AbstractPurposeThe aim of this study was to assess the role of the shear-wave velocity (SWV) value in predicting chemotherapeutic response and progression-free survival (PFS) in patients with colorectal cancer liver metastasis (CRLM).
MethodsIn this prospective single-center study, participants with CRLM scheduled for chemotherapy were enrolled between May 2018 and June 2021. SWV measurements were obtained using shear-wave elastography at the CRLM site before and after initiating chemotherapy. Based on the Response Evaluation Criteria in Solid Tumors, the participants were categorized by chemotherapeutic response into responders (complete remission and partial remission) and non-responders (stable disease and progressive disease). Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the performance of changes in SWV measurements in predicting the chemotherapeutic response of CRLM. In addition, a Cox proportional hazards model was used to identify variables associated with PFS.
ResultsIn total, 67 participants (40 men; mean age, 62.3±10.1 years) were enrolled, including 34 responders and 33 non-responders. The area under the ROC curve, sensitivity, and negative predictive value of the SWV measurement in predicting non-responders were 0.840, 97.0%, and 95.2%, respectively, using a cutoff value of a 13% decrease. Additionally, a change in SWV values was independently associated with PFS (hazard ratio, 1.020), non-responder status, and the presence of five or more CRLMs.
IntroductionColorectal cancer is the third most commonly diagnosed malignancy and the second leading cause of cancer-related deaths worldwide [1]. Over half of patients with colorectal cancer develop colorectal cancer liver metastasis (CRLM) during the course of their disease [2]. The standard treatment for patients with CRLM is chemotherapy [3], and evaluating the treatment response to chemotherapy is important for the selection of an optimal chemotherapeutic regimen to improve patient prognosis. Traditionally, the Response Evaluation Criteria In Solid Tumors (RECIST) standard has been used to assess treatment response to chemotherapy, relying on morphological criteria based on changes in tumor size [4]. However, among patients who receive chemotherapy, morphological changes such as tumor size reduction as assessed by RECIST have been found to occur relatively late, while functional or molecular changes often occur earlier [5,6].
Efforts have been made to overcome the limitations of RECIST and assess treatment response earlier than changes in tumor size may occur, with shear-wave elastography (SWE) reportedly showing promise in this area [7-9]. SWE is a quantitative ultrasound technique that can be used to evaluate tissue stiffness by measuring shear-wave velocity (SWV) [10]. Since shear waves propagate at different velocities in tissues with varying stiffness, SWE has been widely used for characterizing tissues, such as grading liver fibrosis or differentiating malignant from benign tumors [11,12]. When patients with CRLM receive chemotherapy, changes such as tumor necrosis, desmoplastic reaction, or degeneration within CRLM occur [13]. These changes can affect SWV values, and changes in SWV measurements may have the potential to predict treatment response and prognosis in patients with CRLM. The usefulness of SWE in predicting treatment response has been investigated primarily in patients with breast cancer, as the breast is a superficial organ that can be easily assessed using SWE [7,8]. A recent study by Lee et al. [9] demonstrated that SWE can also be applied to the liver, similar to its application in the breast, and indicated that a decrease in SWE predicts better progression-free survival (PFS) in patients with CRLM. However, that study had a small sample size (n=25) and was not a registered clinical trial. Therefore, this study aimed to assess the prognostic role of SWV values for predicting chemotherapeutic response and PFS in a larger cohort of patients with CRLMs.
Materials and MethodsCompliance with Ethical StandardsThis prospective single-center study was approved by the institutional review board of our institution (H-1704-027-843), and informed consent was obtained from all participants. Additionally, this study is registered with the Clinical Research Information Service of Korea (KCT0004611).
ParticipantsBetween May 2018 and June 2021, study participants were consecutively referred from the medical oncology department and enrolled based on the following inclusion criteria: (1) a diagnosis of CRLM confirmed by either pathological or imaging examination [14,15], (2) measurable CRLM(s) on computed tomography (CT) or magnetic resonance imaging, (3) scheduled for chemotherapy treatments, and (4) no previous history of systemic chemotherapy for CRLM or complete remission of CRLM achieved through chemotherapy confirmed at least 1 month prior (Fig. 1). The exclusion criteria included (1) inability to obtain reliable tumor SWV measurements, (2) loss to follow-up, and (3) withdrawal of consent. All participants underwent contrast-enhanced abdominal and chest CT examinations to obtain baseline images.
