Nomogram for prediction of prognosis in patients with initially unresectable colorectal liver metastases

Personalized prediction of postoperative complication and survival among Colorectal Liver Metastases Patients Receiving Simultaneous Resection using machine learning approaches: A multi-center study

BACKGROUND

To predict clinical important outcomes for colorectal liver metastases (CRLM) patients receiving colorectal resection with simultaneous liver resection by integrating demographic, clinical, laboratory, and genetic data.

METHODS

Random forest (RF) models were developed to predict postoperative complications and major complications (binary outcomes), as well as progression-free survival (PFS) and overall survival (OS) (time-to-event outcomes) of the CRLM patients based on data from two hospitals. The models were validated on an external dataset from an independent hospital. The clinical utility of the models was assessed via decision curve analyses (DCA).

RESULTS

There were 1067 patients included in survival prediction analyses and 1070 patients included in postoperative complication prediction analyses. The RF models provided an assessment of the model contributions of features for outcomes and suggested KRAS, BRAF, and MMR status were salient for the PFS or OS predictions. RF model of PFS showed that the Brier scores at 1-, 3-, and 5-year PFS were 0.213, 0.202 and 0.188; and the AUCs of 1-, 3- and 5-year PFS were 0.702, 0.720 and 0.743. RF model of OS revealed that Brier scores of 1-,3-, and 5-year OS were 0.040, 0.183 and 0.211; and the AUCs of 1-, 3- and 5-year OS were 0.737, 0.706 and 0.719. RF model for postoperative complication resulted in an AUC of 0.716 and a Brier score of 0.196. DCA curves clearly demonstrated that the RF models for these outcomes exhibited a superior net benefit across a wide range of threshold probabilities, signifying their favorable clinical utility. The RF models consistently exhibited robust performance in both internal cross-validation and external validation. The individualized risk profile predicted by the models closely aligned with the actual survival outcomes observed for the patients. A web-based tool (https://kanli.shinyapps.io/CRLMRF/) was provided to demonstrate the practical use of the prediction models for new patients in the clinical setting.

CONCLUSION

The predictive models and a web-based tool for personalized prediction demonstrated a moderate predictive performance and favorable clinical utilities on several key clinical outcomes of CRLM patients receiving simultaneous resection, which could facilitate the clinical decision-making and inform future interventions for CRLM patients.

Development and validation of a nomogram for predicting cancer-specific survival in small-bowel adenocarcinoma patients using the SEER database

BACKGROUND

Small bowel adenocarcinoma (SBA) is a rare gastrointestinal malignancy forwhich survival is hampered by late diagnosis, complex responses to treatment, and poor prognosis. Accurate prognostic tools are crucial for optimizing treatment strategies and improving patient outcomes. This study aimed to develop and validate a nomogram based on the Surveillance, Epidemiology, and End Results (SEER) database to predict cancer-specific survival (CSS) in patients with SBA and compare it to traditional American Joint Committee on Cancer (AJCC) staging.

METHODS

We analyzed data from 2,064 patients diagnosed with SBA between 2010 and 2020 from the SEER database. Patients were randomly assigned to training and validation cohorts (7:3 ratio). Kaplan‒Meier survival analysis, Cox multivariate regression, and nomograms were constructed for analysis of 3-year and 5-year CSS. The performance of the nomograms was evaluated using Harrell's concordance index (C-index), the area under the receiver operating characteristic (ROC) curve, calibration curves, decision curve analysis (DCA), net reclassification improvement (NRI), and integrated discrimination improvement (IDI).

RESULTS

Multivariate Cox regression identified sex, age at diagnosis, marital status, tumor site, pathological grade, T stage, N stage, M stage, surgery, retrieval of regional lymph nodes (RORLN), and chemotherapy as independent covariates associated with CSS. In both the training and validation cohorts, the developed nomograms demonstrated superior performance to that of the AJCC staging system, with C-indices of 0.764 and 0.759, respectively. The area under the curve (AUC) values obtained by ROC analysis for 3-year and 5-year CSS prediction significantly surpassed those of the AJCC model. The nomograms were validated using calibration and decision curves, confirming their clinical utility and superior predictive accuracy. The NRI and IDI indicated the enhanced predictive capability of the nomogram model.

CONCLUSION

The SEER-based nomogram offers a significantly superior ability to predict CSS in SBA patients, supporting its potential application in clinical decision-making and personalized approaches to managing SBA to improve survival outcomes.

