Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Aug 1.
Published in final edited form as: Ann Oncol. 2022 May 6;33(8):824–835. doi: 10.1016/j.annonc.2022.04.450

Deciphering radiological stable disease to immune checkpoint inhibitors

J Luo 1,2,3,*, S Wu 4,*, H Rizvi 5, Q Zhang 4, JV Egger 5, JC Osorio 3, AJ Schoenfeld 1,7, AJ Plodkowski 8, MS Ginsberg 8, MK Callahan 7,9,10, C Maher 10, AN Shoushtari 7,10, MA Postow 7,10, MH Voss 7,11, RR Kotecha 7,11, A Gupta 6, R Raja 4, MG Kris 1,7, MD Hellmann 1,7,9,^
PMCID: PMC10001430  NIHMSID: NIHMS1861243  PMID: 35533926

Abstract

Background:

“Stable disease (SD)” per RECIST is a common but ambiguous outcome in patients receiving immune checkpoint inhibitors (ICIs). This study aimed to characterize SD and identify the subset of patients with SD who are benefiting from treatment. Understanding SD would facilitate drug development and improve precision in correlative research.

Patients and methods:

A systematic review was performed to characterize SD in ICI trials. SD and objective response was compared to proliferation index using TCGA gene expression data. To identify a subgroup of SD with outcomes mirroring responders, we examined a discovery cohort of NSCLC. Serial cutpoints of two variables, % best overall response (BOR) and PFS, were tested to define a subgroup of patients with SD with similar survival as responders. Results were then tested in external validation cohorts.

Results:

Among trials of ICIs (59 studies, 14,280 patients), SD ranged from 16–42% in different tumor types and was associated with disease-specific proliferation index (ρ=−0.75, p=0.03), a proxy of tumor kinetics, rather than relative response to ICIs. In a discovery cohort of NSCLC (1220 patients, 313 [26%] with SD to ICIs), PFS ranged widely in SD (0.2 to 49 months, median 4.9 months). The subset with PFS>6 months and no tumor growth mirrored PR-minor (OS HR 1.0) and was proposed as the definition of SD responder. This definition was confirmed in two validation cohorts from trials of NSCLC treated with durvalumab and found to apply in tumor types treated with immunotherapy in which depth and duration of benefit were correlated.

Conclusion:

RECIST-defined SD to immunotherapy is common, heterogenous, and may largely reflect tumor growth rate rather than ICI response. In patients with NSCLC and SD to ICIs, PFS>6 months and no tumor growth may be considered “SD responders.” This definition may improve the efficiency of and insight derivable from clinical and translational research.

Keywords: Immunotherapy/checkpoint blockade, RECIST, lung cancer

Introduction:

Advances in imaging technology and iterative refinements in defining benefit to treatment using reproducible measurements of response1–3 have led to the reliable criteria used in clinical trials today. Radiological response criteria (e.g. response evaluation criteria in solid tumors [RECIST]4) remains the gold standard for antitumor activity. RECIST has largely been able to distinguish response and progression in clinical trials of cytotoxic chemotherapies and targeted therapies. However, in patients receiving immunotherapy, there can be challenges to measuring response using RECIST.

RECIST stable disease (SD) as the best response is a common but imprecise category that limits efficient clinical and translational research. SD is defined as a change in tumor size that is neither radiological response (greater than 30% shrinkage) nor progression (greater than 20% growth). In this radiological intermediate setting, the clinical and biological interpretations are ambiguous. Stable disease is likely comprised of heterogenous patient subgroups reflecting at least three distinct disease scenarios: (1) real benefit from treatment that is radiologically imperceptible (2) no benefit and slow progression, and (3) indolent, non-progressive disease reflecting intrinsic disease kinetics rather than ongoing response to therapy. The frequent uncertainty when it comes to interpreting whether a patient with SD benefited from treatment is a bottleneck in efficient drug development and in deriving reliable conclusions from translational studies.

In clinical trials, the proportion of SD may reflect disease kinetics of the enrolled patient population rather than providing a signal of drug activity; however, it is nevertheless grouped with response in a commonly reported secondary efficacy endpoint (e.g. “disease control rate” [DCR, SD + PR + CR]) despite little contribution to informing continued drug development.

In translational studies, various biomarker reports of immunotherapies have alternatively categorized SD as non-response5, 6, response7, defined an arbitrary cutpoint in-between8–10, or excluded SD altogether due to the uncertainty of its meaning.6, 11 In short, there is general uncertainty and variability about how to interpret or handle SD as an outcome in clinical and translational research; tools to clarify whom with SD is truly benefiting from therapy are crucially needed.

Our group recently examined whether prospectively collected cell-free circulating tumor DNA (ct)DNA is predictive for patients with radiological SD at the first scan. We found ctDNA trajectory could distinguish long-term benefit among patients with SD12 and suggested that at least some fraction of patients with SD experienced clinically relevant benefit to treatment. However, serial ctDNA testing is not routinely performed in practice or in clinical trials and refining the definition of SD needs a broadly applicable solution.

In this report, we sought to bring attention to the common and complex issue of “stable disease” and develop a strategy to use existing radiological measures to define response in this scenario. We hypothesized that existing objective measures of response may be able to define a subgroup of “SD responders” (SD-R), the subset benefiting from therapy. To do so, we systematically tested definitions of SD-R using serial cutpoints of objective response and compared these groups to a subgroup of patients with partial response.

Methods:

Overview

We first sought to characterize the scope of SD in patients who receive ICIs by performing a systematic review of published trials. We then focused on patients with NSCLC with granular data on % best overall response (BOR) and progression-free survival (PFS) to identify a definition for patients with SD who benefited from therapy. We tested our definition in several other cohorts to examine the generalizability of this definition.

Patients

This retrospective study was approved by the Institutional Review Board at Memorial Sloan Kettering Cancer Center (MSK).

The discovery cohort represented consecutive patients with non-small cell lung cancer (NSCLC) seen at MSK who received programmed cell death protein (ligand) 1 [PD-(L)1] blockade with or without another immunotherapy (e.g. cytotoxic T-cell lymphocyte-associated antigen 4 [CTLA-4] blockade) between April 1, 2011 and December 31, 2017 and in whom RECIST v1.14 assessments were available. Database lock occurred on September 30, 2019.