SWE ProtocolBefore the SWE examination, a board-certified radiologist (J.S.B., with 10 years of experience in abdominal imaging) reviewed baseline CT images to identify CRLMs suitable for SWE examination. The same radiologist (J.S.B.) performed point SWE using acoustic radiation force impulse on an ultrasound machine (Acuson S2000, Siemens Healthineers, Erlangen, Germany) equipped with a 6-MHz convex probe. To obtain reliable SWV measurements, a representative CRLM for each patient was selected based on these criteria: (1) larger than 1 cm in diameter; (2) free from artifacts such as motion or reverberation; and (3) well-defined, allowing for reliable follow-up SWV measurement during chemotherapy. Subsequently, SWV measurements were made by assigning regions of interest within the tumor [11,16]. For each tumor, SWV measurements were taken at the periphery of the tumor (Fig. 2). The periphery was chosen for assessment because metastases like CRLM often include central fibrosis or necrosis but contain viable cells at the periphery [17], making chemotherapy-induced changes more pronounced in peripheral areas. SWV measurements were taken 10 times per tumor, and the median SWV measurement was used as a representative value [9,10,18,19]. In each participant, SWV measurements were considered reliable when the interquartile range-to-median ratio did not exceed 15% [20]. SWV measurements were performed twice. The first measurement was taken on the same day as chemotherapy, but before its initiation. The second occurred 48 to 96 hours after the initiation of chemotherapy [9,13]. The 48- to 96-hour time window was chosen to allow flexibility in case a second SWV measurement could not be performed on day 2 due to administrative issues, such as scheduling follow-up appointments for participants. Whenever possible, the same site of the same CRLM was chosen for the subsequent SWV measurement. The change in SWV values between the two sets of measurements (i.e.,
ChemotherapyThe chemotherapeutic regimen utilized was either FOLFOX (5-fluorouracil, leucovorin, and oxaliplatin), which is a combination of infusional 5-fluorouracil, leucovorin, and oxaliplatin, or FOLFIRI (5-fluorouracil, leucovorin, and irinotecan), a combination of infusional 5-fluorouracil, leucovorin, and irinotecan. At the discretion of the referring physician, bevacizumab or cetuximab was added (Table 1). Each regimen was administered every 14 days, and all participants continued with their respective chemotherapeutic regimens until disease progression, complete remission, or hepatic resection occurred.
OutcomesAll participants were scheduled to undergo follow-up contrast-enhanced abdominal and chest CT examinations 8 to 12 weeks after chemotherapy. Treatment responses to chemotherapy on these follow-up CT examinations were evaluated by the same radiologist (J.S.B.) according to the revised RECIST, version 1.1 [4]. Changes in tumor burden, including CRLMs, were assessed as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD) to determine the patient’s response to chemotherapy. In addition, patients were also classified into either the responder group (CR or PR) or the non-responder group (SD or PD). Follow-up CT images were also evaluated to identify the presence of chemotherapy-associated liver injury, such as sinusoidal obstruction syndrome or steatosis [21].
The 12-week follow-up period for CT examination may be insufficient to fully evaluate the treatment response to chemotherapy in patients with CRLM. Thus, this study also assessed PFS, which was defined as the time interval between the initiation of chemotherapy and disease progression or death, whichever occurred first. Participants without disease progression or death were censored at the date of their last assessment before October 25, 2022. Follow-up data for PFS evaluation were obtained by reviewing electronic medical records from our institution and by accessing national statistical data from the Korean Ministry of Government Administration and Home Affairs. All participants were followed up with every 3 months for the first 2 years and every 6 months for an additional 3 years. During each visit, participants underwent a physical examination, contrast-enhanced abdominal and chest CT examinations, and laboratory investigations. The results of the abdominal and chest CT examinations were interpreted according to the revised RECIST, version 1.1.
Statistical AnalysisThe sample size was determined based on a previous study assessing SWV for CRLMs in predicting treatment response to chemotherapy [9]. In that study, the average SWV measurement before chemotherapy was 2.85 m/s, and the average difference in SWV values before and after chemotherapy was 0.17 m/s. Assuming a type I error of 0.05, a type II error of 0.20, and a standard deviation of 0.25, a minimum of 33 participants were required. The response rate of CRLM to first-line chemotherapy has been reported to range from 48%-54% [22,23]. With an assumed 50% response rate for CRLM and a 12% drop-out rate, the final sample size was calculated to be 75.
