A Multicenter Open-Label Randomized Phase II Study of Osimertinib With and Without Ramucirumab in Tyrosine Kinase Inhibitor–Naïve EGFR-Mutant Metastatic Non–Small Cell Lung Cancer (RAMOSE trial)

Abstract

PURPOSE

Preclinical studies demonstrated that dual inhibition of epidermal growth factor receptor (EGFR) and vascular endothelial growth factor (VEGF) pathways delay the emergence of resistance to EGFR tyrosine kinase inhibitors (TKIs), and in trials with first-generation EGFR TKIs, the combination of EGFR VEGF pathway inhibitors prolonged progression-free survival (PFS).

METHODS

The RAMOSE trial (ClinicalTrials.gov identifier: NCT03909334, HCRN LUN-18-335) is a randomized, open-label multicenter phase II study comparing osimertinib with ramucirumab (arm A) to osimertinib (arm B) for initial treatment of metastatic EGFR-mutant non–small cell lung cancer (NSCLC) with 2:1 random assignment. The primary end point is PFS for evaluable patients; secondary end points include objective response rates (ORRs), disease control rate (DCR), overall survival, and safety. The stratification criteria were EGFR mutation type and the presence of CNS metastasis.

RESULTS

At data cutoff on August 29, 2023, 160 patients consented, 147 patients received treatment, and 139 patients were evaluable with at least one scan. In this preplanned interim analysis, the median follow-up was 16.6 months. Among the evaluable patients, 57 PFS events occurred. The median PFS was 24.8 (A) versus 15.6 (B) months (hazard ratio, 0.55 [95% CI, 0.32 to 0.93]; log-rank P = .023), 12-month PFS rate was 76.7% (A) versus 61.9% (B; P = .026). No significant difference was observed in the ORRs and DCRs between arms. Any-grade (G) adverse events (AEs) occurred in 100% (A) and 98% (B) of patients, with no G5 treatment-related AE (TRAE), one G4 TRAE (hyponatremia, A), and 53% (A) versus 41% (B) G3 TRAEs. AE-related discontinuation occurred in 13 patients (9.7% in A and 8.7% in B). The safety profile was in line with known safety of each drug.

CONCLUSION

Ramucirumab plus osimertinib significantly prolonged PFS compared with osimertinib alone in patients with TKI-naïve EGFR-mutant NSCLC. The combination is safe and well tolerated.

INTRODUCTION

Genetic alterations in kinase domain of epidermal growth factor receptor (EGFR) lead to oncogenesis of lung cancer. Those mutations, deletions, and insertions can be classified to four structural function groups with each having different therapeutic options for patients.^1,2 For patients with advanced or metastatic non–small cell lung cancer (NSCLC) harboring classical EGFR mutations, third-generation EGFR tyrosine kinase inhibitors (TKIs), such as osimertinib,³ alone or in combination with platinum-doublet chemotherapy⁴ are the current standard of care as the first-line treatment worldwide. As monotherapy, osimertinib provides a progression-free survival (PFS) of 18.9 months and overall survival (OS) of 38.6 months, with a well-tolerated toxicity profile.^3,5 Osimertinib plus platinum-doublet chemotherapy provides a median PFS of 25.5 months with expected additional toxicities from chemotherapy.⁶ However, there is continued effort to improve upon the current clinical outcomes, including extending PFS. The vascular endothelial growth factor (VEGF)/VEGF receptor (VEGFR) pathway has been recognized as a key mediator of tumor angiogenesis. Both preclinical and clinical studies suggest that EGFR-mutant NSCLC appear to be particularly sensitive to VEGF pathway inhibition. Preclinical studies demonstrated that dual VEGF and EGFR blockade could delay emergence of EGFR TKI resistance in EGFR-mutant models.⁷ Furthermore, in EGFR-mutant NSCLC, constitutively active EGFR signaling increases VEGF through hypoxia-independent mechanisms,⁸ and elevated VEGF, in turn, contributes to the emergence of resistance to EGFR TKIs.⁹ In clinical trials, the addition of anti-VEGF therapy (bevacizumab or ramucirumab) to first-generation EGFR TKIs considerably improved clinical outcomes in patients with TKI-naïve EGFR-mutant NSCLC.^10-12

CONTEXT

• Key Objective

• The combination of anti–vascular endothelial growth factor with first-generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) can prolong progression-free survival (PFS) for metastatic EGFR-mutant non–small cell lung cancer (NSCLC), whether the combination of ramucirumab with osimertinib can also prolong PFS in comparison with osimertinib alone in this frontline setting?

• Knowledge Generated

• In the interim analysis of a multicenter, randomized, open-label phase II study (RAMOSE trial), 139 patients with TKI-naïve metastatic EGFR-mutant NSCLC were treated and evaluated, and the combination of ramucirumab with osimertinib had improved PFS (24.8 months) than osimertinib alone (15.6 months; hazard ratio, 0.55; P = .023). The combination was safe and well tolerated.

