Evaluating the tumor response assessment of modified-RECIST 1.1 in advance non-small cell lung cancer: a post-hoc analysis of 1,147 patients from the EAST-LC trial

Introduction

Response Evaluation Criteria in Solid Tumors (RECIST) is the most widely utilized and standardized guideline in assessing antitumor activity of treatment in current clinical practice. Using data from more than 18,000 target lesions and more than 6,500 patients from sixteen clinical trials (1,2), the Response Evaluation Criteria in Solid Tumors Working Group introduced RECIST 1.1 in 2009. The revision consisted of significant changes, including the use of positron emission tomography/computed tomography (PET/CT), the measurement of lymph nodes (LNs), and the maximum number of target lesions (1-4). Retrospective studies involving 6,512 patients from 16 clinical trials revealed that when using the guidelines set out in RECIST 1.0, which allowed for the assessment of a maximum of ten target lesions, it did not enhance the accuracy of tumor burden calculation. Of note, the sizes of ten target lesions were equally as effective as assessing five lesions for the assessment of the overall response rate (ORR) and progression-free survival (PFS) (2). Consequently, the main change to RECIST 1.1 was that the amount of target lesions should be no more than five, with the maximum of two lesions per organ (1).

And the RECIST 1.1 indicated a high assessment concordance with the RECIST 1.0, along with reduced workload (5-7). Nowadays, the RECIST 1.1 has been the most widely accepted criterion in clinic and oncology clinical trials and play a key role in providing objective, accurate and reproducible evaluation of tumor therapies (8,9). However, the reproducibility and repeatability of the RECIST 1.1 is impacted by several factors, including the choice of target lesions and the measurement of tumor burden (9-15). Previous studies have reported that the selection of target lesions could impact the inter-reader agreement in RECIST 1.1 (12,14). When the target lesions chosen by readers are same, the inter-reader agreement for response assessments in oncology according to RECIST 1.1 is virtually perfect, whereas the target lesions chosen are different, inter-reader discordance roses (16,17). Besides, Tareco Bucho et al. observed a linear increase in disagreement between readers as the total number of lesions increased (18). For the RECIST 1.1, the standard of up to two target lesions per organ is considered to be an arbitrary decision without any supporting objective evidence (19).

Intriguingly, in colorectal cancer (CRC) patients, who had liver metastases, evaluating the size of the largest hepatic lesion alone was equally as predictive of anti-tumor response as measuring up to five lesions (20). One study specifically, found that when using modified RECIST 1.1 (mRECIST 1.1), which involves assessing the single largest lesion per organ, it had a similar tumor response classification as RECIST 1.1 when evaluating two target lesions in each organ in advanced CRC (21), non-small cell lung cancer (NSCLC) (22), gastric cancer (GC) (23), and small cell lung cancer (SCLC) patients (24). Nevertheless, the sample sizes of these studies were small. Thus, larger-scale studies are warranted to help identify the optimum number of target lesions per organ for more improved and more reliable tumor assessment.

In order to investigate the interchangeability of the mRECIST 1.1 (evaluating the single largest lesion per organ), we carried out a retrospective analysis using data from the East Asia S-1 Trial in Lung Cancer (EAST-LC). The dataset comprised of 1,154 patients split into two chemotherapy groups: docetaxel and S-1 (25). The tumor responses were compared using both RECIST 1.1 and mRECIST 1.1 to assess the differences between the criteria. It is the study with the largest sample size to test whether the single-lesion measurement per organ yields the same response classification. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-809/rc).

Methods

Patients

The medical files of advanced NSCLC patients that received platinum-based therapy were randomly assigned 1:1 to receive docetaxel or S-1 therapy in the EAST-LC clinical trial between July 2010 and June 2014. The EAST-LC trial included 1,154 patients (docetaxel 577, and S-1 577). All randomized patients, with the exception of those who had significant protocol deviation, were included into the full analysis set (FAS). The FAS included 577 patients (S-1) and 570 patients (docetaxel). The baseline characteristics were similar between the two cohorts (25). All participating institutes of the EAST-LC approved its protocol, and all patients gave written informed consent.

