Abstract
Background
Mixed response (MR), a scenario featuring discordant tumor changes, has been reported primarily with targeted therapies or immunotherapy. We determined the incidence and prognostic significance of MR in advanced non–small cell lung cancer (NSCLC) treated with cytotoxic chemotherapy.
Patients and Methods
We analyzed patient-level data from ECOG-ACRIN E5508 (carboplatin-paclitaxel + bevacizumab induction followed by randomization to maintenance therapy regimens). For patients with at least 2 target lesions and available measurements after cycle 2, we characterized response as homogeneous response (HR, similar behavior of all lesions), MR (similar behavior but >30% difference in magnitude of best and least responding lesions), or true mixed response (TMR, best and least responding lesions showing different behavior: ≥10% growth versus ≥10% shrinkage). We compared category characteristics using Mann-Whitney U and Chi-square tests, and overall survival (OS) using log-rank test and Cox models.
Results
Among 965 evaluable patients, HR occurred in 609 patients (63%), MR in 208 (22%), and TMR in 148 (15%). Median OS was 13.6 months for HR, 12.0 months for MR, and 7.6 months for TMR (P < .001). Compared to HR, TMR had inferior OS among stable disease cases (HR 1.62; 95% CI, 1.23-2.12; P < .001) and a trend toward inferior OS among progressive disease cases (HR 1.39; 95% CI, 0.83-2.33; P = .2). In multivariate analysis, TMR was associated with worse OS (HR 1.48; 95% CI, 1.22-1.79; P < .001).
Conclusion
True mixed response occurs in a substantial minority of lung cancer cases treated with chemotherapy and independently confers poor prognosis.
Keywords: chemotherapy, lung cancer, mixed response, multivariate analysis, survival
Mixed response (MR), a scenario featuring discordant tumor changes, has been reported primarily with targeted therapies or immunotherapy. This article focuses on the incidence and prognostic significance of MR in advanced non–small cell lung cancer treated with cytotoxic chemotherapy.
Implications for Practice.
This study demonstrates that mixed response (MR) occurs in a substantial minority of lung cancer cases treated with chemotherapy and independently confers poor prognosis. The clinical implications of these observations are substantial and include considerations for clinical and radiographic disease monitoring, as well as second-line therapy. These findings also raise numerous questions for future research, including molecular biologic mechanisms underlying MR as well as the optimal therapeutic approach to this clinical scenario.
Introduction
Unusual patterns of radiological response have received increased attention in recent years. In advanced non–small cell lung cancer (NSCLC) harboring genomic driver alterations, single-site or oligo-progression may represent heterogeneous molecular evolution. In such cases, local therapy with stereotactic radiation therapy or surgical resection, with continuation of prior systemic therapy, may result in prolonged disease control.1,2 In advanced NSCLC and other malignancies treated with immune checkpoint inhibitors, atypical responses such as pseudoprogression occur in a meaningful minority of cases.3-5 This phenomenon, attributed to an early influx of antitumor immune cells, refers to the transient enlargement of tumor sites, followed by subsequent shrinkage. Compared to other systemic therapies, immune checkpoint inhibitors have also been associated with increased rates of hyperprogression, a term most commonly defined as a doubling or more of the pretreatment tumor growth rate.6,7
In the era of molecularly targeted therapy and immunotherapy, clinicians have also reported patterns of mixed response (MR).8 Some studies have identified a positive prognostic effect9; others, detrimental.10,11 Potentially confounding and limiting the interpretation of these reports, some studies describe small, single-center cohorts.12,13 Some include a variety of cancer therapies,10 while others employ atypical imaging modalities, such as serial positron emission tomography-computed tomography (PET-CT).14 Most obtain tumor dimensions and assign response categories retrospectively. Additionally, the MR literature includes several definitions of MR, thereby rendering cross-study comparisons difficult.
Given the growing clinical interest in but limited data informing the approach to MR in lung cancer, we analyzed heterogeneous radiographic response patterns in a large, multicenter phase III trial of chemotherapy-based treatment for advanced NSCLC.
