Outcomes and Prognostic Factors of Patients with Metastatic Colorectal Cancer Who Underwent Pulmonary Metastasectomy with Curative Intent: A Brazilian Experience

Abstract

Background

We aimed to identify clinicopathological and molecular features associated with progression‐free survival (PFS) and overall survival (OS) after pulmonary metastasectomy for metastatic colorectal cancer in a retrospective cohort in Brazil.

Materials and Methods

We did a retrospective review of thoracic surgeries performed in a single large academic hospital in Brazil from January 1985 to September 2019. Demographics, previously described prognostic factors, and clinicopathological and molecular characteristics were abstracted. Univariate Cox regression was performed for each variable, and, when significant, data were dichotomized to provide clinically meaningful thresholds.

Results

Records from 698 patients were reviewed. Fifty‐eight patients underwent pulmonary metastasectomy with curative intent. Of those, 53.4% had a single metastatic lesion. The median size of the largest lesion was 1.5 cm. Results of RAS, RAF, and mismatch repair testing and of cytokeratin 20 (CK20) and CDX2 testing were available for 13.8% and 58.6% of the sample, respectively. Median PFS was 14 months, median OS was 58 months, and 5‐year survival was 49.8%. Unfavorable prognostic factors for OS included disease‐free interval (DFI) <24 months, synchronous presentation, size of the largest lesion ≥2 cm, and loss of CK20 expression. Presenting with more than one lesion was prognostic for PFS but not for OS.

Conclusion

In this Brazilian cohort, our findings corroborate existing data supporting DFI, synchronous presentation, and number and size of lesions as prognostic factors. Furthermore, we found that loss of CK20 expression may be associated with more aggressive disease and shorter OS. Additional molecular prognostic factors after pulmonary metastasectomy for colorectal cancer should be further explored.

Implications for Practice

This study consolidates disease‐free interval, synchronous presentation, and number and size of lesions as clinically relevant data that may help guide therapy for patients with colorectal cancer and lung metastases who are candidates for curative‐intent metastasectomy. Additionally, in this sample, lack of cytokeratin 20 expression in metastases was associated with shorter progression‐free survival and overall survival, suggesting that biomarkers also may have a role in guiding therapy in this setting and that additional biomarkers should be further explored.

Short abstract

Pulmonary metastasectomy is increasingly used for select patients with metastatic colorectal cancer. This article explores potential prognostic factors and outcomes of patients with metastatic colorectal cancer treated with pulmonary metastasectomy in Brazil.

Introduction

Metastasectomy represents a potentially curative approach for metastatic colorectal cancer (mCRC) [1]. Although metastasectomy is an option for only about 20% of patients, given the ability to cure approximately 17% of patients with metastatic disease, metastasectomy should be considered for those who may benefit [2]. This is currently underscored by several guidelines and is considered standard treatment when resection is technically feasible [3, 4, 5]. Although hepatic metastasectomy is more frequent, pulmonary metastasectomy is also supported by guidelines and has been increasingly adopted for select patients [6, 7].

The International Registry of Lung Metastases, a multinational database of lung metastasectomies performed in 18 centers across Europe and North America, laid the foundation of lung metastasectomy and served as a benchmark of both safety and outcomes of these procedures [8]. In mCRC, the adoption of pulmonary metastasectomy was further supported by a meta‐analysis of observational studies that reported a 5‐year overall survival (OS) of 42%, which is far better than the historical 5‐year overall survival of 10% in mCRC [9]. However, thus far all the data for pulmonary metastasectomy in colorectal cancer are based on observational and retrospective studies and are thus subject to limitations of observational data [10].

The PulMiCC (Pulmonary Metastasectomy in Colorectal Cancer) trial (NCT01106261) [11] intended to fill this gap in evidence by comparing pulmonary metastasectomy with active monitoring in patients with mCRC. However, it was stopped because of poor recruitment after the accrual of only 65 patients. Although the small sample precludes definitive conclusions, the estimated 5‐year survival rate in this study was 38% (23%–62%) for patients in the metastasectomy group and 29% (16%–52%) in the control group, suggesting that patient selection may be key for the benefit of this procedure shown in previous retrospective studies. Additionally, an exploratory analysis of the reasons for not randomizing patients demonstrated that when patients made the decision on their own after being fully informed in the consent and assessment periods, 53.6% (22/41) decided to undergo metastasectomy. On the contrary, when clinicians made the decision, 99% (77/78) of patients would undergo metastasectomy [12]. These data support that from the clinicians’ perspective, there may be lack of equipoise, although without randomized data an accurate assessment of the benefit of pulmonary metastasectomy remains uncertain.