RECIST reports performed by thoracic radiologists at MSK were reviewed for categorical BOR and % change from baseline of the sum of target lesions at the time of BOR. PFS and overall survival (OS) were calculated from initiation of therapy. Patients who did not experience progression and/or death were censored at the time of the database lock. Patient baseline characteristics were manually extracted and entered into a clinical data form.

Two external validation cohorts consisted of patients with NSCLC enrolled in phase I and II studies receiving durvalumab with or without tremelimumab. These were Study CP110813 (NCT01693562, durvalumab, n=326) and Study C00614 (NCT02000947, durvalumab with tremelimumab, n=381). These studies were approved by each institution’s Ethical Review Board and patients provided written informed consent.

To examine SD in different solid tumors and treatments, we identified additional cohorts at MSK and patients enrolled in phase III studies. We identified three cohorts of patients who received PD-(L)1 blockade based therapy including patients with melanoma (seen at MSK, n=99), bladder cancer (NCT02516241, durvalumab with or without tremelimumab, n=708), and HNSCC (NCT02369874, durvalumab with or without tremelimumab, n=487). Four additional cohorts were identified of non-immunotherapy treatment. These cohorts consisted of patients with NSCLC (NCT02453282, chemotherapy, n=372), bladder cancer (NCT02516241, chemotherapy, n=317), HNSCC (NCT02369874, chemotherapy, n=249), and RCC (seen at MSK, TKI, n=156). For the MSK RCC cohort, a composite PFS endpoint was used. For individuals who did not experience RECIST assessed progression or death while on treatment, time to start of the subsequent line of therapy was used.

A final cohort was identified examining the relationship between radiological assessment and circulating tumor (ct)DNA analysis in SD. This consisted of 41 patients enrolled in clinical trials who received durvalumab with or without tremelimumab (NCT01693562 and NCT02087423) and opted in for baseline and on-treatment ctDNA testing (6–8 weeks after ICI start) and had SD as BOR.

Systematic Review

A systematic review was performed using MEDLINE and Google Scholar searching for published phase I-III studies of patients with advanced solid tumors treated with PD-(L)1 blockade +/− CTLA-4 blockade on August 8, 2020. Exclusion criteria included tumor types with fewer than 100 total patients treated with therapy, trials that did not report proportion of SD, trials where the number of patients who were not evaluable for RECIST BOR was either unknown or not clearly stated, and trials without published manuscripts. Manuscripts and accompanying supplementary materials were reviewed for characteristics of interest.

Approach to comparing SD with tumor-specific proliferation index

We hypothesized the proportion of SD across solid tumor types may associate with cancer-specific growth kinetics. To test this hypothesis, we compared best overall response (BOR) from the trials included in the systematic review with a proxy of tumor growth rate, proliferative index (PI).15 PI was calculated for each tumor type using RNA-seq data from Cancer Genome Atlas (TCGA) tumors. PI was defined by the meta- proliferating cell nuclear antigen (PCNA) signature from Venet et al. This study defined a proliferation score as the 1% of genes most positively correlated with PCNA expression in normal human organs from a gene expression database (n=131 genes, including markers of proliferation such as MKI67 and MCM2).16 TCGA RNA-seq data was downloaded from the Genomic Data Commons (GDC) Portal (portal.gdc.cancer.gov) between October 19, 2020 and October 21, 2020. HTSeq count data for each file was normalized to counts per million (CPM) prior to calculating PI scores for each individual sample. Individual PIs were averaged for each tumor type. Count matrix assembly and data analysis was performed using Python v3.7.3 (Python Software Foundation, https://www.python.org/) using NumPy17 and matplotlib; and R v3.6.2 (R foundation for Statistical Computing, Vienna, Austria) with RStudio v1.3.776 using Proliferative Index v1.0.1.3.18

Approach to developing SD responder (SD-R) definition

We hypothesized objective measures of response (PFS and BOR) could be used to identify a subpopulation of patients with RECIST SD whose survival outcomes mirrored those with partial response (PR). Our primary analysis was a comparison of outcomes between patients who met a specific definition of “SD responder” (SD-R) and patients with objective response.

Consistent with other reports19, depth of response correlates with duration of progression-free survival. For example, stratifying those who achieved BOR PR into two groups by the median depth of response (−57%, 57% tumor shrinkage), the median PFS was 10.4 months vs 22.4 months in those with 30–57% vs 57–100% reduction, respectively. This reflects the heterogeneity of outcomes within PR. Therefore, in pursuit of identifying responders among those with radiological stable disease, we focused on comparing those with SD to those with minor PR (30–57% tumor shrinkage, PR-minor) as the subgroup likely to be most similar to the hypothesized SD-R.

In the discovery cohort, we systematically examined serial thresholds of %BOR and PFS as candidate definitions of SD-R. Overall survival was estimated using Kaplan-Meier methodology from the time of start of therapy. The long-rank test was used to compare survival curves. The hypothesis and a step by step schema of the methods is shown in Supplementary Figure 1. The proposed definition was subsequently examined in the two external clinical trial validation cohorts.

Statistical Analysis

Statistical analysis and data visualization was performed using Python v3.7.3 (Python Software Foundation, https://www.python.org/) with NumPy17, scipy, matplotlib and seaborn, and R v3.6.2 (R foundation for Statistical Computing, Vienna, Austria) with RStudio v1.3.776 with survival. A multivariable Cox proportional hazards model was used to calculate adjusted hazard ratios (aHR). The chi-squared [χ2] test was used to test for independence among biomarker selected subgroups. NumPy was used to calculate linear least squares fit and Statsmodels v0.12.020 was used to calculate weighted linear least squares fit. Spearman’s rank correlation coefficients (ρ) were used to compare best overall response vs proliferative index and %BOR vs PFS. All statistical tests were two-sided with a significance level of 2.5% for each tail. Seaborn v0.11.021 was used for the violin.

Results:

Disease-specific frequency of stable disease with immunotherapy

Sixty-seven cohorts from 59 published trials (14,280 individual patients) were identified in 12 types of advanced solid tumors treated with PD-1 blockade-based immunotherapy (Supplementary Table 1). The proportion of patients with SD as BOR ranged from 42% in hepatocellular carcinoma to 16% in melanoma and did not appear to correspond to ICI responsive tumor types (Figure 1A). The median proportion of patients with SD in the 3 tumor types enrolling the most patients were 31% in non-small cell lung cancer (NSCLC) (n=4,264 from 13 trials), 16% in melanoma (n=2,802 from 9 trials), and 21% in head and neck squamous cell cancer (n=1,780 from 8 trials).

