5 toxicities or treatment-related deaths were observed during this study.
DISCUSSION
The standard treatment approach for patients with inoperable early-stage NSCLC is RT alone. In recent years, there has also been an increase in the number of patients who refuse surgery, undergoing RT as the sole treatment modality. Although recent large and comprehensive reviews25,41 identified inherent biases in these mostly retrospective reports, RT alone was capable of producing MSTs of up to 30 months and 5-year survival rates of up to 30% and even 40% in T1N0.13 Furthermore, recent years also brought preliminary evidence that high-precision RT in a form of stereotactic radiosurgery42-44 or stereotactic fractionated RT45-53 may lead to an even better outcome in these patients.
A combination of RT and CHT has not been practiced in this patient population, probably because of an assumption that predominating is local disease with low metastatic potential, although reviews of the literature showed that distant metastasis may occur in up to one third of such patients.25,41 Another assumption has been that these patients could not tolerate successful administration of CHT. In the current study, we administered low-dose daily CHT to explore radioenhancing effects of the two drugs. We also limited eligibility criteria to those who would be fit enough to tolerate RT/CHT and excluded patients with weight loss of more than 5% and those with a KPS of less than 70%. Finally, of paramount importance for the success of this study was that more than two thirds of our patients actually refused surgery, which was in striking difference with other studies. Although the percentage of such patients in other studies was usually around 10%, in several studies it was more than 20%.5,14,29,30 It is interesting to observe the highest MSTs (up to 33 months) in studies having more patient refusals, which was coupled with the highest 5-year survival rates (up to 32%).5,14,29,30 Our strict policy of patient selection in the current study was rewarding. Results obtained are among the best ever reported (MST, 35 months; 5-year survival, 36%) and only five patients (9%) died of intercurrent disease. This figure compares favorably with the majority of literature, in which the CSSs, when reported, are approximately 15% to 20% higher than that for overall survivals.9,12,15-18 Furthermore, our RT/CHT did not cause more treatment interruptions than usually seen with Hfx RT alone,29,30 which is proven to have an adverse impact on treatment outcome.54 Finally, both acute and late high-grade toxicity was similar to both studies, which used Hfx RT alone29,30 or other fractionation regimens.4-24 The overall acute-toxicity rate in the current study was 36% and appears lower than previously reported in the literature on concurrent radiochemotherapy. In two concurrent RT/CHT arms of RTOG 9410,55 overall acute toxicity was 48% (standard fractionation) and 62% (bid fractionation). Possible explanations for this discrepancy may include larger RT fields in RTOG 9410,55 because of more advanced disease, and high-dose CHT as administered in RTOG 9410.55
Local progression was predominant, confirming observations from the literature, and only seven patients (12%) had isolated distant metastasis. Positron emission tomography should improve staging and will decrease the number of patients who would progress (locally or distantly) soon after the beginning of RT. Indeed, in our study two (29%) of seven distant failures occurred after 5 and 6 months, respectively, indicating that these cases would have likely been upstaged with the use of positron emission tomography.
During this study we combined Hfx RT and concurrent low-dose, daily paclitaxel/carboplatin. Although we have previously successfully used Hfx RT alone to a total tumor dose of 69.6 Gy using 1.2 Gy bid,29,30 during this study we used a somewhat more accelerated RT regimen. We shortened the total treatment duration for more than 1 week using a slightly higher dose per fraction (1.3 Gy in the current study v 1.2 Gy used previously) to enable better local tumor control, which addressed the issue of possible accelerated tumor clonogen repopulation, expected to occur during the RT course, although no formal phase I study on this fractionation (with or without concurrent CHT) was performed. This was accompanied by our standard approach of using low-dose daily CHT, which was proven to be an effective approach in a number of studies, including those of NSCLC,32-34 small-cell lung cancer,56,57 and head and neck cancer.58,59 In the current study, low-dose carboplatin remained unchanged from our previous studies.27,32,33 Taking care of cell-cycle synchrony, we administered paclitaxel 1 day before the first RT fraction, allowing enough time to achieve this. Then, a daily schedule of both drugs should have continued with synchrony (paclitaxel administered before daily RT) and enhancing effects (of both paclitaxel and carboplatin), possibly fortified by the administration of carboplatin in between the twice-daily fraction. In that way, we tried to enable the presence of both drugs in tumor cells during both daily fractions, although daily fractionation of paclitaxel administration was not based on previous data but rather on the assumption that this may be an adequate daily dose to achieve enhancement, bringing less toxicity than usually used higher doses of that drug. Results of such an intervention are encouraging. In this patient population, this concurrent regimen worked well with low toxicity.
