Bjørn Henning Gronberg, MD, PhD, presented a paper at the ASCO20 Virtual Scientific Program reporting astounding positive results favoring higher-dose, twice-daily radiation therapy in limited-stage small cell lung cancer.1 This was a phase II study (large for phase II but small for phase III) conducted by centers all over Scandinavia. Patients were randomly assigned to either three or four 5-day weeks of twice-daily radiation at 150 cGy per fraction in addition to standard concurrent chemotherapy. The extra 5 days of radiation improved 2-year overall survival from 46% to 70% and the projected median overall survival from 23 to 42 months.
Steven E. Vogl, MD
Study Background
From about 1980 until 2018, the only improvement in the treatment of small cell lung cancer was a refinement in radiation techniques, employing twice-daily radiation to a dose of 45 cGy that improved 5-year overall survival to 26% from 16% for once-daily radiation to the same total dose. The 2-year survival went to 47% from 41%.
In 2018, the first of two studies2,3 was published showing a modest 2-month improvement in median overall survival of extensive-stage small cell lung cancer with PD-L1–targeted drugs; this year, a PD-1–targeted drug came close to a positive result as well.4 That sums up 40 years of limited progress in limited- and extensive-stage small cell lung cancer.
Since the study published in 1999 that established twice-daily radiation to 4,500 cGy with chemotherapy as the standard regimen for limited disease5 (defined in 1970 as disease that could be encompassed in a single radiation field), research into radiation techniques has focused on testing whether a higher total radiation dose given once a day is equal to 4,500 cGy twice a day in outcomes. Once-a-day treatment is a more convenient schedule for patients, so they do not have to sit around waiting 6 hours to get their second daily dose. The results to date have shown no major differences but remain equivocal, with neither equivalence nor noninferiority established for a higher dose given once a day.
In 2007, a group from western Sweden published a series establishing the safety of 6,000 cGy given twice a day over 4 weeks with carboplatin and etoposide for patients without large lung primaries, poor lung function, or poor general condition.6 The present study is an attempt to test whether this dose escalation is worth pursuing.
Is a High Twice-Daily Dose Worth Pursuing?
The answer is, “Yes, it is worth pursuing.” However, this study is insufficient to establish a 4-week course of 150 cGy twice a day as a new standard of therapy.
While the study is small for a practice-changing study, it is not its size that limits its interpretation. Statistical analysis is very good at detecting the overinterpretation of small numbers of events, and the differences here were statistically significant as presented. The issues relate to the details of randomization and exclusions from the analysis of randomly assigned patients.
Do Not Give 100 mg/m2 of Cisplatin With Radiation and Etoposide!
Before going into these details, the dose of chemotherapy in the presentation needs to be corrected. The study abstract did not give the doses of cisplatin and etoposide administered. On the presenter’s slides, the cisplatin dose was given as 100 mg/m2 every 3 weeks with 3 days of etoposide at 100 mg/m2. This is a very toxic cisplatin dose, especially in an elderly population (28% of patients in this study were older than 70) and given together with other chemotherapy and large radiation fields.
The dose actually given in the study was 75 mg/m2 cisplatin once in each chemotherapy cycle. The usual dose is 60 to 75 mg/m2. Efforts to escalate doses of both drugs from 80 mg/m2 of cisplatin and 80 mg/m2 of etoposide daily for 3 days published in 1994 by a National Cancer Institute team showed that dose escalation of cisplatin and etoposide was much more toxic and no more effective.7 Dr. Gronberg has submitted corrected slides to ASCO.
Timing of Randomization Is a Big Issue
The randomization in this study occurred before the first dose of chemotherapy, 3 weeks before radiation therapy was to commence, often before a radiation oncologist had evaluated the patient, and occurred 6 weeks before the treatments actually diverged. This early randomization and some imprecision in the words of the presentation made many listeners question the results.
For a study in which the dose of radiation is key, one would like to ensure that everyone entered can technically be treated with the higher dose. This was not done in this study, in part because some centers did not have a local radiation department. The study was written with detailed parameters of acceptable doses of radiation to critical organs and volumes of heart and lungs. Such dose planning is standard in radiotherapy as practiced after 2010.
When this planning was finally done in the study, 5 of 89 patients already assigned to high-dose radiation were excluded because the treating department could not devise an acceptable treatment plan to a total dose of ≥ 5,400 cGy. These patients were treated to a dose of 4,500 cGy and kept in the intention-to-treat analysis, but not the analysis that was presented at the ASCO meeting. Some of the listeners to the presentation, including me, interpreted it as suggesting that total radiation doses were reduced because of toxicity during radiochemotherapy. That was not the case—the exclusions from the presented analysis were on the basis of an inability to devise safe treatment plans (Gronberg B, personal communication, June 4, 2020).
“The problem is that similar exclusions were not applied to patients assigned to lower-dose radiation, so the groups are no longer defined by similar criteria.”— Steven E. Vogl, MD
Tweet this quote
The problem is that similar exclusions were not applied to patients assigned to lower-dose radiation, so the groups are no longer defined by similar criteria. Had the randomization occurred after radiation consultation and treatment planning, perhaps at the beginning of the second cycle of chemotherapy, the groups would have been comparable and could also have been stratified according to toxicity to the first cycle of chemotherapy. If radiation planning is such that the first 4,500 cGy could be identical in each group, the ideal time to randomize would be at 4,500 cGy (at the beginning of the third chemotherapy cycle for patients on schedule), when half the patients would be randomized to continue radiation for another 10 fractions over 5 more days.
