Toxicity and Outcome Results of RTOG 93-11: A Phase I-II Dose-Escalation Study Using Three-Dimensional Conformal Radiotherapy in Patients with Inoperable Non-small-cell Lung Carcinoma

Background

The RTOG (Radiation Therapy Oncology Group) initiated dose escalation protocol 93-11 for patients with stage I-III lung cancer in order to attempt improvement in local control. Three-dimensional conformal radiotherapy was used to increase the dose but minimize normal tissue toxicity to structures such as the lungs, spinal cord and esophagus. There is no clear data for the optimal dose to treat inoperable non-small cell lung cancer (NSCLC), but local control rates have shown improvement with increasing dose in prior studies. Due to the proximity of important normal structures, the tumor dose has traditionally been limited to 60-70 Gy. Arriagada et al showed only a 17% pathological local control rate after a dose of 65 Gy. From principles advocated by Fletcher, it is thought that doses approaching 100 Gy may be needed.

Methods

• Phase I-II dose escalation study
• The primary objective was to determine the maximal tolerated dose that can be safely delivered to patients
• The secondary objectives were local control and survival rates
• Needed to have Stage I-III NSCLC
• Patients ineligible if they had:
• • supraclavicular disease
  • history of concurrent chemotherapy
  • history of prior radiotherapy (RT)
  • complete tumor resection, or eligibility for definitive surgery
  • recurrent disease
• Staging included chest x-ray, CT (computed tomography) chest and upper abdomen with IV contrast, PFTs (pulmonary function tests)
• Bone scan only for bone pain or elevated alkaline phosphatase
• GTV (gross tumor volume) included the primary tumor and any enlarged lymph nodes (>1cm on short axis)
• PTV (planning target volume) was a 1.0 cm margin around the GTV; however, fluoroscopy was used to check the PTV margin with respect to respiratory variation
• Doses were prescribed to the center of the PTV, ensuring that the 93% isodose line covered the PTV, and that the maximum dose did not exceed 107%
• No corrections for lung heterogeneity were performed for the treatment plan, but a second plan was submitted with heterogeneity corrections
• The total lung volume was defined as the volume of both lungs minus the PTV
• The total lung volume receiving greater than 20 Gy (V 20) was used to group patients:
• • Patients with a V 20 < 25% were in Group 1, and received escalating total doses between 70.9 and 90.3 Gy.
  • Patients with a V 20 of 25-36% received doses between 70.9 and 77.4 Gy
  • Patients with a V 20 of > or = 37% received doses between 64.5 and 77.4 Gy. However, dose escalation in group 3 was closed early because of poor accrual.
• The dose per fraction was 2.15 Gy, but because the PTV was enclosed within the 93% isodose line, this translates to 2.0 Gy fractions if the PTV had been covered by the 100% line
• Acute and late toxicity was scored according to the RTOG (Radiation Therapy Oncology Group) scale
• Follow up scanning included a CXR 4 weeks after treatment, and a chest CT at 6 and 12 months, then yearly afterward
• PFTs were also obtained at 6 and 12 months

