Correspondence re L. J. Wirth et al., Induction Docetaxel and Carboplatin Followed by Weekly Docetaxel and Carboplatin with Concurrent Radiotherapy, Then Surgery in Stage III Non-Small Cell Lung Cancer: a Phase I Study. Clin Cancer Res 2003;9:1698–704.

We have read with great interest the paper from L. J. Wirth and co-workers reporting their experience with a multimodality treatment for locally advanced non-small cell lung cancer (LA-NSCLC; Ref. 1). In this report, the authors described the findings of a complex and multidisciplinary approach to stage III (a and b) NSCLC, exploring its feasibility with a Phase I dose escalation study design. In this setting, however, an analysis of the possible “induction” potential (i.e., tumor clinical and pathological response) of this protocol along with the possible effects on survival have been realized.

This experience is very similar to the one that we realized 2 years ago and reported previously (2, 3). Thus, we have drafted some reflections integrating the substantially common background and outcome with recent evidence and recommendations regarding the general approach to LA-NSCLC (4, 5).

First of all, we see the point that the definition of LA-NSCLC has not been homogeneous in the last 2 decades. Even if this may seem a collateral issue, it is actually the rationale for why the majority of small and large trials exploring therapeutic strategies for LA-NSCLC include patients with very different cancers. For this reason, the interpretation of results has been, and is still, somewhat confused. A functional definition as the one reported by Macchiarini et al. (5) does not resolve the problem of patient stratification inside clinical trials, especially when surgical indication follows chemotherapy and/or radiation treatment.

In an extreme simplification, in fact, it is widely accepted (even in the absence of proper consolidated results from Phase III randomized trials) that LA-NSCLC with clinical stage IIIa for a T₃N₁ or T_1–3N₂ condition could be “ideally” cured by a multimodality chemotherapy or chemo-radiation approach, eventually followed by surgery (and in carefully selected patients, adjuvant chemotherapy). On the other hand, when clinical IIIb cases for a T₄ or N₃ status are at stake, there is much more confusion.

Continuing along the lines of simplification, we could assert that this heterogeneity exists because, whereas in the first group of patients (IIIa), the fact that the induction treatment is chosen on the basis of “operability”, in the second group (IIIb), the treatment is chosen to modify “resectability,” which is a more immediate objective (subjective from the surgeon’s point of view) and strict parameter, and operability stands second in line.

Following the recent “guidelines” of Mactay and Jeremic (4), we can say that a direct surgical indication could exist in very selected T₄N_0–1 cases (with poor long-term results and a 5-year survival rate at <20%) at the price of extended and aggressive surgical approaches, whereas N₃ cases are generally not considered eligible for surgery at all.

The potential impact of a “neoadjuvant” approach for T₄ and/or N₃ cases on clinical tumor response and survival has been investigated in the past. A certain correlation among local control (clinical tumor response and, thus, clinicopathological downstaging) and survival seems to effectively exist (5, 6, 7, 8, 9, 10), especially if downstaging is obtained at the N level (7).

In our experience (2) and in that reported by Wirth et al. (1) several stage IIIb patients were included in the Phase I study (designed as dose finding); clinical response to the multimodality treatment was “incidentally” so good as to have some of those initially judged inoperable and unresectable re-enter the chance for complete surgical resection.

This evidence supported the idea that the two multimodality approaches (1, 2), even if designed as Phase I studies, proved to have an induction potential; moreover, a possible benefit on long-term survival could be hypothesized in the light of previously cited consolidated experiences (5, 6, 7, 8, 9, 10).

If we focus on the very group of IIIb patients studied by Wirth et al. (1), we see that one clinical T₄N₀ case who experienced a partial response to a chemo plus chemoradiation protocol, one clinical T₂N₃ case with partial response as well and one clinical T₂N₃ case with a preoperative question of progression of disease were all deemed to have achieved such satisfactory downstaging at restaging as to have re-entered operability and resectability and were operated on. The first and the second cases experienced significant pathological downstaging [with complete pathological response (clinical partial response) and with complete absence of tumor in the specimen of the latter] whereas the third was upstaged to stage IV because of intra-operative and pathological evidence of a neoplastic nodule in a lung lobe different from that of the primary tumor site.

We would like to amicably invite the authors to clarify some aspects of the following results:

(a) re-staging (T and N statuses in adjunct to the stage) should be disclosed so as to make the reader better understand the process of surgical indication;

(b) in particular, it would be interesting to understand why the T₄ patient got this clinical status and what was her restaging situation. In this setting, in fact, an important difference exists if a patient is T₄ because of an assumed neoplastic nodule in a different lobe of the same lung or because of a direct infiltration of, let’s say, the main branch of the pulmonary artery. Moreover, if this is the case, it is extremely difficult to outline the criteria (essentially attributable to technical heterogeneity and interobserver variability) for upgrading a case from T₃, because of a simple contact between the tumor and the mediastinal pleura around the vessel, to T₄, because of a clear vessel infiltration. This kind of difficulty is magnified when the assessment is made after chemotherapy or chemo-plus-radiation treatment has been administered when, at the moment of the restaging, the morphology of the therapy-induced biological effect is checked and interpreted (11).

(c) furthermore, we assume that the two N₃ cases had had their mediastinal controlateral nodal involvement status assessed via pathological verification of mediastinoscopic biopsy material indicated on a clinical suspicion. A comprehensive definition of the clinical N₃ status (which stations and how many, now many nodes, etc.) would be welcome; moreover, because both cases experienced a pathological downstaging at the N₃ level (one to N₀ and one to N₂) we would like to know how this was assessed: clinically only (computed tomography scan, positron emission tomography, and so forth); by re-do mediastinoscopy? pathologically by controlateral lymphadenectomy at operation? If re-do mediastinoscopy was performed, it would be of interest if the authors could discuss this procedure after such a complex chemotherapy plus chemoradiation protocol in the light of very important issues such as technical feasibility, biopsy reliability, and so forth, which, in our turn, we have reported (3) and discussed (12).

In this setting, in fact, we have experienced significant technical difficulty in re-do mediastinoscopy procedures performed after concurrent chemoradiation treatment (two patients): in both cases, the pre- and paratracheal plans were nullified, and the senior surgeon who performed the operations was not entirely sure to have re-biopsied the same area as in the first step (at the moment of mediastinoscopy); the pathologist disclosed a significant difficulty in assessing the nature of the specimen (neoplastic versus scar) at the frozen-section examination and could confirm the eventual absence of tumor only after a complete and definitive examination of the specimen.

We would like to close this short and friendly comment by asking the authors whether, in light of their experience, they agree that selected T₄ and/or N₃ LA-NSCLC patients should be offered a multimodality approach with neoadjuvant intent and whether surgery should be indicated in all cases in which resectability could be reasonably predicted on a coherent restaging: this may include, in the N₃ patients, a complex procedure (clinical and pathological) and, eventually, an extensive surgical approach.

In this setting, we feel strongly that an evidence-based consensus through the analysis and discussion of proper and focused clinical research experiences is still needed.