As strong supporting evidence on which patients can be cured by resection is still lacking, decisions are made on a case‐by‐case basis weighing individual surgical risk against clinical features that may be associated with better outcomes after lung metastasectomy such as disease‐free interval (DFI), number of lesions, size of the largest lesion, preoperative carcinoembryonic antigen (CEA) levels, and presence of positive lymph nodes [8, 13, 14, 15, 16, 17, 18]. Molecular prognostic factors in this context have not yet been thoroughly explored, and additional prognostic factors are still being described. In this study, we further explore potential prognostic factors and outcomes of patients treated with pulmonary metastasectomy for mCRC in our institution.

Materials and Methods

We did a retrospective review of thoracic surgeries described as nodulectomy, segmentectomy, or lobectomy performed in an academic hospital in Porto Alegre, Brazil, from January 1985 to September 2019. We identified patients who underwent pulmonary metastasectomy for mCRC, then abstracted data from the electronic medical record, including patient demographics, histologic type, pathologic grade, lymphovascular invasion, Crohn's‐like lymphoid reaction, number of positive and total resected lymph nodes in primary disease, use of adjuvant chemotherapy for initially localized disease, microsatellite stability status, RAS and BRAF status, primary tumor stage, DFI, surgical techniques, surgical time, surgical complications, preoperative forced expiratory volume in 1 second (FEV1) and forced vital capacity (FVC), readmissions, pathologic margins, size and number of metastatic lesions, preoperative CEA both in primary disease and relapse, staining of cytokeratin 20 (CK20) and CDX2 in metastasis, presence of lymph node metastases, use of adjuvant or perioperative chemotherapy, relapse, and mortality.

Use of complementary ablation techniques, such as percutaneous ablation or stereotactic body radiation therapy (SBRT), was allowed. Use of 18‐fluorodeoxyglucose positron emission tomography or magnetic resonance imaging was not mandatory for inclusion. For the purposes of this study, patients had their staging reclassified according to the 8th edition of the American Joint Committee on Cancer/International Union Against Cancer's Cancer Staging Manual to assure applicability of data. Performance status was recorded according to the Eastern Cooperative Oncology Group scale. Only patients who underwent surgery with curative intent and for whom resection of all metastatic lesions was attempted were included in this analysis. Patients who had already been discharged from follow‐up were contacted by phone to assure progression and survival data accuracy.

DFI was defined as the time between resection of the primary disease and the first clinical or radiological evidence of relapse. Progression‐free survival (PFS) was defined as the time interval between resection of metastatic disease and clinical or radiological progression or death. Comorbidities were quantified using the Charlson Comorbidity Index (CCI), a risk score that has already demonstrated impact on both long‐term survival and complications after thoracic surgery [19]. The CCI was calculated at the time of decision‐making process on whether to perform a metastasectomy. Weighted comorbidity classes were attributed based on scores and defined as low if 0 points, medium if 1–2 points, high if 3–4 points, and very high if ≥5 points. Postoperative complications were recorded according to the Clavien‐Dindo classification [20], ranging from minor deviations of ideal postoperative course without the need of pharmacological therapies (grade I) to those that need pharmacologic therapies (grade II), those that need either surgical, endoscopic, or radiological intervention (grade III), life‐threatening complications (grade IV), and death (grade V). Grade III was further subdivided based on the need for regional or local anesthesia (grade IIIa) or general anesthesia (grade IIIb), as well as grade IV, based on single‐organ failure (grade IVa) or multiorgan failure (grade IVb). Death was considered postoperative mortality if within 90 days of surgery and/or in the same hospital stay.

