• clinical trials;
  • phase 1;
  • oncology;
  • pharmacodynamic markers


  1. Top of page
  2. Abstract
  6. Acknowledgements


Increasingly, early phase clinical trials involve pharmacodynamic (PD) and pharmacokinetic (PK) assays as well as frequent imaging studies. The authors conducted a prospective study examining patients' willingness to undergo such tests and the number of tests the patients would tolerate.


A prospective, correlative study was conducted using a self-reported questionnaire to measure patients' willingness on a scale from 1 (not willing) to 10 (very willing) to undergo various procedures (eg, tumor and skin biopsies, blood tests) and imaging studies (eg, magnetic resonance imaging, echocardiogram). In addition, correlations were assessed between the number and type of tests and demographics, previous test experience, inconvenience, and insurance coverage. Sixty-one patients (22 women and 39 men) with advanced malignancies were enrolled. Descriptive, nonparametric, and parametric inferential statistics were used.


Overall willingness to undergo study-required tests was very high. Patients were most willing to undergo urine, blood, ultrasound, x-rays, echocardiogram, and computed tomography studies and were least willing to undergo tumor and skin biopsies and magnetic resonance imaging (all P ≤ .01). Significant inverse relations were observed between the frequency of a particular test and patient's willingness to undergo such tests. Inconvenience and prior negative experiences for more invasive tests (eg, skin biopsies) modestly affected willingness to undergo these tests again. College education, insurance coverage, and the requirement of tests for enrollment were correlated positively with willingness to undergo tests.


The current findings provide the first prospectively collected data on patients' willingness to undergo PK/PD tests and imaging studies associated with early stage oncology drug trials and can serve as basis for further exploration toward the design of patient-friendly, biomarker-driven clinical studies in oncology. Cancer 2011. © 2011 American Cancer Society.

In the era of biomarker-driven, clinical translational research, an increasing number of clinical trials involve optional or mandatory pharmacodynamic (PD) and pharmacokinetic (PK) assays, imaging studies, and invasive procedures, such as skin and tumor biopsies.1, 2 These tests serve to guide dose escalation, assess target modulation, and determine early response; and they may be useful for the development of biomarkers.3, 4 Patient adherence to these tests is important for safety and success of the study, especially because determining the optimal biologic dose and other biomarker-driven endpoints are becoming increasingly important in drug development.5

Previous studies have addressed patient perceptions and expectations of treatment outcomes, potential benefits and side effects, informed consent,6, 7 and ethical aspects of phase 1 oncology trials.8 To our knowledge, no study has prospectively assessed patients' willingness to undergo PD/PK tests, such as skin biopsy or hair follicle collection; has asked about patients' willingness to participate in frequent imaging studies, including functional imaging; or has assessed the number of tests/studies that patients would tolerate in a single trial within the context of evaluating biomarkers and the biologic effects of new agents.

Therefore, the primary objectives of the current study were to assess patients' willingness to undergo such tests and studies and to determine the number of tests patients would tolerate in a single trial. A secondary purpose was to correlate the number and type of tests with patient demographics, prior test experience, test inconvenience, and cost coverage. Results from our study may be used in the design of biomarker-driven clinical trials while recognizing patient's motivation and readiness to participate in increasingly complex clinical trials in oncology.9-13


  1. Top of page
  2. Abstract
  6. Acknowledgements

In brief, before designing this study and its questionnaire, an extensive review of the literature revealed only a few relevant studies,4, 14 and none addressed the questions in which we were interested. Therefore, a questionnaire was developed by the participating authors based on their combined extensive expertise and experience in oncology drug development. Items were developed based on commonly encountered questions and procedures in early phase oncology trials.

Study Design

We conducted a prospective, exploratory, correlational patient survey/questionnaire study. Data were collected at a single clinical research center that was conducting early phase oncology trials.


Eligibility criteria included age ≥ 18 years; ability to read, write, and speak English; and referral to, current enrollment in, or evaluation for an early phase (phase 1/2) clinical trial with a diagnosis of advanced treatment-refractory malignancy. Patients who did not meet these criteria were excluded.