The independent t-test or Wilcoxon rank-sum test was employed to compare continuous variables, while the chi-square test or Fisher exact test was utilized for categorical variables, as appropriate. Univariable and multivariable logistic regression analyses were conducted to identify variables associated with treatment response. Receiver operating characteristic curve analysis was carried out to assess the predictive performance of changes in SWV measurements for distinguishing non-responders from responders. Cutoff values for changes in SWV values were determined using the Youden index and by maximizing specificity for detecting non-responders. PFS curves were computed using the Kaplan-Meier method. Univariable and multivariable Cox proportional hazards models were employed to identify predictors of poor PFS. Furthermore, subgroup analyses were conducted based on the type of chemotherapeutic regimens. Statistical analyses were performed using commercially available software (SPSS version 27, IBM Corp., Armonk, NY, USA; Medcalc, Medcalc Software Ltd., Ostend, Belgium). A P-value of <0.050 was considered to indicate a statistically significant difference.
ResultsParticipantsOf the 75 participants who met the inclusion criteria, eight were excluded due to unreliable SWV measurements (n=5), loss to follow-up (n=2), or withdrawal of consent (n=1). Ultimately, 67 participants (mean age, 62.3±10.1 years; 40 men) were included (Fig. 1). The characteristics of the participants are detailed in Table 1. The majority of participants had synchronous CRLMs (92.5% [62/67]), with a median CRLM size of 3.6 cm (range, 2.1 to 5.1 cm). Most SWV measurements were conducted on CRLMs located in the right lobe of the liver (86.6% [58/67]). The mean SWV measurement depth was 4.6±1.2 cm. Regarding treatment response, CR was achieved in one participant (1.5%), PR in 33 participants (49.3%), SD in 31 participants (46.2%), and PD in two participants (3.0%). Based on the binary categorization, the participants included 34 responders (50.8%) and 33 non-responders (49.2%). The mean interval between baseline and follow-up CT examinations was 69.6±11.8 days. No participants exhibited chemotherapy-associated liver injury on follow-up CT examinations.
SWV MeasurementsThe median initial SWV value of the CRLMs was 2.49 m/s (range, 0.57 to 4.55 m/s). Table 2 displays the SWV values and changes between the first and second sets of CRLMs by participant group. The SWV values did not differ significantly between the two groups for either the first or second set of measurements. However, the SWV measurements substantially decreased in responders, whereas they increased in non-responders, reflecting a significant between-group difference (-18.4% vs. 7.3%; P<0.001). A comparison between participants with increased SWV values and those with decreased SWV values revealed that non-responders were more prevalent among participants with increased measurements (Supplementary Table 1). This pattern of changes in SWV values was also observed in subgroup analyses according to the type of chemotherapeutic regimen (Supplementary Tables 2, 3).
Prediction of Treatment ResponseThe sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of change in SWV values for the prediction of non-responders were 97.0% (32/33; 95% confidence interval [CI], 84.2% to 100.0%), 58.8% (20/34; 95% CI, 40.7% to 75.4%), 69.6% (32/46; 95% CI, 60.4% to 77.4%), 95.2% (20/21; 95% CI, 74.0% to 99.3%), and 77.6% (52/67; 95% CI, 65.8% to 88.9%), respectively. The corresponding area under the curve was 0.840 (95% CI, 0.731 to 0.919), with a cutoff value of change in SWV values of -13.0% obtained using the Youden index (Fig. 3). Utilizing the cutoff value of 10.5% that was calculated to maximize specificity, a change in SWV values predicted non-responders with sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 42.4% (14/33; 95% CI, 25.5% to 60.8%), 97.1% (1/33; 95% CI, 84.7% to 99.9%), 93.3% (14/15; 95% CI, 66.1% to 99.0%), 63.5% (33/52; 95% CI, 56.3% to 70.1%), and 70.1% (47/67; 95% CI, 57.7% to 80.7%), respectively. Subgroup analyses based on chemotherapeutic agents also revealed that the change in SWV values demonstrated an area under the curve of 0.798-0.983, sensitivity of 92.3%-100.0%, and negative predictive value of 90.0%-100.0% (Supplementary Table 4).
