Severe hepatic sinusoidal obstruction and oxaliplatin-based chemotherapy in patients with metastatic colorectal cancer: a real entity?

We have read with interest the study of Rubbia-Brandt et al. [1] entitled ‘Severe hepatic sinusoidal obstruction associated with oxaliplatin-based chemotherapy in patients with metastatic colorectal cancer’. In this study, the investigators identified in the specimens from liver resection for colorectal metastases, some histological lesions of the non-tumoral liver parenchyma attributed to oxaliplatin-based chemotherapy. The results of this study raise some comments. First, it is well known that a variability in the assessment of the lesions and their severity may exist when examination is performed by multiple observers. In this study, the review of the slides was made by four pathologists, blinded to the administration of preoperative chemotherapy. In terms of histological evaluation, information is not given regarding the interobserver variation and the resolution of discrepancies. Second, ‘severe hepatic sinusoidal obstruction’ described by the authors refers to electron microscopic evaluation. However, no information was available concerning the number of cases studied by this method and comparison with the control group was not performed. This information is mandatory to attribute this entity to all the veno-occlusive lesions. Third, while many histological items have been evaluated, the authors mentioned that no lesions at all were found in the group without chemotherapy. This is surprising, considering the usual frequency of histological lesions in the general population, particularly in those over 60 years old. For example, steatosis is not included in Table 1. However, the authors themselves mentioned a 50% rate of steatosis in the group without chemotherapy. By ‘no hepatic lesion’, did the authors mean ‘no centrilobular vascular lesion’? Among the histological findings, the authors did not evaluate the portal fibrosis, which may be observed in old patients and in the non-tumoral liver parenchyma adjacent to tumors (due to the compressive effect of the tumor). As the mean age of patients in the study was 62 years, and the specimen usually involves a limited amount of non-tumoral liver, it is likely that some samples contained fibrosis. Fourth, a 74% incidence of vascular lesions (20/27) was observed in the oxaliplatin group. This result is not in keeping with our experience for both the frequency and the type of lesions observed in our oxaliplatin-treated patients. In a series of 52 patients treated with oxaliplatin-based chemotherapy, we found 38% (20/52) of severe vascular lesions and we never observed veno-occlusive disease (unpublished data). Whether this difference could be related to a lower toxicity of chronomodulated infusion of chemotherapy in our patients compared with continuous infusion, deserves further investigation. Fifth, regarding the evolution of histological lesions, only three cases presenting with histological lesions were assessable for long-term follow-up. One of them had a progression of the fibrosis. We think it is premature to draw any conclusion from this observation. Finally, the clinical impact of histological lesions after chemotherapy on the post-operative outcome is of particular interest. We understand that this was not the aim of this study. However, this merits further investigation. From our extensive experience, including over 700 patients treated mainly by oxaliplatin-based chemotherapy, we did not observe any increased operative mortality in those 95 patients who underwent further liver resection [2]. In summary, while we agree that oxaliplatin-based chemotherapy can induce histological lesions of the liver, the entity ‘severe hepatic sinusoidal obstruction’ described by the authors requires more detailed information. The reported high frequency and its clinical relevance merit further exploration before considering that it could constitute a limit to the increasing use of chemotherapy. Indeed, the strategy of chemotherapy followed by liver surgery is the only one that provides long-term survival and even cure to patients with initially unresectable colorectal cancer [2].