Methods
Data Source and Collection
For this analysis, we analyzed patient-level data from the Eastern Cooperative Oncology Group-American College of Radiology (ECOG-ACRIN) E5508 clinical trial.15 This trial (NCT01107626) enrolled patients with previously untreated advanced nonsquamous NSCLC not known to harbor activating EGFR or ALK alterations with performance status ECOG 0-1. Patients received up to 4 cycles of induction carboplatin-paclitaxel + bevacizumab treatment at standard dose and schedule (carboplatin AUC 6, paclitaxel 200 mg/m2, bevacizumab 15 mg/kg every 3 weeks). Baseline CT scans were performed within 28 days of trial registration. CT scans were then performed every 2 cycles during this phase to assess treatment efficacy by Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Patients tolerating treatment and experiencing clinical benefit (stable or responding disease) after 4 cycles were randomized to one of 3 maintenance arms: (A) bevacizumab, (B) pemetrexed, and (C) bevacizumab plus pemetrexed. Treatment was continued until disease progression or intolerable toxicity, with CT scans performed every 3 cycles during maintenance therapy. We selected this trial for the present study due to its large size, use of contemporary RECIST parameters, and availability of long-term follow-up data.
A total of 1516 patients were enrolled in E5508. Clinical data, including RECIST measurements, were entered into a Medidata Rave database. RECIST measurements were performed at the treating site without central input or confirmation. For the primary endpoint of overall survival (OS), there was no significant difference among the 3 maintenance treatment arms.
For each patient, we prospectively collected tumor measurements of the target non-nodal and target nodal lesions at baseline and after cycle 2. The analysis included patients who had at least 2 lesions for which measurements were available at baseline and after cycle 2. The increase or decrease for each individual lesion was calculated in percentages. We also collected demographic, tumor, treatment, and OS data.
The maximum difference in response between individual lesions was determined as a continuous variable. As described previously,11 based on the observed differences between the best-responding lesion and the least-responding lesion, we classified radiographic response heterogeneity on a patient level as follows: MR: >30% difference in the best-responding lesion and the least-responding lesion, with all lesions demonstrating similar behavior (either responding or progressing). True mixed response (TMR): the best-responding lesion and the least-responding lesion showing different behavior (≥10% growth versus ≥10% shrinkage). Homogeneous response (HR): similar behavior of all lesions. To arrive at these thresholds and definitions, the authors tested various cutoff values for each category. The model showing the best distribution of patients, the most significant differences in OS curves, and the largest absolute difference in median OS (mos) was deemed most discriminative. Nontarget lesions were not factored into these categories (although they were incorporated into RECIST categories).
We selected this approach—previously evaluated in patients with metastatic colorectal cancer—because it (a) can be performed using patient-level clinical trial data, (b) does not require rereview of radiographic images, (c) relies on standard radiographic assessment (eg, CT), and (d) had been developed using one of the largest cohorts (N = 290) ever evaluated for MR.11 By contrast, other published approaches to MR either examined only longitudinal dynamics of MR cases without comparison to other response categories,8 incorporated PET imaging,10 considered only cases with RECIST progressive disease (PD),9 or limited the analysis to lesions in different organs.16
Using RECIST version 1.1 criteria, the best overall radiographic response was categorized as complete response (CR), partial response (PR), stable disease (SD), and PD.
Statistical Analysis
Baseline characteristics were compared among the HR, MR, and TMR groups using Mann-Whitney U tests for continuous data and Pearson’s Chi-square for categorical data. We calculated the incidence of HR, MR, and TMR. We estimated the median OS for each response category using the Kaplan-Meier method and used the log-rank test to compare OS curves among the 3 response groups. We performed univariate and multivariate analyses using a Cox proportional hazards model. Variables with P < .05 in univariate analyses were selected for multivariate analyses.
Results
In total, 965 patients had 2 or more RECIST target lesions with baseline and post-cycle 2 measurements and were therefore considered for analysis (Fig. 1). Mean age was 64 years, and 52% were women. Additional case characteristics, overall and according to response pattern, are shown in Table 1. HR occurred in 609 patients (63%), MR in 208 patients (22%), and TMR in 148 patients (15%). Response heterogeneity pattern was significantly associated with stage, mediastinal nodal metastases, maintenance treatment arm, and RECIST response. Specifically, patients with stage IV M1b, mediastinal nodal involvement, and PR were more likely to have MR. Patients with TMR were less likely to receive maintenance therapy than were patients with HR or MR.
Figure 1.
CONSORT diagram for survival and RECIST analyses.
Table 1.