PFS after Chemotherapy for CRLMsDuring the median follow-up period of 9 months (range, 0 to 48 months), 53 patients experienced disease progression, and three patients died. The estimated 1-, 2-, and 3-year PFS rates after chemotherapy were 42.9%, 12.0%, and 10.0%, respectively. The prognostic factors associated with PFS are presented in Table 3. Univariable and multivariable Cox proportional hazards regression analyses revealed that a change in the SWV value after chemotherapy, relative to the initial SWV value, was an independent predictor of PFS (hazard ratio [HR], 1.020; 95% CI, 1.002 to 1.037; P=0.029). Non-responder status (i.e., SD or PD) (HR, 2.861; 95% CI, 1.408 to 5.817; P=0.004) and having five or more CRLMs (HR, 2.550; 95% CI, 1.125 to 5.779; P=0.025) were also associated with PFS. The results of subgroup analyses are presented in Supplementary Tables 5-9.
DiscussionThis study found a sensitivity of 97.0% and a negative predictive value of 95.2% for predicting non-responders, with a cutoff value of a 13.0% decrease in SWV values in patients with CRLM. By employing a cutoff value of a 10.5% increase in SWV values, non-responders were predicted with specificity and positive predictive values of 97.1% and 93.3%, respectively. This predictive performance of SWV values can be attributed to the change in tumor stiffness caused by chemotherapy. During chemotherapy, cellular damage occurs, leading to tumor necrosis [24,25]. As a result, the cellular density of the tumor decreases, which can cause the tumor to soften, as detected by a decrease in SWV values. In other words, a decrease in SWV measured in the CRLM indicates that the tumor has become softer due to necrosis development after chemotherapy, implying that the chemotherapeutic regimen was effective. Furthermore, non-responders were more prevalent among participants with increased SWV values. This early prediction of non-responders to chemotherapy treatment could benefit patients with CRLM by providing the opportunity to try alternative chemotherapeutic regimens that may be effective, rather than continuing an initial regimen that would be ineffective at the follow-up examinations performed approximately 2 months later. Therefore, the authors cautiously suggest that a change in SWV values after starting chemotherapy in patients with CRLM has the potential to be used as a biomarker to predict the response to chemotherapy, which can aid in earlier assessments of chemoresponsiveness. In addition, the change in SWV values differed significantly between responders and non-responders in our study (-18.4% vs. 7.3%; P<0.001). This result contradicts that of a previous study by Lee et al. [9], which reported no significant association between changes in SWV values during chemotherapy and treatment response. The reason for this discrepancy may be due to the different methods of analyzing the change in SWV values: our study used the exact values of SWV changes during chemotherapy, whereas the study by Lee et al. [9] dichotomized changes in SWV values as decreased or increased. Furthermore, the present sample size was larger than that in the previous study (67 vs. 25) and was derived by calculating a statistically reasonable number of participants, which would support the higher validity of our results.
Significant changes in tumor burden, such as PD, may occur over only a single cycle of chemotherapy according to the revised RECIST 1.1 criteria. However, if a significant change takes place over an extended period, the treatment response after only one cycle of chemotherapy would be assessed as SD, not PD. To account for this limitation in evaluating treatment response after a single chemotherapy cycle, this study assessed PFS, which could reflect a slower response to chemotherapy. A change in SWV values during chemotherapy was an independent variable associated with PFS (HR, 1.020; P=0.029), as well as response in the initial cycle after chemotherapy and the presence of five or more CRLMs. This finding highlights the relatively long-term prognostic value of changes in SWV values for patients with CRLM undergoing chemotherapy and is consistent with the results of a previous study [9].
Liver injury induced by chemotherapy, such as sinusoidal obstruction syndrome, has the potential to impact liver stiffness, which in turn can influence SWV measurements in tumors [21]. However, it is important to acknowledge a limitation of this study concerning the absence of an assessment of the effect of changes in background liver stiffness associated with chemotherapy. The authors recognize that liver injury resulting from chemotherapy is commonly observed in patients undergoing longer treatment regimens, typically encompassing six cycles [26,27]. In this study, the second SWV measurement was performed within 48 to 96 hours after chemotherapy, representing an early post-chemotherapy period. Consequently, the potential impact of chemotherapy on liver stiffness in our study may not have been significant. Additionally, follow-up CT images were examined to determine whether participants had liver injury, although the authors acknowledge that subtle injuries might not be detected on CT.