• Relevance (T.E. Stinchcombe)

• The combination of osimertinib and ramucirumab is a potential treatment for patients with EGFR-mutant NSCLC.*

• *Relevance section written by JCO Associate Editor Thomas E. Stinchcombe, MD.

Ramucirumab is a monoclonal antibody blocking VEGFR2 and widely used in treating metastatic lung cancer and other solid tumors in combination with different anticancer therapies.^13-15 In lung cancer, it was first approved as a second-line treatment after platinum-based therapy.¹³ The combination of ramucirumab with erlotinib was approved by US Food and Drug Administration (FDA) as first-line therapy for EGFR-mutant NSCLC on the basis of a randomized phase III RELAY trial (ClinicalTrials.gov identifier: NCT02411448), which demonstrated significantly longer PFS for the combination (19.4 months) compared with erlotinib monotherapy (12.4 months; hazard ratio [HR] 0.59 [95% CI, 0.46 to 0.76]).¹⁶ More recently, a phase I study demonstrated safety and preliminary efficacy of combination of osimertinib with ramucirumab in patients with EGFR-mutant T790M NSCLC.¹⁷ On the basis of preclinical and clinical safety evidence, we hypothesized that the addition of ramucirumab to osimertinib, a third-generation EGFR TKI, can prolong PFS in patients with TKI-naïve advanced NSCLC and classical EGFR mutations, compared with osimertinib alone, with a tolerable safety profile. A every-3-week ramucirumab infusion schedule was chosen¹³ to improve convenience to patients and potentially enhance compliance.

METHODS

Study Design and Patients

The RAMOSE trial is an open-label, randomized, multisite, phase II trial (ClinicalTrials.gov identifier: NCT03909334) conducted at 11 sites in the United States through Hoosier Cancer Research Network (HCRN LUN18-335). The protocol was approved by participating centers' institutional review boards (IRBs) and was reviewed annually by the institution's data and safety monitoring board. Patients age 18 years or older were eligible if they had locally advanced or metastatic NSCLC with documented EGFR exon 19 deletion or L858R mutations. Patients must have measurable disease (per RECIST Guidelines, v1.1). Baseline brain magnetic resonance imaging (MRI) was required. Patients with known brain metastases were eligible if asymptomatic and stable (not requiring steroids or on a stable or decreasing dose of ≤10 mg prednisone or equivalent). Previous EGFR TKI or previous antiangiogenic drugs were not allowed, but previous chemotherapy was allowed. Adequate organ functions were required. The study was conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki and in discussion with the US Food and Drug Administration. All patients provided written informed consent.

Study Procedures

Eligible patients were randomly assigned in a 2:1 ratio to arm A (ramucirumab 10 mg/kg intravenous [IV] once every 3 weeks plus osimertinib 80 mg oral once daily) versus arm B (osimertinib 80 mg oral once daily). The randomization was stratified by type of EGFR mutation (exon 19 deletion v L858R) and the presence versus absence of brain metastasis. For ramucirumab, dose could be reduced to 8 mg/kg or 6 mg/kg IV once every 3 weeks, and for osimertinib, dose could be reduced to 40 mg once daily (arms A and B), if necessary, in the presence of toxicity. Dose interruption was allowed. Tumor assessments (using computed tomography, positron emission tomography-computed tomography, or MRI) were performed and response evaluation was conducted using RECIST v1.1 per investigator at baseline, at 6 weeks after initiation of treatment, and then at 9-12-week intervals. Brain MRIs were required for all patients. AEs were monitored throughout the study and for 30 days after treatment discontinuation and, along with laboratory abnormalities, were graded by investigators, according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.

Study End Points

The primary end point was PFS in the evaluable patient population, evaluated by investigators, defined as disease progression per RECIST 1.1 or death due to any cause. Secondary outcome end point measures were objective response rate (ORR), disease control rate (DCR), duration of response, OS, and safety and tolerability. Exploratory end points were circulating tumor DNA dynamics, cytokine profiling, and peripheral blood mononuclear cell analyses and their association with clinical outcomes.

Statistical Analysis

Previous data estimated a median PFS of 18.9 months for patients receiving osimertinib alone. This study expected to prolong the median PFS to 29.7 months, corresponding to a HR of 0.64. Using a one-sided log-rank test with type 1 error of 0.10 lead to a sample size of 150 patients (100 on A and 50 on B) to provide 80% power to detect a HR of 0.64 or less. An interim analysis was planned once 54 events occurred among patients on the trial. Interim monitoring using the method of Lan-DeMets with O'Brien-Fleming type stopping boundaries provided the following nominal P < .020 for efficacy and P > .469 for futility. The final analysis for superiority will be conducted using a nominal P < .094. We anticipated an accrual rate of roughly 6.25 patients per month, therefore 24 months to complete the entire enrollment of 150 patients, upon which, patients would be followed for an additional 3 years, leading to a study duration of 5 years. The final analysis will occur after the last randomly assigned patient has been followed for 3 years, or 108 PFS events have transpired. The study was allowed to continue meeting randomization goals as planned by the Data and Safety Monitoring Committee at MD Anderson Cancer Center.