Clinical Prediction Models for Prognosis of Colorectal Liver Metastases: A Comprehensive Review of Regression-Based and Machine Learning Models

Colorectal liver metastasis (CRLM) is a disease entity that warrants special attention due to its high frequency and potential curability. Identification of "high-risk" patients is increasingly popular for risk stratification and personalization of the management pathway. Traditional regression-based methods have been used to derive prediction models for these patients, and lately, focus has shifted to artificial intelligence-based models, with employment of variable supervised and unsupervised techniques. Multiple endpoints, like overall survival (OS), disease-free survival (DFS) and development or recurrence of postoperative complications have all been used as outcomes in these studies. This review provides an extensive overview of available clinical prediction models focusing on the prognosis of CRLM and highlights the different predictor types incorporated in each model. An overview of the modelling strategies and the outcomes chosen is provided. Specific patient and treatment characteristics included in the models are discussed in detail. Model development and validation methods are presented and critically appraised, and model performance is assessed within a proposed framework.

Recurrence-risk stratification using the Beppu score and selection of perioperative chemotherapy for colorectal liver metastases

The annual postoperative disease-free survival for colorectal liver metastases can be easily estimated by weighting six preoperative clinical parameters (Beppu score). We identified three recurrence-risk stratification groups: the low (≤6 points), moderate (7-10 points), and high-risk (≥11 points). For low-, moderate-, and high-risk patients, hepatectomy alone, hepatectomy with adjuvant chemotherapy, and hepatectomy with preoperative chemotherapy are recommended, respectively. The Beppu score enables the decision on the necessity and timing of perioperative chemotherapy.

Prediction of survival in patients with colorectal liver metastases- development and validation of a prognostic score model

BACKGROUND

Metastatic spread of colorectal cancer to the liver impacts prognosis. Advances in chemotherapy have resulted in increased resectability rates and thereby improved survival in patients with colorectal liver metastases (CRLM). However, criteria are needed to ensure that patients selected for hepatic resection benefit from the invasive therapy. The study aimed to construct a predictive model for overall survival (OS) in patients with CRLM, based on preoperatively available information.

METHODS

The retrospective cohort study reviewed all patients with CRLM discussed at multidisciplinary team conference at Karolinska University Hospital, Stockholm, Sweden, 2013-2018. Independent prognostic factors for OS were identified, based on which a score model was generated. The model was validated on patients treated for CRLM at Hôpital Universitaire Paul Brousse, Villejuif, France, 2007-2018. Calibration and discrimination methods were used for internal and external validation.

RESULTS

The Swedish development cohort included 1013 patients, the French validation cohort 391 patients. Poor OS was significantly associated with age>60years (hazard ratio (HR) 3.57 (95%CI 2.18-9.94)), number of CRLM (HR 4.59 (2.83-12.20)), diameter of largest CRLM>5 cm (HR 2.59 (1.74-5.03)), right-sided primary tumour (HR 2.98 (2.00-5.80)), extrahepatic disease (HR 4.14 (2.38-15.87)) and non-resectability (HR 0.77 (0.66-0.90)). The C-statistic for prediction of OS was .74, in the development cohort and 0.69 in the validation cohort.

CONCLUSION

The presented predictive score model can adequately predict OS for patients at the initial diagnosis of CRLM. The prognostic model could be of clinical value in the management of all patients with CRLM, by predicting individualized survival and thereby facilitating treatment recommendations.

Radiomics artificial intelligence modelling for prediction of local control for colorectal liver metastases treated with radiotherapy

•

Computed tomography imaging contains quantifiable information to characterize colorectal liver metastases.

•

Shape, texture, and intensity statistical features quantified the computed tomography liver volume.

•

An artificial intelligence model to predict local progression from radiomic features was developed with high accuracy.

•

Maximum dosage and textural coarseness of liver volume were features with highest predictive value.

Background and Purpose

Prognostic assessment of local therapies for colorectal liver metastases (CLM) is essential for guiding management in radiation oncology. Computed tomography (CT) contains liver texture information which may be predictive of metastatic environments. To investigate the feasibility of analyzing CT texture, we sought to build an automated model to predict progression-free survival using CT radiomics and artificial intelligence (AI).

Materials and Methods

Liver CT scans and outcomes for N = 97 CLM patients treated with radiotherapy were retrospectively obtained. A survival model was built by extracting 108 radiomic features from liver and tumor CT volumes for a random survival forest (RSF) to predict local progression. Accuracies were measured by concordance indices (C-index) and integrated Brier scores (IBS) with 4-fold cross-validation. This was repeated with different liver segmentations and radiotherapy clinical variables as inputs to the RSF. Predictive features were identified by perturbation importances.

Results

The AI radiomics model achieved a C-index of 0.68 (CI: 0.62–0.74) and IBS below 0.25 and the most predictive radiomic feature was gray tone difference matrix strength (importance: 1.90 CI: 0.93–2.86) and most predictive treatment feature was maximum dose (importance: 3.83, CI: 1.05–6.62). The clinical data only model achieved a similar C-index of 0.62 (CI: 0.56–0.69), suggesting that predictive signals exist in radiomics and clinical data.