Response assessment

Magnetic resonance imaging (MRI) and CT of the head, chest, and abdomen were conducted within 21 days prior to randomization to assess the patient’s baseline disease status. Baseline bone scanning was also required if any of the following was observed: a known history of bone metastasis, evidence of bone metastasis at baseline CT or other modalities, and pain suggestive of bone metastasis. Imaging studies included a 6-week follow-up of the chest, abdomen, and head lesions identified at baseline, performed under the same conditions (slice thickness, use of contrast agent) as at baseline. Any baseline lesions other than those mentioned above or any new lesions later suspected to have arisen were assessed when appropriate. The allowable window for the scheduling of imaging studies was ±1 week. Unscheduled imaging was arranged when disease progression was suspected (e.g., worsening of symptoms) or partial or complete response (CR) was confirmed (at least 4 weeks after the initial response). Reasons patients withdrew from the study were: radiographically-confirmed progressive disease (PD), withdrawal of consent, and drug intolerance.

Tumor response in reference to RECIST 1.1 was referred to from the original paper (25). The tumor response according to mRECIST 1.1 was defined as the largest single lesion per organ to be re-evaluated in all patients, and the maximum number of target lesions to be assessed was five (19). For mRECIST 1.1 the response definitions of CR, partial response (PR), stable disease (SD) and PD, were consistent with original RECIST 1.1. The time from the randomization to death or disease progression by any cause, which came first, was determined as the PFS. The time from the randomization to death by any cause was determined as the overall survival (OS).

Evaluation of the RECIST 1.1 and the mRECIST 1.1

We reassessed patients’ objective response rate by mRECIST 1.1, and classified the objective response as modified CR (mCR), modified PR (mPR), modified SD (mSD) and modified PD (mPD). The results of mRECIST 1.1 reassessment were used to calculate the modified PFS and OS, to see if there was a discrepancy between the two criteria, and to evaluate the concordance between the mRECIST 1.1 and the RECIST 1.1. And all patients were divided into mCR, mPR, mSD and mPD groups, and the median OS of the different groups was calculated. We also calculated the median OS of the four groups using the original RECIST 1.1 criteria and made a comparison with mRECIST 1.1. In addition, patients were split into four groups by quartiles using an absolute PFS or modified PFS value, and the OS was compared between the groups.

Statistical analysis

The ORR was evaluated by the RECIST 1.1 and the mRECIST 1.1. The Chi-squared test was utilized to compare the response rate for the two criteria. The kappa coefficient was used to assess the concordance of tumor responses evaluated by RECIST 1.1 and mRECIST 1.1. And a κ value of >0.75 was regarded as showing strong concordance. A paired Student’s t-test was used to evaluated the statistical significance of changes in the number of target lesions at baseline between the two criteria. The concordance index (C-index) was used to evaluate the prognostic performance of treatment response for OS, when using RECIST 1.1 and mRECIST 1.1 guidelines (7,19). And the C-index ranges from 0 to 1, where 0.5 indicates random estimation, 0.51–0.70 denotes low accuracy, 0.71–0.90 represents intermediate accuracy, and 0.91–0.99 reflects high accuracy. The Kaplan-Meier method, and the log-rank test were used to compare survival between responders (patients with CR or PR) and non-responders (patients with SD or PD). All statistical tests were two-sided and a P value<0.05 was considered significant. For statistical analyses, SPSS25 (IBM, Armonk, NY, USA) and R software (version 4.4.3) were used.

Ethical statement

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Institutional Review Board of Sun Yat-sen University Cancer Center’s (SYSUCC) approved this retrospective analysis (No. A2012-043-01) and waived the need for informed consent for the retrospective analysis. In order to protect patients’ privacy, only deidentified data was used in this analysis.

Results

Patient characteristics

One thousand one and fifty-four patients out of 1,255 screened patients were registered and randomized between July 2010 and June 2014, and 577 patients were randomized to each cohort. There were 577 patients received S-1 treatment and 570 patients received docetaxel treatment, included in the FAS. A total of 1,129 patients (docetaxel 560, and S-1 569) were administered an experimental drug. Table 1 shows the well-balanced baseline characteristics between the two cohorts. As of November 20, 2015, there were 487 (85.4%) deaths in the docetaxel cohort, and 479 (83.0%) deaths in the S-1 cohort. The median follow-up time was 30.75 months.