Human Investigations

The authors certify that informed consent from each participant was obtained by the investigators. The trial was performed after the approval of IRB at each institution.

RESULTS

Patient Population

Between October 2019 and July 2023, 161 patients were enrolled and 159 were randomly assigned at 11 sites in the United States. At data cutoff on August 29, 2023, for this interim analysis, seven patients were in screening phase, and 11 patients never started treatment and two patients immediately withdrew consent after first dose of protocol therapy because of adverse events (AEs) before any subsequent tumor assessment and therefore were considered nonevaluable (Appendix Table A1, online only). A total of 139 patients were allocated, treated, and received at least one tumor assessment on treatment, with 93 in arm A and 46 in arm B (Fig 1). Median age was 65 years, 71% were female, and 32% had smoked cigarettes. There were 62% White patients, 24% Asian patients, and 4% Black patients. Overall, baseline characteristics were balanced between two arms. At study entry, 46% patients had stable CNS metastasis (43% in A and 52% in B); 69% tumors had exon19 deletion and 31% had L858R mutations (Table 1). All patients were naïve to EGFR TKI therapies, and 13 (9.4%) patients received chemotherapy before starting on the trial for their advanced NSCLC.

FIG 1.

CONSORT diagram for the RAMOSE trial screening, randomization, allocation, treatment, and follow-up. AE, adverse event; EGFR, epidermal growth factor receptor; NSCLC, non–small cell lung cancer; PFS, progression-free survival.

TABLE 1.

Patient Demographics and Tumor Characteristics

Characteristic	Arm A (ramucirumab + osimertinib)	Arm B (osimertinib)	Total
Age, years, median (range)	65 (37-83)	65 (41-83)	65 (37-83)
Sex, No. (%)
Male	27 (29.0)	13 (28.3)	40 (28.8)
Female	66 (71.0)	33 (71.7)	99 (71.2)
Race, No. (%)
White	60 (64.5)	26 (56.5)	86 (61.9)
Asian	24 (25.8)	10 (21.7)	34 (24.5)
Black	3 (3.2)	2 (4.4)	5 (3.6)
Other	6 (6.5)	8 (17.4)	14 (10.1)
EGFR mutations, No. (%)
Del 19	64 (68.8)	32 (69.6)	96 (69.1)
L858R	29 (31.2)	14 (30.4)	43 (30.9)
CNS metastasis, No. (%)
No	53 (57.0)	22 (47.8)	75 (54.0)
Yes	40 (43.0)	24 (52.2)	64 (46.0)
Previous chemotherapy, No. (%)
No	84 (90.3)	42 (91.3)	126 (90.7)
Yes	9 (9.7)	4 (8.7)	13 (9.4)
ECOG PS, No. (%)
PS 0	31 (33.3)	18 (39.1)	49 (35.3)
PS 1	62 (66.7)	28 (60.9)	90 (64.8)
Smoking history, No. (%)
Never	64 (68.8)	29 (63.0)	93 (66.9)
Current	1 (1.1)	1 (2.2)	2 (1.4)
Former	28 (30.1)	15 (32.6)	43 (30.9)
Not collected	0	1 (2.2)	1 (0.7)
Tumor histology, No. (%)
Adenocarcinoma	83 (89.3)	41 (89.1)	124 (89.2)
Adenosquamous	1 (1.1)	1 (2.2)	2 (1.4)
Not otherwise specified	9 (9.7)	4 (8.7)	13 (9.4)

Abbreviations: ECOG, Eastern Cooperative Oncology Group; EGFR, epidermal growth factor receptor; PS, performance status.

Efficacy

At data cutoff, the median follow up was 16.6 months, and there were 57 PFS events: 32 in arm A and 25 in arm B. The primary end point is PFS by investigators. The median PFS was 24.8 (17.9-NR) months in ramucirumab with osimertinib arm A and 15.6 (11.7-22.8) months in the osimertinib monotherapy arm B (HR, 0.55 [95% CI, 0.32 to 0.93]; Cox model P = .026; log-rank P = .023). One-year PFS rates were 77% (95% CI, 65.0 to 84.9) in arm A and 62% (95% CI, 43.8 to 75.6) in arm B, and 2-year PFS rates were 51% and 30% in arms A and B, respectively (Fig 2A). The ORRs were similar in two arms, 76.3% (95% CI, 67.7 to 85.0) in arm A and 80.4% (95% CI, 69.0 to 91.9) in arm B (χ² P = .59). DCRs were also similar in both arms (96.8% [95% CI, 93.2 to 100] in A and 95.7% [95% CI, 89.8 to 100] in B, χ² P = .74; Fig 2B). The OS data are immature at time of this interim analysis, and a total of 20 deaths were observed (11/93 [11.8%] on arm A and 9/46 [19.6%] on arm B; χ² P = .221).

FIG 2.