Statistical Analysis

This study was exploratory, and the sample comprised all available and eligible cases from our institution. Normality of data was assessed using visual inspection and the Shapiro‐Wilk test. Survival time distributions were plotted using the Kaplan‐Meier method for both OS and PFS, and 95% confidence intervals (CIs) are presented according to Greenwood's formula. Median PFS and median OS with 95% CIs were estimated using the Brookmeier and Crowley method. To explore prognostic factors associated with prognosis after lung metastasectomy, we performed univariate Cox proportional hazard regressions for each variable of interest. Continuous variables that demonstrated a statistically significant association with either PFS or OS in univariate Cox regression are presented dichotomized to ease data interpretation. Two‐sided p values for both Cox regression and log‐rank tests are provided. A value of p ≤ .05 is considered significant. The statistical calculations were performed using STATA version 16.0 for Windows software.

Results

Records from 698 patients were reviewed. Fifty‐eight were identified as having had a pulmonary metastasectomy for metastatic colorectal cancer. In total, 140 metastatic lesions were resected in 86 procedures. Metastasectomy was performed once for 63.8% of patients, and 25.9%, 8.6%, and 1.7% of patients had two, three, or four surgeries in different operative times, respectively. Demographic, clinical, and surgical information are shown in Table 1.

Table 1.

Demographic and clinicopathological characteristics

Characteristics	% (n = 58)
Age, median (IQR), years	64 (58–70)
Age, years
<50	5.1
50–59	25.9
60–69	41.4
70–79	24.1
≥80	3.4
Sex (men)	55.2
Active or previous smoker	39.6
Primary site
Colon
Ascending	6.9
Transverse	3.4
Descending	10.3
Sigmoid	22.4
Rectum
Upper	8.6
Middle	29.3
Lower	13.8
CEA at diagnosis, median (IQR), ng/mL	5 (2.17–17)
CEA at diagnosis
<5 ng/mL	48.9
≥5 ng/mL	51.1
Lesion obstructive at diagnosis	39.7
Stagea
I	3.5
II	28.1
III	49.2
IV	19.2
Lymph nodes resected, median (IQR)	19 (14–25)
Grade
1	5.4
2	87.3
3	7.3
Adjuvant chemotherapy
Colon
No	36.3
Adjuvant	63.7
Rectum
No	15.2
Neoadjuvant only	3
Adjuvant only	27.3
Both neoadjuvant and adjuvant	54.5
Disease‐free interval, median (IQR), months	13 (6–26)
Disease‐free interval
<24 months	74.1
≥24 months	25.9
CEA at first presentation of lung metastasis, median (IQR), ng/mL	2.3 (1.3–4.7)
Synchronous resectable liver disease	25.9
Charlson Comorbidity Index (relapse)
Low (0)	3.4
Medium (1–2)	46.6
High (3–4)	44.8
Very high (≥5)	5.2

^a One missing.

Abbreviations: CEA, carcinoembryonic antigen; IQR, interquartile range.

Median age was 64 years, and 55.2% of the patients were male. Primary tumor was left‐sided in 89.7% and located in the rectum in 56.9% of patients. Median CEA at diagnosis was 5 ng/mL (interquartile range [IQR], 2.17–17). Metastatic disease was synchronous in 19.2% of patients.

Fifty‐three percent of patients had a single metastatic lesion, and the number of lesions ranged from one to eight. Median size of largest lesion was 1.5 cm (range, 0.4–4.2 cm). Median CEA at relapse was 2.3 (IQR, 1.3–4.7), and median DFI was 13 months (IQR, 6–26). At the time of metastasectomy, 25.9% of patients had synchronous resectable liver disease, and 3.4%, 46.6%, 44.8%, and 5.2% had low, medium, high, and very high CCI, respectively.

Data regarding surgical outcomes and pathological analysis of metastases are described in Table 2. A wedge resection was performed in 87.9% of patients, whereas 12.1% had a lobectomy. Nodal sampling was performed in 27.3% of patients, and nodal disease was found and resected in 5.2%. Thirty‐day readmission rate was 5.2%, and only one patient had a grade IIIb or higher complication according to Clavien‐Dindo classification. Margins were negative (R0) in 89.1% of patients, whereas a grossly positive (R2) margin was found in one patient. Results of RAS, BRAF, and mismatch repair (MMR) testing were available for 13.8% of patients, and results of CK20 and CDX2 testing in the resected lesion were available for 58.6% of patients.

Table 2.

