Characteristics of Patients With ROS1+ Cancers: Results From the First Patient-Designed, Global, Pan-Cancer ROS1 Data Repository

Abstract

PURPOSE:

The discovery of driver oncogenes, such as ROS1, has led to the development of targeted therapies. Despite clinical advancements, gaps remain in our understanding of characteristics of patients with ROS1-positive (ROS1+) cancers. The purpose of this study was to comprehensively assess demographic, clinical, and environmental characteristics associated with ROS1+ cancers worldwide.

METHODS:

In collaboration with a panel of patients with ROS1+ cancer, we designed and conducted a 204-question online assessment regarding the demographic, clinical, and environmental factors of patients with ROS1+ cancers. We invited patients with ROS1+ cancers to participate in the study from May 2016 to December 2018.

RESULTS:

A total of 277 patients from 18 countries worldwide responded and completed at least 90% of the survey. The majority of respondents were female (n = 191; 69%), non-Hispanic white (n = 202; 73%), never-smokers (n = 180/240; 75%). Most were diagnosed with lung cancer (n = 261/277; 94%) and stage IV disease (n = 201/277; 76%). The majority received chemotherapy in first (n = 137/199; 69%) and second (n = 103/199; 52%) lines of therapy. For patients diagnosed with lung cancer after the availability of crizotinib (n = 199), only a minority (n = 55/199; 28%) reported receiving crizotinib in the first line of therapy.

CONCLUSION:

This study is the first global, patient-designed approach, to our knowledge, to comprehensively assess demographic, clinical, and environmental characteristics associated with ROS1+ cancers. Future efforts include assessing these characteristics as well as patient-reported outcomes and treatment responses longitudinally.

INTRODUCTION

Over the past decade, the identification of driver oncogenes and development of novel therapies that target these oncogenes have significantly improved clinical outcomes and patient-reported quality of life.^1-3 The discovery of ROS1, a receptor tyrosine kinase, and its associated gene rearrangements, is one such example. ROS1 rearrangements are less-common driver oncogenes, typically present in only 1%-2% of patients with cancer worldwide.⁴ Although ROS1 occurs mostly in non–small-cell lung cancer (NSCLC), this driver oncogene is also found, although to a lesser degree, in other cancers, such as colon,⁵ glioblastoma,⁶ and cholangiocarcinoma.⁷ On the basis of small, often single-institution studies of patients with NSCLC,⁸ ROS1+ cancers appear to be more common among younger, nonsmoking patients and, because of treatments that directly target ROS1 through inhibition of tyrosine kinase, such as crizotinib, are associated with significantly improved survival.^9,10 Given the difficulty in conducting large prospective clinical trials among patients with this less-common driver oncogene, most data to date regarding demographic and clinical characteristics are from smaller institution-based studies.¹¹ Specifically, no single registry describes the demographic, clinical, and environmental exposure characteristics of patients and the outcomes from these cancers worldwide and longitudinally.¹²

In response to these persistent gaps in knowledge, we embarked on a novel, longitudinal, patient-led collaboration to enhance our understanding of the demographic, clinical, and environmental factors that may contribute to the development of ROS1+ cancers. Together with an international panel composed of 10 patients with ROS1+ cancers, we designed a patient-reported assessment focused on three distinct stages of a patient’s journey with a ROS1+ cancer to enhance our understanding of demographic, clinical, and environmental characteristics associated with the development of and outcomes from these cancers: prediagnosis (such as environmental exposures, diet, lifestyle, and hormone use); diagnosis (stage, comprehensive molecular profiling, comorbidities); and postdiagnosis (treatments, response to treatments, duration of response). In this article, we present the early findings from this comprehensive assessment.