Figure 1. Stable disease (SD) in patients receiving immune checkpoint inhibitors (ICIs).

Figure 1.

Open in a new tab

A. Percentage of patients with SD as their RECIST v1.1 best overall response (BOR) at the trial level by solid tumor type in the published literature. Non-small cell lung cancer (NSCLC) has the greatest number of patients (n=4,264) and one of the highest %SD (31%). Each bubble represents one arm from a phase I-III clinical trial of ICIs (PD-[L]1 monotherapy or in combination with CTLA-4 therapy). The area of the bubble is proportional to the number of patients enrolled in the study arm. Horizontal bars represent the weighted median %SD for that tumor type. Horizontal jitter within each column was introduced to better visualize differences among bubbles. The denominator for calculating %SD was the number of patients who received at least one dose of therapy. Exclusion criteria: tumor types with fewer than 100 total patients treated with therapy, trials where the number of patients who were not evaluable for RECIST BOR was either unknown or not clearly stated, and trials without published manuscripts. Search performed: August 8, 2020. Abbreviations: HCC = hepatocellular carcinoma; RCC = renal cell cancer; NSCLC = non-small cell lung cancer; dMMR = mismatch repair deficiency; HNSCC=head and neck squamous cell cancer; UCC = urothelial cell carcinoma; SCLC = small cell lung cancer, *Cervical includes cervical, vulvar, and vaginal cancer. B and C. Best overall response of the tumor types from phase I-III clinical trials of ICIs compared to proliferative index (PI), a commonly used signature calculated from RNA-seq data from TCGA tumors reflective of tumor kinetics. Among comparisons made between SD, PR/CR, and PI, only SD significantly associates with PI (ρ = −0.75, p=0.03). The area of the bubble is proportional to the number of patients enrolled in trials investigating the tumor type. The shading of the bubble is proportional to the number of TCGA tumors used to calculate mean PI. ^cpm = counts per million was used to normalize PI, *Cervical includes cervical, vulvar, and vaginal cancer D. Trials of patients with NSCLC from 1A ordered by %SD. %SD ranged from 22–39%. Abbreviations: TPS = PD-L1 expression by tumor proportion score, TC = PD-L1 expression on tumor cells, PR = partial response, SD = stable disease, PD = progressive disease E. %SD compared to either %PR/CR (purple) or %PD (orange) in PD-L1 expression unselected trials in patients with NSCLC. Lines are weighted linear least squares fit. F. %SD compared to either %PR/CR (purple) or %PD (orange) in 3 trials that had BOR data grouped by PD-L1 expression Keynote 042 (top), CheckMate 227 (middle), CheckMate 568 (bottom). Lines show weighted linear least squares fit.

We hypothesized the heterogenous distribution of SD across tumor types may largely reflect the intrinsic kinetics of the disease, rather than relative responsiveness to ICIs. To test this hypothesis, we examined the relationships among these three entities: SD, a commonly used gene signature representing disease kinetics15, 18, 22, 23, and objective response. To estimate disease kinetics, we calculated the average proliferative index [PI] (or meta-PCNA transcriptomic signature) using TCGA RNA-seq data for the tumor types that overlapped with our systematic review. We found that SD significantly associates with disease kinetics across tumor types (ρ = −0.75, p=0.03) (Figure 1B). By contrast, partial response/complete response (PR/CR) does not significantly associate with disease kinetics across tumor types (ρ = −0.04, p=0.9). Additionally, the proportion of SD did not significantly associate with the proportion of PR/CR across tumor types (ρ = 0.13, p=0.7) (Figure 1C). Among these three possible comparisons, the most significant one is that %SD associates with disease kinetics. This suggests %SD to immunotherapy treatment is more reflective of the proportion of slower growing tumors rather than the relative ICI-responsiveness in a given cancer type.

Stable disease in NSCLC

Although SD overall may most reflect tumor-intrinsic growth kinetics, we hypothesized that within the heterogeneous composition of SD there is a subpopulation that is benefitting from treatment. Given the range of the proportion of SD across tumors, we first focused on NSCLC in detail to further understand this radiological category. We selected NSCLC because the proportion of SD was among the highest of tumor types examined in our systematic review (Figure 1A). Moreover, given an annual incidence of ~190,000 cases in the US24 and the routine use of PD-(L)1 blockade-based therapy, NSCLC has the highest absolute number of patients with SD to ICIs among solid tumors.

The proportion of patients with SD within NSCLC trials ranged from 22–39% (Figure 1D). There was no clear pattern between the proportion of SD examining trials of PD-(L)1 blockade monotherapy vs. combination with CTLA-4 blockade therapy or by line of therapy.

We examined the relationship between SD and PR/CR or PD. We hypothesized that comparing the proportions of patients in each RECIST category at the trial level among NSCLC ICI trials would reveal the relative underlying composition of degree of benefit among patients with SD. Specifically, the observation of a greater “population shift” or steeper slope when comparing SD and PR/CR or SD and PD would inform whether patients with SD were more likely to consist of patients with responsive or non-responsive disease. We approached this by examining weighted linear least squares best fit lines comparing %SD and %PR/CR as well as %SD and %PD.

We found the relationship between SD and PD was stronger (slope = −1.17, R2=0.36, p=0.07) than the relationship between SD and PR/CR (slope = 0.17, R2=0.01, p=0.8) (Figure 1E). Specifically, variation between trials was largely related to fluctuations in the proportions of SD and PD, while the proportion of PR/CR was largely stable. This suggests that there is more fluidity and overlap between the SD and PD categories, and SD is likely populated by more patients with non-responsive than responsive disease.

We next examined the impact of patient selection using a predictive biomarker on the proportion of SD and relationship with PR/CR or PD. We hypothesized that if biomarker selection enriches for patients more likely to benefit, it will affect the proportion of patients with SD with responsive vs non-responsive disease. We tested this by comparing the only three trials that published BOR data stratified by PD-L1 expression: Keynote 04225, CheckMate 22726, and CheckMate 56827. The overall proportion of SD significantly diminished with increasing biomarker thresholds defining subgroups more likely to benefit (χ2 test of independence: KN042 χ2 =15.4, p=0.004; CM227 χ2 =16.7, p=0.002; CM568 χ2 =30.2, p<0.0001). In contrast to the relationship seen in the biomarker unselected studies (Figure 1E), there was a stronger population shift seen between SD and PR/CR rather than SD and PD in the setting of examining biomarker selected cohorts (Figure 1F). These findings suggest a predictive biomarker not only enriches for response but also decreases the proportion of patients with ambiguous response. Notably, these results support the hypothesis that some with SD are experiencing benefit from ICIs.