Number of target lesions

Using mRECIST 1.1, the amount of target lesions was 1,776 in 1,147 advanced NSCLC patients, which was less than those assessed by RECIST 1.1, which was 1,932. According to mRECIST 1.1, the median number of target lesions was 1 (range, 1–5), demonstrating a statistically significant reduction compared to the median number of target lesions identified using RECIST 1.1 (2, range, 1–5) (P<0.001). When adopting mRECIST 1.1, no patients had an increase in the number of target lesions, and 153 patients had a reduction in target lesions compared to RECIST 1.1. Compared to the RECIST 1.1, 153 patients (13.3%) demonstrated a reduction in the percentage changes (range, 7.88–50%) of the sum of the tumor measurements. In the 153 patients, the median [range] number of target lesions was 3 [2–5] by the RECIST 1.1 and 2 [1–4] by the mRECIST 1.1, and we found that the tumor response showed excellent level of concordance between the two criteria [κ=0.940; 95% confidence internal (CI): 0.896–0.983].

Tumor responses by mRECIST 1.1 versus RECIST 1.1

The best tumor responses according to the two criteria are compared in Table 2. The assessment of tumor responses showed a significant agreement between the two criteria, as illustrated by the kappa value of 0.989 (95% CI: 0.981–0.997). Only seven patients revealed a discordance between two criteria (Table 2). Besides, with using mRECIST 1.1, rather than using RECIST 1.1, one patient was reclassified from PR to SD, three patients from SD to PR, one patient from SD to PD, and two patients with PD to SD. There was no distinct difference in ORR between the two criteria (9.1% for RECIST 1.1 versus 9.3% for mRECIST 1.1, P=0.94).

The accuracy of mRECIST 1.1 for survival

mRECIST 1.1 reassessment results were used to set the new disease progression, and calculate the modified PFS (Figure 1). The C-index was calculated to compared the prognostic accuracy of the RECIST 1.1 and the mRECIST 1.1 in predicting survival outcomes. The RECIST 1.1 and mRECIST 1.1 showed a similar accuracy for estimation of OS, with C-index values of 0.709 (95% CI: 0.700–0.718) and 0.708 (95% CI: 0.699–0.717), respectively. For responders, C-indexes by the RECIST 1.1 and mRECIST were 0.727 (95% CI: 0.690–0.763) and 0.727 (95% CI: 0.690–0.763), respectively. In the cohort of non-responders, C-indexes by RECIST 1.1 and mRECIST were 0.704 (95% CI: 0.693–0.714) and 0.702 (95% CI: 0.692–0.712), respectively (Figure 2).

Figure 1 Progression-free survival according the RECIST 1.1 versus mRECIST 1.1. mRECIST, modified Response Evaluation Criteria in Solid Tumors; RECIST, Response Evaluation Criteria in Solid Tumors.

Figure 2 Forest plot shows the prognostic value of the RECIST 1.1 and the mRECIST 1.1 for OS. C-index, concordance index; CI, confidence internal; mRECIST, modified Response Evaluation Criteria in Solid Tumors; OS, overall survival; RECIST, Response Evaluation Criteria in Solid Tumors.

The survival analysis showed that OS and PFS of non-responders by the two criteria were significantly shorter than responders, median OS: 12.5 versus 34.1 months for the RECIST 1.1 [hazard ratio (HR) 2.423, 95% CI: 2.006–2.926; P<0.001] and 12.5 versus 31.4 months for the mRECIST 1.1 (HR 2.332, 95% CI: 1.933–2.813; P<0.001); median PFS: 2.8 versus 7.7 months for the RECIST 1.1 (HR 2.367, 95% CI: 2.022–2.770; P<0.001) and 2.8 versus 7.7 months for the mRECIST 1.1 (HR 2.338, 95% CI: 1.999–2.735; P<0.001) (Figure 3).

Figure 3 Kaplan-Meier analysis of OS and PFS. OS (A) and PFS (B) according to the RECIST 1.1. OS (C) and PFS (D) according to the mRECIST 1.1. CI, confidence internal; HR, hazard ratio; mRECIST, modified Response Evaluation Criteria in Solid Tumors; OS, overall survival; PFS, progression-free survival; RECIST, Response Evaluation Criteria in Solid Tumors.