METHODS

Study Design

In 2016, three patients who had ROS1+ NSCLC approached the investigator team to develop research studies to accelerate the development of new treatments for ROS1+ cancer. After a series of interviews with ROS1+ patients and investigators in academia and industry, we identified the need for a global, longitudinal ROS1+ cancer data repository to better understand demographic characteristics and epidemiologic factors associated with the development of and outcomes from these rare and understudied cancers.¹³ In collaboration with a patient advocacy group¹⁴ and 10 volunteer patients and their caregivers from across the globe, we designed and deployed a 204-question survey internationally.¹⁵

Study Population and Outreach

We used a variety of outreach sources to provide patients, caregivers, and health care providers with our online questionnaire link. Specifically, we used a website to describe the initiative in lay terms¹⁵; newsletters; social media such as twitter and Facebook pages¹⁶; flyers and information booths at national and international cancer-related conferences; e-mail solicitation to oncology providers across the world; joint outreach efforts through patient-focused advocacy groups such as the Melanoma Research Alliance,¹⁷ Gastric Cancer Foundation,¹⁸ and National Brain Tumor Society¹⁹; and finally through publicized efforts and by word of mouth through an international ROS1+ cancer support group composed of patients and their caregivers.²⁰

Data Collection

Data were collected through an online questionnaire designed by the 10-patient panel and study investigators and administered in English through a Stanford University license with Qualtrics (Provo, UT).²¹ The questionnaire, composed of a total of 204 questions, was divided into sections. Section 1 has 22 patient demographic questions that assessed sex, race and ethnicity, marital status, current residence, occupation, income, medical conditions, and family history of cancer; 80 questions that assessed exposures to exogenous hormones, dietary factors, exercise, and substance abuse history; and 64 questions regarding environmental and occupational exposures. Section 2 has 17 questions regarding diagnostic procedures, stage at diagnosis, tumor histology, and genomic testing. Section 3 has 21 questions regarding postdiagnosis treatments and outcomes.

Ethics and Informed Consent

Voluntary informed consent was obtained from all participants, who were assured that confidentiality of those involved in the study would be maintained and no identifiable data would be shared outside of the study team. All procedures performed in our study involving human participants were reviewed and approved by the Stanford University institutional research review board.

Statistical Analysis

All data were entered, tabulated, and analyzed using Stata15 software²² by two study investigators (D.A.P. and Principal Investigator M.I.P.). Questionnaire responses from participants from May 8, 2016 until December 20, 2018 were extracted and analyzed. Missing data were dropped from analysis for those specific questions, as noted in the results section for each question.

RESULTS

Patient Demographics

For the 18 months of the questionnaire posted online, a total of 277 patients with ROS1+ cancer from > 18 different countries worldwide responded to the survey, with at least 90% of all survey questions completed by each respondent. Figure 1 depicts the geographic location of all of the respondents. The majority of participants were from the United States, representing 28 different states (n = 182/277; 66%). Respondents from other countries included Australia (n = 19/277; 7%), the Netherlands (n = 14/277; 5%), Germany (n = 11/277; 4%), and Singapore (n = 3/277; 1%).

Fig 1.

Geographic locations of respondents.

Table 1 lists demographic factors of all of the survey participants. The median age at diagnosis was 50 years (range, 25-69 years) and the majority were female (n = 191/277; 69%). The majority of respondents reported their race and ethnicity as non-Hispanic white (n = 202/277; 73%), followed by Asian (n = 41/277; 15%), Hispanic (n = 16/277; 6%), black or African (n = 8/277; 3%), and other (n = 5/277; 2%). Among Asian respondents, 56% self-identified as Chinese (n = 23/41) and 20% identified as Indian (n = 8/41). Of the 250 participants who responded to the question regarding marital status, the majority were presently married (n = 170/250; 68%). Many of the participants reported working professionally (n = 94/277; 34%) and had pursued higher education, such as Masters or advanced graduate degrees (n = 105/277; 38%). Of the 258 participants who responded to the question regarding annual household income, approximately one-third reported levels > $100,000 (US dollars; n = 77/258; 30%). Response rate for the question regarding smoking status was 87%, and of the 240 respondents, 75% stated that they had never smoked (n = 180/240; 75%).

TABLE 1.