The present study had a few limitations. First, it was a single-center study, and SWV measurements were performed only once for each set, before and after chemotherapy, by a single radiologist. As a result, it was not possible to assess interobserver and intraobserver agreement. Second, pathological correlations were not conducted because obtaining tissue samples of CRLMs in patients undergoing chemotherapy is not feasible in real-world clinical practice. Third, only one representative CRLM was evaluated for each participant rather than all CRLMs, which could introduce selection bias. However, performing SWV measurements for all CRLMs would be time-consuming and may not be feasible in real-world clinical practice. Lastly, there was heterogeneity in the chemotherapeutic regimens used.
In conclusion, an observed change in SWV values during the early stages of chemotherapy showed significant diagnostic potential for predicting non-responders in patients with CRLM. Additionally, the change in SWV value was independently correlated with PFS.
NotesAuthor Contributions Conceptualization: Bae JS, Lee JY, Lee DH. Data acquisition: Bae JS, Lee JY, Lee DH, Han SW, Lim Y, Kim TY. Data analysis or interpretation: Bae JS, Lee JY, Lee DH. Drafting of the manuscript: Bae JS, Lee JY. Critical revision of the manuscript: Bae JS, Lee JY, Lee DH, Han SW, Lim Y, Kim TY. Approval of the final version of the manuscript: all authors. Conflict of InterestDong Ho Lee serves as Editor for the Ultrasonography, but has no role in the decision to publish this article. All remaining authors have declared no conflicts of interest. References1. Keum N, Giovannucci E. Global burden of colorectal cancer: emerging trends, risk factors and prevention strategies. Nat Rev Gastroenterol Hepatol 2019;16:713–732.
2. Manfredi S, Lepage C, Hatem C, Coatmeur O, Faivre J, Bouvier AM. Epidemiology and management of liver metastases from colorectal cancer. Ann Surg 2006;244:254–259.
3. Sperti E, Faggiuolo R, Gerbino A, Magnino A, Muratore A, Ortega C, et al. Outcome of metastatic colorectal cancer: analysis of a consecutive series of 229 patients. The impact of a multidisciplinary approach. Dis Colon Rectum 2006;49:1596–1601.
4. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228–247.
5. Koh DM, Padhani AR. Diffusion-weighted MRI: a new functional clinical technique for tumour imaging. Br J Radiol 2006;79:633–635.
6. Lee DH, Kim SH, Im SA, Oh DY, Kim TY, Han JK. Multiparametric fully-integrated 18-FDG PET/MRI of advanced gastric cancer for prediction of chemotherapy response: a preliminary study. Eur Radiol 2016;26:2771–2778.
7. Jing H, Cheng W, Li ZY, Ying L, Wang QC, Wu T, et al. Early evaluation of relative changes in tumor stiffness by shear wave elastography predicts the response to neoadjuvant chemotherapy in patients with breast cancer. J Ultrasound Med 2016;35:1619–1627.
8. Gu J, Polley EC, Denis M, Carter JM, Pruthi S, Gregory AV, et al. Early assessment of shear wave elastography parameters foresees the response to neoadjuvant chemotherapy in patients with invasive breast cancer. Breast Cancer Res 2021;23:52.
9. Lee DH, Lee JY, Joo I, Kim TY, Han SW, Lee KH. Shear-wave velocity for colorectal cancer liver metastases as a potential prognostic factor after chemotherapy: a preliminary study. Clin Radiol 2021;76:224–232.
10. Dietrich CF, Bamber J, Berzigotti A, Bota S, Cantisani V, Castera L, et al. EFSUMB guidelines and recommendations on the clinical use of liver ultrasound elastography, update 2017 (long version). Ultraschall Med 2017;38:e16–e47.
11. Kim JE, Lee JY, Bae KS, Han JK, Choi BI. Acoustic radiation force impulse elastography for focal hepatic tumors: usefulness for differentiating hemangiomas from malignant tumors. Korean J Radiol 2013;14:743–753.
12. Zheng J, Guo H, Zeng J, Huang Z, Zheng B, Ren J, et al. Two-dimensional shear-wave elastography and conventional US: the optimal evaluation of liver fibrosis and cirrhosis. Radiology 2015;275:290–300.