Clinical efficacy. Bold indicates P = .026 is statistically significant. (A) Kaplan-Meier curves for PFS in arm A (ramucirumab with osimertinib) and arm B (osimertinib); (B) waterfall plots for responses per RECIST from baseline two arms. CR, complete response; HR, hazard ratio; NE, not evaluable; NR, not reached; PD, progressive disease; PFS, progression-free survival; PR, partial response; SD, stable disease.

The PFS benefit of osimertinib with ramucirumab was observed across most of the major subgroups, regardless of age, sex, race, types of EGFR mutations, and presence of CNS metastasis (Fig 3). Among patients with exon 19 deletion, the median PFS was 20.5 months in the arm A and 14.1 months in the arm B; among those with L858R mutation, it was 24.8 months and 18.4 months, respectively. Among the 64 patients with CNS metastases at baseline, the median PFS was 17.9 months in arm A and 13.8 months in arm B; among those without CNS metastases at baseline, the median was not reached in arm A and 18.4 months in arm B.

FIG 3.

Subgroup analysis for efficacy. EGFR, epidermal growth factor receptor; PH, proportional hazard.

Safety and Tolerability

The combination of ramucirumab with osimertinib was overall safe and tolerable with good patient adherence. On the basis of a median follow-up of 16.6 months, the ramucirumab on-treatment duration was 14.2 months; during the on-treatment time, the ramucirumab dose intensity was 86% (Fig 4). In the entire trial, there was no grade 5 AE, and one grade 4 AE, of hyponatremia in arm A (not attributed to ramucirumab; Appendix Table A2). Grade 3 AEs were observed in 53% in ramucirumab with osimertinib in arm A versus 41% in osimertinib alone in arm B. Diarrhea, fatigue, headache, and cough were the most common AEs in arm A, whereas diarrhea, fatigue, and rash were the most common AEs in arm B. There were no new safety signals.

FIG 4.

Treatment-emerged adverse events in the RAMOSE trial by treatment arms.

Treatment-related AEs (TRAEs) that were considered by the investigator to have a causal relationship to any trial treatment are reported in Appendix Table A3. TRAEs of special interest potentially related to ramucirumab were hypertension (13% G1-2 and 23% G3 in A, and 11% G1-2 and 2% G3 in B), epistaxis (41% G1-2, no G3 in A, and 3% G1-2 no G3 in B), and proteinuria (17% G1-2, 2% G3 in A, and 2% G1-2, no G3 in B). ILD or pneumonitis was very uncommon in our study. There were two patients on arm A (two of 93; 2.2%) and four patients on arm B (four of 46; 8.7%). The AE-related discontinuation rates were similar in two arms at 9.7% (9/93) in ramucirumab with osimertinib arm A versus 8.7% (4/46) in osimertinib monotherapy arm B. For ramucirumab, 56 patients had at least one dose interruptions (skipping of a dose), and nine patients had one dose level reduction (8 mg/kg IV once every 3 weeks) and seven patients had dose reduction to 6 mg/kg IV once every 3 weeks. The most common reasons for dose interruptions were proteinuria, hypertension, infusion reaction, anemia, and thrombocytopenia (Appendix Table A4).

DISCUSSION

The RAMOSE trial is a US-only, multicenter, phase II study, which enrolled patients with NSCLC with EGFR classical mutations naïve to EGFR TKIs. The RAMOSE trial interim analysis showed that the addition of ramucirumab to osimertinib significantly prolonged PFS to 24.8 months compared with osimertinib monotherapy at 15.6 months with a HR of 0.55 (95% CI, 0.32 to 0.93; P = .026), and the combination is overall safe and well tolerated, with excellent patient compliance.

These data aligned with previous randomized clinical trials combining VEGF pathway inhibitors with first-generation EGFR TKIs, such as the large phase III trials NEJ026¹¹ and ARTEMIS¹²: combining bevacizumab with erlotinib prolonged PFS with HR 0.55-0.63, compared with erlotinib monotherapy. In the RELAY study,¹⁶ the addition of ramucirumab to erlotinib resulted in a HR of 0.59 (95% CI, 0.46 to 0.76; P < .0001), and the PFS was 19.4 months versus 12.4 months, respectively. Our RAMOSE trial confirmed that the benefit of adding anti-VEGFR2 antibody to EGFR TKIs can be generalized from first- to third-generation TKI, providing further support that VEGF/EGFR pathway inhibition may delay the emergence of therapeutic resistance. Some trials combining osimertinib with VEGF inhibitors did not demonstrate significant improvement of PFS with the combination compared with osimertinib monotherapy. In both the WJOG9717 trial¹⁸ and the TORG1833 (OSIRAM) trial,¹⁹ the patients' exposure to VEGF inhibitors were quite short at 8 months and 4.2 months, respectively, which is likely the primary reason for the lack of superiority in PFS with combination.

