This study is a retrospective analysis of a prospectively maintained database of all patients undergoing multimodality treatment for CRLM in the form of surgery or other non-surgical LDTs at an eminent tertiary referral centre in India. The period of the study included all patients with CRLM treated from November 2011 to December 2020. Only the patients with a follow-up period of 6 months after the intervention for CRLM were included in the study. The study included all patients with CRLM managed with liver-directed therapies, i.e., surgical resection, radiofrequency ablation (RFA), transarterial chemoembolization (TACE), or transarterial radioembolization (TARE) with or without systemic chemotherapy. Patients treated with palliative intent with only systemic therapy at the index diagnosis as well as patients with extrahepatic metastasis were excluded from the study.

All potential patients of CRLMs were discussed in a multidisciplinary “Liver Clinic” meeting comprising surgical, medical, and radiation oncologists along with interventional radiologists. Demographic parameters and clinico-radiologic data were recorded. Characteristics of the primary tumour, viz., location, stage, treatment modalities, tumour differentiation, and serum carcinoembryonic antigen (CEA) levels, were also noted.

Contrast-enhanced computed tomography (CECT) scan of the thorax, abdomen, and pelvis or whole-body positron emission tomography and computed tomography (PET-CT) was performed to stage and characterize the liver metastases such as size, distribution, i.e., unilobar versus bilobar, and the number of lesions. Depending upon the extent of the disease, tumour location, and liver function, patients were categorized as resectable, borderline resectable, or unresectable ( Table 1 ).

Surgery was performed with curative intent to obtain a margin negative (R0) resection for patients with resectable disease. For superficial lesions, a non-anatomical resection was performed. For lesions situated deep in the liver parenchyma, a formal liver resection such as segmentectomy or hepatectomy was performed as per the International Hepatopancreatobiliary Association Brisbane 2000 Classification (13).

Synchronous liver metastases were managed with simultaneous resection or a staged approach as per discretion of the surgical team depending on the extent of the disease, the site of primary, and the nature of the liver resection that would be required. The metachronous liver lesions were confirmed as colorectal metastasis by immunohistochemistry.

Some patients, though deemed resectable, underwent ablation, as the tumour location would have entailed the loss of a large amount of normal liver parenchyma if resected. Some were treated with a combined approach of surgery with intraoperative or perioperative ablations.

Patients with borderline resectable and unresectable diseases were treated with downsizing conversion chemotherapy (in combination with other LDTs in few cases) followed by curative intent surgery if adequate downsizing was achieved by chemotherapy to facilitate R0 resection. Those who were still unresectable after chemotherapy received non-surgical LDTs with systemic chemotherapy.

Ablation by RFA or microwaves was performed in few resectable tumours as described above and for unresectable diseases or if patients were either unfit or refused surgery. TACE was used in patients with unresectable liver disease as a means of disease control and also as a downsizing tool in few patients. TARE using yttrium-90 (a pure beta emitter)-impregnated microspheres to the hepatic tumours via the tumour arterial feeders were used mainly for those patients who had progressive but liver-limited unresectable disease post-chemotherapy.

For patients undergoing ablation, response assessment was performed using a regional PET scan 4–6 weeks and 3 months from the day of ablation using the Response Evaluation Criteria in Solid Tumors (RECIST) criteria (14). Patients treated with TACE and TARE were followed up with PET-CT (with triphasic CECT) after 4 weeks of the procedure. Treatment response was assessed using Positron Emission Tomography Response Criteria in Solid Tumours (PERCIST) criteria and RECIST criteria (14, 15).

All patients were offered post-intervention adjuvant chemotherapy and followed up every 3 months with serum CEA and routine blood investigations and an ultrasound examination of the abdomen with a chest X-ray. A CECT scan of the thorax, abdomen, and pelvis was performed every 6 months for the first 2 years and annually thereafter or when there was suspicion of recurrence. The last follow-up was performed telephonically.

The data of the present study were collected in the course of common clinical practice, and accordingly, signed informed consent was obtained from each patient. Data collection was in accordance with the Declaration of Helsinki. (16) As per institutional protocol, a formal board review was not taken, as it is a retrospective analysis.

SPSS version 25 was used for statistical analysis. Student’s t-test was used for quantitative data. The chi-square test was used for categorical variables. Overall survival (OS) and event-free survival (EFS) were calculated using the Kaplan–Meier curves. OS was calculated from the diagnosis date till death due to any cause or last follow-up, and EFS was calculated from the date of intervention for CRLM to either recurrence or progression of the disease. Univariate and multivariate analyses for prognostic variables were performed using the Cox regression analysis.

A total of 627 cases of CRLM were discussed in the multidisciplinary meeting. Out of these, 252 cases were excluded, as they had extrahepatic metastasis or were treated with palliative chemotherapy without any LDT. Finally, approximately 375 eligible patients undergoing multimodality treatment for CRLM including surgical or non-surgical LDTs were included in the study ( Figure 1 ). The total cohort consisted of 245 (65.3%) male patients and 130 (34.7%) female patients with a male-to-female ratio of 1.8:1. The median age of the population was 55 (range, 22–82) years.