Baseline characteristics according to response heterogeneity category.
| Mean ± SD or number (%) | P-value | |||||
|---|---|---|---|---|---|---|
| HR | MR | TMR | Total | |||
| (N = 609) | (N = 208) | (N = 148) | (N = 965) | |||
| Age | Mean (SD) | 63.7 ± 9.4 | 62.9 ± 9.0 | 63.3 ± 8.8 | 63.5 ± 9.2 | .47 |
| Gender | Male | 313 (51) | 110 (53) | 83 (56) | 506 (52) | .59 |
| Female | 296 (49) | 98 (47) | 65 (44) | 459 (48) | ||
| RECIST best response | CR | 3 (0) | 2 (1) | 0 (0) | 5 (1) | <.001 |
| PR | 221 (36) | 115 (55) | 23 (16) | 359 (37) | ||
| SD | 182 (30) | 21 (10) | 63 (43) | 266 (28) | ||
| PD | 46 (8) | 13 (6) | 27 (18) | 86 (9) | ||
| Unevaluable/Unknown | 157 (26) | 57 (27) | 35 (24) | 249 (26) | ||
| ECOG | 0 | 254 (42) | 76 (37) | 57 (39) | 387 (40) | .39 |
| 1 | 355 (58) | 132 (63) | 91 (61) | 578 (60) | ||
| Stage | IIIB | 18 (3) | 4 (2) | 0 (0) | 22 (2) | .04 |
| IV M1a | 166 (27) | 36 (17) | 33 (22) | 235 (24) | ||
| IV M1b | 398 (65) | 155 (75) | 110 (74) | 663 (69) | ||
| Recurrent | 26 (4) | 13 (6) | 5 (3) | 44 (5) | ||
| Unknown | 1 (0) | 0 (0) | 0 (0) | 1 (0) | ||
| Smoking history | Current | 268 (44) | 101 (49) | 79 (53) | 448 (46) | .32 |
| Former | 288 (47) | 90 (43) | 58 (39) | 436 (45) | ||
| Never | 53 (9) | 17 (8) | 11 (7) | 81 (8) | ||
| Maintenance therapy arm | Bevacizumab | 135 (22) | 45 (22) | 34 (23) | 214 (22) | <.001 |
| Pemetrexed | 149 (24) | 55 (26) | 16 (11) | 220 (23) | ||
| Bevacizumab + pemetrexed |
136 (22) | 50 (24) | 26 (18) | 212 (22) | ||
| None | 189 (31) | 58 (28) | 72 (49) | 319 (33) | ||
| Histology | Other | 60 (10) | 24 (12) | 15(10) | 99 (10) | .77 |
| Adenocarcinoma | 549 (90) | 184 (88) | 133 (90) | 866 (90) | ||
| Hilar nodal metastases | Absent | 254 (42) | 82 (39) | 64 (43) | 400 (41) | .96 |
| Present | 342 (56) | 121 (58) | 81 (55) | 544 (56) | ||
| Unknown | 13 (2) | 5 (2) | 3 (2) | 21 (2) | ||
| Mediastinal nodal metastases | Absent | 211 (35) | 53 (25) | 58 (39) | 322 (33) | .01 |
| Present | 390 (64) | 153 (74) | 85 (57) | 628 (65) | ||
| Unknown | 8 (1) | 2 (1) | 5 (3) | 15 (2) | ||
| Brain metastases | Absent | 527 (87) | 175 (84) | 121 (82) | 823 (85) | .61 |
| Present | 68 (11) | 26 (12) | 22 (15) | 116 (12) | ||
| Unknown | 14 (2) | 7 (3) | 5 (3) | 26 (3) | ||
Abbreviations: CR, complete response; ECOG, Eastern Cooperative Oncology Group; HR, homogeneous response; MR, mixed response; PR, partial response; PD, progressive disease; SD, stable disease; TMR, true mixed response.
Overall survival according to response heterogeneity category is shown in Fig. 2. Four patients with missing survival time data were excluded from the analysis. Median OS for the entire cohort was 12.0 (95% CI, 11.3-13.0) months. By category, median OS was 13.6 months (95% CI, 12.3-15.5 months) for HR, 12.0 months (95% CI, 10.7-15.2 months) for MR, and 7.6 months (95% CI, 5.8-9.2 months) for TMR.
Figure 2.
Overall survival according to response heterogeneity category. Abbreviations: HR, homogeneous response; MR, mixed response; TMR, true mixed response.
Overall survival according to both response heterogeneity and RECIST response is shown in Table 2 and Fig. 3. This analysis included 759 patients with available response heterogeneity, post-cycle 2 RECIST response, and OS data. We observed certain differences in RECIST response according to response heterogeneity. For instance, RECIST PD was more common in patients with TMR (23%) than among patients with MR (9%) or HR (8%). Conversely, RECIST PR occurred more frequently among patients with MR (69%) than in patients with HR (38%) and TMR (7%). Within each RECIST response category (apart from CR, of which there were only 2 cases), median OS was numerically worse for MR and TMR cases than for HR cases. In the RECIST SD category, survival was significantly shorter for TMR than for HR cases: HR 1.62 (95% CI, 1.23-2.12; P < .001). In the RECIST PD category, this difference did not reach statistical significance: HR 1.39 (95% CI, 0.83-2.33; P = .2).