Osimertinib and other third-generation EGFR TKIs have been established as the first-line treatment for EGFR-mutant NSCLC worldwide,³ but there is a continued need to improve the clinical outcome. Other than our RAMOSE trial, two large, randomized phase III trials also evaluated different first-line intensification approaches. In the global phase III FLAURA2 trial, the osimertinib-chemotherapy combination improved median PFS to 25.5 months, with a HR of 0.62 (95% CI, 0.49 to 0.79; P < .001).⁶ Furthermore, osimertinib with chemotherapy (FLAURA2 regimen) was approved by FDA for patients with advanced NSCLC harboring EGFR exon 19 deletion or L858R mutation.⁴ The MARIPOSA trial is also a global phase III trial, adding amivantamab, a bispecific antibody to EGFR and MET, to a third-generation EGFR TKI, lazertinib, which prolonged median PFS to 23.7 months, with an HR of 0.70 (95% CI, 0.58 to 0.85; P < .001).²⁰ Currently, those combinatory options are in different phases of clinical development and approval worldwide.

Since osimertinib monotherapy is a convenient oral medication with a very well-tolerated toxicity profile, it becomes critical to take a more tailored treatment recommendation to offer the appropriate intensification regimen to each patient. The major factors for this clinical decision making are subgroup-specific benefits, toxicity profiles, and financial logistical burden.

In previous studies, the L858R mutation subgroup was associated with a lesser-degree benefit from EGFR TKIs, including both early-generation TKIs, such as gefitinib or erlotinib,²¹ as well as third-generation osimertinib.³ In the FLAURA trial, a median PFS of 12.4 months was observed for osimertinib monotherapy³ compared with 21.4 months for the exon 19 deletion subgroup. It therefore represents another subgroup that might benefit from intensification. In our RAMOSE trial, PFS benefit was also observed for this subgroup as well, with a median PFS of 24.8 months. Our L858R subgroup had numerically longer PFS than exon 19 deletion subgroup in both arms, but the difference was not statistically different (Appendix Table A5), likely because of relatively small patient population size. Additional studies of osimertinib-based intensification strategies (or new treatments) for this subgroup are warranted.

Ramucirumab has a toxicity profile that is generally nonoverlapping with osimertinib and, like other VEGF pathway inhibitors, is associated with AEs including hypertension, proteinuria, and epistaxis. Overall, the AEs observed here were consistent with this profile. Fifty-four percent of patients had grade 3 or higher (≥G3) treatment-emerged AEs in arm A, compared with 41% in arm B. For side effects of special interest related to ramucirumab, there was hypertension G1-2 at 23% and G3 at 23%, proteinuria G1-2 at 17% and G3 at 2%, and epistaxis G1-2 at 41%. This observation was quite consistent with the RELAY study toxicity profile, which demonstrated very similar grades and incidences of hypertension, proteinuria, and epistaxis.¹⁶ There were no major bleeding or clotting events in the RAMOSE trial, and there was no new safety signal. Overall, patients were highly compliant with 86.6% dose intensity, also indicating excellent safety and tolerability. In comparison, the FLAURA2 study had 64% grade 3 or higher toxicities in the osimertinib-chemotherapy group without new safety signals. Cytopenia from all three lineages was as expected.⁶ The MARIPOSA trial had 75% ≥G3 toxicities in the amivantamab and lazertinib combination arm. In the MARIPOSA trial, venous thrombosis events were observed in 37% of patients, including 11% at grades 3, 4 and 5. Therefore, prophylaxis with anticoagulation was recommended for the first 4 months of amivantamab with lazertinib in studies with this combination.²⁰ Taken together, each intensification strategy has its own toxicity features, and patients' comorbidities and organ function need to be considered to select the suitable regimen.

The RAMOSE trial is an open-label, randomized, phase II study and has several limitations. The study size was moderate, therefore, not powered for additional exploratory analysis; for example, the interaction between smoking status and benefit of antiangiogenic therapies could not be evaluated.²² The current analysis was a preplanned interim analysis, with investigator-evaluated PFS as the primary end point, which could introduce bias, especially in an open-label study. The study is not powered for OS analysis and the data are immature at the interim analysis. Therefore, additional time for events to accumulate for final analysis will be important to confirm these findings, and further validation of the ramucirumab with osimertinib combination in the first-line setting in a larger study will be necessary. The trial exclusively enrolled patients in the United States, which may be an important consideration regarding the generalizability of findings. Being an open-label study, there is potential for patient and physician bias, which can lead to imbalance in two study arms. At the time of random assignment, the CNS metastatic rates were balanced; however, because of more arm B patients deciding to withdraw consent, the CNS metastasis rates were numerically imbalanced, although there was no statistical difference (Appendix Table A6). Moreover, the protocol was amended to reduce clinical visits for the osimertinib monotherapy group, while patients in the osimertinib plus ramucirumab group continued to have visits every 3 weeks. This amendment in theory has an early-detection effect on the combination arm, and therefore, it should not undermine the positive PFS outcome of the study.

In summary, data from the phase II, randomized, multicenter study RAMOSE trial indicate that the addition of ramucirumab to osimertinib significantly prolongs PFS in TKI-naïve patients with NSCLC harboring EGFR classical mutations. The combination demonstrated a safety profile in line with each drug without new safety signals, presenting a promising intensification strategy.

APPENDIX

TABLE A1.

