The present retrospective cohort study included 36 patients who received RAM plus FOLFIRI treatment for unresectable CRC as second-line or salvage chemotherapy. The study was conducted between September 15, 2016, and May 5, 2019, at the Kawasaki Medical School Hospital (Kurashiki, Japan) and Okayama University Hospital (Okayama, Japan). Relevant clinical data, including patient demographics, laboratory parameters, imaging findings, treatment toxicities and prognostic outcomes, were retrospectively collected.

The RAM plus FOLFIRI regimen was administered intravenously every 2 weeks. RAM was administered at a dose of 8 mg/kg over 1 h on day 1, folinic acid was administered at a dose of 400 mg/m² over 2 h on day 1, irinotecan was administered at 150 mg/m² over 150 min on day 1 of each cycle; fluorouracil was administered intravenously as a bolus of 400 mg/m² on day 1, followed by continuous infusion of fluorouracil at a dosage of 2,400 mg/m² over 46 h on days 1–3. Dose adjustments or interruptions were made in accordance with the institutional clinical practice guidelines. Treatment response was evaluated using the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 (9). Regular imaging evaluations, predominantly using computed tomography, were performed every 2 months, corresponding to once every four treatment cycles, as per standard institutional practice.

The patients were divided into two treatment groups: The second-line group consisted of 58% (21 patients) of the patients and included patients who received RAM plus FOLFIRI as a second-line treatment, whereas the salvage-line group consisted of 42% (15 patients) of the patients, who were treated with RAM plus FOLFIRI as salvage therapy (third-line or later). The duration of BEV administration was calculated before the administration of RAM, excluding the BEV withdrawal period. In the second-line group, among 21 patients, 4 did not use BEV in the first-line treatment, whereas the remaining 17 received BEV plus oxaliplatin with a fluoropyrimidine (5-FU or capecitabine). In the salvage-line group, the duration of BEV was calculated from the treatment history. Among the 15 patients in the salvage-line group, 13 were treated with BEV plus oxaliplatin with a fluoropyrimidine (5-FU, capecitabine or S1) as first-line therapy. Of the remaining 2 patients, 1 patient received BEV plus irinotecan with capecitabine; the other received panitumumab plus oxaliplatin with 5-FU. BEV was used in 8 of the 15 patients in the second-line treatment and was administered again in 3 patients in the third-line or later chemotherapy before RAM administration.

The criteria for hepatic resection were as follows: i) LMs must be considered surgically resectable (regardless of the number of LMs); ii) the disease must be controlled by chemotherapy [partial response (PR) to stable disease (SD) for ≥2 months]; iii) the only other organ with metastasis, other than the liver, was the lungs.

Formalin-fixed paraffin-embedded tissue specimens were used for immunohistochemical (IHC) staining. Serial 3-µm tissue sections were prepared and subjected to either hematoxylin and eosin (H&E) staining or IHC staining using mouse monoclonal anti-CD42b (cat. no. NBP1-28457; clone MM2/174; Novocastra; Leica Biosystems). Immunostaining was performed using an automated Bond-III Stainer (Leica Microsystems GmbH) according to the manufacturer's instructions.

Briefly, tissue specimens were fixed in 10% neutral-buffered formalin at room temperature for 24–48 h. The sections were first dewaxed with BOND Dewax Solution, after which, antigen retrieval was performed by heating the tissue sections in BOND Epitope Retrieval ER1 Solution (citrate-based buffer, pH 6.0) at 95–100°C for 20 min. Endogenous peroxidase activity was quenched using the Refine Detection Kit Peroxide Block (Leica Biosystems) for 5 min at room temperature and blocking was performed using normal goat serum (cat. no. 5425; Cell Signaling Technology, Inc.) at room temperature for 20 min. Subsequently, the sections were incubated with anti-CD42b (1:100) at room temperature for 30 min. Secondary detection was performed using the Refine Detection Kit Polymer (cat. no. DS9800; Leica Biosystems) at room temperature for 10 min. Chromogen detection was carried out using the Refine Detection Kit Mixed DAB Refine (Leica Biosystems), according to the manufacturer's instructions. For H&E staining, the sections were automatically stained using the Bond-III Stainer with 0.1% hematoxylin for 5 min at room temperature, followed by 0.5% eosin staining for 2 min at room temperature.

Images were captured using a light microscope (Olympus BX53; Olympus Corporation) with a 40× objective lens. CD42b immunostaining results were categorized into four stages based on the observed staining pattern: Score 0, absence of deposition; score 1, presence of fine granular deposition; score 2, presence of coarse granular deposition; and score 3, presence of linear deposition.

Statistical analyses were performed using JMP Pro 17 (SAS Institute, Inc.). PFS was defined as the time from the initiation of RAM plus FOLFIRI treatment to disease progression. OS was defined as the time from the initiation of RAM plus FOLFIRI treatment to death. Cumulative survival probabilities were estimated using the Kaplan-Meier method. The ORR was calculated as the proportion of patients achieving a complete response (CR) or PR, and the disease control rate (DCR) was defined as the proportion of patients achieving CR, PR or SD. In cases where the RECIST evaluation differed from the attending physician's assessment, the latter's judgment was included in the evaluation. Differences in CD42b scores, presented as median (IQR), and the mean duration of anti-VEGF antibody treatment before liver resection between the groups were evaluated using a Mann-Whitney U test. The correlation between the CD42b score and the duration of anti-VEGF antibody treatment (months) before liver resection was examined using logistic regression analysis with Spearman's coefficient (ρ). Categorical variables were examined by Fisher's exact test. P<0.05 was considered to indicate a statistically significant difference.