Patient-level heterogeneity among NSCLC with SD

Given group-level heterogeneity of response in SD, we examined this challenge more closely at the patient level. We identified a discovery cohort of 1220 patients with NSCLC who received ICI therapy in whom objective radiological assessment had been performed using RECIST v1.1 (Table 1). Of the 313 patients with SD as their best overall response (BOR), the distribution of progression free survival (PFS) ranged widely, from 0.2 to 49 months (median 4.9 months) (Figure 2A). The depth of tumor shrinkage (% BOR, defined as the percent change of the sum of diameters of target lesions at the time of best overall response) ranged from 30% shrinkage to 20% growth (median 2% shrinkage) (Figure 2B).

Table 1:

Baseline patient characteristics and treatment details

Characteristic MSK cohort (n=1220) CP1108 (n=326) C006 (n=381)
Median age [IQR] - yr 67 [59–74] 65 [56–71] 65 [57–71]
Gender - no. (%)
 Female 626 (51%) 140 (43%) 161 (42%)
 Male 594 (49%) 186 (57%) 220 (58%)
Smoking status - no. (%)
 Ever 1035 (85%) 275 (84%) 319 (84%)
 Never 185 (15%) 51 (16%) 62 (16%)
Histology - no. (%)
 Non-squamous 1029 (84%) 161 (49%) 355 (93%)
 Squamous 191 (16%) 165 (51%) 26 (7%)
Line of therapy - no. (%)
 1 373 (31%) 67 (21%) 49 (13%)
 2 575 (47%) 104 (32%) 259 (68%)
 3 or higher 272 (22%) 155 (48%)* 73 (19%)
 Median line of therapy [IQR] 2 [1–2] 2 [2–3] 2 [2–2]
PD-L1 expression - no./total no. (%)
 0% 243/612 (40%) 52/301 (17%) 130/332 (39%)
 1–49% 135/612 (22%) 128/301 (43%) 139/332 (42%)
 ≥ 50% 234/612 (38%) 121/301 (40%) 63/332 (19%)
 Unknown 608 25 49
ECOG PS - no./total no. (%)
 0–1 1128/1220 (92%) 325/325 (100%) 379/379 (100%)
 2 or higher 92/1220 (8%) 0 (0%) 0 (0%)
 Unknown 0 1 2
Treatment - no. (%)
 Anti-PD-(L)1 monotherapy 1030 (84%) 326 (100%) 0 (0%)
 Anti-PD-(L)1+CTLA-4 combination 137 (11%) 0 (0%) 381 (100%)
 Anti-PD-(L)1+other (non-chemo) 53 (4%)* 0 (0%) 0 (0%)
Best overall response - no. (%)
 CR, complete response 27 (2%) 2 (1%) 3 (1%)
 PR, partial response 210 (17%) 54 (17%) 61 (16%)
 SD, stable disease 313 (26%) 116 (36%) 136 (36%)
 PD, progressive disease 670 (55%) 154 (47%) 181 (48%)
Open in a new tab
*

Percentages may not add up to 100 due to rounding. IQR = interquartile range, PD-(L)1 = programmed cell death protein (ligand) 1, ECOG PS = Eastern Cooperative Oncology Group performance status, CTLA-4 = cytotoxic T-lymphocyte-associated protein 4, chemo = chemotherapy

Figure 2. SD is heterogenous and objective response criteria may identify SD responders.

Figure 2.

Open in a new tab

A. Violin showing distribution of progression free survival in 313 patients with BOR SD seen at MSK. PFS ranged from 0.2–49 months (median 4.9 months, interquartile range 3.4–8.6 months). Vertical dashed lines and percentages represent the median and interquartile range B. Distribution of patients with SD by best % change of the sum of target lesions per RECIST v1.1 (% best overall response, [%BOR]). %BOR ranged from −30 to +20% (median −2%, 2% shrinkage). C. PFS vs %BOR in patients with SD (n=313) (blue) and PR/CR (n=235) (purple). Each circle represents one patient. Stratifying the PR/CR group by the median (57% shrinkage), the darker purple circles represent the PR group (PR-minor) that is most similar in objective response to SD. Lines are linear least squares fit of SD and PR/CR. D. Flow diagram outlining approach to identifying responders in SD. We used serial cutpoints of %BOR and PFS as candidate definitions of SD responder. Each proposed definition was compared to the PR minor group by calculating log-rank hazard ratios (HR) for overall survival. Supplementary Figure 1 is an illustrated schema of the methods in greater detail. E. Matrix of PFS (horizontal) and %BOR (vertical) cutpoints used as candidate definitions of SD responder and resulting HRs when compared to PR minor. A PFS of > 6 months and no tumor growth defined the population of SD responder most similar to PR. The heatmap is shaded by HR using an inverse linear scale.

Identifying stable disease responders (SD-R)

Using known objective measures of response, we asked whether we could identify a subset within SD benefiting from treatment, termed stable disease responders (SD-R). To identify this subgroup within SD, we first examined the distribution of % BOR and PFS in patients with SD (n=313) and PR/CR (n=235). We observed that depth and duration of benefit correlated (Figure 2C). Given the heterogeneity of response within PR, to optimally identify SD-R, we compared candidate definitions of “SD responder” to the subgroup within PR/CR most similar to SD (PR-minor, % shrinkage < median among responders [−30 to −57%]) (Figure 2D). A schema of the approach is provided in Supplementary Figure 1. Each candidate definition was compared to the PR-minor group using log-rank hazard ratios (HR) for overall survival (OS). A PFS of >6 months and no growth identified the group of SD that best tracked with PR-minor (HR=1.0, 95% CI 0.7–1.5, Figure 2E). This population represented 28% (n=87/313) of patients in the discovery cohort with stable disease and 7% of all patients treated (Figure 3A) and led to significant differences in survival among SD (HR=0.43, 95% CI 0.33–0.57) (Supplementary Figure 2A). This finding remained significant after adjusting for baseline features of derived neutrophil to lymphocyte ratio (dNLR) and tumor mutation burden (TMB) (aHR = 0.42, 95% CI 0.28–0.63) (Supplementary Table 2).