Table 2.
Overall survival according to response heterogeneity and RECIST response category.
| Response heterogeneity | RECIST response | N | Median OS (95% CI) |
|---|---|---|---|
| HR | PR | 175 | 16.1 (13.8-19.2) |
| SD | 253 | 14.7 (12.7-17.6) | |
| PD | 38 | 3.7 (2.5-5.7) | |
| MR | PR | 126 | 12.9 (11.5-16.7) |
| SD | 40 | 11.2 (8.2-18.9) | |
| PD | 16 | 3.4 (2.6-7.3) | |
| TMR | PR | 8 | 13.3 (7.2-NA) |
| SD | 77 | 8.2 (6.6-11.9) | |
| PD | 26 | 2.8 (1.8-4.6) |
Abbreviations: CI, confidence interval; CR, complete response; HR, homogeneous response; MR, mixed response; OS, overall survival; PR, partial response; PD, progressive disease; SD, stable disease; TMR, true mixed response.
Figure 3.
Overall survival according to response heterogeneity and RECIST response categories. (a) RECIST PR, (b) RECIST SD, (c) RECIST PD. Abbreviations: HR, homogeneous response; MR, mixed response; PD, progressive disease; PR, partial response; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease; TMR, true mixed response.
In both univariate and multivariate analysis, male gender, M1b staging, ECOG 1 (versus 0), nonadenocarcinoma histology, mediastinal nodal metastasis, no administration of maintenance therapy, and TMR were significantly associated with worse OS (Table 3). There was no significant difference in OS between HR and MR.
Table 3.
Univariate and multivariate analyses for OS.
| Characteristic | Univariate analysis | Multivariate analysis | ||
|---|---|---|---|---|
| HR (95% CI) | P-value | HR (95% CI) | P-value | |
| Gender | ||||
| Male | Reference | Reference | ||
| Female | 0.74 (0.65-0.85) | <.001 | 0.78 (0.68-0.90) | .001 |
| Smoking history | ||||
| Current | Reference | |||
| Former | 0.90 (0.78-1.03) | .13 | ||
| Never | 0.79 (0.61-1.02) | .08 | ||
| Stage | ||||
| IIIB | Reference | Reference | ||
| IV M1a | 1.44 (0.85-2.44) | .18 | 1.70 (1.00-2.90) | .05 |
| IV M1b | 2.21 (1.32-3.70) | .003 | 2.34 (1.39-3.94) | .001 |
| Recurrent | 1.69 (0.92-3.09) | .09 | 1.62 (0.88-2.98) | .12 |
| Unknown | 7.48 (0.98-56.91) | .05 | 7.20 (0.91-57.31) | .06 |
| ECOG | ||||
| 0 | Reference | Reference | ||
| 1 | 1.45 (1.3-1.7) | <.001 | 1.38 (1.20-1.59) | <.001 |
| Histology | ||||
| Other | Reference | Reference | ||
| Adenocarcinoma | 0.738 (0.59-0.92) | .007 | 0.74 (0.59-0.93) | .009 |
| Hilar nodal metastases | ||||
| Absent | Reference | |||
| Present | 1.09 (0.97-1.2) | .17 | ||
| Mediastinal nodal metastases | ||||
| Absent | Reference | Reference | ||
| Present | 1.14 (1-1.3) | .04 | 1.15 (1.00-1.31) | .05 |
| Brain metastases | ||||
| Absent | Reference | |||
| Present | 1.04 (0.88-1.2) | .66 | ||
| Maintenance therapy arm | ||||
| Bevacizumab | Reference | Reference | ||
| Pemetrexed | 0.87 (0.71-1.07) | .17 | 0.85 (0.69-1.04) | .12 |
| Bevacizumab + pemetrexed | 0.89 (0.72-1.09) | .26 | 0.89 (0.73-1.10) | .30 |
| None | 2.36 (1.96-2.84) | <.001 | 2.30 (1.90-2.77) | <.001 |
| Response heterogeneity category | ||||
| HR | Reference | Reference | ||
| MR | 0.99 (0.83-1.18) | .91 | 0.91 (0.76-1.09) | .3 |
| TMR | 1.71 (1.41-2.06) | <.001 | 1.48 (1.22-1.79) | <.001 |
We also specifically examined cases featuring the emergence of new nontarget lesions. Although not considered in MR definitions, this scenario is considered disease progression by RECIST. Overall, 76 cases (8%) had new lesions documented. Among these, 41 cases (54%) were HR, 13 cases (17%) were MR, and 22 cases (29%) were TMR—approximately twice the frequency of TMR cases in the overall study population (15%).