OS analysis in response groups

Patients were classified as a CR and PR group, an SD group, or a PD group, the Kaplan-Meier curve of OS according to the two criteria is shown in Figure 4. The median OS in reference to RECIST 1.1 was 33.6 months (95% CI: 23.8–40.8), 18.6 months (95% CI: 16.5–20.3) and 7.3 months (95% CI: 6.7–8.5) in the CR and PR group, SD group and PD group, respectively (Figure 4A). Based on mRECIST 1.1 reassessment, the median OS was 30.9 months (95% CI: 23.8–40.0), 18.3 months (95% CI: 16.5–20.2) and 7.3 months (95% CI: 6.7–8.5) in the mCR and mPR group, mSD group and mPD group, respectively (Figure 4B). The log-rank test revealed that the P values were not significant in median OS between the PD group and mPD group (P=0.98), the SD group and the mSD group (P=0.94), CR and PR group and mCR and mPR group (P=0.83).

Figure 4 Kaplan-Meier curve of OS according to RECIST 1.1 (A) and mRECIST 1.1 (B). CR, complete response; mCR, modified complete response; mPD, modified progressive disease; mPR, modified partial response; mRECIST, modified Response Evaluation Criteria in Solid Tumors; mSD, modified stable disease; OS, overall survival; PD, progressive disease; PR, partial response; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease.

OS surrogate value comparison of mRECIST 1.1

According to the PFS evaluated by the two criteria respectively, patients were classified into four groups basing on quartiles of the absolute value of PFS. The Kaplan-Meier curve of OS of the four groups is shown in Figure 5. The median OS according to the RECIST 1.1 was 6.0 months (95% CI: 4.8–7.1), 8.4 months (95% CI: 7.0–9.2), 14.6 months (95% CI: 13.0–16.5) and 25.6 months (95% CI: 23.7–30.5) in the lower quartile group, second quartile group, third quartile group and fourth quartile group, respectively (Figure 5A). According to the mRECIST 1.1 reassessment results, the median OS was 5.8 months (95% CI: 4.8–7.1), 8.3 months (95% CI: 7.0–9.2), 14.5 months (95% CI: 13.0–16.4) and 25.6 months (95% CI: 23.7–30.5) in the lowest, 2^nd, 3^rd and 4^th quartile groups respectively (Figure 5B). The log-rank test revealed that the P values were not significant in median OS between the lower quartile group (P=0.98), the second quartile group (P=0.94), third quartile group (P=0.94) and fourth quartile group (P=0.97) according to RECIST 1.1 and mRECIST 1.1.

Figure 5 Kaplan-Meier curve of OS by PFS based on RECIST 1.1 (A) and mRECIST 1.1 (B). mRECIST, modified Response Evaluation Criteria in Solid Tumors; OS, overall survival; PFS, progression-free survival; RECIST, Response Evaluation Criteria in Solid Tumors.

Discussion

Here, we examined whether mRECIST 1.1 might be used as an alternative to RECIST 1.1 for evaluating tumor responses among NSCLC patients. mRECIST 1.1 measures the single largest lesion per organ, whereas, RECIST 1.1 assesses two target lesions per organ, and the maximum number of target lesions to be assessed was five in total for the two criteria. Our results found no marked difference in ORR and PFS between two criteria. The C-index by these two criteria was similar for OS. Furthermore, we classified patients into four groups based on quartiles of absolute PFS values. Interestingly, the median OS when using mRECIST 1.1 was similar to the median OS calculated when using RECIST 1.1. This suggests that mRECIST 1.1 yields comparable outcomes to RECIST 1.1 and may be advantageous insofar as convenience and time saving when assessing tumor response in advanced NSCLC. Our findings are consistent with several other studies (21-24,26). One study that compared the RECIST 1.1 and the mRECIST 1.1 in advanced NSCLC patients who were administered first-line chemotherapy, found no marked difference between ORR and disease control rate (DCR) regarding the two criteria. Other similar retrospective studies that assessed patients with SCLC, GC, and CRC, found that mRECIST 1.1 was similar with RECIST 1.1 when evaluating tumor response (21,23,24,26). All of the above studies concluded that there is a high degree of agreement between the mRECIST 1.1 and the RECIST 1.1 in evaluating tumor response. However, the sample size of these studies was small, lacked a balanced control group and did not compare OS.

