We would like to comment on the very interesting, albeit controversial, article by Lautz et al titled “Successful nontransplant resection of POST-TEXT III and IV hepatoblastoma.”1 The authors state that 93% (13 of 14 patients) of POST-TEXT III and IV hepatoblastomas referred to their institution were successfully resected by conventional means, despite the patients being potential candidates for liver transplantation. However, the diagnostic staging of some of these tumors raises concerns because the original PRETEXT was unknown and based on local assessment only in 6 patients referred from other institutions, whereas we know from the SIOPEL 4 trial analysis that 20% of patients classified as PRETEXT IV locally were reclassified on central review (unpublished data). In addition, in our opinion, figure 1A in Lautz et al, which is presented as POST-TEXT IV, appears to be POST-TEXT III on the images shown and hence resectable by conventional left trisegmentectomy. Moreover, 3 of the 8 remaining cases were PRETEXT III (only 1 of them was associated with complete portal involvement: P2), which suggests that at least two of them were potentially resectable from the beginning of treatment. Therefore, comparison of this study with other series and trials is difficult.
The authors report 93% survival after nontransplant resection in patients who might have otherwise been considered for primary transplantation, but their follow-up is less than 2 years in 4 patients, and the 5-year overall survival seems to be 88% with a 5-year event-free survival of 75%. In addition, survival in the last case is uncertain taking into account the long interval from the resection to the first relapse (48 months) and insufficient follow-up after the second resection (30 months). Thus, in our opinion, the 5-year event-free survival of 75%, as stated in the report, is a more useful measure. This is similar to the SIOPEL experience.2, 3
We are concerned that the message from this article might be misleading and might encourage inappropriate surgical strategy in less experienced hands. In particular, “shaving a margin or performing a separate wedge resection from the remaining sector of liver” and “using an ultrasonic device to clear residual hepatocytes from the vessel when the abutting tumor is truly adherent” are not standard practices and should not be recommended when liver transplantation is available.4, 5
Ten patients had a POST-TEXT III, centrally located tumor with encasement, invasion, or close proximity of the main hilar structures or all 3 hepatic veins after neoadjuvant chemotherapy. The details of vascular involvement reported in table 1 of Lautz et al (P0V2, P1, P1, P0V1, P1, P0, P1V1, P1V1, P0V1) suggest that the majority of these patients would not require transplantation under current surgical guidelines. The exception is the 2 PRETEXT IV multifocal cases downstaged to POST-TEXT III by neoadjuvant chemotherapy. The policy of the SIOPEL group remains that these patients should be treated by transplantation, until there is evidence that this is unnecessary.4 Ideally, a randomized, controlled trial would be performed, but this is unlikely to be practical.
The authors state that “Although inconclusive, existing evidence does not link microscopic residual disease with tumor recurrence or death.” Indeed, in most patients in the SIOPEL studies, a microscopic positive margin was not associated with worse outcome.2 Various explanations may be proposed for this finding. It is possible that microscopic residual disease in a chemosensitive tumor will reliably succumb to adjuvant chemotherapy. Alternatively, it may be that microscopic residual disease is a risk factor for recurrence, but that existing studies have failed to detect this effect. The actual surgical margin may be difficult to interpret, because cautery artifact may result in a thick eschar of burned tissue. Microscopic residual tumor found in the operative specimen may have no counterpart on the preserved liver. Also, the ultrasonic dissector aspirates approximately 5 mm of liver tissue at the margin of resection, and so may have removed all of the residual tumor.
Despite this, the general rule should remain a complete, surgical resection with a margin of at least 1 cm of normal parenchyma. The risk of recurrence is real if this rule is not respected.
Finally, we have a major problem when the article states “There is no substitute for operative exploration to make an ultimate determination of resectability.” The authors do not explain how to perform this surgical exploration. In our opinion, the identification of cases where extended resection is possible should not be based on surgical exploration alone, but also rely on preoperative imaging confirmed by intraoperative ultrasonography.
In our international experience, we have seen cases where surgeons attempted extended liver resection, starting to cut through the parenchyma to end in a cul-de-sac, with the child running into major bleeding-related liver failure. Despite the good results achieved, we think that the authors' message might be misleading and result in undesired risks for patients.