Surgical and pathological characteristics

Characteristics	% (n = 58)
Resection
Wedge resection	87.9
Lobectomy	12.1
Margins
R0	89.1
R1	9.2
R2	1.7
Numberof procedures
1	63.8
2	26.2
3	8.6
4	1.7
FEV1, median (IQR), L	2.33 (1.9–2.9)
FEV1, median (IQR), % predicted	93.5 (84.3–101)
FVC, median (IQR), L	2.97 (2.5–3.4)
FVC, median (IQR), % predicted	91.6 (82–99.6)
Days until discharge, median (range)	4.5 (2–20)
Thirty‐day readmission rate	5.2
Clavien‐Dindo complication IIIb or higher	1.7
Number of lesions, median (range)	1 (1–8)
Number of lesions
1	53.4
2	25.9
3	8.6
4	6.9
≥ 5	5.2
Size of largest lesion, median (IQR), cm	1.5 (0.4–2.2)
Nodal sampling performed	27.3
Nodal disease found	5.2
Adjuvant chemotherapy
Perioperative	12.1
Adjuvant	39.7
RAS/RAF/MMR testing available	13.8
CK20/CDX2 testing available	58.6

Abbreviations: CK20, cytokeratin 20; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; IQR, interquartile range; MMR, mismatch repair.

Kaplan‐Meier estimates of PFS and OS with 95% confidence intervals are provided in Figures 1 and 2. Median PFS was 14 months (95% CI, 10.4–17.5), and median OS was 58 months (95% CI, 33.5–82.4). The estimated 5‐year survival was 49.8% (95% CI, 33.5–64.0). At the time of data cutoff, 67.2% of patients had progressed, and 48.2% had died. Two patients were lost to follow‐up and could not be contacted. The remaining patients were censored at the time of data cutoff.

Figure 1.

Progression‐free survival and 95% confidence interval for patients with colorectal cancer who underwent pulmonary metastasectomy with curative intent. Progression‐free survival was defined as the time from surgery to first clinical or radiologic evidence of relapse or death.

Figure 2.

Overall survival and 95% confidence interval for patients with colorectal cancer who underwent pulmonary metastasectomy with curative intent. Overall survival was defined as the time from metastasectomy to death.

The size of the largest lesion, number of metastatic lesions, DFI, synchronous presentation, and CK20 expression were significant after univariate Cox proportional hazard regression model and are presented in Table 3. Neither age, smoking status, primary site, laterality, obstructive lesion (defined as the presence of signs and symptoms of obstruction at diagnosis), CEA at both diagnosis or relapse, grade, margin, BRAF, RAS, MMR, total positive and total resected lymph nodes in primary, CCI, VEF1, FVC, nor use of adjuvant chemotherapy either in primary or at relapsed disease was associated with PFS or OS. Lack of CK20 expression remained prognostic for OS even after multivariate Cox regression adjustment controlling for disease‐free interval and synchronous disease, two widely recognized factors associated with prognosis confirmed by our analyses, with a hazard ratio of 23.5 (95% CI, 1.9–314; p = .014).

Table 3.

Prognostic factors

Variable	Progression‐free survival (95% CI), months	HR (95% CI)a	p valueb	Overall survival (95% CI), months	HR (95% CI)a	p valueb
Disease‐free interval
<24 mo	13 (9.8–16.1)	1.4 (0.7–3.1)	.29c	40 (31.8–48.1)	4.2 (1.4–12.5)	<.010c
≥24 mo	19 (15.9–22.0)			85 (75.7–96.2)
Presentation
Synchronous	7 (0.9–13.0)	4.8 (2.2–10.4)	<.001	33 (23.9–42.0)	4.0 (1.7–9.4)	<.001
Metachronous	19 (11.8–26.1)			77 (50.7–103.2)
Largest lesion size
<2 cm	13 (6.3–19.7)	1.1 (0.6–2.2)	.603	81 (33.7–128.2)	2.4 (1.1–5.3)	.024
≥2 cm	14 (9.6–18.3)			37 (22.9–51.0)
CK20 expression
Lacking	4 (0.8–7.2)	4.3 (1.2–15.2)	.012	19 (12.1–27.2)	37 (3.2–430)a	.004
Present	19 (9.0–28.9)			83 (46.9–119.0)
Number of lesions
1	23 (0.1–59.2)	2.5 (1.3–4.8)	.005d	81 (48.8–113.1)	1.8 (0.8–3.9)	.121d
≥2	11 (7.6–14.3)			37 (23.7–50.2)

^a Hazard ratio calculated through Cox regression using data as dichotomous variables.