Conclusions

The AI model achieved good prediction accuracy for progression-free survival of CLM, providing support that radiomics or clinical data combined with machine learning may aid prognostic assessment and management.

Construction of a nomogram to predict the survival of metastatic gastric cancer patients that received immunotherapy

Background

Immunotherapy has shown promising results for metastatic gastric cancer (MGC) patients. Nevertheless, not all patients can benefit from anti-PD-1 treatment. Thus, this study aimed to develop and validate a prognostic nomogram for MGC patients that received immunotherapy.

Methods

Herein, MGC patients treated with anti-PD-1 between 1 October 2016 and 1 June 2022 at two separate Chinese PLA General Hospital centers were enrolled and randomly divided into training and validation sets (186 and 80 patients, respectively). The nomogram was constructed based on a multivariable Cox model using baseline variables from the training cohort. Its predictive accuracy was validated by the validation set. The consistency index (C-index) and calibration plots were used to evaluate the discriminative ability and accuracy of the nomogram. The net benefit of the nomogram was evaluated using decision curve analysis (DCA). Finally, we stratified patients by median total nomogram scores and performed Kaplan–Meier survival analyses.

Results

We developed the nomogram based on the multivariate analysis of the training cohort, including four parameters: surgery history, treatment line, lung immune prognostic index (LIPI), and platelet-to-lymphocyte ratio (PLR). The C-index of the nomogram was 0.745 in the training set. The calibration curve for 1- and 2-year survival showed good agreement between nomogram predictions and actual observations. In the validation group, the calibration curves demonstrated good performance of the nomogram, with a C-index for overall survival (OS) prediction of 0.713. The OS of patients with a score greater than the median nomogram score was significantly longer than patients with a score lower or equal to the median (p < 0.001).

Conclusion

We constructed a nomogram to predict the outcomes of MGC patients that received immunotherapy. This nomogram might facilitate individualized survival predictions and be helpful during clinical decision-making for MGC patients under anti-PD-1 therapy.

Development and validation of prognostic nomograms for patients with metastatic small bowel adenocarcinoma: a retrospective cohort study

We aimed to explore factors associated with prognosis in patients with metastatic small bowel adenocarcinoma (SBA) as well as to develop and validate nomograms to predict overall survival (OS) and cancer-specific survival (CSS). Relevant information of patients diagnosed between 2004 and 2016 was extracted from the Surveillance, Epidemiology, and End Results (SEER) database. Nomograms for predicting 1- and 3-year OS and CSS were established with potential risk factors screened from multivariate cox regression analysis. The discrimination and accuracy of the nomograms were assessed by concordance index (C-index), calibration plots, and the area under receiver operating characteristic curve (AUC). In total, 373 SBA patients with M1 category were enrolled. Multivariate analysis revealed that age, size and grade of primary tumor, primary tumor surgery, and chemotherapy were significant variables associated with OS and CSS. The C-index values of the nomogram for OS were 0.715 and 0.687 in the training and validation cohorts, respectively. For CSS, it was 0.711 and 0.690, respectively. Through AUC, decision curve analysis (DCA) and calibration plots, the nomograms displayed satisfactory prognostic predicted ability and clinical application both in the OS and CSS. Our models could be served as a reliable tool for prognostic evaluation of patients with metastatic SBA, which are favorable in facilitating individualized survival predictions and clinical decision-making.

A prognostic nomogram for intrahepatic progression-free survival in patients with colorectal liver metastases after ultrasound-guided percutaneous microwave ablation

PURPOSE

The present study identified predictors of intrahepatic progression-free survival (ihPFS) after ultrasound (US)-guided percutaneous microwave ablation (MWA) for colorectal liver metastases (CRLMs) and developed a nomogram to predict ihPFS.

METHODS

From January 2013 to December 2018, a total of 314 patients (224 men and 90 women; mean age, 57.1 ± 11.8 years; range, 23-86 years) with 645 CRLMs (mean diameter, 16.6 ± 6.5 mm; range, 6.0-45.0 mm) treated with US-guided percutaneous MWA were included and analyzed. The average number of CRLMs per patient treated was 2 (range, 1-8). A nomogram to predict ihPFS was developed based on a multivariable Cox model. Validation of the nomogram was performed using the concordance index (C-index), calibration curves, and decision curve analyses.