Reasons for Patients Not on Treatment After Random Assignment

Subject ID	Arm	Reason
335-1006	A	Patient withdrawal before therapy start
335-1008	A	Screen failure
335-1017	A	AE/side effects/complications. Patient had allergic reaction to ramucirumab, come off study completely
335-1021	A	Screen failure because of new hemoptysis
335-1030	A	Patient withdrawal before therapy start
335-2008	B	Patient declined trial participation after random assignment
335-2032	B	Patient withdrawal before therapy start
335-2046	B	Patient withdrawal before therapy start
335-3001	B	Patient withdrawal before therapy start
335-3002	A	Screen failure because of new pulmonary emboli
335-3021	A	AE/side effects/complications. Headaches because of ramucirumab. Coming off treatment completely
335-4020	B	Patient withdrawal before therapy start
335-4027	B	Patient withdrawal before therapy start

Abbreviation: AE, adverse event.

TABLE A2.

TEAEs by Treatment Arm

Description	Arm A (ramucirumab + osimertinib), No. (%)					Arm B (osimertinib), No. (%)
	G1	G2	G3	G4	Any Grade	G1	G2	G3	Any Grade
Diarrhea	52 (55.9)	11 (11.8)	3 (3.2)	0	66 (71.0)	25 (54.4)	5 (10.9)	0	30 (65.2)
Skin rash	46 (49.5)	10 (10.8)	1 (1.1)	0	57 (61.3)	25 (54.4)	1 (2.2)	1 (2.2)	27 (58.7)
Fatigue	42 (45.2)	10 (10.8)	3 (3.2)	0	55 (59.1)	15 (32.6)	4 (8.7)	0	19 (41.3)
Headache	37 (39.8)	9 (9.7)	0	0	46 (49.5)	7 (15.2)	1 (2.2)	0	8 (17.4)
Cough	36 (38.7)	10 (10.8)	0	0	46 (49.5)	8 (17.4)	2 (4.4)	0	10 (21.7)
Epistaxis	38 (40.9)	3 (3.2)	0	0	41 (44.1)	3 (6.5)	0	0	3 (6.5)
Paronychia	30 (32.3)	8 (8.6)	0	0	38 (40.9)	13 (28.3)	0	0	14 (30.4)
Anorexia	24 (25.8)	10 (10.8)	0	0	34 (36.6)	3 (6.5)	1 (2.2)	1 (2.2)	5 (10.9)
Hypertension	3 (3.2)	10 (10.8)	21 (22.6)	0	34 (36.6)	1 (2.2)	4 (8.7)	1 (2.2)	6 (13.0)
Platelet count decreased	22 (23.7)	11 (11.8)	1 (1.1)	0	34 (36.6)	7 (15.2)	0	1 (2.2)	8 (17.4)
Dry skin	30 (32.3)	2 (2.2)	0	0	32 (34.4)	11 (23.9)	1 (2.2)	0	12 (26.1)
Nausea	28 (30.1)	4 (4.3)	0	0	32 (34.4)	6 (13.0)	1 (2.2)	0	7 (15.2)
Alopecia	25 (26.9)	3 (3.2)	0	0	28 (30.1)	10 (21.7)	0	0	10 (21.7)
Anemia	21 (22.6)	2 (2.2)	3 (3.2)	0	26 (28.0)	6 (13.0)	1 (2.2)	2 (4.4)	9 (19.6)
Pain in extremity	19 (20.4)	7 (7.5)	0	0	26 (28.0)	5 (10.9)	2 (4.4)	0	7 (15.2)
Back pain	15 (16.1)	8 (8.6)	2 (2.2)	0	25 (26.9)	8 (17.4)	1 (2.2)	0	9 (19.6)
Dyspnea	16 (17.2)	7(7.5)	1 (1.1)	0	24 (25.8)	4 (8.7)	5 (10.9)	1 (2.2)	10 (21.7)
WBC decreased	19 (20.4)	5 (5.4)	0	0	24 (25.8)	3 (6.5)	2 (4.4)	0	5 (10.9)
Pruritus	22 (23.7)	2 (2.2)	0	0	24 (25.8)	3 (6.5)	0	1 (2.2)	4 (8.7)
Vomiting	19 (20.4)	5 (5.4)	0	0	24 (25.8)	2 (4.4)	1 (2.2)	0	3 (6.5)
Hyponatremia	13 (14.0)	4 (4.3)	2 (2.2)	1 (1.1)	20 (21.5)	7 (15.2)	0	1 (2.2)	8 (17.4)
Proteinuria	5 (5.4)	11 (11.8)	2 (2.2)	0	18 (19.4)	0	1 (2.2)	0	1 (2.2)

Abbreviations: G, grade; TEAEs, treatment-emergent adverse events.

TABLE A3.