For this study, we used an investigator-developed, 2-part (Parts A and B), multiple-item, patient-reported questionnaire. Both parts were preceded by an introduction, instructions, and study-test definitions. Part A (n = 56 items) was divided into 2 sections. Section 1 included questions pertaining to demographics, clinical characteristics, and treatment history. Section 2 included questions about diagnostic studies, such as positron emission tomography (PET), computed tomography (CT), x-ray, magnetic resonance imaging (MRI), ultrasound (US), and echocardiogram (ECHO). Other items assessed the amount of time patients were willing to spend in healthcare settings per week and test coverage. Section 1 of Part B (n = 59 items) contained questions about PD assays, such as skin and tumor biopsies and blood, hair follicle, stool, and urine samples. At the end of the questionnaire, a short section (Section 2, Part B) asked patients to give feedback on the questionnaire's quality and design, and this section was intended for use in developing future studies that address similar questions. Table 1 provides an example of the questions that were asked. The questionnaire was pilot tested on patients and healthcare providers to assess its clarity and ease of use. Feedback was incorporated, and the questionnaire was modified slightly before full application.

Table 1. Examples of Questions Asked in the Survey
Diagnostic Scans: Positron emission tomography (PET)
1. Have you ever undergone a PET scan? (yes/no/unsure)
2. No. of PET scans: 1, 2, 3, ≥4
3. Inconvenience: Numeric rating scale [NRS] from 0 (minimal inconvenience) to 10 (severe inconvenience)
4. Previous scan influenced another scan (positive/negative/no influence)
5. Willingness to undergo PET scan if mandatory for enrollment (yes/no)
6. Willingness to undergo 1, 2, 3, or 4 PET scans in a single trial: NRS from 0 (not willing) to 10 (very willing)


The study and the questionnaire were approved by the local institutional review board (IRB), and written informed consent was obtained from all study participants before enrollment. Patients completed questionnaires either in the clinic or at home. Study coordinators were available and assisted patients as needed for clarification and questions on the survey/questionnaire. Data were collected between June 2008 and December 2008. Two study coordinators/investigators entered and verified (against the raw data) all data entered into Microsoft Excel spreadsheets (Microsoft Corp., Redmond, Wash).

Data Analysis

Data were transferred to an SPSS data file (SPSS version 17; SPSS, Inc., Chicago, Ill) and were analyzed using descriptive and nonparametric and parametric, inferential statistics. Nonparametric tests were used when data did not meet distributional requirements for parametric statistical tests. To determine whether demographic variables were related to patients' willingness to undergo the various tests, Mann-Whitney or Kruskal-Wallace tests were used, as appropriate, and Spearman rho coefficients were used for ranked data.

The primary analysis was conducted using generalized estimating equations.15 Because most answers were marked in the higher range on the 0-to-10 survey scale (with 0 indicating “not willing” and 10 indicating “very willing”), the data were skewed toward the higher numbers of the survey scale. To accommodate this skew and accurately model the data, an inverse Gaussian (Wald) distribution analysis method was used in which the 1-to-10 scale was transformed into an inverse scale from 1 to 11 (to remove the zero); a 1 was added to each score to set the minimum value at 1, with the maximum at 11 in the inverse order (ie, 1 = 11, 2 = 10, …, 11 = 1). To accommodate the tendency for within-subject responses for the contiguous dataset to be more highly intercorrelated than noncontiguous responses (eg, willingness to undergo 1 test and willingness to undergo 2 tests were more highly correlated than willingness to undergo 1 test and willingness to undergo 4 tests), the working correlation matrix was specified as autoregressive at lag 1. Finally, data were transformed so that the ordering of values was identical to that in the raw data with the addition of 1 to scores to allow calculation of geometric means and standard deviations and to minimize the impact of extreme responses (eg, “1” and “2”) on estimates of central tendency.16

Linear trend tests (polynomial contrast) were conducted under the generalized estimating equations model to assess the relation between the number of tests requested and willingness to undergo the tests. The main effect of test type also was assessed to determine whether participants differed, on average, in their willingness to undergo different tests. The urine sample was used as the baseline comparator in this analysis, because it was identified as the 1 test participants were most willing to undergo.