Of previously investigated prognostic factors, sex seems not to play a major role in the outcome of patients,12,14,18,20,22,24,29,30 and the results of current study confirm this. Age, too, was frequently observed not to have an influence on treatment outcome,11,12,15,16,18-20,22,24,29,30 and only an occasionally14,17 detrimental effect of advanced age was found. Recent analyses focusing on elderly patients with early-stage NSCLC60-62 showed similar outcome for this patient population when treated with RT alone, which may go as high as 36% for a 5-year CSS.60 Our study also showed no influence of age on treatment outcome. Investigation of the influence of the KPS on treatment outcome is controversial. Although Dosoretz et al,7 Slotman and Karim,18 Kaskowitz et al,15 and Gauden et al22 found no influence of performance status on either overall survival or disease-specific survival, Rosenthal et al,24 Hayakawa et al,20 Kupelian et al,8 and Jeremić et al29,30 did note its effect on either overall survival and/or disease-specific survival/relapse-free survival. Similarly, in the current study we identified KPS as the strongest prognosticator of survival. Regarding histology, only Sibley et al17 found an improvement in CSS for squamous histology, and Gauden et al22 observed the same for the mixed (adenocarcinoma/SCC) histology with both overall survival and relapse-free survival as end points; all other studies observed no such effect.12,18,19,20,24,29,60 The latter is in concordance with the findings of the current study. Although SCC histology carried an improvement in survival in the Kaplan-Meier analysis, it disappeared when using a multivariate analysis. Finally, only Hayakawa et al20 observed an influence of tumor location (better for tumors located in the upper lobes or the superior segment of the lower lobes) on outcome of these patients; all other studies excluded its possible effect when comparing central versus peripheral locations.9,13,18,19,29 In the current study, patients with peripheral tumor location had significantly better survival than those with tumors located centrally. No rational explanation exists for this observation except perhaps that centrally located tumors had more propensity for earlier seeding than those located peripherally. However, this cannot be clearly established without an adequate randomized trial investigating a number of concurrent issues such as the elective nodal treatment and patterns of failure. The latter was infrequent in the current study and, therefore, prevented us of from making a hypothesis.
In conclusion, this large phase II study with strict criteria including favorable patients with early-stage NSCLC showed that Hfx RT to a total dose of 67.6 Gy and concurrent low-dose daily paclitaxel/carboplatin was safe and feasible in this patient population. Long-term results obtained with this approach are among the best ever reported, and this approach should continue to be investigated for this disease. The effectiveness and low toxicity of this regimen in early-stage NSCLC confirm recent observations in locally advanced NSCLC, in which the same regimen achieved an MST of 28 months and 5-year survival of 26%.63
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed discription of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
| Authors | Employment | Leadership | Consultant | Stock | Honoraria | Research Funds | Testimony | Other |
|---|---|---|---|---|---|---|---|---|
| Branislav Jeremić | Bristol-Myers Squibb (A) |
Dollar Amount Codes (A) < $10,000 (B) $10,000–99,000 (C) ≥ $100,000 (N/R) Not Required
Kaplan-Meier survival curves for overall (dashed line) and cause-specific (solid line) survival.
Kaplan-Meier survival curves for local progression-free (dashed line) and distant metastasis-free (solid line) survival.
Characteristics of Patients
Overall Survival
Acute High-Grade (≥ 3) Toxicity
Late High-Grade (≥ 3) Toxicity
Footnotes
-
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
- Received April 1, 2005.
- Accepted June 8, 2005.