Discussion
Given the heightened awareness of and attention to MR in patients treated with targeted therapies and immune checkpoint inhibitors, it is critical to place this clinical and radiographic phenomenon in context. We therefore conducted, to our knowledge, the largest study to date of MR in lung cancer or any malignancy. In the present analysis, we analyzed almost 1000 patients with advanced nonsquamous NSCLC treated initially with carboplatin-paclitaxel plus bevacizumab therapy. We observed that MR (target lesions with similar behavior but > 30% difference in individual lesion responses) occurred in over 20 percent of cases. True mixed response (ie, different behavior [growth and shrinkage] of target lesions) occurred in 15% of patients overall and in almost 30% of cases featuring the appearance of new (nontarget) lesions. In the overall population study population, these individuals had inferior OS. Significantly worse outcomes were also seen among the subset of cases with RECIST SD, with a clear but nonsignificant trend noted for the subset of cases with RECIST disease progression. In multivariate analysis controlling for numerous patient and tumor characteristics, TMR was associated with a nearly 50% increase in the rate of death compared to HR and MR cases.
The incidence of MR and TMR in this chemotherapy-treated population is noteworthy. Mixed or dissociated responses have largely been considered phenomena restricted to patients treated either with immune checkpoint inhibitors or with molecularly targeted therapies. Indeed, the 15% rate of TMR in the present study is approximately twice the rate of dissociated response observed in patients with NSCLC receiving single-agent anti-PD1/PDL1 therapy (8%).5 We recognize the limitations of comparing our current findings with studies using different definitions of response heterogeneity, as the case for this immune checkpoint inhibitor-treated population. Nevertheless, one might surmise that the actual difference is even greater. The “dissociated response” used in the immunotherapy study requires only a different directionality of tumor behavior (growth vs. shrinkage). However, the “true mixed response” used in the present analysis requires a minimum change (≥10%) in each direction. Notably, our reported rate of TMR also numerically exceeds that in a chemotherapy-treated colorectal cancer population using the same parameter (9%).11
The underlying mechanism of MRs in the current study of cytotoxic therapy remains unclear. Previously reported series in lung cancer have identified discordant molecular evolution of classical EGFR mutations (eg, EGFR T790M, wild-type EGFR, and MET alterations) during the process of metastasis and among cases with MRs,10,17 but such genomic considerations seem unlikely to apply to the present study’s population of EGFR wild-type NSCLC treated with cytotoxic and antiangiogenic agents. Mixed responses to immunotherapy have been attributed to tissue-dependent tumor microenvironments.18 Among surgically resected localized lung adenocarcinomas, those with greater intratumoral heterogeneity (assessed by multiregion whole exome sequencing) were more likely to recur,19 consistent with a worse prognosis independent of medical therapy. However, the association between intratumoral heterogeneity—a metric based on a single tumor—and MR among 2 or more lesions is not known.
The adverse prognostic impact of MR, particularly TMR, is striking. In a study of NSCLC cases treated with various targeted and cytotoxic therapies, MR characterized by PET was associated with worse outcomes.10 What is particularly striking about our current findings is that MR, specifically TMR, tended to have inferior survival across RECIST response categories. One might expect this observation among cases with RECIST PR; an enlarging lesion among others that are clearly shrinking or unchanged might indicate near-term emergence of broader therapeutic resistance. But the worse outcomes of TMR cases within the RECIST SD and PD categories is a more complex consideration. For instance, why a progressing cancer with some shrinking lesions would have inferior survival to a progressing cancer featuring uniformly enlarging lesions is not known. One potential explanation is that, due to biologic heterogeneity, cancers demonstrating MR are difficult to treat with any medical therapy, including subsequent lines of treatment. This could be particularly true when the initial MR occurs in the setting of a multiagent regimen—in this study, a platinum, taxane, and antiangiogenic agent. Nonmedical, local treatment modalities such as radiation therapy or surgical resection, either in the setting of oligoprogression or persistence at the time of expected best response,20-23 may help address these challenges.