Percentiles (50th = median) and interquartile ranges (25th and 75th percentiles) are reported for non-normally distributed data. To assess the significance of trends within the patient (eg, willingness to undergo 1, 2, 3, etc, tests), data were normalized using square roots transformation, and a repeated-measures analysis of variance was conducted on the transformed data. All significance tests were 2-tailed, and α = .05 was used as the criterion for significance. The study was approved by the IRB at Scottsdale Healthcare Hospitals.


  1. Top of page
  2. Abstract
  6. Acknowledgements


Patients were approached consecutively during clinic visits by study coordinators and were asked to participate. Of 70 patients who were approached for the study, 61 patients agreed to participate, and 9 patients refused. Comments and reasons for refusals included mostly time commitments or other conflicting study-related matters (eg, other tests). For patients who declined to participate, there was no bias toward patients from any individual physician or toward being treated on a particular protocol. The average time for patients to complete the survey was 30 minutes (range, 12-67 minutes). Table 2 displays the sample's demographic characteristics.

Table 2. Demographic Characteristics (n=61)
CharacteristicNo. of Patients (%)
  • AHCCS indicates Arizona Health Care Cost Containment System; KPS, Karnofsky performance status.

  • a

    Some patients preferred not to provide certain information; therefore, numbers do not always equal 100%.

Age: Mean [range], y59.1 [19-84]
 Women22 (36)
 Men39 (64)
Marital status
 Married48 (79)
 Not married13 (21)
 White/Caucasian56 (92)
 Minority5 (8)
Religious preferencea
 Christian41 (68)
 Jewish2 (3)
 Mormon1 (2)
 Other4 (7)
 None12 (20)
 College graduate30 (49)
 Noncollege graduate29 (47)
Income, $a
 ≤74,99924 (61)
 ≥75,00015 (39)
 Arizona resident45 (74)
 Non-Arizona resident13 (21)
 Private26 (42)
 AHCCS4 (7)
 Medicare11 (18)
 Combined private and federal insurance18 (30)
KPS, %
 ≥80: Some signs and symptoms45 (74)
 ≤70: Unable to carry out normal activity16 (26)

Table 3 displays the patients' clinical characteristics. Some patients reported more than 1 comorbidity, and 27 different tumor types were represented in the sample. Although all patients were considered to have refractory, advanced (mostly metastatic) malignancies, an unexpected finding was that approximately 54% of patients (n = 33) stated that they did not know “their stage of disease.” Karnofsky performance status, knowledge of disease stage, and presence of comorbidities, for the most part, were not correlated with the outcome variables/tests (data not shown).

Table 3. Clinical Characteristics (n=61)
CharacteristicNo. of Patients (%)
  • GERD indicates gastroesophageal reflux disease; GIST, gastrointestinal stromal tumor.

  • a

    Other comorbidities were anemia, back pain, Barrett esophagus, clots, depression, hypothyroidism, scleroderma, and tachycardia.

  • b

    Percentages may be more than or less than 100%, because some patients did not provide answers to all questions, and some participants were diagnosed with more than 1 cancer (ie, totals >100% than the expected maximum of n=61).

  • c

    Other types of cancer were carcinomas or sarcomas, including adrenocortical carcinoma, alveolar soft-part sarcoma, basosquamous carcinoma, brain carcinoma, carcinoma (unknown), carcinoid, choriosarcoma, endometrial carcinoma, GIST, head/neck, leiomyosarcoma, hepatocellular carcinoma, mucinous adenocarcinoma, parotid carcinoma, penile carcinoma, renal carcinoma, sinus carcinoma, skin cancer (n=2), and tongue cancer (n=2).

Any comorbidities
 No34 (56)
 Yes25 (41)
 Arthritis5 (14)
 Asthma4 (11)
 Diabetes6 (17)
 GERD2 (6)
 Hypertension11 (30)
 Othera8 (22)
Type of cancerb
 Basal cell carcinoma7 (12)
 Breast6 (10)
 Colon5 (8)
 Melanoma5 (8)
 Leukemia3 (5)
 Lung7 (12)
 Ovarian3 (5)
 Pancreatic10 (16)
 Prostate4 (7)
 Otherc20 (32)
Diagnosed with more than 1 cancer
 No45 (74)
 Yes15 (24)
Current stage of cancer
 III4 (7)
 IV22 (36)
 Unknown33 (54)
 Missing2 (3)
Family members with cancer
 No15 (25)
 Yes46 (75)

Analysis by Research Question

Research Question 1: Perceived willingness to undergo specific tests

This question and analysis addressed overall willingness to undergo a specific test across all tests. Geometric means and standard deviations are provided for the main effect of tests (collapsed across number of tests; Table 4) and are ranked in ascending order, with the patients who were least willing to undergo the test at the top. Willingness to undergo tumor biopsy, skin biopsy, and MRI and willingness to provide a hair follicle sample differed significantly (P < .05) from the baseline comparator (urine sample).