Demographic Characteristics of ROS1 Respondents

Diagnostic and Treatment Factors

The overwhelming majority of respondents were diagnosed with lung cancer (n = 261/277; 94%), followed by gastric cancer (n = 5/277; 2%), cervical cancer (n = 3/277; 1%), ovarian cancer (n = 3/277; 1%), glioblastoma (n = 3/277; 1%), and colon cancer (n = 2/277; 1%). A total of 224 respondents (81%) had histologic diagnosis of adenocarcinoma, 25 (9%) with squamous cell carcinoma, and 8 (3%) with large-cell carcinoma. The overwhelming majority of patients were diagnosed with stage IV disease (n = 201/277; 76%), with lymph node involvement as the most commonly reported metastatic site (n = 70/201; 35%), followed by lung (n = 54/201; 27%) and bone (n = 26/201; 13%). Of the 27% of respondents (n = 76/277) who were diagnosed with earlier stages of disease, 86% (n = 65/76) experienced progression of their disease to a distant metastatic site, with a median time to progression of 3 months after their initial diagnosis. The majority of patients reported that their ROS1 gene rearrangement was not detected at diagnosis (n = 174/277; 63%), and 95% (n = 263/277) did not know the specific gene rearrangement. Among the 5% of respondents who knew their specific rearrangement, CD74 was most commonly reported (n = 10/13; 77%). Although the majority reported that their physician requested molecular testing of their tumors (n = 180/277; 65%), one-third (n = 91/277; 33%) self-requested the molecular and genome sequencing of their tumors (Fig 2A). In the United States, Australia, Canada, Germany, and the Netherlands, the countries representing the majority of survey respondents, approximately 37%-73% reported that ROS1 was not detected at diagnosis (Fig 2B).

Fig 2.

Patient-reported ROS1+ cancer clinical factors. (A) Reason for ROS1 testing. (B) ROS1 testing at diagnosis by country. (C) First-line systemic therapy including crizotinib, chemotherapy, or other therapy.

Of the 253 patients who responded to the questions regarding treatment, most patients (n = 199/253; 79%) received crizotinib at some point during their treatment. Among respondents whose lung cancers were diagnosed after the US Food and Drug Administration approval in April 2016 and for whom crizotinib was available (n = 199), only a minority (n = 55/199; 28%) reported receiving this treatment in the first line. The majority of patients who received crizotinib in the first line were from the United States (n = 35/55; 63%), and the trend of crizotinib usage increased over time, with patients surveyed in 2018 more likely to report receiving crizotinib in the first line (n = 20) compared with 2016 (n = 10) and 2017 (n = 5).

The majority of respondents stated that they received chemotherapy in their first (n = 137/199; 69%) and second (n = 103/199; 52%) line of therapy (Fig 2C). Twenty-eight percent (n = 55/199) reported receiving chemotherapy in the third line of treatment. Thirteen percent (n = 25/199) of patients participated in ROS1-directed therapy clinical trials, the majority of whom were residents of the United States (n = 19/25; 72%). Other countries from which a single patient reported participation in a ROS1-directed therapy clinical trial included Australia, Belgium, Barcelona, Germany, Hong Kong, and Japan. Only one patient reported receiving immunotherapy at any point during their cancer care. The majority of respondents (n = 159/199; 80%) reported that crizotinib was “the most effective” and noted “very few side effects” with this treatment (n = 179/199; 90%).

Family History, Lifestyle, and Environmental Exposures

A total of 160 respondents (58%) reported cancer diagnoses in a first-degree relative. Of the women in the sample, 70% (n = 133/191) reported having used oral contraceptives. Thirty percent of respondents (n = 83/277) reported having used hormones (estrogen/progesterone/testosterone) for > 12 months, with 2% (n = 5/277) having used recreational anabolic steroids. The overwhelming majority of respondents (n = 243/277; 88%) reported no environmental or occupational exposures to asbestos, radon, coal dust, silica, radiation, uranium, automobile exhaust, insulation, or molten materials. The overwhelming majority of participants also reported no comorbidities before their diagnosis (n = 254/277; 92%).

DISCUSSION

This study, to the best of our knowledge, represents the largest, global patient-led assessment of demographic, clinical, and environmental exposure characteristics of patients with ROS1+ cancers. Our study was organized organically through a unique collaboration composed of an international panel of patients diagnosed with ROS1+ cancers, an academic institution, and a nonprofit patient advocacy group. The study resulted in a total of 277 voluntary patient responses, with a high percentage of completion, demonstrating the relevance of this work among the patients diagnosed with these rare diseases.

Many of our results validate those previously reported in small single-institution–based studies among patients with ROS1+ lung cancer. For example, the majority of respondents in our study were nonsmokers, diagnosed at younger ages and with more advanced stages of disease.⁸ Our study is also in agreement with those that report on the aggressive nature of these diseases²³ and further highlights that almost all respondents with earlier stages of cancers experienced a rapid progression to metastatic disease. Similar to data reported in prior clinical trials,⁹ our study cohort also reported that targeted tyrosine kinase inhibitors appear to be the most effective and tolerated therapeutic.