Compared with previous small sample research, to improve statistical power, we selected a larger sample size comprising 1,147 patients from the EAST-LC clinical trial to confirm the precise use of mRECIST 1.1. Our investigation was based on complete data from a structured clinical trial with high reliability, in which we revealed that the only a discrepancy between mRECIST 1.1 and RECIST 1.1 was regarding the tumor response of only seven patients. And that there was no statistical difference in PFS and ORR between the two criteria. The 1,147 NSCLC patients all received chemotherapy and had good comparability, full patient survival data, and two balanced and homogenous cohorts from the large-scale clinical trial, which are crucial for reducing statistical bias.

Our study revealed that mRECIST 1.1 was comparable to RECIST 1.1, and that mRECIST 1.1 plays a vital role in accurate and reproducible tumor response assessment, which provides more precise surrogate endpoints in clinical medicine. Compared with RECIST 1.1, mRECIST 1.1 measures the largest single lesion per organ which may reduce the intra-observer and inter-observer variability, and increase reproducibility. As the RECIST 1.1 relies heavily on the variation of tumor size to reflect anti-tumor therapy effectiveness, the target lesion is the key point of tumor assessment (12,15). In the clinic however, not all tumor lesions respond uniformly to anticancer therapy, as evidenced by varied changes in tumor size after receiving treatment (27-29). Hence, measuring all target lesions is idealistic in the assessment of anticancer reactions, which is time consuming (19). With this in mind, the selection of an adequate amount of target lesions that precisely indicate how the whole-body tumor load and long-term survival change is necessary and urgently required. As shown in our results, and in other investigations, assessing the single largest lesion in each organ seems to be sufficient in predicting OS. In summary, our data supports that evaluating the single largest lesion per organ can potentially improve the convenience and practicability of RECIST 1.1.

Additionally, it is reported that tumor measurement is one of critical factors for contributing to increased workload (30,31). Compare to the RECIST 1.0, the RECIST 1.1, which reduced the maximum number of target lesions from ten to five, showed minimal impact on outcome, along with requiring less evaluation time and reducing workload (5-7). In our study, the number of target lesions for mRECIST 1.1 is lower than that for RECIST 1.1, which suggests that the mRECIST 1.1 has the potential to reduce the workload for radiologists in clinical practice.

There are several limitations in this study. Firstly, this was a retrospective study only in NSCLC patients without performing more subgroup analyses, which need to be validated by more patients with primary cancer in other sites in prospective multi-centers clinical trials and future real-world studies. Secondly, patients in this study received ≥second-line treatment with chemotherapy. While, several developed target agents can’t result in tumor shrinkage and can induce cystic change and necrosis (32,33). AS the primary outcome in this study, OS could be confounded by various factors (such as previous and post systemic treatment regimens). The mRECIST 1.1 needs to be verified in treatment-naïve patients receiving targeted therapies and other novel treatment options to minimize such biases. Thirdly, given the retrospective nature of this study, the analyses of special scenarios such as pseudo-progression and mixed tumor response were not feasible. Future investigations exploring and validating these phenomena in affected patient cohorts would be of significant interest. However, we are confident that our investigation has clinical significance in the age of precision therapy and is worthy of further exploratory analysis.

Conclusions

In sum, assessing NSCLC tumor response using mRECIST 1.1 is comparable with the RECIST 1.1. Our results reveal that it might be advisable to assess tumor response by evaluating the single largest lesion in each organ in the clinic. Subsequent directions may involve confirming the ability of the mRECIST 1.1 to evaluate the anti-tumor response among different types of treatment. The potential of mRECIST 1.1 in other cancers needs to be explored in more prospective trials with larger samples.