TRAEs by Treatment Arm

Description	Arm A (ramucirumab + osimertinib), No. (%)				Arm B (osimertinib), No. (%)
	G1	G2	G3	Any Grade	G1	G2	G3	Any Grade
Diarrhea	49 (52.7)	11 (11.8)	3 (3.2)	63 (67.7)	25 (54.4)	4 (8.7)	0	29 (63.0)
Skin rash	46 (49.5)	10 (10.8)	0	56 (60.2)	22 (47.8)	1 (2.2)	1 (2.2)	24 (52.2)
Paronychia	27 (29.0)	8 (8.6)	0	35 (37.6)	13 (28.3)	0	0	14 (30.4)
Epistaxis	30 (32.3)	3 (3.2)	0	33 (35.5)	1 (2.2)	0	0	1 (2.2)
Fatigue	23 (24.7)	9 (9.7)	1 (1.1)	33 (35.5)	9 (19.6)	2 (4.4)	0	11 (23.9)
Dry skin	29 (31.2)	2 (2.2)	0	31 (33.3)	10 (21.7)	1 (2.2)	0	11 (23.9)
Platelet count decreased	20 (21.5)	11 (11.8)	0	31 (33.3)	4 (8.7)	0	0	4 (8.7)
Hypertension	2 (2.2)	9 (9.7)	20 (21.5)	31 (33.3)	0	0	0	0
Headache	23 (24.7)	7 (7.5)	0	30 (32.3)	0	0	0	0
Alopecia	19 (20.4)	3 (3.2)	0	22 (23.7)	9 (19.6)	0	0	9 (19.6)
AST increased	18 (19.4)	3 (3.2)	0	21 (22.6)	2 (4.4)	2 (4.4)	1 (2.2)	5 (10.9)
Anorexia	15 (16.1)	6 (6.5)	0	21 (22.6)	2 (4.4)	1 (2.2)	1 (2.2)	4 (8.7)
WBC decreased	17 (18.3)	4 (4.3)	0	21 (22.6)	2 (4.4)	2 (4.4)	0	4 (8.7)
Pruritus	19 (20.4)	2 (2.2)	0	21 (22.6)	4 (8.7)	0	0	4 (8.7)
Nausea	19 (20.4)	1 (1.1)	0	20 (21.5)	3 (6.5)	0	0	3 (6.5)
Mucositis oral	14 (15.1)	5 (5.4)	0	19 (20.4)	3 (6.5)	1 (2.2)	0	4 (8.7)
Proteinuria	5 (5.4)	11 (11.8)	2 (2.2)	18 (19.4)	0	0	0	0
Edema limbs	15 (16.1)	2 (2.2)	0	17 (18.3)	2 (4.4)	0	0	2 (4.4)
Anemia	13 (14.0)	3 (3.2)	1 (1.1)	17 (18.3)	5 (10.9)	0	0	5 (10.9)
Infusion-related reaction	6 (6.5)	8 (8.6)	2 (2.2)	16 (17.2)	0	0	0	0

Abbreviations: G, grade; TRAEs, treatment-related adverse events.

TABLE A4.

Reasons for Ramucirumab Dose Interruption or Reduction

Adverse Event Term	Frequency	%
Abdominal distension	3	1.78
Anemia	12	7.1
Anorexia	1	0.59
Arthralgia	1	0.59
Blood and lymphatic system disorders	2	1.18
Colitis	1	0.59
Colonic hemorrhage	1	0.59
Constipation	1	0.59
Creatinine increased	10	5.92
Diarrhea	1	0.59
Dyspnea	2	1.18
Ejection fraction decreased	1	0.59
Eye disorders—other, specify	1	0.59
Fall	2	1.18
Fatigue	5	2.96
Flank pain	1	0.59
GI disorders—other	2	1.18
General disorders and administration	3	1.78
Hemorrhoidal hemorrhage	2	1.18
Hypertension	16	9.47
Hyponatremia	1	0.59
Hypotension	1	0.59
Infections and infestations—other	3	1.78
Infusion-related reaction	14	8.28
Injury, poisoning, and procedural complications	3	1.78
Localized edema	1	0.59
Lung infection	1	0.59
Neutrophil count decreased	10	5.92
Platelet count decreased	13	7.69
Pleural effusion	1	0.59
Pneumothorax	3	1.78
Proteinuria	24	14.2
Renal and urinary disorders—other	1	0.59
Respiratory, thoracic, and mediastinal	3	1.78
Surgical and medical procedures—other	4	2.37
Thrombotic thrombocytopenic purpura	8	4.73
Tooth infection	1	0.59
Vascular disorders—other, specify	1	0.59
Weight loss	3	1.78
Wound complication	5	2.96
Total	169	100

TABLE A5.