Randomly Assigned Patients Were Excluded for Events After Randomization
Such randomization closer to the divergence of treatment would have avoided five patient exclusions on the lower-dose radiation arm of the study for events that occurred after chemotherapy started but before any radiation was given—one each of withdrawal of consent, death, myocardial infarction, stroke, and renal failure. Apparently, none of these things happened in the higher-dose arm, which nevertheless also had five exclusions because of withdrawal of consent (one patient) or catastrophic intervening illness (four patients), some of it toxic.
The lower-dose arm also had more hematologic toxicity than the higher-dose arm, likely due to a failure of randomization in such a small study to balance risk factors for toxicity among the two arms, such as renal function, liver function, comorbidities, frailty, and weight loss. Evidence for an effect of the latter three is poor. The prediction of toxicity of chemotherapy is complex and uncertain, as are modifications of the initial dose for impairment of general health or specific organ dysfunction.
Random assignment of a large number of patients to each arm likely balances these variables, which are difficult to use for stratification. This study was probably not large enough (170 patients randomized and 160 analyzed) for this to occur.
Dr. Gronberg has done an analysis by the “intention-to-treat” principle without the 10 exclusions (5 from the high-dose arm who could not be safely treated to doses > 5,399 cGy and 5 from the low-dose arm who never got radiation because of withdrawal of consent or catastrophic intervening illness, some of which was toxic from the chemotherapy). He says the results are similar to the data he presented, and the difference in overall survival remains significant.
Repeat the Study
I conclude that, before we adopt the use of 6,000 cGy in twice-daily fractions starting early in chemotherapy as the standard treatment for limited-stage small cell lung cancer, this study should be repeated with the entry confined to patients in whom an acceptable plan can be devised to reach 6,000 cGy twice a day in 20 fractions over 26 days. Given the high rate of major toxicity in this study, consideration should be given to reducing the cisplatin dose to 60 mg/m2 per cycle in both arms.
“Given the high rate of major toxicity in this study, consideration should be given to reducing the cisplatin dose to 60 mg/m2 per cycle in both arms.”— Steven E. Vogl, MD
Tweet this quote
The inclusion of a PD-1 or PD-L1 monoclonal antibody will have to be addressed in the study design, given either during and after chemoradiation or given entirely after its completion, but should be identical in both arms. Since most patients with limited-stage small cell lung cancer still die of their disease, extrapolating from extensive disease suggests there will prove to be a survival advantage to including a PD-1 or PD-L1 antibody in initial therapy. Such a study using atezolizumab during and after chemotherapy and radiation is underway in the United States. Totally excluding immunotherapy would likely make the study design—a phase III trial comparing standard with higher doses of twice-daily radiation—out of date before it completes its accrual.
A One-Sided Test Is Not Appropriate in a Study Looking at Dose
The design of the successor study brings up another point. The statistical design of the current study employed a one-sided test of significance, implying that higher doses of radiation could not possibly be inferior to lower doses. The justification is that only superiority was of interest. I find this argument hard to accept, because the statistical question is whether the groups are different. If survival is lower at higher doses, this would be important and worth knowing.
That higher doses cannot give worse results was emphatically not the case in a recent RTOG study in non–small cell lung cancer, in which overall survival was substantially worse at a dose of 7,400 cGy compared to 6,000 cGy. Some investigators think this was the result of excessive radiation doses to the heart at the higher dose. Studies of high vs low doses of radiation or chemotherapy should be analyzed with two-sided tests, acknowledging the sometimes severe toxicities of the modalities we use to treat cancer. Happily, the overall survival analysis presented by Dr. Gronberg gave a P value of .027 using a two-sided test. Nonetheless, I think we still need to repeat this study.
DISCLOSURE: Dr. Vogl reported no conflicts of interest.
REFERENCES
1. Gronberg BH, Killingberg KT, Fløtten Ø, et al: Randomized phase II trial comparing the efficacy of standard-dose with high-dose twice-daily thoracic radiotherapy in limited disease small-cell lung cancer. ASCO20 Virtual Scientific Program. Abstract 9007.
2. Schmid P, Adams S, Rugo H, et al: Atezolizumab and nab-paclitaxel in advanced triple-negative breast cancer. N Engl J Med 379:2108-2121, 2018.
3. Paz-Ares L, Dvorkin M, Chen Y, et al: Durvalumab plus platinum-etoposide vs platinum-etoposide in first-line treatment of extensive-stage small-cell lung cancer (CASPIAN): A randomised, controlled, open-label, phase 3 trial. Lancet 394: 1929-1939, 2019.
4. Rudin C, Awad M, Navarro A, et al: KEYNOTE-604: Pembrolizumab or placebo plus etoposide and platinum as first-line therapy for extensive-stage small-cell lung cancer. ASCO20 Virtual Scientific Program. Abstract 9001.
5. Turrisi AT, Kim K, Blum R, et al: Twice-daily compared with once-daily thoracic radiotherapy in limited small-cell lung cancer treated concurrently with cisplatin and etoposide. N Engl J Med 340:265-271, 1999.
6. Hallqvist A, Rylander H, Björk-Eriksson T, et al: Accelerated hyperfractionated radiotherapy and concomitant chemotherapy in small cell lung cancer limited-disease: Dose response, feasibility and outcome for patients treated in western Sweden, 1998-2004. Acta Oncol 46:969-974, 2007.
7. Ihde DC, Mulshine J, Kramer BS, et al: Prospective randomized comparison of high-dose and standard-dose etoposide and cisplatin chemotherapy in patients with extensive-stage small-cell lung cancer. J Clin Oncol 12:2022-2034, 1994.