Our study also uncovered several previously unreported findings that warrant additional investigation. First, we assessed clinical factors, such as information regarding tumor testing for ROS1. Although the vast majority had their tumors tested for molecular markers, a high proportion had self-requested this testing. Furthermore, we uncovered that despite the availability and approval to use targeted therapeutics in the initial course of therapy,²⁴ many did not receive this therapy in the first line of treatment. These findings, taken together, could be due to the absence, during our study period, of universal, professional guidelines that recommend the routine testing and targeted treatment of tumors for ROS1+.²⁵ Future efforts should investigate whether these findings persist, given more recent efforts to disseminate guidelines regarding the testing and treatment of these cancers. Nonetheless, these findings acknowledge the sheer importance of patient education, activation, and engagement in improving the delivery of guideline-based cancer care among patients with rare diseases.

Our ongoing study, like many other registry studies for patients with rare diseases, provides an example of a successful patient-driven research approach. Our unique model involving self-organized patients, across international borders, allowed us to tailor the study to reflect challenges that patients may face prediagnosis, at diagnosis, and postdiagnosis in real-world settings. Our team’s success in conceiving, piloting, and recruiting an unexpectedly large number of participants transpired because of the active support of the patients who were personally affected by this rare condition. Despite the extensive and lengthy nature of the online assessment at 204 questions, our high response rate, with few missing data, demonstrates the singular motivation and focus of patients with little-understood conditions through a natural organizing framework in a social networking site²⁶ with support from community advocacy foundations and researchers to advance knowledge in this field. Since development of our study, several patient-initiated groups whose tumors have rare molecular and genetic rearrangements, including the ALKies,²⁷ EGFRResisters,²⁸ and Exon20Group,²⁹ have emerged to advance knowledge in the field similar to our study.

Limitations of our ongoing study, like many volunteer registries, include reporting bias and sampling bias. Although there is great utility in using patient-reported data, complex medical information such as tumor histology and testing results cannot be confirmed. Despite the relative rareness of these cancers, our study resulted in a large number of respondents among patients globally with pan-cancer diagnoses. Future studies, including our own efforts, should continue to assess these important contributing factors and others that become salient as more patients are diagnosed and living with these cancers.

In conclusion, our study demonstrates the results of an online assessment of demographic, clinical, and environmental characteristics of patients whose tumors harbor ROS1 rearrangements. This study represents the largest global effort to enhance our understanding of the demographic, clinical, and environmental exposure that may lead to development and outcomes of ROS1+ cancers. Our patient-informed collaboration involving community groups and researchers serves as a model to improve understanding of other rare and relatively understudied conditions.

AUTHOR CONTRIBUTIONS

Conception and design: Divya A. Parikh, Guneet Walia, Janet Freeman-Daily, Tori Tomalia, Lisa Goldman, Bonnie Addario, Manali I. Patel

Provision of study material or patients: Janet Freeman-Daily, Merel Hennink, Manali I. Patel

Collection and assembly of data: Divya A. Parikh, Guneet Walia, Janet Freeman-Daily, Merel Hennink, Lysa Buonanno, Manali I. Patel

Data analysis and interpretation: Divya A. Parikh, Guneet Walia, Manali I. Patel

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Characteristics of Patients With ROS1+ Cancers: Results From the First Patient-Designed, Global, Pan-Cancer ROS1 Data Repository

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/site/ifc/journal-policies.html.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

Guneet Walia

Employment: Genentech

Stock and Other Ownership Interests: Genentech

Janet Freeman-Daily

Honoraria: NeoGenomics Laboratories, AstraZeneca

Travel, Accommodations, Expenses: NeoGenomics Laboratories, Questex, Turning Point Therapeutics

Merel Hennink

Travel, Accommodations, Expenses: Roche, Takeda, AstraZeneca, Pfizer

Lysa Buonanno

Honoraria: Pfizer

Speakers' Bureau: Lung Cancer Foundation of America

Travel, Accommodations, Expenses: Pfizer, Lung Cancer Foundation of America

Manali I. Patel

Consulting or Advisory Role: Celgene

No other potential conflicts of interest were reported.