Figure 3. SD responders among patients who received ICIs.

Figure 3.

Open in a new tab

For each figure in this series, the donut (left) represents the % of patients in each response category and the Kaplan Meier (right) shows estimated overall survival of SD-responder compared to the subset of PR/CR most similar to SD, PR minor (PR and shrinkage of less than the median within PR/CR). The wedge within the donut is the % of patients who met our definition of SD responder (a PFS of > 6 months and no tumor growth). The three cohorts were A. the MSK discovery cohort (n=1220), B. the external validation cohort of patients with NSCLC who received ICIs in study CP1108 (n=326), and C. the external validation cohort of patients with NSCLC who received ICIs in study C006 (n=381). PR major = PR and shrinkage of at least the median within PR/CR.

We proposed PFS of > 6 months and no growth as our definition for SD-R to examine in validation. We next examined two external, independent validation cohorts of patients with NSCLC who were enrolled in phase I and II studies of durvalumab +/− tremelimumab (CP110813 and C00614). Despite some differences in baseline characteristics (Table 1), when applying our definition of SD-R, we found nearly identical proportions of this group compared to the discovery cohort (24% of SD and 9% overall in CP1108; 31% of SD and 11% overall in C006) (Figure 3B, 3C). Consistent with results from the discovery cohort, SD-R had similar overall survival compared to PR-minor (CP1108 HR=0.7, 95% 0.3–1.7; C006 HR=0.7, 95% CI 0.3–1.7) and SD-R was significantly different from the remainder of SD (CP 1108 HR=0.43, 95% CI 0.26–0.75; C006 HR=0.36, 95% CI 0.21–0.61) (Supplementary Figure 2B, 2C). Although the hazard ratios are not as close to 1.0 seen in the discovery cohort, this is expected with the smaller validation sets, reflected by the wider 95% confidence interval that spans 1.0.

Examination of the generalizability of SD-R among patients treated with immunotherapy

We next asked whether this definition of SD-R could be applied to other disease and treatment settings.

To explore the applicability of SD-R across tumor types treated with immunotherapy, we identified three additional cohorts of solid tumors (melanoma, n=99; bladder cancer, n=708; and HNSCC, n=487). Given our approach was based in part upon the observation that depth of response and duration of benefit (i.e. %BOR vs PFS) were correlated in NSCLC, we first examined whether this correlation was present in other tumor types. Similar to our discovery NSCLC cohort (spearman ρ=−0.36, p <0.0001) and most solid tumors in the published literature 28–31, depth and duration were correlated in melanoma (spearman ρ=−0.52, p<0.001) and bladder cancer (spearman ρ=−0.32, p<0.001) (Supplementary Figure 3A) but not in HNSCC (spearman ρ=0.04, p=0.7) (Supplementary Figure 3B).

We applied the definition of SD-R (>6 months PFS and no tumor growth) derived from NSCLC to these additional immunotherapy-treated cohorts. In tumor types in which depth and duration correlated as expected (melanoma and bladder cancer), we found promising results in which SD-R mirrored PR-minor and was distinct from the remainder of SD (Figure 4A). In patients with HNSCC, an exception where depth and duration were not correlated, SD-R was able to distinguish survival among SD but these patients did not have as positive outcomes as PR-minor (Figure 4B).

Figure 4. SD responders among patients who received ICIs in multiple solid tumor types.

Figure 4.

Open in a new tab

For each figure in this series, the line (top) represents linear least squares fit and bootstrapped 95% CI for %BOR vs PFS for PR/CR. Also see Supplementary Figure 3. Kaplan-Meier estimated overall survival curves show SD-R vs PR-minor (middle) and SD-R vs non-SD-R (bottom). The cohorts include A. melanoma (MSK, n=99; SD-R vs PR-min HR=2.07, 95% CI 0.47–9.09, SD-R vs non-SD-R HR=0.26, 95% CI 0.10–0.71), bladder cancer (Danube study, durvalumab+/−tremelimumab, n=708, SD-R vs PR-min HR=1.49, 95% CI 1.03–2.15, SD-R vs non-SD-R HR=0.39, 95% CI 0.26–0.59), B. HNSCC (Eagle study, durvalumab+/−tremelimumab, n=487, SD-R vs PR-min HR=2.94, 95% CI 1.49–5.82, SD-R vs non-SD-R HR=0.44, 95% CI 0.29–0.66). ρ denotes the spearman rank correlation coefficient of PR/CR %BOR vs PFS. Hazard ratios (HR) and 95% confidence intervals (CIs) were calculated using the log-rank test.

Examination of the generalizability of SD-R across treatment types

We also explored the application of this SD-R definition across other treatment types, acknowledging that mechanisms of action differ between immunotherapy and other therapies. We did this in four cohorts of solid tumors treated with chemotherapy (NSCLC, n=372; bladder cancer, n=317; and HNSCC, n=249) or tyrosine kinase inhibitors (TKIs) (RCC, n=156). Depth of response associated with duration of benefit in patients with RCC who received TKIs and NSCLC who received chemotherapy (spearman ρ=−0.42, p<0.001; and spearman ρ=−0.30, p=0.001, respectively) (Supplementary Figure 3C), but not bladder cancer or HNSCC treated with chemotherapy (Supplementary Figure 3D). When applying the definition of SD-R, we found there was reasonable performance in distinguishing survival of patients with SD in RCC with TKIs and NSCLCs with chemotherapy (Supplementary Figure 4A) and less so in bladder cancer and HNSCC treated with chemotherapy (Supplementary Figure 4B).

Discussion:

We examined RECIST-defined stable disease to ICI therapy to highlight the common challenge of assigning individual-level treatment benefit within this category and to create a framework with which to decipher responders and non-responders in retrospective analyses. Improving our understanding of this radiological category is essential for deriving confident conclusions in research and drug development.

SD was the best response to ICI therapy in ~30% and had a wide range of outcomes from no benefit to exceptionally long-term benefit across and within tumor types. Differences in SD among tumor types likely reflects the variation in the underlying biology in response to ICIs. Across solid tumor types, a higher proportion of SD to ICIs significantly associated with a lower proliferation index. We infer that a higher proportion of SD within a cancer type may reflect a slower growing tumor type rather than ICI response. Future analysis with patient level gene expression data of a variety of tumors will be needed to explore this hypothesis further, such as teasing apart the contributions of stage and line of therapy.