RESULTS

The 1-, 2-, and 3-year cumulative ihPFS rates were 59.0%, 38.9%, and 30.8%, respectively. Maximal CRLM size, number of CRLMs, ablative margin, primary tumor lymph node status, and chemotherapy were five independent prognostic factors for ihPFS. The C-index of the nomogram was 0.702 (CI: 0.681-0.723). A risk classification system that perfectly classified the patients into three risk groups was constructed. The median ihPFS of patients in the low-, intermediate-, and high-risk groups was 36.3 months (95% CI: 21.4-51.1), 13.4 months (95% CI: 12.1-14.6), and 3.8 months (95% CI: 2.3-5.3), respectively.

CONCLUSION

The nomogram and risk classification system will facilitate the personalized assessment of ihPFS for patients receiving US-guided percutaneous MWA for CRLMs.

Limited Prognostic Value of KRAS in Patients Undergoing Hepatectomy for Colorectal Liver Metastases

BACKGROUND

RAS mutation status is considered a powerful prognostic factor in patients undergoing hepatectomy for colorectal liver metastases (CLM). However, whether its prognostic power is robust regardless of administration of preoperative chemotherapy or tumor burden remains unclear.

METHODS

Consecutive patients who underwent initial hepatectomy for CLM from April 2010 through March 2017 in two hospitals were included. The prognostic value of KRAS was compared based on whether patients received preoperative chemotherapy and their tumor burden score (TBS).

RESULTS

We included 409 patients (median follow-up 38 months). In the preoperative chemotherapy group, patients with mutant KRAS (mt-KRAS) CLM had poorer overall survival (OS) than those with wild KRAS (wt-KRAS; 5-year OS: 37.7% vs 53.8%, p = 0.024), although their OS was not different from patients undergoing upfront surgery. Similarly, patients with mt-KRAS had poorer OS than those with wt-KRAS in TBS of 3-9 (5-year OS: 33.1% vs 63.2%, p = 0.001), although their OS was not different from patients with TBS < 3 or ≥ 9. In multivariate analysis, mt-KRAS was an independent prognostic factor of OS among patients receiving preoperative chemotherapy (hazard ratio [HR] 1.61, 95% confidence interval [CI]: 1.034-2.491; p = 0.035) and patients with TBS of 3-9 (HR 1.836, 95% CI 1.176-2.866; p = 0.008). However, it was not a prognostic factor in patients who underwent upfront surgery or with TBS > 3 or ≥ 9.

CONCLUSIONS

In patients undergoing hepatectomy for CLM, the prognostic value of KRAS depends on their history of preoperative chemotherapy or tumor burden.

Recurrent colorectal liver metastasis patients could benefit from repeat hepatic resection

Background

Local treatment remains the best option for recurrent colorectal liver metastasis (CRLM). The current study aimed to investigate predictive factors of survival outcomes and select candidates for local treatment for CRLM at first recurrence.

Methods

Data were collected retrospectively from CRLM patients who underwent hepatic resection and developed first recurrence between 2000 and 2019 at our institution.
A nomogram predicting overall survival was established based on a multivariable Cox model of clinicopathologic factors. The predictive accuracy and discriminative ability of the nomogram were determined by the concordance index and calibration curve.

Results

Among 867 patients who underwent curative hepatic resection, 549 patients developed recurrence. Three hundred patients were evaluated and had resectable and liver-limited disease. Among them, repeat liver resection and percutaneous radiofrequency ablation were performed in 88 and 85 patients, respectively. The other 127 patients received only systemic chemotherapy. Multivariable analysis identified primary lymph node positivity, tumor size > 3 cm, early recurrence, RAS gene mutation and no local treatment as independent risk factors for survival outcomes. Integrating these five variables, the nomogram presented a good concordance index of 0.707. Compared with patients who received only systemic chemotherapy, radical local treatment did not significantly improve survival outcomes (median OS: 21 vs. 15 months, p = 0.126) in the high-risk group (total score ≥ 13).

Conclusion

Radical local treatment improved the survival of recurrent CRLM patients. The proposed model facilitates personalized assessments of prognosis for patients who develop first recurrence in the liver.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12893-021-01323-y.

A Prognostic Scoring System to Predict Survival Outcome of Resectable Colorectal Liver Metastases in this Modern Era

BACKGROUND

An individualized treatment decision is based on the accurate evaluation of clinical risk factors and prognosis for resectable colorectal liver metastases. The current study aimed to develop an effective nomogram to predict progression-free survival (PFS) and to design a treatment schedule preoperatively.

METHODS

The study enrolled a primary cohort of 532 patients with resectable colorectal liver metastases (CRLM) who underwent hepatic resection at two institutions and a validation cohort of 237 patients at two additional institutions with resectable CRLM between 1 January 2008 and 31 December 2018. A nomogram was created based on the independent predictors in the multivariable analysis of progression-free survival in the primary cohort. The predictive accuracy and discriminative ability of the nomogram were determined by the concordance index (C-index) and the calibration curve. The score was compared with the current standard Fong score and validated with an external cohort.