Strengths of this study include the large sample size, the prospective collection of RECIST tumor dimensions, the uniform nature and timing of serial imaging studies, the duration of follow-up, and the availability of detailed clinical data such as sites of metastases. Main weaknesses include lack of tumor genomic characterization (beyond the absence of activating EGFR and ALK alterations), inclusion of only cytotoxic and antiangiogenic therapies, lack of information on postprogression therapy, and exclusion of squamous tumors. Although patients in the E5508 trial we analyzed were assigned to one of 3 maintenance therapy regimens, there was no difference in survival among these treatments; accordingly, maintenance therapy assignment seems unlikely to influence our findings. Because our classification of response heterogeneity incorporated only target lesion data, it is also possible that we underestimated the true incidence of MR, as some cases might have shrinking target lesions concurrent with the appearance of new or unequivocal growth of nontarget lesions. Finally, we recognize that the present analysis only applies to cases with multiple RECIST target lesions, a population that might be expected to have worse outcomes than cases with only a single lesion based on the association between radiographic tumor burden and survival in advanced NSCLC.24
We also acknowledge that our approach to define MR, while based on previously published methodology,11 may not represent the optimal characterization of this phenomenon. For instance, in cases of very small target lesions, differences of 10% in longest diameters would be expected to have relatively large margins of error—a shortcoming that might be addressed by mandating minimum absolute changes in addition to proportional changes, as required by RECIST. It is also possible that different category definitions (eg, discordant changes other than ±10%, or differences in concordant changes other than 30%) may have had greater discriminative capacity in this study population. The focus of the present study, however, was not the development of a new metric. Instead, we sought to apply a previously established method to a new clinical scenario.
Conclusion
Although divergent radiographic responses have gained attention primarily among tumors treated with molecularly targeted therapies or immune checkpoint inhibitors, they also occur in a substantial proportion of cancers treated with conventional, cytotoxic chemotherapy. True mixed response—a scenario in which some tumor lesions shrink while others grow—is independently associated with poor prognosis. Within RECIST categories, it has worse outcomes than HRs. These findings highlight the clinical importance of divergent radiographic response patterns across treatment types. Further research into the biologic underpinnings and optimal treatment of MR is warranted.
Acknowledgments
The E5508 study was coordinated by the ECOG-ACRIN Cancer Research Group (Peter J. O’Dwyer, MD, and Mitchell D. Schnall, MD, PhD, Group Co-Chairs) and supported by the National Cancer Institute (NCI) of the National Institutes of Health under the following award numbers: U10CA180820, U10CA180794, UG1CA233196, UG1CA233247, and UG1CA233302. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors thank Ms. Dru Gray for assistance with manuscript preparation.
Contributor Information
David E Gerber, Harold C. Simmons Comprehensive Cancer Center, UT Southwestern Medical Center, Dallas, TX, USA; Department of Internal Medicine, UT Southwestern Medical Center, Dallas, TX, USA; Peter O’Donnell Jr. School of Public Health, UT Southwestern Medical Center, Dallas, TX, USA.
Yating Wang, Eastern Cooperative Oncology Group Statistical Center, Boston, MA, USA; Dana-Farber Cancer Institute, Boston, MA, USA.
Suresh S Ramalingam, Winship Cancer Institute, Emory University, Atlanta, GA, USA.
Sheena Bhalla, Harold C. Simmons Comprehensive Cancer Center, UT Southwestern Medical Center, Dallas, TX, USA; Department of Internal Medicine, UT Southwestern Medical Center, Dallas, TX, USA.
Zhuoxin Sun, Eastern Cooperative Oncology Group Statistical Center, Boston, MA, USA; Dana-Farber Cancer Institute, Boston, MA, USA.
Hossein Borghaei, Fox Chase Cancer Center, Philadelphia, PA, USA.
Julie R Brahmer, University of Virginia, Charlottesville, VA, USA.
Joan H Schiller, University of Virginia, Charlottesville, VA, USA.
