Is prior cancer history a hindrance for non-small cell lung cancer patients to participate in clinical trials?

With the development of novel treatment modalities, cancer patients’ survivals are improved, which contributes to a growing number of cancer survivors [1]. Therefore, patients with a prior cancer are not rare. Lung cancer is still the leading cause of cancer-related mortality worldwide [2, 3]. It is estimated that one in five lung cancer patients has a prior cancer [1, 4].

It is a longstanding practice that the non-small cell lung cancer (NSCLC) patients with a prior cancer are excluded from clinical trials [5, 6]. The researchers deemed that previous malignancies and treatments may have an adverse impact on the survivals of this population subset [7], and thus, interfere with study outcomes. However, little evidence has been found to support this notion. Whether this exclusion criterion is evidence-based remains unclear [8]. In the real-world setting, this criterion has prevented many potential candidates from participating in clinical trials which might improve their survival outcomes [9, 10]. In addition, this criterion could halt the pace and generalizability of clinical trials [11, 12]. So, there is a pressing need to revisit this so-called eligible criterion and explore the exact role that it plays in NSCLC patients’ survivals.

In such instances, the current study analyzed the data of resected NSCLC patients from a large Chinese cohort to determine the effect of a prior cancer on the prognosis of NSCLC patients. We aimed to answer the question that whether NSCLC patients with a prior cancer should be excluded from clinical trials.

In this study, we investigated the effect of a prior cancer history on the survivals of resected NSCLC patients from a large Chinese cohort. The results showed that patients with and without prior cancer had similar survival rates both before and after PSM. The LASSO-penalized multivariable Cox analysis demonstrated that a prior cancer history was not a prognostic factor for survivals. Therefore, our data supported the notion that it might be unreasonable to exclude the NSCLC patients with a prior cancer history from clinical trials.

In our study, the incidence of patients with a prior cancer was 8.2%, which was lower than the previous studies [7, 8, 18, 19] where the authors reported that the frequency ranged from 15 to 20%, but was similar to Aguilo et al.’s study [20]. Our data also showed that patients in the prior cancer ( +) group were younger than those in the prior cancer (-) group. The reason behind the difference could be explained by the fact that patients with a prior cancer could be more engaged in screening program [7], which might contribute to early diagnosis of NSCLC. In addition, our finding that patients in the prior cancer ( +) group had more early-stage tumors when compared with the patients in the prior cancer (-) group also supported the assumption. The early-stage tumors allowed more patients in the prior cancer ( +) group received VATS and sublobectomy. To date, there are scant data to compare the baseline clinicopathological characteristics between these two group patients [6‐8, 18‐22]. Our study provided comprehensive knowledge of these patients to address this gap.

Several studies have attempted to identify the role that a prior cancer history plays in the survivals of NSCLC patients. In the study by Monalve et al. [7], the authors reviewed the data of 821,323 NSCLCs from the National Cancer Database (NCDB) and demonstrated that the OS of the patients with a prior cancer was better than those without cancer history (9.8% vs. 9.5%, P < 0.001). They further clarified that a prior cancer history posed a negative impact on patients with early-stage tumors but posed a favorable impact on those with advanced-stage tumors [7]. The strength of their study was the large data set of the NCDB. However, although subgroup analyses were conducted in their study, the baseline imbalance of groups might compromise their conclusions. In addition, the 6^th edition of TNM staging system was used in their study, which might lend inconvenience to clinical practice. Pruitt et al. [18] presented a large study of 42,910 early-stage lung cancer patients of whom 21% had a prior cancer from the Surveillance, Epidemiology and End Results (SEER) database. Their data showed that patients with and without prior cancer had similar survivals which was akin to our results. Nevertheless, DFS was not an endpoint of their study. It is known that many clinical trials select DFS as the primary endpoint. Therefore, lacking the data of DFS might limit clinical reference value. Laccetti et al. [8, 19] also explored the effect of a prior cancer history on the NSCLC patients’ survivals from the SEER database and obtained a similar conclusion that a prior cancer history had no effect on patients’ survivals. In our view, the data recorded in the public databases is rough, and other potentially important factors such as preoperative comorbidity, postoperative complications and novel pathologic findings (LVI and VPI) was not thoroughly explored, which might lead to an unreliable conclusion.

We conducted a relatively large study of 4,102 resected NSCLC patients. To the best of our knowledge, this is the first study using the PSM method to reduce the bias caused by the observed confounding variables in the baseline characteristics. Thus, our results were more reliable and persuasive. The patients included in this study received standard treatments. In most previous studies however, the inclusive criteria were not rigor, and patients who received no treatment were also included, which could lead to bias [7, 8, 18, 19, 21]. Additionally, the dataset that we used was well-managed, which included well-annotated clinicopathologic data and complete follow-up information. So, our data was more comprehensive.

To date, whether a prior cancer history affects NSCLC patients’ survivals has not been clearly elucidated. On one hand, some researchers demonstrated that prior cancers and previous treatments may have an adverse impact on the survivals of this population subset [7]. On the other hand, patients with a prior cancer might be more compliant with regular reexaminations. It is no doubt that early detection and timely management of the NSCLC could improve these patients’ survivals. After balancing the baseline differences such as tumor stages, treatment modalities and comorbidities, our results found that patients with and without prior cancer still had similar survivals. Our findings were meaningful. In the current clinical practice, the prior cancer history has prevented many patients from having access to new therapies, which might improve their survivals [9, 10]. Furthermore, exclusion of these patients could halt the pace and generalizability of clinical trials [11, 12]. Our study added to the existing body of evidence on the topic that it was reasonable for clinical trials to enroll the NSCLC patients with a prior cancer.

Our study had several limitations. First, our database lacked the information of previous cancer status and prior cancer treatments modalities and timings, which were important prognosis factors for survivals. Therefore, well-managed individual databases are warranted to validate our results. Second, only resected patients were included in this study because the data about unresected patients is unavailable. Therefore, patient selection bias might exist in our study. In addition, our results should be cautiously interpreted owing to the relatively small number of patients with a prior cancer and the retrospective design of this study. Multicenter studies with large volume are necessary to validate our results.

In conclusion, our results demonstrated that resected NSCLC patients with and without a prior cancer had comparable survival rates. Herein, we proposed that it might be reasonable for clinical trials to enroll the NSCLC patients with a prior cancer.