^b p value based on log‐rank test.

^c p value for Cox regression using variable as continuous: p = .019 for progression‐free survival and p = .002 for overall survival.

^d p value for Cox regression using variable as continuous: p < .001 for progression‐free survival and p = .010 for overall survival.

Abbreviations: CI, confidence interval; CK20, cytokeratin 20; HR, hazard ratio.

Discussion

Pulmonary metastasectomy in metastatic colorectal cancer is an accepted treatment that may lead to prolonged progression‐free survival and cure in select patients, but high‐quality evidence supporting its use is lacking. There has been a lack of prospective randomized trials providing data to guide patient selection, given poor accrual. Additionally, recent evidence suggests that the outcomes observed in retrospective studies of lung metastasectomy are not necessarily due to the procedure alone but that selection bias may play a greater role than previously thought [11]. Two other recently published trials also contribute to the idea that selection bias has a large contribution to estimated survival also after ablative therapies in mCRC: the CLOCC (Chemotherapy and Local Ablation versus Chemotherapy) trial [21], which investigated radiofrequency ablation for liver metastases, and the SABR‐COMET (Stereotactic Ablative Radiotherapy for the Comprehensive Treatment of Oligometastases) trial [22], which tested SBRT for any primary site and any secondary site except for the central nervous system. The control groups of these trials consisted of traditional systemic therapy, and their reported 5‐year survival rates were 30% and 25%, respectively, which are numerically very similar to the results of the PulMiCC trial control group. These data further support the idea that the benefit of pulmonary metastasectomy is overestimated by the simple comparison of the outcomes of retrospective cohorts and historical unselected patients [12].

In contrast, larger series from cancer centers show results that are consistently better than the control groups of these trials. An analysis of MD Anderson Cancer Center, as well as a joint analysis of Duke University and Memorial Sloan‐Kettering Cancer Center, showed 5‐year overall survival rates of 55.4% and 56%, respectively [23, 24]. Another cohort of 785 patients from 46 Japanese hospitals who underwent pulmonary metastasectomy for mCRC with curative intent also demonstrated a surprising 5‐year overall survival rate of 68.1% (95% CI, 64.6%–71.8%) [25]. Yet, to which extent these impressive results can be attributed to surgical resection alone without the contribution of selection bias will remain unknown in the absence of data from prospective randomized studies.

Despite the uncertainties associated with pulmonary metastasectomy, it is well known that patients who experience very long relapse‐free periods are unlikely to relapse again and are likely cured [2]. Therefore, pursuing long‐term relapse‐free survival and eventually cure, clinicians may feel compelled to offer metastasectomies despite limited evidence. Moreover, we still do not have a method to readily identify patients who may benefit from this invasive procedure, and decisions are often made based on the presence of previously described favorable prognostic factors. In this study, we characterized a population of patients with metastatic colorectal cancer from our institution in Brazil who underwent pulmonary metastasectomy with curative intent and explored potential prognostic factors that, by previous description in literature or by additional pathologic information, could be associated with favorable long‐term outcomes.

Our long‐term relapse‐free survival and overall survival are similar to those from previously published studies [26, 27, 28]. Variables that showed significance were longer DFI, large lesion size, number of lesions, metachronous presentation, and CK20 expression in metastasis, demonstrating their importance as prognostic factors. Laterality did not have a statistically significant impact in our sample, which is consistent with a smaller effect size previously described [29, 30, 31, 32]. Similarly, the effect size of perioperative chemotherapy seems to be small, recently described as a hazard ratio of 0.83 (95% CI, 0.75–0.92) in univariate analysis of a systematic review and meta‐analysis [33]. In our study, no statistically significant effect of either adjuvant or perioperative chemotherapy was demonstrated.