Table 4. Geometric Means and Standard Deviations for Willingness to Undergo Testsa
TestGeometric Mean±SD
  • SD indicates standard deviation; MRI, magnetic resonance imaging; PET, positron emission tomography; CT, computed tomography.

  • a

    The willingness of patients to undergo the various tests was compared with their answers for urine samples. A scale from 1 (not willing) to 11 (very willing) was converted from the 0-to-10 scale that was used in the questionnaire as described in the text (see Materials and Methods, Data Analysis).

  • b

    Differed significantly from urine sample as a comparator (P<.05).

Tumor biopsyb5.3±2.3
Skin biopsyb5.9±2.2
Hair follicleb7.8±1.9
Stool sample7.9±1.7
Blood sample9.2±1.5
Urine sample9.6±1.4
Research Question 2: To determine the number of tests patients would tolerate in a single trial

Question 1 addressed the overall tendency of willingness for a specific test across all tests, whereas Question 2 asked how often patients would be willing to undergo a particular test. In general, the more often a test was required (the survey asked for 1 to > 4 tests), the less willing patients were to undergo such tests (all tests for linear trend; P < .01). Willingness was high to undergo ultrasound and x-rays examinations on 1 to 3 occasions during a study, closely followed by CT and PET scans (Table 5). MRI was the least favored imaging modality. Urine, blood, and stool samples achieved high willingness for 1 to 4 tests, followed by hair follicle pull and even skin biopsy. Patients also were very willing to undergo 1 tumor biopsy, but there was a sharp decline in willingness to undergo ≥ 2 tumor biopsies on a study. Geometric means and standard deviations reflecting patients' willingness to undergo 1 to 4 tests are provided in Table 5.

Table 5. Willingness to Undergo Specific Tests and the Number of Tests Patients Would Tolerate in a Single Trial
 Geometric Mean±SDa
No. of TestsPET ScanCT ScanX-RayMRIUltrasoundEchocardiogram
  • SD indicates standard deviation; PET, positron emission tomography; CT, computed tomography; MRI, magnetic resonance imaging.

  • a

    Answers were scored on a scale from 1 (not willing) to 11 (very willing). The scale range was recoded from a 0-to-10 scale to a 1-to-11 scale by adding 1 to each score to accommodate calculation of the geometric mean and SD.

P test for trend<.001<.001<.001<.001.003.010
 Geometric Mean±SDa
No. of TestsSkin BiopsyTumor BiopsyBlood SampleHair FollicleStool SampleUrine Sample
P test for trend.
Research Question 3: To correlate willingness to undergo tests with demographic variables

++Most demographics, such as age, marital status, sex, religion, insurance, and income level, were not associated with willingness to undergo the various tests. However, college graduates (n = 30) indicated that they were more willing to undergo CT scans, MRI scans, ultrasound studies, tumor biopsies, urine samples, echocardiograms, and skin biopsies (P < .05) than noncollege graduates (n = 29). A formal breakdown of data for minority patients was not performed, because there were only 5 minority participants (also see Discussion, below).

Research Question 4: To correlate willingness to undergo tests with prior test experience

Prior test experience can have a positive or negative influence on future willingness to participate in such tests again. Most participants reported prior experience with most of the tests (see Table 6). For those patients who were unfamiliar with certain tests/studies, explanations of study procedures and tests were provided in the surveys. Average willingness reported to undergo from 1 to > 4 tests was calculated and then correlated with each patient's “inconvenience” rating (Spearman rho). Among patients who had a particular test in the past, the more inconvenient/painful they found the earlier tests, the less willing they were to undergo these tests in the future. It is noteworthy that these correlations were rather weak and were not significant for many tests, including correlations for skin and tumor biopsies. Negative experience with a previous MRI had the strongest negative correlation with being unwilling to undergo MRI studies in the future.