The primary cancer was addressed with surgical resection in all but four (1.06%) patients who either had progressive metastatic disease even after undergoing liver-directed treatment along with chemotherapy (n = 3) or refused surgery for primary disease (n = 1). The rectum was the most prevalent primary site, with 221 (58.9%) patients. More than four metastatic lesions were seen in 64 (17.1%) patients. Table 2 describes the characteristics of the primary tumour and liver metastatic disease.

Synchronous liver metastasis was seen in 247 (65.9%) patients, while 128 (34.1%) developed metachronous liver metastasis. The median time interval between primary surgery and identification of metachronous lesion was 15 months (range, 8 to 123 months). Amongst patients presenting with the synchronous disease, the most common primary site was the rectum at 122/247 (49.34%), followed by the ascending and transverse colon (87; 35.22%) and the left colon (38; 15.38%). Of the patients, 64 (17.1%) developed extrahepatic disease during the follow-up period with the lung (28 patients; 43.75%) as the most common site.

A total of 237 (58.7%) patients received some form of neoadjuvant chemotherapy before any surgical or non-surgical LDT intervention. Most of them were fluoropyrimidine-oxaliplatin-based regimens. Four cycles of chemotherapy were administered prior to surgery in most cases. All patients received adjuvant chemotherapy post-intervention. In addition, based on K-RAS status, targeted cetuximab was given in the later part of the study.

Surgical resection was performed in 217 (57.9%) patients as a definitive modality for the CRLM, and 158 (42.1%) patients underwent non-surgical LDTs. Out of 217 patients undergoing surgical modality, 191 (88%) patients underwent only surgery, whereas 21 (9.7%) patients were treated by a combination of surgical resection and post-operative ablation by either RFA or microwave energy. Five (2.3%) patients underwent surgery after being downsized by TACE. Of the total, 92 (42.3%) patients underwent parenchyma-preserving non-anatomical liver resections, with 125 (57.6%) patients requiring a formal hepatic resection.

In the surgical cohort, median blood loss was 750 ml (range, 30–13,000 ml). The median hospital stay was 7 days (range, 3–65 days). The major (Clavien–Dindo Grade III and above) post-operative morbidity was 9.8%, and mortality was 1.8%. Post-operative liver insufficiency and bile leaks were the commonest complications occurring in 6.4% (n = 14) and 5.5% (n = 12) patients, respectively. A total of 60 patients had simultaneous resection of primary tumour along with liver metastasis and had similar post-operative major morbidity rates as compared to staged resection. All patients received chemotherapy after surgery.

Amongst patients receiving non-surgical LDT, 100 (63.29%) patients received ablation alone. Of the patients, 49 received TACE either as a standalone treatment option [n = 21; (13.92%)] or in combination with ablation [n = 28 (17.77%)]. Seven (4.4%) patients were managed by TARE as a standalone treatment option, while two were managed with a combination of TARE with ablation. There was no major complication or mortality in the non-surgical group with a median hospital stay of 2.5, 3, and 2.8 days for those undergoing ablation, TACE, and TARE, respectively ( Figure 2 ).

At a median follow-up period of 39 (range, 6–163) months, the median OS for the entire study population was 52 months (range, 6–163) months. At the last follow-up, 106 (28.3%) patients were alive and disease free, whereas 70 (18.7%) patients were alive with disease. Of the patients, 158 (42.1%) died due to the disease, and 30 were lost to follow-up (8%).

The median EFS was 16 months (range, 0–138 months). The 5-year overall survival for the cohort was 49.8%, whereas the 5-year EFS was 21.4% (refer to Figures 3A , 4A ).

The median OS for the patients undergoing surgical resection was 78 months (range, 6–163 months) compared to 39 months (range, 6–142) for those patients who underwent non-surgical LDTs, which was statistically significant (p < 0.05) (refer to Figure 3B ). Similarly, the median EFS was significantly better in the surgical group (20 months) as compared with the non-surgical LDT group (15 months) (p < 0.005) (refer to Figure 4B ).

Recurrence was seen in 133 patients out of the 217 undergoing surgical resection (61.2%), while 132 patients had disease progression (83.5%) after non-surgical LDT.

Better OS was observed in patients considered to have a resectable disease (median OS, 78 months) in the original multidisciplinary meeting as compared to patients with borderline resectable (median OS, 39 months) and unresectable disease (median OS, 29 months). The unresectable group had the worse survival (refer to Figure 5 ).

On univariate analysis, the OS was significantly worse in women as compared to men (0.002). Also, metastatic lesions more than 4 (p = 0.001) bilobar distribution of the metastatic lesions (p = 0.0001), borderline resectable (p = 0.039) as well as an unresectable disease at (p = 0.0001), elevated CEA levels more than 5 ng/ml (p = 0.016), and non-surgical LDT (p = 0001) were associated with poor OS (refer Table 3 ) Age, site of primary, size of tumour >5 cm, KRAS mutational status, synchronous or metachronous liver metastasis, neo-adjuvant or pre-procedural conversion chemotherapy, and T and N stage of disease did not affect OS.

A stepwise multivariate analysis, however, showed that only female gender (HR 1.79, 95% CI 1.30–2.47, p = 0.002), borderline resectable disease (HR 1.83 95% CI 1.26–2.64, p = 0.0001), unresectable disease (HR 2.91 95% CI 1.80 –4.69, p = 0.0001), and non-surgical LDTs as the modality of treatment (HR 1.43, 95% CI 1.01–2.02, p = 0.042) were independently associated with poor OS.