RESULTS

The independent risk factors for CRLM patients identified in the multivariable analysis were tumor larger than 5 cm, more than one tumor, RAS mutation, primary lymph node metastasis, and primary tumor located on the right side. All five factors were considered in the nomogram. The C-index of the nomogram for predicting survival was 0.696. With external validation, the C-index of the nomogram for the prediction of the PFS was 0.682, which demonstrated that this model has a good level of discriminative ability. For high-risk patients (score > 16), neoadjuvant chemotherapy improved PFS and overall survival (OS) after hepatic resection.

CONCLUSION

The current nomogram demonstrated an accurate performance in predicting PFS for resectable CRLM patients with liver-limited disease. Based on the current nomogram, high-risk patients (nomogram score > 16) might benefit from neoadjuvant chemotherapy.

Accuracy of Modern Clinical Risk Score Including RAS Status Changes Based on Whether Patients Received Perioperative Chemotherapy for Colorectal Liver Metastases

BACKGROUND

A modified Fong clinical score (m-Fong CS) that includes the RAS mutation status has recently been proposed and offered an improved survival stratification of patients who undergo surgery and systemic chemotherapy for colorectal liver metastases (CLM). The aim of this study is to assess whether a CS that includes RAS status is influenced by whether patients receive perioperative chemotherapy.

METHODS

We created a new CS using multivariate analysis of data of patients who underwent hepatectomy for CLM for the first time between 2010 and 2016 at a single hospital (n = 341, 79% received perioperative chemotherapy). The resulting CS and m-Fong CS were then validated in the patient cohort at three other hospitals (n = 309). Furthermore, the applicability of the two CS in the total cohort (n = 650) was tested according to whether the patients received perioperative chemotherapy.

RESULTS

The new CS comprised mutant RAS status, ≥4 CLMs, and a CA19-9 level ≥100 U/mL (1 point per factor). Both the new CS and m-Fong CS failed to stratify the survival of the 309 patients in the validation cohort, including those who did not receive perioperative chemotherapy (29%). Both of the CS accurately stratified the survival of patients who underwent perioperative chemotherapy but not of those who underwent surgery alone.

CONCLUSION

A CS that includes the RAS mutation status can stratify the survival of patients who undergo hepatectomy combined with perioperative chemotherapy, but it has limited value for patients who undergo surgery alone.

A novel prognostic nomogram for colorectal cancer liver metastasis patients with recurrence after hepatectomy

PURPOSE

We aimed to construct a nomogram to predict personalized post-recurrence survival (PRS) among colorectal cancer liver metastasis (CRLM) patients with post-hepatectomy recurrence.

METHODS

Colorectal cancer liver metastasis patients who received initial hepatectomy and had subsequent recurrence between 2001 and 2019 in Sun Yat-sen University Cancer Center from China were included in the study. Patients were randomly assigned to a training cohort and a validation cohort on a ratio of 2:1. Univariable analysis was first employed to select potential predictive factors for PRS. Then, the multivariable Cox regression model was applied to recognize independent prognostic factors. According to the model, a nomogram to predict PRS was established. The nomogram's predictive capacity was further assessed utilizing concordance index (C-index) values, calibration plots, and Kaplan-Meier curves.

RESULTS

About 376 patients were finally enrolled, with a 3-year PRS rate of 37.3% and a 5-year PRS rate of 24.6%. The following five independent predictors for PRS were determined to construct the nomogram: the largest size of liver metastases at initial hepatectomy, relapse-free survival, CEA level at recurrence, recurrent sites, and treatment for recurrence. The nomogram displayed fairly good discrimination and calibration. The C-index value was 0.742 for the training cohort and 0.773 for the validation cohort. Patients were grouped into three risk groups very well by the nomogram, with 5-year PRS rates of 45.2%, 23.3%, and 9.0%, respectively (p < 0.001) in the training cohort and 36.0%, 9.2%, and 4.6%, respectively (p < 0.001) in the validation cohort.

CONCLUSION

A novel nomogram was built and validated to enable the prediction of personal PRS in CRLM patients with post-hepatectomy recurrence. The nomogram may help physicians in decision making.

Upgraded nomograms for the prediction of complications and survival in patients with colorectal liver metastases treated with neoadjuvant chemotherapy followed by hepatic resection

Background

To establish upgraded nomograms incorporating neoadjuvant chemotherapy (NAC)-related factors and preoperative testing markers to predict postoperative complications, progression-free survival (PFS) and overall survival (OS) in patients with colorectal liver metastases (CRLM).

Methods

Multivariate regression analyses were used to reveal independent predictors for postoperative complications, PFS and OS. Nomograms incorporating independent predictors were constructed, and discrimination and calibration were evaluated. Survival was estimated by the Kaplan-Meier method and compared using the log-rank test.