Table 6. Correlation of Patients' Previous Experience With Tests and Willingness to Undergo Tests Againa
TestNo. of Patients Who Previously Had the TestP (Spearman Rho)
  • PET indicates positron emission tomography; CT, computed tomography; MRI, magnetic resonance imaging.

  • a

    The first column represents the number of patients who previously underwent each test. The second column correlates patients' report of inconvenience/pain and their willingness to undergo these tests in the future. Negative coefficients indicate that the more inconvenient/painful the test was for patients in the past, the less willing they were to undergo that test in the future. The higher the absolute value of the coefficient, the stronger the correlation.

  • b

    P < .05: Negative valence indicates that the worse the previous experience, the less willing the patient is to undergo the test in the future.

  • c

    Pain of biopsy, rather than inconvenience, was measured.

PET scan54−.34
CT scan60−.41b
Skin biopsyc35−.18
Tumor biopsyc56−.23
Blood sample61−.21
Hair follicle11−.09
Stool sample42−.19
Research Question 5: To correlate willingness to undergo tests with financial/insurance coverage of tests

Patients were categorized into 2 groups: those who would enroll in a trial only if it was fully paid by a sponsor (n = 31) and those who would enroll if it was partially paid for by a sponsor or insurance (n = 28). Without exception, across all tests, the groups differed (P < .001) in their willingness to undergo tests. Members of the partial payment group (n = 28) were more willing to participate than those in the group who would participate only if the trial was fully paid (n = 31). Very few participants (n = 4) indicated that they would not be willing to undergo some of the various tests even if they were mandatory for enrollment in a clinical trial. This indicates that most patients were willing to undergo most tests regardless of whether the test was voluntary or mandatory and whether or not the test was paid for by a sponsor or insurance.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Early phase clinical trials in oncology have become increasingly complex and require patients to participate in a rising number of study procedures, including PD/PK tests and imaging studies. The impact on patients of compliance with study procedures frequently is discussed, but prospective data similar to those presented here have not been published. Patients with refractory malignancies have limited treatment options and may be more likely to comply with study requirements even if these cause inconvenience. Therefore, data examining patients' willingness to participate in the often invasive tests and to tolerate the burden of frequent imaging studies (eg, time commitment) are of essential value to investigators, patients, sponsors, advocacy groups, and regulatory agencies. The objective of our approach of asking about willingness and frequency of tests was to obtain a more quantitative readout and analysis from the posed questions.

In the current study, we have demonstrated that the overall willingness of patients with refractory, advanced malignancies to undergo study specific tests was very high. Very few patients (≤ 4 patients for all tests combined) were unwilling to undergo a/any particular test (and thus, to risk not being able receive treatment on a particular trial) even if a test was mandatory for trial enrollment, indicating that patients indeed would undergo most study-related procedures. Reasons may be obvious, such as limited therapy options remaining or wanting to participate in a trial with a particular agent. Another explanation is that invasive and noninvasive tests are perceived as less intrusive by patients than is commonly assumed. Regardless, patients with advanced-stage cancer are faced with a decision whether to obtain a new therapy regimen/agent on study versus the overall inconvenience. Understanding and, for the first time, actually asking patients with advanced-stage cancer themselves what they would be willing to do to enroll on study must be investigated. Thus, our main objective was to gain an initial insight into and a better understanding of what such patients would be willing to tolerate to receive treatment on a clinical trial.

Intuitively, patients were less willing to undergo more invasive tests, such as biopsies. Somewhat surprisingly, prior experience with tumor biopsies did not have a significant negative effect on patients' willingness to undergo additional tumor biopsies, indicating that most tumor biopsies are acceptable to patients. However, a subanalysis indicated that willingness declined sharply for more than 1 tumor biopsy in a single study, and undergoing serial biopsies was not met with great acceptance (“willingness”) by patients. These results need to be confirmed in additional studies; nonetheless, physicians, sponsors, and regulatory agencies should be cognitive about the number of tumor biopsies asked for from patients.