13. Cui Y, Zhang XP, Sun YS, Tang L, Shen L. Apparent diffusion coefficient: potential imaging biomarker for prediction and early detection of response to chemotherapy in hepatic metastases. Radiology 2008;248:894–900.
14. Danet IM, Semelka RC, Leonardou P, Braga L, Vaidean G, Woosley JT, et al. Spectrum of MRI appearances of untreated metastases of the liver. AJR Am J Roentgenol 2003;181:809–817.
15. Lee KH, Lee JM, Park JH, Kim JH, Park HS, Yu MH, et al. MR imaging in patients with suspected liver metastases: value of liver-specific contrast agent gadoxetic acid. Korean J Radiol 2013;14:894–904.
16. Friedrich-Rust M, Wunder K, Kriener S, Sotoudeh F, Richter S, Bojunga J, et al. Liver fibrosis in viral hepatitis: noninvasive assessment with acoustic radiation force impulse imaging versus transient elastography. Radiology 2009;252:595–604.
17. Consul N, Sirlin CB, Chernyak V, Fetzer DT, Masch WR, Arora SS, et al. Imaging features at the periphery: hemodynamics, pathophysiology, and effect on LI-RADS categorization. Radiographics 2021;41:1657–1675.
18. Takahashi H, Ono N, Eguchi Y, Eguchi T, Kitajima Y, Kawaguchi Y, et al. Evaluation of acoustic radiation force impulse elastography for fibrosis staging of chronic liver disease: a pilot study. Liver Int 2010;30:538–545.
19. Piscaglia F, Salvatore V, Di Donato R, D'Onofrio M, Gualandi S, Gallotti A, et al. Accuracy of VirtualTouch Acoustic Radiation Force Impulse (ARFI) imaging for the diagnosis of cirrhosis during liver ultrasonography. Ultraschall Med 2011;32:167–175.
20. Barr RG, Wilson SR, Rubens D, Garcia-Tsao G, Ferraioli G. Update to the Society of Radiologists in Ultrasound Liver Elastography consensus statement. Radiology 2020;296:263–274.
21. Gangi A, Lu SC. Chemotherapy-associated liver injury in colorectal cancer. Therap Adv Gastroenterol 2020;13:1756284820924194.
22. Pozzo C, Basso M, Cassano A, Quirino M, Schinzari G, Trigila N, et al. Neoadjuvant treatment of unresectable liver disease with irinotecan and 5-fluorouracil plus folinic acid in colorectal cancer patients. Ann Oncol 2004;15:933–939.
23. Ho WM, Mok TS, Ma BB, Lim B, Lai PB, Chui A, et al. Liver resection after irinotecan (CPT-11), 5-fluorouracil (5FU) and folinic acid (FA) for patients (pts) with unresectable liver metastases (Mets) from colorectal cancer (CRC). Proc Am Soc Clin Oncol 2003;22:312.
24. Chenevert TL, McKeever PE, Ross BD. Monitoring early response of experimental brain tumors to therapy using diffusion magnetic resonance imaging. Clin Cancer Res 1997;3:1457–1466.
25. Thoeny HC, De Keyzer F, Chen F, Ni Y, Landuyt W, Verbeken EK, et al. Diffusion-weighted MR imaging in monitoring the effect of a vascular targeting agent on rhabdomyosarcoma in rats. Radiology 2005;234:756–764.
Table 1.Values are presented as mean±standard deviation (range), median (IQR), or number (%). CEA, carcinoembryonic antigen; CRLM, colorectal liver metastasis; ARFI, acoustic radiation force impulse; SWV, shear-wave velocity; FOLFOX, 5-fluorouracil, leucovorin, and oxaliplatin; FOLFIRI, 5-fluorouracil, leucovorin, and irinotecan; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease. Table 2.Table 3.
CI, confidence interval; CEA, carcinoembryonic antigen; SWV, shear-wave velocity; SD, stable disease; PD, progressive disease; KRAS, Kirsten rat sarcoma virus; CRLM, colorectal liver metastasis; FOLFOX, 5-fluorouracil, leucovorin, and oxaliplatin; FOLFIRI, 5-fluorouracil, leucovorin, and irinotecan. |