Conflict of Interest
David E. Gerber reported research funding (institutional) from Astra-Zeneca, BerGenBio, Karyopharm, and Novocure; stock ownership in Gilead, Medtronic, and Walgreens; consulting/advisory boards for Astra-Zeneca, Catalyst Pharmaceuticals, Daiichi-Sankyo, Elevation Oncology, Janssen Scientific Affairs, LLC, Jazz Pharmaceuticals, Regeneron Pharmaceuticals, and Sanofi; and serving as co-founder and Chief Scientific Officer, OncoSeer Diagnostics, Inc. Suresh S. Ramalingam reported research funding (institutional) from AstraZeneca, Amgen, Merck, Bristol Myers Squibb, and Pfizer. Sheena Bhalla has served in consulting/advisory roles for Takeda, Mirati, and Merus. Hossein Borghaei reported research support (clinical trials) from BMS, Lilly, and Amgen; advisory board/consultanting for BMS, Lilly, Genentech, Pfizer, Merck, EMD-Serono, Boehringer Ingelheim, AstraZeneca, Novartis, Genmab, Regeneron, BioNTech, Amgen, Axiom, PharmaMar, Takeda, Mirati, Daiichi, Guardant, Natera, Oncocyte, Beigene, iTEO, Jazz, Janssen, Da Volterra, Puma, BerGenBio, Bayer, and Iobiotech; Data and Safety Monitoring Board for University of Pennsylvania: CAR T Program, Takeda, Incyte, Novartis, and Springworks; employment with Fox Chase Cancer Center; Scientific Advisory Board: Sonnetbio (Stock Options), Inspirna (formerly Rgenix, Stock Options), and Nucleai (stock options); honoraria from Amgen, Pfizer, Daiichi, and Regeneron; and travel expenses from Amgen, BMS, Merck, Lilly, EMD-Serono, Genentech, and Regeneron. Julie R. Brahmer reported consulting/advisory relationship: Amgen, AstraZeneca, BMS, Eli Lilly, Genentech/Roche, GlaxoSmithKline, Merck, Sanofi, Janssen/Johnson and Johnson, Regeneron, Incyte; Research Funding: BMS, AstraZeneca, and RAPT Therapeutics. Joan H. Schiller is on the Data Monitoring Board of the following pharmaceutical organizations: Merck, AstraZeneca, and Roche and consults for Noetic. The other authors indicated no financial relationships.
Author Contributions
Conception/design: D.E.G. Provision of study material or patients: D.E.G., S.S.R., H.B., J.R.B., J.H.S. Collection and/or assembly of data: Y.W., S.S.R., Z.S., J.H.S. Data analysis and interpretation: D.E.G., Y.W., S.B., Z.S. Manuscript writing: D.E.G. Final approval of manuscript: All authors.
Data Availability
The data underlying this article were provided by ECOG-ACRIN under licence/by permission. Data will be shared on request to the corresponding author with permission of ECOG-ACRIN.
References
- 1. Gan GN, Weickhardt AJ, Scheier B, et al. Stereotactic radiation therapy can safely and durably control sites of extra-central nervous system oligoprogressive disease in anaplastic lymphoma kinase-positive lung cancer patients receiving crizotinib. Int J Radiat Oncol Biol Phys. 2014;88(4):892-898. 10.1016/j.ijrobp.2013.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Weickhardt AJ, Scheier B, Burke JM, et al. Local ablative therapy of oligoprogressive disease prolongs disease control by tyrosine kinase inhibitors in oncogene-addicted non-small-cell lung cancer. J Thorac Oncol. 2012;7(12):1807-1814. 10.1097/JTO.0b013e3182745948 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Gonugunta AS, von Itzstein MS, Gerber DE.. Pseudoprogression in advanced non-small cell lung cancer treated with combination chemoimmunotherapy: a case report. J Med Case Rep. 2022;16(1):289. 10.1186/s13256-022-03485-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Jia W, Gao Q, Han A, Zhu H, Yu J.. The potential mechanism, recognition and clinical significance of tumor pseudoprogression after immunotherapy. Cancer Biol Med. 2019;16(4):655-670. 10.20892/j.issn.2095-3941.2019.0144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Tazdait M, Mezquita L, Lahmar J, et al. Patterns of responses in metastatic NSCLC during PD-1 or PDL-1 inhibitor therapy: comparison of RECIST 11, irRECIST and iRECIST criteria. Eur J Cancer. 2018;88:38-47. 10.1016/j.ejca.2017.10.017 [DOI] [PubMed] [Google Scholar]
- 6. Popat V, Gerber DE.. Hyperprogressive disease: a distinct effect of immunotherapy? J Thorac Dis. 2019;11(Suppl 3):S262-S265. 10.21037/jtd.2019.01.97 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Borcoman E, Kanjanapan Y, Champiat S, et al. Novel patterns of response under immunotherapy. Ann Oncol. 2019;30(3):385-396. 10.1093/annonc/mdz003 [DOI] [PubMed] [Google Scholar]