Detailed Comparison of L858R and Exon19 Deletion Subgroup Characteristics and Clinical Outcomes

Characteristic	L858R	Exon 19 Deletion	P	Total
Treatment, No. (%)
Arm A	29 (67.4)	64 (66.7)	.928	93 (66.9)
Arm B	14 (32.6)	32 (33.3)		46 (33.1)
Sex, No. (%)
Male	11 (25.6)	29 (30.2)	.578	40 (28.8)
Female	32 (74.4)	67 (69.8)		99 (71.2)
Age, years, No. (%)
<50	3 (7)	13 (13.5)	.145	16 (11.5)
50-59	13 (30.2)	20 (20.8)		33 (23.7)
60-69	16 (37.2)	37 (38.5)		53 (38.1)
70-79	7 (16.3)	24 (25)		31 (22.3)
≥80	4 (9.3)	2 (2.1)		6 (4.3)
Race, No. (%)
White	27 (62.8)	59 (61.5)	.924	86 (61.9)
Asian	10 (23.3)	24 (25)		34 (24.5)
Black	1 (2.3)	4 (4.2)		5 (3.6)
Other	5 (11.6)	9 (9.4)		14 (10.1)
CNS, No. (%)
No brain metastases	25 (58.1)	50 (52.1)	.508	75 (54.0)
Brain metastases	18 (41.9)	46 (47.9)		64 (46.0)
Smoking status, No. (%)
Never	33 (76.7)	60 (62.5)	.099	93 (66.9)
Current/former	10 (23.3)	36 (37.5)		46 (33.1)
Median PFS	22.8	19.0	.913
95% CI	17.9 to 37.1	15.6 to 35.9

TABLE A6.

CNS Metastasis Rates per Arm for All Randomly Assigned Patients and for Evaluable Patients

All Randomly Assigned Patients	Arm A, No. (%)	Arm B, No. (%)	Total, No. (%)
Exon 19 deletion and brain metastases	35 (33.3)	18 (33.3)	53 (33.3)
Exon 19 deletion and no brain metastases	38 (36.2)	19 (35.2)	57 (35.8)
L858R and brain metastases	14 (13.3)	8 (14.8)	22 (13.8)
L858R and no brain metastases	18 (17.1)	9 (16.7)	27 (17.0)
Total	105 (100)	54 (100)	159 (100)
CNS
No	56 (53.3)	28 (51.8)	84 (52.8)
Yes	49 (46.7)	26 (48.2)	75 (47.2)
Total	105 (100)	54 (100)	159 (100)

Evaluable Randomly Assigned Patients	Arm A, No. (%)	Arm B, No. (%)	Total, No. (%)
Exon 19 deletion and brain metastases	29 (31.2)	17 (37.0)	46 (33.1)
Exon 19 deletion and no brain metastases	35 (37.6)	15 (32.6)	50 (36.0)
L858R and brain metastases	11 (11.8)	7 (15.2)	18 (13.0)
L858R and no brain metastases	18 (19.4)	7 (15.2)	25 (18.0)
Total	93 (100)	46 (100)	159 (100)
CNS
No	53 (57.0)	22 (47.8)	75 (54.0)
Yes	40 (43.0)	24 (52.2)	64 (46.0)
Total	93 (100)	46 (100)	139 (100)

DATA SHARING STATEMENT

A data sharing statement provided by the authors is available with this article at DOI https://doi.org/10.1200/JCO.24.00533. The authors certify that this manuscript reports original clinical trial data. Data reported in this manuscript are available within the article or posted publicly at www.clinicaltrials.gov, according to the required timelines. Additional data from the study are available upon reasonable request.

AUTHOR CONTRIBUTIONS

Conception and design: Xiuning Le, Jyoti D. Patel, Catherine A. Shu, Mike Hernandez, Jhanelle E. Gray, John V. Heymach

Administrative support: Andreas Saltos, John V. Heymach

Provision of study materials or patients: Xiuning Le, Elaine Shum, Catherine A. Shu, Mary Jo Fidler, Richard Hall, Don Gibbons, Nisha A. Mohindra, Joshua Sabari, Rafael Santana-Davila, Ben Creelan, Bruna Pellini, Tawee Tanvetyanon, Jhanelle E. Gray, Andreas Saltos, John V. Heymach

Collection and assembly of data: Xiuning Le, Jyoti D. Patel, Elaine Shum, Christina Baik, Rachel E. Sanborn, Catherine A. Shu, Chul Kim, Mary Jo Fidler, Yasir Y. Elamin, Janet Tu, George Blumenschein, Don Gibbons, Carl Gay, Young Chae, Yanis Boumber, Joshua Sabari, Rafael Santana-Davila, Shane Rogosin, Benjamin Herzberg, Ben Creelan, Tawee Tanvetyanon, Mike Hernandez, Andreas Saltos, John V. Heymach

Data analysis and interpretation: Xiuning Le, Jyoti D. Patel, Christina Baik, Catherine A. Shu, Chul Kim, Mary Jo Fidler, Richard Hall, Yasir Y. Elamin, George Blumenschein, Jianjun Zhang, Nisha A. Mohindra, Joshua Sabari, Rafael Santana-Davila, Benjamin Herzberg, Ben Creelan, Bruna Pellini, Simon Heeke, Mike Hernandez, Jhanelle E. Gray, Andreas Saltos, John V. Heymach

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

A Multicenter Open-Label Randomized Phase II Study of Osimertinib With and Without Ramucirumab in Tyrosine Kinase Inhibitor–Naïve EGFR-Mutant Metastatic Non–Small Cell Lung Cancer (RAMOSE trial)

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).