Additionally, despite wide confidence intervals, our results suggest that the lack of CK20 expression in metastases may be an additional unfavorable prognostic factor. This conflicts with previous data describing poorly differentiated adenocarcinomas lacking CK20 expression as associated with better prognosis compared with CK20‐positive disease [34, 35]. A possible explanation is that these observations were performed in localized instead of metastatic colorectal cancer, thus limiting generalizability of this finding to our sample. Remarkably, microsatellite instability–high status is also associated with contrasting effects on prognosis depending on stage, being associated with better prognosis in localized colorectal cancer and worse prognosis in metastatic disease [36, 37, 38, 39]. All our patients had metastatic disease, mainly from left‐sided primary tumors and disease presentation, or evolution was such that a pulmonary metastasectomy was performed, which makes this population a very specific one.

Lack of CK20 expression in localized disease is associated with right colon disease, microsatellite instability, CpG island methylator phenotype (CIMP), NTRK fusions, and BRAF mutations [40, 41, 42]. However, BRAF mutant tumors do not seem to alter their CK20 expression in the absence of microsatellite stability, thus suggesting that it is unlikely that an association with BRAF mutations is the driver of the worse prognosis seen in patients whose metastases lack CK20 expression [43]. Additionally, microsatellite instability and CIMP are uncommon in the left colon, which also weakens their chance of playing a major role in our cohort. As this association is based on a purely exploratory analysis, the true relevance of this finding is yet to be clarified by further confirmatory, ideally prospective cohorts.

Our study has several limitations. Firstly, this is a single‐institution retrospective cohort, which is prone to selection bias. Although they were underpowered, previous studies have already made clear that modern control groups outperform historical survival rates, implying that selection alone could explain most of the benefit previously fully attributed to pulmonary metastasectomy. Also, RAS, BRAF, and MMR test results were available in only a few patients, as the cohort included patients from the last 35 years. It is likely that CK20 expression is only a surrogate marker for tumor biology in our patients and that the true molecular driver of this finding is still to be clarified. As an exploratory analysis, results of this study are in no means conclusive and warrant further investigation.

In this cohort of patients with metastatic colorectal cancer who underwent pulmonary metastasectomy, we demonstrated a 5‐year survival rate of 49.8% (95% CI, 33.5–64.0), which is consistent with results previously described in literature. We further consolidated the prognostic importance of DFI, synchronous presentation, and number and size of metastatic lesions. Lack of CK20 expression was also prognostic and associated with a statistically and clinically meaningful shorter overall survival after multivariate Cox regression in our study. This finding is exploratory and demands further investigation but suggests that biomarkers may have a role in guiding patient selection for pulmonary metastasectomy.

Conclusion

Pulmonary metastasectomy is a potentially curative treatment for metastatic colorectal cancer and is associated with long‐term survival for select patients. In the absence of high‐quality evidence, therapy is currently guided by the presence of clinicopathological factors associated with better outcomes. Molecular prognostic factors had not yet been thoroughly explored in this scenario and should be further investigated.

Author Contributions

Conception/design: Gustavo C. L. Gössling, Márcio F. Chedid, Gilberto Schwartsmann, Aparna R. Parikh

Provision of study material or patients: Gustavo C. L. Gössling, Márcio F. Chedid, Fernando S. Pereira, Rafaela K. da Silva, Leonardo B. Andrade, Nícolas Peruzzo, Maurício G. Saueressig

Collection and/or assembly of data: Gustavo C. L. Gössling, Fernando S. Pereira, Rafaela K. da Silva, Leonardo B. Andrade, Nícolas Peruzzo

Data analysis and interpretation: Gustavo C. L. Gössling, Márcio F. Chedid, Maurício G. Saueressig, Gilberto Schwartsmann, Aparna R. Parikh

Manuscript writing: Gustavo C. L. Gössling, Márcio F. Chedid, Nícolas Peruzzo, Gilberto Schwartsmann, Aparna R. Parikh

Final approval of manuscript: Gustavo C. L. Gössling, Márcio F. Chedid, Fernando S. Pereira, Rafaela K. da Silva, Leonardo B. Andrade, Nícolas Peruzzo, Maurício G. Saueressig, Gilberto Schwartsmann, Aparna R. Parikh

Disclosures

Aparna R. Parikh: Checkmate, Eli Lilly & Co., Pfizer, Natera (C/A), Bristol‐Myers Squibb, Eli Lilly & Co., Guardant, Array, Pfizer, Plexxicon (RF—institution), Roche (SAB). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board