This study evaluated 36 patients with mCRC who were treated with RAM plus FOLFIRI. The characteristics of the patients enrolled in the present study are shown in Table I. Of these, 58% were female. With respect to primary tumor location, 31% of primary tumors were on the right side. Additionally, 67% presented with LMs and 94% had an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0–1 (10). Moreover, 64% of the patients had a mutant RAS status.

In the second-line group, the median age was 57 years (range, 16–73 years) and 57% of patients were female. All patients in this group had an ECOG PS of 0–1. Of the patients in this group, 29% had right-sided primary tumors located from the cecum to the transverse colon, 62% presented with LMs and 67% had a mutant RAS status. In the salvage group, the median age was 63 years (range, 37–76 years) and 60% were female. Among these patients, 87% had an ECOG PS of 0–1. Additionally, 33% had right-sided primary colon tumors, 60% had a mutant RAS status and 73% had LMs.

The present study also reviewed the duration of BEV treatment prior to RAM administration, categorizing it into the following: No use, <6 months, 6–12 months, and ≥13 months. Across the entire cohort, 4 patients had not been treated with BEV prior to RAM, 9 were treated for <6 months, 7 for 6–12 months and 16 for ≥13 months. In the second-line group, 4 patients had no prior BEV treatment, 6 were treated for <6 months, 6 for 6–12 months and 5 for ≥13 months. In the salvage group, all had received prior BEV treatment, with 3 treated for <6 months, 1 for 6–12 months and 11 for ≥13 months.

On average, patients underwent 7.4 cycles of RAM plus FOLFIRI treatment [95% confidence interval (CI), 5.3–9.5]. Those in the second-line group averaged 8.2 cycles (95% CI, 5.0–11.4), whereas the salvage group averaged 6.2 cycles (95% CI, 3.7–8.7). The number of chemotherapy lines after RAM administration was not significantly different between the second-line and salvage groups for 0–2 lines (excluding cases with unknown data), although there were slightly different percentages of patients undergoing 1 or 2 lines of chemotherapy in each group. Notably, only patients in the second-line group had ≥3 chemotherapy lines.

Finally, clinical responses were assessed. Neither the second-line nor salvage treatment groups exhibited CR. The ORR was 10% in the second-line group and 7% in the salvage group. The DCR was notably higher in the second-line group at 81% compared to 47% in the salvage group. In the second-line group, 71% of patients displayed SD, whereas in the salvage group, this was observed in 40%. PD was observed in 14% of the second-line group and 53% of the salvage group.

PFS and OS were evaluated following the initiation of RAM plus FOLFIRI treatment. For the second-line group, the median PFS was 5 months (95% CI, 3–9 months; Fig. 1A). Conversely, the median PFS in the salvage group was shorter, at 2 months (95% CI, 1–4 months; Fig. 1C). The median OS in the second-line group was 23 months (95% CI, 12–34 months; Fig. 1B), which is in contrast to the salvage group, where the median OS was 8 months (95% CI, 5–19 months; Fig. 1D).

Of the 36 patients, 14 (39%) underwent surgical resection of LMs during chemotherapy. In this subset, 6 patients underwent their first surgical resection for LMs during second-line RAM plus FOLFIRI treatment, forming the RAM-LM group. Of the remaining 8 patients, 2 underwent their first resection during first-line chemotherapy before BEV initiation and thus these cases were excluded, and 6 underwent their first resection during BEV-based first-line chemotherapy, constituting the BEV-LM group (Fig. 2A). The details of patients who underwent LM resection can be found in Table II. The RAM-LM group showed a trend toward a shorter duration of BEV administration, except in one case; this was due to the early failure of front-line BEV plus oxaliplatin-based chemotherapy. Notably, 1 patient did not receive BEV but was treated with RAM plus FOLFIRI as a second-line regimen because of relapse during oxaliplatin-based adjuvant chemotherapy.

Since H&E staining showed no hepatic sinusoid injury in any LM samples resected after BEV treatment or those resected after BEV followed by RAM administration (data not shown), the association between anti-VEGF antibody exposure duration and platelet aggregation (CD42b) score was subsequently assessed. While CD42b score in the liver sinusoids of the BEV-LM group was influenced by the duration of BEV treatment only, 1 of the 6 patients in the RAM-LM group was BEV-naive (CD42b score=3), while the remaining 5 patients received BEV prior to RAM; therefore, the CD42b score of the RAM-LM group was influenced by the duration of both BEV and RAM treatment. Although there was no significant difference in the mean duration of anti-VEGF antibody treatment between the BEV-LM and RAM-LM groups (Table II), the RAM-LM group had a notably higher median platelet aggregation (CD42b) score (median value, 3; range, 0–3) compared with the BEV-LM group (median value, 1; range, 0–3; P=0.01; Fig. 2B and C). Additionally, a tendency for a negative correlation between the duration of anti-VEGF antibody exposure and the CD42b score was detected in both the BEV-LM and RAM-LM groups (Fig. 2D).