Acknowledgments

We are grateful to the patients and their families who were included in this retrospective study. We also thank Taiho for conducting the analysis and providing the relevant data in this study.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-809/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-809/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-809/prf

Funding: This work was supported by the National Natural Science Foundation (No. 82473346 to Hongyun Zhao, Nos. 82272789 and 82241232 to L.Z.), the Guangzhou Key Research and Development Plan (No. 202206010141 to Hongyun Zhao), the Guangdong Basic and Applied Basic Research Foundation (No. 2021A1515110697 to W.H.), the Young Talents program of Sun Yat-sen University Cancer Center (No. YTP-SYSUCC-0094 to Y.M.), the Cancer Innovative Research Program of Sun Yat-sen University Cancer Center (No. CIRP-SYSUCC-0028 to Hongyun Zhao). And the EAST-LC clinical trial was sponsored by Taiho Pharmaceuticals Co., Ltd.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-809/coif). All authors report that the EAST-LC clinical trial was sponsored by Taiho Pharmaceuticals Co., Ltd. Also, Taiho assisted in conducting the analysis and providing the relevant data in this study. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Institutional Review Board of Sun Yat-sen University Cancer Center’s (SYSUCC) approved this retrospective analysis (No. A2012-043-01) and waived the need for informed consent for the retrospective analysis. In order to protect patients’ privacy, only deidentified data was used in this analysis.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. 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;45:228-47. [Crossref] [PubMed]
2. Bogaerts J, Ford R, Sargent D, et al. Individual patient data analysis to assess modifications to the RECIST criteria. Eur J Cancer 2009;45:248-60. [Crossref] [PubMed]
3. Schwartz LH, Bogaerts J, Ford R, et al. Evaluation of lymph nodes with RECIST 1.1. Eur J Cancer 2009;45:261-7. [Crossref] [PubMed]
4. Dancey JE, Dodd LE, Ford R, et al. Recommendations for the assessment of progression in randomised cancer treatment trials. Eur J Cancer 2009;45:281-9. [Crossref] [PubMed]
5. Litière S, Collette S, de Vries EG, et al. RECIST - learning from the past to build the future. Nat Rev Clin Oncol 2017;14:187-92. [Crossref] [PubMed]
6. McBride D. New criteria for measuring tumor size ease workloads in clinical trials. ONS Connect 2009;24:21.
7. He LN, Chen T, Fu S, et al. Reducing number of target lesions for RECIST1.1 to predict survivals in patients with advanced non-small-cell lung cancer undergoing anti-PD1/PD-L1 monotherapy. Lung Cancer 2022;165:10-7. [Crossref] [PubMed]
8. Gouel P, Callonnec F, Levêque É, et al. Evaluation of the capability and reproducibility of RECIST 1.1. measurements by technologists in breast cancer follow-up: a pilot study. Sci Rep 2023;13:9148. [Crossref] [PubMed]
9. Fournier L, de Geus-Oei LF, Regge D, et al. Twenty Years On: RECIST as a Biomarker of Response in Solid Tumours an EORTC Imaging Group - ESOI Joint Paper. Front Oncol 2021;11:800547. [Crossref] [PubMed]
10. Zamboni GA, Cappello G, Caruso D, et al. ESR Essentials: response assessment criteria in oncologic imaging-practice recommendations by the European Society of Oncologic Imaging. Eur Radiol 2025;35:674-83. [Crossref] [PubMed]
11. Li H, Shen J, Shou J, et al. Exploring the Interobserver Agreement in Computer-Aided Radiologic Tumor Measurement and Evaluation of Tumor Response. Front Oncol 2021;11:691638. [Crossref] [PubMed]
12. Iannessi A, Beaumont H, Ojango C, et al. RECIST 1.1 assessments variability: a systematic pictorial review of blinded double reads. Insights Imaging 2024;15:199. [Crossref] [PubMed]
13. Dahm IC, Kolb M, Altmann S, et al. Reliability of Automated RECIST 1.1 and Volumetric RECIST Target Lesion Response Evaluation in Follow-Up CT-A Multi-Center, Multi-Observer Reading Study. Cancers (Basel) 2024;16:4009. [Crossref] [PubMed]