Results

A nomogram predicting postoperative complications was constructed based on preoperative serum gamma-glutamyl transpeptidase (GGT) ≥36 U/L, major liver resection, intraoperative blood loss ≥300 mL, primary site located in the right hemicolon and primary lymph node metastasis, with an area under the receiver operating characteristic curve (AUROC) of 0.750. The calibration curves and Hosmer-Lemeshow test revealed desirable model calibration (chi-square: 4.47, P=0.88). Moreover, a nomogram for the prediction of PFS was constructed based on tumour regression grade (TRG), primary lymph node metastasis, R0 resection and NAC cycles ≥5, with good discrimination (C-index: 0.663±0.024) and calibration, and one for predicting OS was constructed based on preoperative GGT ≥36 U/L, NAC toxicity, NAC cycles ≥5, primary lymph node metastasis and R0 resection, with favourable discrimination (C-index: 0.684±0.030) and calibration. Significant differences in PFS and OS were observed among patients stratified into three different risk groups (P<0.001) according to total scores based on the nomograms.

Conclusions

This study is the first to establish novel predictive nomograms specifically incorporating TRG, NAC toxicity and serum GGT level for the prediction of postoperative complications, PFS and OS in CRLM patients. The nomograms exhibit favourable discrimination and calibration to guide personalized CRLM management and therapy.

Meta-analysis of the association between primary tumour location and prognosis after surgical resection of colorectal liver metastases

BACKGROUND

Primary tumour location is emerging as an important prognostic factor in localized and metastatic colorectal cancers. However, its prognostic role in colorectal liver metastasis (CRLM) after hepatectomy remains controversial. A systematic review and meta-analysis was undertaken to evaluate its prognostic value.

METHODS

References were identified through searches of PubMed, Embase, Web of Science and the Cochrane Library comparing overall or disease-free survival after hepatic resection between patients with CRLM originating from right- or left-sided colorectal cancers. Data were pooled using hazard ratios (HRs) and 95 per cent confidence intervals according to a random-effects model. Meta-regression and subgroup analyses were conducted to assess the effect of underlying confounding factors on HR estimates and to adjust for this.

RESULTS

The final analysis included 21 953 patients from 45 study cohorts. Compared with left-sided primary tumour location, right-sided location was associated with worse overall survival (HR 1·39, 95 per cent c.i. 1·28 to 1·51; P < 0·001; prediction interval 1·00 to 1·93), and also tended to have a negative impact on disease-free survival (HR 1·18, 1·06 to 1·32; P = 0·004; prediction interval 0·79 to 1·75). Subgroup analysis showed that the negative effect of right-sided primary tumour location on overall survival was more prominent in the non-Asian population (HR 1·47, 1·33 to 1·62) than the Asian population (HR 1·18, 1·05 to 1·32) (P for interaction <0·01).

CONCLUSION

This study demonstrated a prognostic role for primary tumour location in patients with CRLM receiving hepatectomy, especially regarding overall survival. Adding primary tumour location may provide important optimization of prognosis prediction models for CRLM in current use.

A Nomogram for the Prediction of Prognosis in Patients With Distant Metastases of Nasopharyngeal Carcinoma

Background: Patients with metastatic nasopharyngeal carcinoma (NPC) have heterogeneous survival outcomes. This study aimed to establish an effective prognostic nomogram for patients with NPC with distant metastases using easily determined factors.

Methods: The nomogram was based on a retrospective study of 103 patients with metastatic NPC at the First Affiliated Hospital of Xiamen University during January 2009–March 2016. Nomogram performance was evaluated using a concordance index (C-index) and assessed using calibration plot. Bootstraps with 1,000 resamples were applied to these analyses.

Results: In univariate and multivariate Cox proportional hazards model analyses, chemotherapy, metastatic liver involvement, number of tumor metastases, N stage and derived neutrophil–lymphocyte ratio correlated with overall survival (OS). The recurrence probability calibration curve indicated good agreement between nomogram-based predictions and actual observations. For OS predictions, the nomogram had a C-index of 0.824 (95% confidence interval, 0.74–0.91). The stratification by nomogram score of patients into different subgroups showed significant distinction.

Conclusion: This novel nomogram comprises factors that are easily determined at most hospitals and can predict survival in patients with distant metastases of NPC. This model can precisely estimate the survival of individual patients and identify subgroups of patients requiring specific therapeutic strategies.

Implication of primary tumor location for the indication of preoperative chemotherapy in patients with colorectal liver metastases

BACKGROUND

The benefit of preoperative chemotherapy for colorectal liver metastases (CRLM) remains uncertain. The aim was to clarify the effect of preoperative chemotherapy on CRLM according to the primary tumor location.