To parse SD, we found a potential underappreciated virtue of predictive biomarkers is to reduce the proportion of radiologically ambiguous outcomes to ICIs. In biomarker unselected studies, SD is likely composed of more individuals with non-response than benefit to ICIs. By contrast, biomarker selection significantly lowered the proportion of SD and increased the proportion of PR/CR.

Within a tumor type, SD likely comprises of patients with multiple biological states – radiologically undetectable benefit, radiologically undetectable non-response, and indolent disease. To identify those with SD and treatment benefit (SD-R), we initially focused our analysis on NSCLC. We then examined the applicability of SD-R in broader settings including other solid tumor types receiving immunotherapy. We found SD-R is particularly well suited for distinguishing responders from non-responders to immunotherapy when depth of response in objective responders (%BOR in PR/CR) correlated with duration of benefit (Figure 4, Supplementary Figure 3). This correlation is generally true among solid tumors in the literature28–31 and in our analysis. HNSCC was an outlier potentially due to a tropism for lymph nodes and association with prior radiation that contributes to challenges in RECIST v1.1 assessment in this disease. Consistent with this hypothesis, SD-R also did not generalize to HNSCC and chemotherapy (Supplementary Figure 4B).

For exploratory purposes, we tested SD-R in chemotherapy and targeted therapy treated cohorts and found it to be less applicable (Supplementary Figure 4). We are cautious to apply this definition to these treatments without further refinement. By contrast to chemotherapy/targeted therapy, the long-term indirect impact of immunotherapy leads to dichotomous and durable clinical responses that is better encapsulated by using PFS as a feature to define SD-R. This exploratory work highlights the need to consider the treatment type when addressing response assessment in SD. Future directions include chemotherapy-specific, targeted therapy-specific and combination-therapy specific investigations.

Our report is the first to use a systematic approach to refine the concept of durable clinical benefit (DCB)32 and propose a preliminary definition of SD responders to immunotherapy; we ultimately aim to identify broadly accepted definitions for application in prospective and retrospective reports that currently struggle with how to interpret this category. This effort is a part of the broader challenge of accurately determining immunotherapy response using radiological tools (e.g. iRECIST, irRECIST).33–37 Mechanistically, the particular difficulty with immunotherapy may be related to the indirect mechanism of action on tumor cells and impact on immunity leading to survival not fully quantifiable by imaging alone. As an example, a responding tumor newly infiltrated with immune cells could appear larger on radiological imaging38, but the scenario of clinically detectable pseudoprogression is rare in NSCLC (1% in our published series)39 and these newer RECIST criteria are not used as primary response assessment in registrational clinical trials. By contrast, SD represents a common and profound inefficiency for clinical trials of new therapies and translational analyses seeking to identify predictors of response.

While other metrics such as PFS measure treatment response, they are imperfect and do not take depth of response into account. Additionally, multiple metrics of response should be used as sensitivity analyses to test the robustness of responder-non-responder analyses.

We propose a definition of “SD responder” to immunotherapy: greater than 6 months PFS and no tumor growth. This definition reflected ~1/3rd of patients with SD and ~10% of patients with NSCLC receiving ICIs alone. These findings are consistent with our results demonstrating a closer association between SD and PD, but still identify a sufficiently large proportion of patients to be useful in clinical and translational studies. As an example, in the published trials of NSCLC and ICIs (n=4,264), this would reclassify ~440 patients.

Clarifying benefit among patients with radiologic stability will ultimately be crucial to real-time clinical decision making, to guide continuation of effective therapy and avoid unnecessary toxicity or cost of ineffective treatment. One promising way to prospectively identify SD responders is tracking changes in plasma cell-free circulating tumor DNA (ct)DNA from apoptotic tumor cells using assays optimized for detection from a routine blood draw.40–43 There remain clinical and technical optimization hurdles before serial ctDNA assessment becomes routine, but we anticipate this tool will help characterize the disease state of patients with radiological stable disease.

In summary, RECIST stable disease in patients receiving ICIs is common but usually uninformative. Biomarkers are needed in this setting to define true benefit. Using a discovery, two external validation cohorts in NSCLC, and applying across other tumor types treated with immunotherapy, we propose those with BOR of SD using RECIST criteria and additionally a PFS>6 months and no tumor growth (%BOR≤0%) may be considered “SD responders” (SD-R). This is the first step towards a unified definition of response in this ambiguous RECIST category. Identification and application of an accepted definition of SD responders will enhance the signal with respect to the noise in drug development and improve precision in translational research. Further refinement of this definition is critical for improving our ability to tailor treatments and identify better therapies for patients with cancer.

Supplementary Material

1

Highlights:

  • SD to ICIs is common; disease-specific frequency may reflect tumor-type disease kinetics rather than treatment response

  • SD encompasses a wide range of outcomes and biomarkers enrich for individuals who benefit (stable disease responders, SD-R)

  • We propose and show PFS >6 mo. and no tumor growth as definition of SD-R to ICI, a group with SD and OS mirroring responders

  • Identifying “responders” among those with SD will be essential for improving precision of research and drug discovery

Acknowledgements:

The authors thank M.J. Andreason and V. Avutu for helpful discussions related to this manuscript.

Funding Source:

Supported by Memorial Sloan Kettering Cancer Center Support Grant/Core [P30-CA008748] and the Druckenmiller Center for Lung Cancer Research at Memorial Sloan Kettering Cancer Center; the National Institutes of Health [T32-CA009207 to J.L., K30-UL1TR00457 to J.L.]; Conquer Cancer Foundation of the American Society of Clinical Oncology [YIA to J.L.]; Damon Runyon Cancer Research Foundation [CI-98–18 to M.D.H.], M.D.H. is a member of the Parker Institute for Cancer Immunotherapy.

Conflict of interest disclosure statement:

J.L. has received honoraria from Targeted Oncology and Physicians’ Education Resource.

S.W. and Q.Z. are former employees of AstraZeneca.

M.K.C. receives institutional research funding from Bristol-Myers Squibb; and has received personal fees from Merck, InCyte, Moderna, ImmunoCore, and AstraZeneca.

A.N.S. reports advisory board positions with Bristol-Myers Squibb, Immunocore, Novartis, and Castle Biosciences; and institutional research support from BMS, Immunocore, and Xcovery, Polaris, Novartis, Pfizer, Checkmate Pharmaceuticals, and Foghorn Therapeutics.