14. Zimmermann M, Kuhl C, Engelke H, et al. Volumetric measurements of target lesions: does it improve inter-reader variability for oncological response assessment according to RECIST 1.1 guidelines compared to standard unidimensional measurements? Pol J Radiol 2021;86:e594-600. [Crossref] [PubMed]
15. Tareco Bucho TM, Petrychenko L, Abdelatty MA, et al. Reproducing RECIST lesion selection via machine learning: Insights into intra and inter-radiologist variation. Eur J Radiol Open 2024;12:100562. [Crossref] [PubMed]
16. Kuhl CK, Alparslan Y, Schmoee J, et al. Validity of RECIST Version 1.1 for Response Assessment in Metastatic Cancer: A Prospective, Multireader Study. Radiology 2019;290:349-56. [Crossref] [PubMed]
17. Kuhl CK, Alparslan Y, Sequeira B, et al. Effect of target lesions selection on between-reader variability of response assessment according to RECIST 1.1. J Clin Oncol 2017;35:2528.
18. Tareco Bucho TM, Tissier RLM, Groot Lipman KBW, et al. How Does Target Lesion Selection Affect RECIST? A Computer Simulation Study. Invest Radiol 2024;59:465-71. [Crossref] [PubMed]
19. He LN, Chen T, Fu S, et al. Tumor response assessment by measuring the single largest lesion per organ in advanced non-small cell lung cancer patients treated with PD-1/PD-L1 inhibitor. Ther Adv Med Oncol 2023;15:17588359231200463. [Crossref] [PubMed]
20. Zacharia TT, Saini S, Halpern EF, et al. CT of colon cancer metastases to the liver using modified RECIST criteria: determining the ideal number of target lesions to measure. AJR Am J Roentgenol 2006;186:1067-70. [Crossref] [PubMed]
21. Kim HS, Kim JH. Ideal number of target lesions per organ to measure in metastatic colorectal cancer. Oncol Lett 2014;8:1896-900. [Crossref] [PubMed]
22. Kim HS, Kim JH, Yang I. Tumor response assessment by measuring the single largest lesion per organ in patients with advanced non-small cell lung cancer. Lung Cancer 2014;85:385-9. [Crossref] [PubMed]
23. Kim HS, Kim JW, Kim JH, et al. Single-lesion measurement per organ for assessing tumor response in advanced gastric cancer. Oncology 2015;88:69-75. [Crossref] [PubMed]
24. Jung SG, Kim JH, Kim HS, et al. Tumor response assessment by the single-lesion measurement per organ in small cell lung cancer. Chin J Cancer Res 2016;28:161-7. [Crossref] [PubMed]
25. Nokihara H, Lu S, Mok TSK, et al. Randomized controlled trial of S-1 versus docetaxel in patients with non-small-cell lung cancer previously treated with platinum-based chemotherapy (East Asia S-1 Trial in Lung Cancer). Ann Oncol 2017;28:2698-706. [Crossref] [PubMed]
26. Jang HJ, Cho JW, Park B, et al. The assessment of tumor response by measuring the single largest lesion per organ in metastatic tumors: a pooled analysis of previously reported data. J Cancer 2015;6:169-76. [Crossref] [PubMed]
27. Zheng X, Lu T, Wu S, et al. A novel approach to evaluation of tumor response for advanced pulmonary adenocarcinoma using the intertumoral heterogeneity response score. MedComm (2020) 2024;5:e493.
28. Ou FS, Ahn DH, Dixon JG, et al. Evaluation of Intratumoral Response Heterogeneity in Metastatic Colorectal Cancer and Its Impact on Patient Overall Survival: Findings from 10,551 Patients in the ARCAD Database. Cancers (Basel) 2023;15:4117. [Crossref] [PubMed]
29. Kumar R, Qi T, Cao Y, et al. Incorporating lesion-to-lesion heterogeneity into early oncology decision making. Front Immunol 2023;14:1173546. [Crossref] [PubMed]
30. Roche K, Paul N, Smuck B, et al. Factors affecting workload of cancer clinical trials: results of a multicenter study of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 2002;20:545-56. [Crossref] [PubMed]
31. Husband JE, Schwartz LH, Spencer J, et al. Evaluation of the response to treatment of solid tumours - a consensus statement of the International Cancer Imaging Society. Br J Cancer 2004;90:2256-60. [Crossref] [PubMed]
32. Bai JW, Qiu SQ, Zhang GJ. Molecular and functional imaging in cancer-targeted therapy: current applications and future directions. Signal Transduct Target Ther 2023;8:89. [Crossref] [PubMed]
33. Castagnoli F, Mencel J, Ap Dafydd D, et al. Response Evaluation Criteria in Gastrointestinal and Abdominal Cancers: Which to Use and How to Measure. Radiographics 2024;44:e230047. [Crossref] [PubMed]