METHODS

Among a total cohort of 163 patients who underwent curative hepatectomy for CRLM, 36 patients had a right-sided and 127 had a left-sided primary tumor. According to the performance of preoperative chemotherapy, survival analysis was conducted and prognostic factors were identified.

RESULTS

Preoperative chemotherapy was administered to 17 patients (47.2%) with a right-sided and 74 (58.3%) with a left-sided primary tumor (P = 0.24). Among the patients who received preoperative chemotherapy, overall survival (OS) and disease-free survival (DFS) were similar between patients with right- and left-sided primary tumors (P = 0.36 and P = 0.44, respectively). Among the patients who underwent upfront hepatectomy, the OS and DFS of patients with a right-sided primary tumor were worse than those with a left-sided primary tumor (P = 0.02 and P = 0.025, respectively). Among the patients who underwent upfront surgery, the right-sided primary tumor was identified as an independent poor prognostic factor for OS (hazard ratio 3.44, P = 0.021).

CONCLUSION

The existence of a right-sided primary tumor may be an indication of preoperative chemotherapy for patients with CRLM.

Association between Primary Tumor Location and Prognostic Survival in Synchronous Colorectal Liver Metastases after Surgical Treatment: A Retrospective Analysis of SEER Data

The prognostic and predictive role of primary tumor location (PTL) in localized and metastatic colorectal cancer (CRC) is a hotspot issue in recent years. However, its prognostic role is still unclear in synchronous colorectal liver metastases (sCRLM), especially in those receiving surgical treatment of primary tumor and liver metastases. Here, a retrospective survival analysis was performed using the Surveillance, Epidemiology, and End Results Program (SEER) database between 2010 and 2014, on patients who were pathologically confirmed sCRLM, and received surgical treatment of both primary tumor and liver metastases. After stringent exclusive procedure, a total of 1508 patients with sCRLM (872 men [57.8%] and 636 women [42.2%]; mean age, 60.9 years) were eligible for the final study. Compared with sCRLM with left-sided PTL, cases with right-sided PTL were more likely to be T4 (31.3% vs. 20.1%, p<0.001), N2 (42.5% vs. 31.8%, p<0.001) and poorly differentiated (30.5% vs. 15.1%, p<0.001). Furthermore, right-sided sCRLM showed significantly shorter cancer specific survival (CSS) than those from left-side (p<0.001). After Cox hazard regression analysis, right-sided PTL still remained to be a strong independent predictor for poor prognosis in this cohort of sCRLM patients (OS, HR=1.75, 95% CI 1.34-2.29; CSS, HR=1.76, 95% CI 1.33-2.35). In conclusion, according to this population-based cohort from the SEER database, PTL was a critical prognostic factor that affect long-term OS and CSS in patients with sCRLM after surgical treatment of primary tumor and liver metastases.

Long-term quality of life after liver transplantation for non-resectable colorectal metastases confined to the liver

Background

Liver transplantation for patients with non‐resectable colorectal liver metastases offers increased survival, with median overall survival of more than 5 years. The aim of this study was to compare quality of life before and up to 3 years after liver transplantation for colorectal liver metastases.

Methods

Quality of life was assessed using the European Organisation for Research and Treatment of Cancer QLQ‐C30 questionnaire version 3.0. The patients received the questionnaire before and up to 3 years after liver transplantation.

Results

Some 23 patients were included in the analysis. Three months after liver transplantation they reported reduced quality of life (global health status scale), physical function and role function, and increased dyspnoea. At 6 months, global health status, physical function and role function had returned to pretransplant values. Three years after liver transplantation all symptom and function scores were comparable to baseline values. Patients with high scores for fatigue, pain and appetite loss at baseline had reduced 3‐year overall survival.

Conclusion

Patients with non‐resectable colorectal liver‐only metastases receiving liver transplantation had good long‐term quality of life. Patients with high symptom scores before transplantation had reduced 3‐year overall survival.

[How to manage a rectal cancer with synchronous liver metastases? A question of strategy]

The prognosis of patients with rectal cancer and synchronous liver metastasis has improved thanks to chemotherapy and rectal and liver surgery progresses. However, there is no consensus about optimal management and practices remain heterogeneous. A curative treatment may be considered for 20 to 30% of patients with complete resection of metastasis and primary tumor after induction chemotherapy. To this end, a primary optimal evaluation by a multidisciplinary board including hepatic and colorectal surgeons is crucial. The therapeutic strategy associates chemotherapy, radiotherapy, hepatic and rectal surgery. The most threatening site guides the sequence of treatments. If hepatic resectability is uncertain, a "liver first" strategy associating induction chemotherapy and hepatic surgery is preferred. In non-resectable metastatic cases, chemotherapies with targeted therapies might lead to secondary resection for 30% of patients (conversion). This has changed our practice and triggers reconsidering resectability after chemotherapy. When metastases remain non-resectable, additional treatment focusing on primary tumor should control pelvic symptoms otherwise hardly impacting quality of life. Rectal surgery, short-course radiotherapy (5×5Gy), conformational long-course chemoradiotherapy or intensity-modulated radiation therapy with dose escalation are options discussed in this review.