Results

• 179 patients were enrolled between 1995 and 2001
• 3 patients were excluded: 2 patients were ineligible, and eligibility could not be determined on the third
• Group 3 was closed in 1999 after enrolling only 2 patients
• No Group 2 patients were enrolled on the highest dose level because a chemoradiation study, the RTOG L-0117, was started at that time for the same patient population
• Most patients were at least 60 years old with a KPS of 70-80
• Most patients had either squamous cell Ca (41%) or adenoCa (34%)
• There were more patients with Stage I cancer in Group 1, and more with Stage III in Group 2. There were few Stage II cancers in either group.
• 14% received neoadjuvant chemotherapy, and all but one of these were Stage III patients
• The pre-treatment FEV1 (forced expiratory volume in 1 second) ranged from as low as 0.42 to as high as 5.95 liters. The mean values for both groups were around 1.4-1.6 liters.
• The median f/u ranged from 13 to 24 months, depending on the subgroup
• 9 patients had grade 3 or worse acute toxicity, but 4 of these were attributed to chemotherapy
• 5 others developed acute pneumonitis: 3 in the high 90.3 Gy dose in Group 1, and 2 in the high 77.4 Gy dose in Group 2.
• Late toxicities were more common than acute events: 12 patients in Group 1 developed late grade 3 or worse lung toxicity, including one case each of grade 4 and 5 toxicity in the highest dose (90.3 Gy) group. The grade 5 patient had a peripheral lung cancer treated with AP/PA fields and developed fatal hemoptysis 2 years after RT. There were 3 grade 3 toxicities in both of Group 2’s dose arms, with 1 grade 4 toxicity in the higher dose arm.
• On multivariate analysis, they studied dose, V20, age, gender, histology, stage, and neoadjuvant chemo, and only mean lung dose (p=0.004) and V20 (p=0.042) were found to be significant factors for grade 3 or worse lung toxicity
• Late grade 3 esophageal toxicity occurred in 5 patients total. There was no grade 4, but one case of grade 5 toxicity occurred in a patient who was in the high dose arm of 90.3 Gy in Group 1. He only had grade 1 esophagitis acutely, but later developed a stricture and then died of pneumonia presumably due to a tracheoesophageal fistula. On review of his plan, over half of his esophagus received around 70 Gy and over 11 cm were treated to over 90 Gy, with the max dose of 104 Gy with heterogeneity corrections.
• Note the following: When looking at the survival curves by stage, in general the overall survival starts out better for the lower stages, but it never really plateaus. However, this is beyond the median follow-up, so it’s difficult to interpret. The local-regional control looks very similar among the stages. When looking at the numerical figures for locoregional control and overall survival according to dose grouping, the only potential difference is that the control rates seem slightly better for the higher dose arm in group 2, although this is probably not statistically significant. There does not appear to be a benefit in group 1 with dose escalation
• 31 patients, or 18%, developed isolated regional nodal failure, (remember that elective nodal radiation was not performed). The location of failure was known in 28 of the 31 patients. Of these, about half had isolated failures inside the radiation field, while the other half were outside.
• The dyspnea index scores were compared between baseline and the first follow-up visit. Most (63%) of the patients showed no change, about a quarter (23%) had worsening of symptoms, and 14% actually had improvement

Author's Conclusions

• Most patients had small tumors (median 31.8 cc and 79.5 cc), and lung function typical of lung CA pts (FEV 1 50-65% predicted)
• Acute toxicity was minimal, but late toxicities were more common; however, they were acceptable for all accept the 90.3 Gy arm in group 1, where the two fatal toxicities occurred
• Locoregional failure remained a significant problem, and there was no apparent dose-response relationship. However, the authors warn that this study was not robust enough to compare results between the dose arms. Given that, they did suggest that the lack of dose response could be due to the prolonged treatment times needed to deliver the higher doses, and they mention the L-0177 as a possible solution to this, which uses hypofractionated RT along with chemotherapy.
• They remark that the overall survival for Stage I was typical, and for Stage III was superior to historical controls without concurrent chemotherapy.
• They conclude that elective nodal failure was <10%

Discussion

This study focused on the long-term toxicity associated with radiation dose escalation in non-small cell lung cancer. In order to escalate the dose safely, the study authors omitted the traditional treatment of clinically negative (but high risk) nodal regions, or so-called "elective nodal irradiation." They mention that the highest dose arm (90.3Gy) in Group 1 was too toxic. They also note that there is no obvious benefit seen with dose escalation, but remark that the trial was not designed to detect a difference, but rather only to evaluate safety.

The authors mention that one problem with dose escalation demonstrating a lack of efficacy is that the treatment is too prolonged (up to 42 fractions), and thus hypofractionation might address this problem when used with concurrent chemotherapy, as in the ongoing RTOG L-0177 trial. However, they do not mention that the only hypofractionated arm in this trial had to close early because of unacceptable lung toxicity. Also, they do not mention that one of the potential reasons for the locoregional failures might be the omission of elective nodal irradiation, as half of the isolated locoregional failures did in fact occur outside of the radiation field.

The study shows that the overall survival for the Stage III patients was superior to historical controls, but note that these patients all had had relatively low-volume disease compared to typical Stage III patients, and therefore likely represent a more favorable subset. The reported elective nodal failure was <10%, but this is for isolated failure only, and there were another 16% of patients who had locoregional failure in addition to distant metastases. Locoregional failure in itself arguably contributes to the distant failure, but there is no further data presented here as to whether the patients with combined locoregional and distant failure had locoregional failure inside or outside the radiation field.

As a follow-up to this study, the future results of the L-0177 trial evaluating dose escalation and concurrent chemotherapy will be very informative, although there is serious concern that the toxicities of this protocol will prove unacceptable. A potential new trial design might compare elective nodal irradiation to dose escalation in order to test whether "the juice is worth the squeeze", and whether omitting the elective nodes affects locoregional control and distant metastases. This comparison would be most meaningful both with and without concurrent chemotherapy, as many patients are simply not able to tolerate concurrent chemotherapy. Finally, incorporating targeted agents along with radiation may provide significant benefit and perhaps even obviate the need for dose intensification in the future.