- 8. Rauwerdink DJW, Molina G, Frederick DT, et al. Mixed response to immunotherapy in patients with metastatic melanoma. Ann Surg Oncol. 2020;27(9):3488-3497. 10.1245/s10434-020-08657-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Tozuka T, Kitazono S, Sakamoto H, et al. Dissociated responses at initial computed tomography evaluation is a good prognostic factor in non-small cell lung cancer patients treated with anti-programmed cell death-1/ligand 1 inhibitors. BMC Cancer. 2020;20(1):207. 10.1186/s12885-020-6704-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Dong ZY, Zhai HR, Hou QY, et al. Mixed responses to systemic therapy revealed potential genetic heterogeneity and poor survival in patients with non-small cell lung cancer. Oncologist. 2017;22(1):61-69. 10.1634/theoncologist.2016-0150 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. van Kessel CS, Samim M, Koopman M, et al. Radiological heterogeneity in response to chemotherapy is associated with poor survival in patients with colorectal liver metastases. Eur J Cancer. 2013;49(11):2486-2493. 10.1016/j.ejca.2013.03.027 [DOI] [PubMed] [Google Scholar]
- 12. Park KY, Jung JW, Nam SB, et al. Two lung masses with different responses to pemetrexed. Korean J Intern Med. 2010;25(2):213-216. 10.3904/kjim.2010.25.2.213 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ryoo BY, Na IIYang SH, Koh JS, Kim CH, Lee JC.. Synchronous multiple primary lung cancers with different response to gefitinib. Lung Cancer. 2006;53(2):245-248. [DOI] [PubMed] [Google Scholar]
- 14. Humbert O, Chardin D.. Dissociated response in metastatic cancer: an atypical pattern brought into the spotlight with immunotherapy. Front Oncol. 2020;10:566297. 10.3389/fonc.2020.566297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Ramalingam SS, Dahlberg SE, Belani CP, et al. Pemetrexed, bevacizumab, or the combination as maintenance therapy for advanced nonsquamous non-small-cell lung cancer: ECOG-ACRIN 5508. J Clin Oncol. 2019;37(26):2360-2367. 10.1200/JCO.19.01006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Furubayashi N, Negishi T, Sakamoto N, et al. Clinical outcomes of mixed response to pembrolizumab in advanced urothelial carcinoma after platinum-based chemotherapy. In Vivo. 2021;35(5):2869-2874. 10.21873/invivo.12575 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Park S, Holmes-Tisch AJ, Cho EY, et al. Discordance of molecular biomarkers associated with epidermal growth factor receptor pathway between primary tumors and lymph node metastasis in non-small cell lung cancer. J Thorac Oncol. 2009;4(7):809-815. 10.1097/JTO.0b013e3181a94af4 [DOI] [PubMed] [Google Scholar]
- 18. Oliver AJ, Lau PKH, Unsworth AS, et al. Tissue-dependent tumor microenvironments and their impact on immunotherapy responses. Front Immunol. 2018;9:70. 10.3389/fimmu.2018.00070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Zhang J, Fujimoto J, Zhang J, et al. Intratumor heterogeneity in localized lung adenocarcinomas delineated by multiregion sequencing. Science. 2014;346(6206):256-259. 10.1126/science.1256930 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Hannan R, Christensen M, Hammers H, et al. Phase II trial of stereotactic ablative radiation for oligoprogressive metastatic kidney cancer. Eur Urol Oncol. 2022;5(2):216-224. 10.1016/j.euo.2021.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Alomran R, White M, Bruce M, et al. Stereotactic radiotherapy for oligoprogressive ER-positive breast cancer (AVATAR). BMC Cancer. 2021;21(1):303. 10.1186/s12885-021-08042-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Iyengar P, Wardak Z, Gerber DE, et al. Consolidative radiotherapy for limited metastatic non-small-cell lung cancer: a phase 2 randomized clinical trial. JAMA Oncol. 2018;4(1):e173501. 10.1001/jamaoncol.2017.3501 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Iyengar P, Kavanagh BD, Wardak Z, et al. Phase II trial of stereotactic body radiation therapy combined with erlotinib for patients with limited but progressive metastatic non-small-cell lung cancer. J Clin Oncol. 2014;32(34):3824-3830. 10.1200/JCO.2014.56.7412 [DOI] [PubMed] [Google Scholar]
- 24. Gerber DE, Dahlberg SE, Sandler AB, et al. Baseline tumour measurements predict survival in advanced non-small cell lung cancer. Br J Cancer. 2013;109(6):1476-1481. 10.1038/bjc.2013.472 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data underlying this article were provided by ECOG-ACRIN under licence/by permission. Data will be shared on request to the corresponding author with permission of ECOG-ACRIN.