M.A.P. reports consulting fees from Bristol-Myers Squibb, Merck, Array BioPharma, Novartis, Incyte, NewLink Genetics, Pfizer, and Aduro; honoraria from BMS and Merck; and institutional research support from Rgenix, Infinity, BMS, Merck, Array BioPharma, Novartis, and AstraZeneca.

M.H.V reports receiving commercial research support from Bristol-Myers Squibb, Pfizer and Genentech/Roche; honoraria from Novartis and Bristol-Myers Squibb; travel/accommodation from Astra Zeneca, Eisai, Novartis and Takeda; consultant/advisory board member for Aveo, Calithera Biosciences, Corvus Pharmaceuticals, Exelixis, Eisai, Merck, Onquality Pharmaceuticals, Novartis and Pfizer.

M.S.G. has been a compensated consultant for Ultimate Opinions in Medicine LLC and MORE Health, Inc.

A.G. and R.R. are employees of AstraZeneca.

R.R. has a patent pending related to tumor mutation burden.

M.G.K. receives personal fees from AstraZeneca, Pfizer, Regeneron, and Daiichi-Sankyo; received honoraria for participation in educational programs from WebMD, OncLive, Physicians Education Resources, Prime Oncology, Intellisphere, Creative Educational Concepts, Peerview, i3 Health, Paradigm Medical Communications, AXIS, Carvive Systems, AstraZeneca, and Research to Practice; received travel support from AstraZeneca, Pfizer, Regeneron, and Genentech. Dr. Kris is an employee of Memorial Sloan Kettering. Memorial Sloan Kettering has received research funding from The National Cancer Institute (USA), The Lung Cancer Research Foundation, Genentech/Roche, and PUMA Biotechnology for research conducted by Dr. Kris. MSK has licensed testing for EGFR T790M to MolecularMD.

M.D.H. as of November 2021, is an employee of AstraZeneca; has received personal fees from Achilles, Adagene, Adicet, Arcus, Blueprint Medicines, Bristol-Myers Squibb, DaVolterra, Eli Lilly, Genentech/Roche, Genzyme/Sanofi, Janssen, Immunai, Instil Bio, Mana Therapeutics, Merck, Mirati, Natera, Pact Pharma, Shattuck Labs, and Regeneron; has options from Factorial, Shattuck Labs, Immunai, and Arcus; has a patent filed by Memorial Sloan Kettering related to the use of tumor mutation burden to predict response to immunotherapy (PCT/US2015/062208), which has received licensing fees from PGDx.

