Pulmonary metastasectomy: the discussion continues

The authors of the PulMiCC trial (PT) have been very passionate and dedicated to the topic of pulmonary metastasectomy. The fact that so many papers were published prior to and after the results of the PT itself demonstrates how much work and effort they have put into this topic. Overall, the PT results were published in 3 steps: first, a paper including 65 randomized patients [1]; second, a publication including 93 randomized patients [2]; and third, an observational publication including 512 mostly nonrandomized patients [3]. In their actual review, the PulMiCC authors (PA) are evaluating the response to their trial and criticize that 40 of 65 studies refer only to their first ‘preliminary’ publication that included 65 patients [4]. The question arises why the PA first published data on 65 patients, then 18 months later published data on 93 patients, and finally published another study including nonrandomized patients. It is obvious that this would have led to confusion and raise questions.

We are well aware that performing a randomized controlled trial is difficult in such a wide field, which is influenced by many different factors such as different types and cycles of chemotherapy given, additional lymph node metastases, number of metastases, molecular genetics of the tumour proven by certificated pathologists, and other factors of which some were evolving factors and can not be included in a trial. Furthermore, the surgeon influences the type of resection and lymph node dissection. Despite these difficulties, randomized controlled studies are considered the best study design, which makes a mutual problem-oriented understanding between surgeon and statistician all the more important.

However, it is interesting that although pulmonary metastasectomy and hepatic metastasectomy are associated with a mortality rate of <1%, oncolgists and thoracic surgeons are discussing the resection of 3 lung metastases, for example. In liver surgery, multiple metastases are resected by default, or further techniques are used to resect liver metastases such as associating liver partition and portal vein ligation for staged hepatectomy, biembolization and up to liver transplantation [5]. However, this discussion is beyond this invited commentary paper.

However, through the PT and further publications, the authors were able to show that even nonoperated patients can achieve a 5-year survival of ∼20–30% [4]. Unfortunately, the long-expected answer of the PT, whether pulmonary metastasectomy provides a survival benefit in these patients, could not be answered [1–3].

Here are a few excerpts from the trial conclusions:

Because of poor and worsening recruitment, the study was stopped. [1]

No inference can be drawn about the effect of metastasectomy on survival in this observational study. [3]

It is understandable that the PA who has invested so much work might be disappointed. We think that all investigators working in this field regret that the results are not clearer in either direction.

However, in the current paper, the authors are now investigating the responses to the PT [4]. In our opinion, this study gives the impression that the cause of the inconclusive result and the supposed lack of acceptance are now being sought in the thoracic surgery and oncology communities. The authors write about a ‘widespread disregard for controlled trials’ [4] and a ‘widespread reluctance to seriously consider the results of a carefully conducted randomized trial’ [4]. We think that the authors are not doing justice to the colleagues who have also invested a lot in their patients and studies. We believe that many colleagues deal with the PT and their results for their patients and out of self-interest. The fact that some studies do not lead to the hoped results, or that the results are interpreted differently, is not a new phenomenon.

One question that arises is why out of 512 recruited patients, only 93 could be randomized, which is only 18.16% of the total. It has to be questioned if this is truly a good recruitment and randomization rate. As an explanation, the PA wrote that in the 3 largest of their centres in England, the patients' wish to participate in a randomized controlled trial was overridden [4]. Here, our opinion is that the PA should perhaps talk to their clinical study partners.

However, the authors are writing in the current paper that the publication of Handy et al. [6] and a publication from our clinic [7] in particular have affected the recruitment [4]. From available information, the recruitment of the PT was between 2010 and 2016 and was stopped because of ‘poor recruitment’ [1, 2]. How can 2 papers published in 2017 and 2019 influence recruitment? In addition to 1 centre from Serbia, which included 28 patients, 65 patients from 13 centres were randomized in England between 2010 and 2016. This means that in England, on average, 0.83 patients were randomized per centre per year. We know and appreciate that the colleagues have invested a lot of work, but even before the 2 studies were published, the randomization rate was low.

In the current paper, the authors complain that their studies are not properly cited. Let us stay with the 2 publications by Handy et al. and Schirren et al.; the PA is writing that these 2 papers are giving the impression of a gap in 5-year survival of ∼60% between nonoperated patients and operated patients. However, neither of the 2 publications is writing this. Furthermore, the PA are writing that the publication by Handy et al. assumed a 5-year survival of 0% for patients without metastasectomy. If one reads the quoted sentence of this publication completely, it is written: ‘Historical controls are used or metastatic disease survival is assumed to be zero, a contention not supported by the literature’ [6]. In addition, in our publication, we never wrote that untreated patients have no chance of 5-year survival. We wrote: ‘Of course, it is possible that, within the selected patient group, there is a portion that will arrive at long-term survival without surgery and therapy’. We only point out that it is not known which patient will benefit from metastasectomy and which will not. Furthermore, we emphasize that pulmonary metastasectomy, in addition to supposed prolonging of survival, might also be a tool for well-deserved breaks in systemic therapy for the patient [7]. However, anyone doing and practicing science should take care to cite papers correctly and fairly.

Likewise, the authors are writing the trial was stopped because of ‘poor and worse recruitment’ [1, 2]. What kind of response did the authors expect to their long-expected and much-promoted trial? Surely it is understandable that the community would immediately engage with their publication. We also did and included it in our most recent study at the time [8]. Is it truly a mistake to cite a conclusion? We do not think it is in bad faith or that the authors are writing in a ‘widespread disregard’. The current study included 40 publications that cited their ‘preliminary’ results, again which they themselves summarized as suffering from poor recruitment. We do not understand what response the authors expected to this sentence. Doing a study now, where more than half of the included studies have contained this conclusion, can in our opinion only lead to a ‘negative response’. The authors are often writing about bias, and we think here might be bias on their part.

Finally, we would like to emphasize that we appreciate the work and commitment of the PA in the field of pulmonary metastasectomy, but overall, the current study gives the impression that the cause of the inconclusive results or a supposed unacceptance is now being sought by others. In our view, the PA should rather be as critical with their own work as they are in addressing the issue of pulmonary metastasectomy and the responses to their work. In this way, the discussion of pulmonary metastasectomy might someday come to an end.

REFERENCES