Long-term postoperative survival prediction in patients with colorectal liver metastasis

Numerous factors affect the prognosis of colorectal liver metastasis (CRLM) patients after hepatic resection. We investigated several factors related to overall survival in patients with CRLM to identify those most likely to benefit from hepatic resection, and produced a rational tumor biology score system. Three hundred CRLM patients treated with preoperative chemotherapy followed by hepatic resection between 2006 and 2016 were enrolled in our study. Clinicopathologic and long-term survival data were collected and assessed. Patient 1-, 3-, and 5-year overall survival rates were 92.7%, 58.3%, and 45.8%, respectively, while 1-, 3-, and 5-year disease-free survival rates were 44.7%, 28.6%, and 24.2%, respectively. Multivariate Cox regression analysis revealed poor preoperative chemotherapy response, Fong clinical risk score > 2, and KRAS mutation to be independent prognostic indicators in CRLM patients. As part of a preoperative staging system in which one point was assigned for each factor, a total score (out of 3) was predictive of long-term survival following surgery. These factors facilitate personalized prognostic assessments in CRLM patients planning for resection.

Resection after preoperative chemotherapy versus synchronous liver resection of colorectal cancer liver metastases: A propensity score matching analysis

This study aimed to determine the prognostic effects of preoperative chemotherapy for colorectal cancer liver metastasis (CLM).We retrospectively evaluated 2 groups of patients between January 2006 and August 2012. A total of 53 patients who had ≥3 hepatic metastases underwent resection after preoperative chemotherapy (preoperative chemotherapy group), whereas 96 patients who had ≥3 hepatic metastases underwent resection with a curative intent before chemotherapy for CLM (primary resection group). A propensity score (PS) model was used to compare the both groups.The 3-year disease-free survival (DFS) rates were 31.7% and 20.4% in the preoperative chemotherapy and primary resection groups, respectively (log-rank = 0.015). Analyzing 32 PS matched pairs, we found that the DFS rate was significantly higher in the preoperative chemotherapy group than in the primary resection group (3-year DFS rates were 34.2% and 16.8%, respectively [log-rank = 0.019]). Preoperative chemotherapy group patients had better DFSs than primary resection group patients in various multivariate analyses, including crude, multivariable, average treatment effect with inverse probability of treatment weighting model and PS matching.Responses to chemotherapy are as important as achieving complete resection in cases of multiple hepatic metastases. Preoperative chemotherapy may therefore be preferentially considered for patients who experience difficulty undergoing complete resection for multiple hepatic metastases.

Long-term outcomes of radiofrequency ablation combined with hepatectomy compared with hepatectomy alone for colorectal liver metastases

Background

Combining radiofrequency ablation (RFA) with hepatectomy may enable treatment with curative intent for patients with colorectal liver metastasis (CRLM). However, the oncological outcomes in comparison with resection alone remain to be clarified.

Methods

Patients who underwent a first hepatectomy between 2001 and 2012 for CRLM were enrolled. Short- and long-term outcomes of patients who underwent hepatectomy plus RFA were compared with those of patients who had hepatectomy alone using propensity score matching.

Results

Of a total of 553 patients, hepatectomy + RFA and hepatectomy alone were performed in 37 and 516 respectively. Before matching, patients in the hepatectomy + RFA group were characterized primarily by a larger tumour burden. After matching of 31 patients who underwent hepatectomy + RFA with 93 who had hepatectomy alone, background characteristics were well balanced. In the matched cohort, overall and disease-free survival in the hepatectomy + RFA group were no different from those among patients who had hepatectomy alone (5-year overall survival rate 57 versus 61 per cent, P = 0·649; 5-year disease-free survival rate 19 versus 17 per cent, P = 0·865). Local recurrence at the ablated site was observed in four of 31 patients (13 per cent). Although overall local recurrence (ablated site and/or cut surface) was more frequent in the hepatectomy + RFA group (9 of 31 (29 per cent) versus 11 of 93 (12 per cent); P = 0·032), there was no difference in intrahepatic disease-free survival between the two groups (P = 0·705).

Conclusion

Hepatectomy + RFA achieved outcomes comparable to hepatectomy alone. Combining RFA with hepatectomy should be considered as an option to achieve cure.