In general, imaging studies were well perceived. Ultrasound was the most accepted modality followed by echocardiogram, x-rays, PET scans, and CT scans. It is noteworthy that MRI was the least favored imaging test, which suggests that MRI is more inconvenient to patients than may be assumed by healthcare providers. With regard to skin biopsies, most patients were very willing to undergo these at least twice during a study. Other tests, such as urine and blood samples, were accepted very well regardless of frequency. Not surprisingly, the worse the prior experience with a test, the less willing patients were to undergo that test again. However, for half of the tests, the correlations were quite modest and were not significant (Table 6), including those for skin and tumor biopsies, which often are thought of as the 2 “most invasive” tests. This indicates that prior experience with a test is not necessarily a strong determinant of future willingness, which is an important finding from our study, because it is frequently perceived that invasive procedures potentially may dissuade patients from trial participation.

Demographics overall had little influence, apart from college graduates and patients with higher income, who had a higher positive association with willingness to undergo most tests. Comparing the only 5 minority participants with the white/Caucasian participants indicated that there was a trend for minority participants to be less willing to undergo PET scans, CR scans, MRI scans, skin biopsies, and stool samples than Caucasian participants (P < .05; data not shown). We did not want to include this exploratory analysis along with the formal results; however, we believe it is worth mentioning in this discussion, because it emphasizes the need to pay attention to providing support for minorities and less privileged socioeconomic groups, and early phase oncology trials can be very time consuming4, 17-19 and costly (eg, time off from work for relatives/caregivers) even if insurance coverage is provided. These observation needs to be confirmed in future studies.

The finding that 54% of patients with advanced, refractory disease did not know their stage of disease might be surprising. This may indicate a lack of communication between patients and their healthcare providers about the true status of their disease or that patients may not place as much emphasis on stage of disease, its meaning, and its treatment implications as healthcare providers. It also may reflect patients' need to exhibit a healthy and “appropriate amount of denial” to maintain a sense of hope8 and therapeutic optimism under difficult, life-threatening circumstances, as described previously.20 This latter aspect was not intended to be tested in our questionnaire.

The current study was conducted at a single clinical research center that conducts early stage (mainly phase 1 and 2) clinical trials. Patients were in the clinic to be evaluated and considered for new agent trials, which, in itself presents a selection. Thus, our results should not be generalized to all study centers, to phase 3 and 4 cancer trials, or to patients with early stage disease, because this study was not intended to cover those patient populations.

The statistical analysis approach, although somewhat complex, was chosen to address the dependant (within-participant) nature and skew (trend toward higher numbers on the survey scale) of the data. It allowed us to avoid simplifying assumptions (eg, compound symmetry in the covariance matrix) and took into account the correlations across repeated measures taken from the same participant.

Surprisingly, little is known about which specific aspects of clinical trial participation patients perceive as burdensome.21, 22 Our findings are consistent with other reports.21, 23 Physicians may have a lower threshold for acceptable risk than patients and frequently anticipate more anxiety in their patients than the patients actually report.4 Results from our study lend support to investigators and referring nononcology physicians for the idea that the informed patient is highly motivated and very willing to undergo invasive and time-consuming tests on clinical trials.

In retrospect, the questionnaire and the questions asked could have been modified with “posterior knowledge.” We hope that the presented data are valuable to the field and will motivate other investigators to perform similar patient-oriented research in an extension of our approach. Future investigations in this arena should obtain data from multiple centers to allow greater generalization; investigate the correlation between patient willingness and investigator-perceived willingness; and examine prospective correlations between self-reported willingness and actual participation in specific procedures. In summary, our prospectively collected data provide physicians, investigators, study sponsors, and regulatory agencies with an improved understanding of patients' willingness to undergo trial-required tests to enhance the design of biomarker-driven clinical trials in oncology.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We thank the patients who graciously participated in this study as well as the staff at the Translational Genomics Research Institute Clinical Research Services (TCRS) clinic and Ronald Korn for input and discussions regarding imaging.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors gratefully acknowledge support for this study from Scottsdale Clinical Research Institute, the Translational Genomics Research Institute, and the Helios Educational Foundation.


  1. Top of page
  2. Abstract
  6. Acknowledgements