Footnotes

The remaining authors have no disclosures.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Moertel CG, Hanley JA. The effect of measuring error on the results of therapeutic trials in advanced cancer. Cancer 1976; 38 (1): 388–394. [DOI] [PubMed] [Google Scholar]
  • 2.World Health O. WHO handbook for reporting results of cancer treatment. Geneva: World Health Organization; 1979. [Google Scholar]
  • 3.Therasse P, Arbuck SG, Eisenhauer EA et al. New Guidelines to Evaluate the Response to Treatment in Solid Tumors. JNCI: Journal of the National Cancer Institute 2000; 92 (3): 205–216. [DOI] [PubMed] [Google Scholar]
  • 4.Eisenhauer EA, Therasse P, Bogaerts J et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009; 45 (2): 228–247. [DOI] [PubMed] [Google Scholar]
  • 5.Hellmann MD, Nathanson T, Rizvi H et al. Genomic Features of Response to Combination Immunotherapy in Patients with Advanced Non-Small-Cell Lung Cancer. Cancer Cell 2018; 33 (5): 843–852 e844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Tumeh PC, Harview CL, Yearley JH et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 2014; 515 (7528): 568–571. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.de Vries R, Muller M, van der Noort V et al. Prediction of response to anti-PD-1 therapy in patients with non-small-cell lung cancer by electronic nose analysis of exhaled breath. Ann Oncol 2019; 30 (10): 1660–1666. [DOI] [PubMed] [Google Scholar]
  • 8.Braun DA, Hou Y, Bakouny Z et al. Interplay of somatic alterations and immune infiltration modulates response to PD-1 blockade in advanced clear cell renal cell carcinoma. Nat Med 2020; 26 (6): 909–918. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Miao D, Margolis CA, Gao W et al. Genomic correlates of response to immune checkpoint therapies in clear cell renal cell carcinoma. Science 2018; 359 (6377): 801–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Roh W, Chen PL, Reuben A et al. Integrated molecular analysis of tumor biopsies on sequential CTLA-4 and PD-1 blockade reveals markers of response and resistance. Sci Transl Med 2017; 9 (379). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Miao D, Margolis CA, Vokes NI et al. Genomic correlates of response to immune checkpoint blockade in microsatellite-stable solid tumors. Nat Genet 2018; 50 (9): 1271–1281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Zhang Q, Luo J, Wu S et al. Prognostic and Predictive Impact of Circulating Tumor DNA in Patients with Advanced Cancers Treated with Immune Checkpoint Blockade. Cancer Discovery 2020; 10 (12): 1842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Antonia SJ, Balmanoukian A, Brahmer J et al. Clinical Activity, Tolerability, and Long-Term Follow-Up of Durvalumab in Patients With Advanced NSCLC. J Thorac Oncol 2019; 14 (10): 1794–1806. [DOI] [PubMed] [Google Scholar]
  • 14.Antonia S, Goldberg SB, Balmanoukian A et al. Safety and antitumour activity of durvalumab plus tremelimumab in non-small cell lung cancer: a multicentre, phase 1b study. Lancet Oncol 2016; 17 (3): 299–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Venet D DJ, Detours V (2011) Most Random Gene Expression Signatures Are Significantly Associated with Breast Cancer Outcome. PLoS Comput Biol 7(10): e1002240. 10.1371/journal.pcbi.1002240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Ge X, Yamamoto S, Tsutsumi S et al. Interpreting expression profiles of cancers by genome-wide survey of breadth of expression in normal tissues. Genomics 2005; 86 (2): 127–141. [DOI] [PubMed] [Google Scholar]
  • 17.Harris CR, Millman KJ, van der Walt SJ et al. Array programming with NumPy. Nature 2020; 585 (7825): 357–362. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Ramaker RC, Lasseigne BN, Hardigan AA et al. RNA sequencing-based cell proliferation analysis across 19 cancers identifies a subset of proliferation-informative cancers with a common survival signature. Oncotarget; Vol 8, No 24 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Blumenthal GM, Zhang L, Zhang H et al. Milestone Analyses of Immune Checkpoint Inhibitors, Targeted Therapy, and Conventional Therapy in Metastatic Non–Small Cell Lung Cancer Trials: A Meta-analysis. JAMA Oncology 2017; 3 (8): e171029-e171029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Seabold Skipper, Perktold. statsmodels: Econometric and statistical modeling with python. Proceedings of the 9th Python in Science Conference 2010. [Google Scholar]
  • 21.Waskom M, Botvinnik O, O’Kane D et al. mwaskom/seaborn v0.11.0 (September 2020). Zenodo; 2020. [Google Scholar]
  • 22.Endesfelder D, Burrell R, Kanu N et al. Chromosomal instability selects gene copy-number variants encoding core regulators of proliferation in ER+ breast cancer. Cancer Res 2014; 74 (17): 4853–4863. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Tomas G, Tarabichi M, Gacquer D et al. A general method to derive robust organ-specific gene expression-based differentiation indices: application to thyroid cancer diagnostic. Oncogene 2012; 31 (41): 4490–4498. [DOI] [PubMed] [Google Scholar]
  • 24.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA: A Cancer Journal for Clinicians 2020; 70 (1): 7–30. [DOI] [PubMed] [Google Scholar]
  • 25.Mok TSK, Wu YL, Kudaba I et al. Pembrolizumab versus chemotherapy for previously untreated, PD-L1-expressing, locally advanced or metastatic non-small-cell lung cancer (KEYNOTE-042): a randomised, open-label, controlled, phase 3 trial. Lancet 2019; 393 (10183): 1819–1830. [DOI] [PubMed] [Google Scholar]
  • 26.Hellmann MD, Ciuleanu T-E, Pluzanski A et al. Nivolumab plus Ipilimumab in Lung Cancer with a High Tumor Mutational Burden. New England Journal of Medicine 2018; 378 (22): 2093–2104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Ready N, Hellmann MD, Awad MM et al. First-Line Nivolumab Plus Ipilimumab in Advanced Non–Small-Cell Lung Cancer (CheckMate 568): Outcomes by Programmed Death Ligand 1 and Tumor Mutational Burden as Biomarkers. Journal of Clinical Oncology 2019; 37 (12): 992–1000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.McCoach CE, Blumenthal GM, Zhang L et al. Exploratory analysis of the association of depth of response and survival in patients with metastatic non-small-cell lung cancer treated with a targeted therapy or immunotherapy. Ann Oncol 2019; 30 (3): 492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Osgood C, Mulkey F, Mishra-Kalyani PS et al. FDA analysis of depth of response (DpR) and survival across 10 randomized controlled trials in patients with previously untreated unresectable or metastatic melanoma (UMM) by therapy type. Journal of Clinical Oncology 2019; 37 (15_suppl): 9508–9508. [Google Scholar]
  • 30.Wang M, Chen C, Jemielita T et al. Are tumor size changes predictive of survival for checkpoint blockade based immunotherapy in metastatic melanoma? J Immunother Cancer 2019; 7 (1): 39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Fucà G, Corti F, Ambrosini M et al. Prognostic impact of early tumor shrinkage and depth of response in patients with microsatellite instability-high metastatic colorectal cancer receiving immune checkpoint inhibitors. J Immunother Cancer 2021; 9 (4): e002501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Rizvi H, Sanchez-Vega F, La K et al. Molecular Determinants of Response to Anti–Programmed Cell Death (PD)-1 and Anti–Programmed Death-Ligand 1 (PD-L1) Blockade in Patients With Non–Small-Cell Lung Cancer Profiled With Targeted Next-Generation Sequencing. Journal of Clinical Oncology 2018; 36 (7): 633–641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wolchok JD, Hoos A, Day S et al. Guidelines for the Evaluation of Immune Therapy Activity in Solid Tumors: Immune-Related Response Criteria. Clinical Cancer Research 2009; 15 (23): 7412. [DOI] [PubMed] [Google Scholar]
  • 34.Nishino M, Giobbie-Hurder A, Gargano M et al. Developing a Common Language for Tumor Response to Immunotherapy: Immune-Related Response Criteria Using Unidimensional Measurements. Clinical Cancer Research 2013; 19 (14): 3936. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Nishino M, Gargano M, Suda M et al. Optimizing immune-related tumor response assessment: does reducing the number of lesions impact response assessment in melanoma patients treated with ipilimumab? J Immunother Cancer 2014; 2: 17–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Bohnsack O, Hoos A, Ludajic K. 1070P - Adaptation of the Immune Related Response Criteria: Irrecist. Annals of Oncology 2014; 25: iv369. [Google Scholar]
  • 37.Seymour L, Bogaerts J, Perrone A et al. iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics. The Lancet Oncology 2017; 18 (3): e143–e152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Di Giacomo AM, Danielli R, Guidoboni M et al. Therapeutic efficacy of ipilimumab, an anti-CTLA-4 monoclonal antibody, in patients with metastatic melanoma unresponsive to prior systemic treatments: clinical and immunological evidence from three patient cases. Cancer Immunology, Immunotherapy 2009; 58 (8): 1297–1306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Osorio JC, Arbour KC, Le DT et al. Lesion-Level Response Dynamics to Programmed Cell Death Protein (PD-1) Blockade. J Clin Oncol 2019; 37 (36): 3546–3555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Goldberg SB, Narayan A, Kole AJ et al. Early Assessment of Lung Cancer Immunotherapy Response via Circulating Tumor DNA. Clin Cancer Res 2018; 24 (8): 1872–1880. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Anagnostou V, Forde PM, White JR et al. Dynamics of Tumor and Immune Responses during Immune Checkpoint Blockade in Non-Small Cell Lung Cancer. Cancer Res 2019; 79 (6): 1214–1225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Nabet BY, Esfahani MS, Moding EJ et al. Noninvasive Early Identification of Therapeutic Benefit from Immune Checkpoint Inhibition. Cell. [DOI] [PMC free article] [PubMed]
  • 43.Tran HTN, Ang KS, Chevrier M et al. A benchmark of batch-effect correction methods for single-cell RNA sequencing data. Genome Biology 2020; 21 (1): 12. [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.

Supplementary Materials

1
NIHMS1861243-supplement-1.pdf (8MB, pdf)

RESOURCES