AMDL-ELISA DR-70 (FDP) IMMUNOASSAY, MODEL DR2101

K072901 · Amdl, Inc. · NTY · Jul 1, 2008 · Immunology

Device Facts

• Record ID: K072901
• Device Name: AMDL-ELISA DR-70 (FDP) IMMUNOASSAY, MODEL DR2101
• Applicant: Amdl, Inc.
• Product Code: NTY · Immunology
• Decision Date: Jul 1, 2008
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 866.6010
• Device Class: Class 2

Indications for Use

The DR-70® (FDP) ELISA is designed for IN VITRO DIAGNOSTIC USE ONLY for the quantitative measurement of DR-70® (FDP) in human serum. Serial testing using the AMDL-ELISA DR-70® (FDP) is to be used as an aid in monitoring the disease progression of patients who have been diagnosed previously with colorectal cancer. Results of DR-70® (FDP) testing should be used in conjunction with other clinical modalities that are standard of care for monitoring disease progression in these patients.

Device Story

The AMDL-ELISA DR-70® (FDP) is an enzyme-linked immunosorbent assay (ELISA) used to quantify Fibrin/ogen Degradation Products (FDP) in human serum. The device utilizes a 96-well microtiter plate coated with affinity-purified rabbit anti-DR-70 antibodies. Patient serum (diluted 1:200) is added to the wells; captured FDP is then bound by horseradish peroxidase-conjugated anti-DR-70 antibodies, forming an immunological sandwich. After washing, TMB substrate is added, and the reaction is stopped with 0.1N HCl. A microplate reader measures absorbance at 450 nm. The concentration is determined by interpolating from a standard curve. The test is performed in a clinical laboratory setting. Clinicians use the results, specifically a 15% or greater increase in DR-70 levels between serial visits, as an aid in monitoring disease progression in colorectal cancer patients, alongside standard clinical modalities.

Clinical Evidence

Clinical study evaluated 112 colorectal cancer patients with 446 paired observations. Performance was assessed by comparing DR-70 serial measurements to clinical disease status. A 15% increase in DR-70 levels was defined as a positive test for progression. Per-visit analysis (n=335) showed 65.19% sensitivity and 67.34% specificity. Per-patient analysis (n=112) showed 66.21% sensitivity and 68.18% specificity. PPV was 57.52% and NPV was 74.03%.

Technological Characteristics

ELISA-based sandwich immunoassay. Components: affinity-purified rabbit anti-DR-70 antibody-coated 96-well microtiter plates, horseradish peroxidase-conjugated anti-DR-70 antibodies, TMB substrate. Energy source: microplate reader (450 nm). Sample: 10uL human serum. Analytical sensitivity: 0.06 µg/ml. Functional sensitivity: 0.052-0.063 µg/ml. No hook effect up to 250 µg/ml.

Indications for Use

Indicated for serial monitoring of disease progression in patients previously diagnosed with colorectal cancer. Prescription only.

Regulatory Classification

Identification

A tumor-associated antigen immunological test system is a device that consists of reagents used to qualitatively or quantitatively measure, by immunochemical techniques, tumor-associated antigens in serum, plasma, urine, or other body fluids. This device is intended as an aid in monitoring patients for disease progress or response to therapy or for the detection of recurrent or residual disease.

Special Controls

*Classification.* Class II (special controls). Tumor markers must comply with the following special controls: (1) A guidance document entitled “Guidance Document for the Submission of Tumor Associated Antigen Premarket Notifications (510(k)s) to FDA,” and (2) voluntary assay performance standards issued by the National Committee on Clinical Laboratory Standards.

Predicate Devices

• TOSOH BioScience AIA-PACK™ CEA (P910053)

Submission Summary (Full Text)

{0}

1

# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY

A. 510(k) Number:
k072901

B. Purpose for Submission:
New device

C. Measurand:
Fibrin and Fibrinogen Degradation Products (FDP)

D. Type of Test:
Quantitative, ELISA Immunoassay

E. Applicant:
AMDL Inc.

F. Proprietary and Established Names:
DR-70® FDP

G. Regulatory Information:
1. Regulation section:
21 CFR §866.6010, Tumor associated antigen immunological test system
2. Classification:
Class II
3. Product codes:
NTY, System, Test, Fibrin/Fibrinogen degradation products for monitoring of colorectal cancer
4. Panel:
Immunology (82)

H. Intended Use:
1. Intended use:
The DR-70® (FDP) ELISA is designed for IN VITRO DIAGNOSTIC USE ONLY for the quantitative measurement of DR-70® (FDP) in human serum.
2. Indication(s) for use:
Serial testing using the AMDL-ELISA DR-70® (FDP) is to be used as an aid in monitoring the disease progression of patients who have been diagnosed previously with colorectal cancer. Results of DR-70® (FDP) testing should be used in conjunction with other clinical modalities that are standard of care for monitoring disease progression in these patients.
3. Special conditions for use statement(s):
For prescription only.
4. Special instrument requirements:
ELISA microwell plate reader that reads the 450 nm (kinetic) wavelengths

I. Device Description:
The DR-70® (FDP) Kit consists of DR-70® (FDP) Antibody-Coated Wells for a 96-well Plate; Enzyme Antibody Conjugate; Low Control; High Control; DR-70® (FDP) calibrators at concentrations of: 0, 0.625, 2.5, 5.0, and 10.0 µg/mL; Diluent Buffer Concentrate (5X): Wash Buffer Concentrate (20X): 3,3',5,5'-tetramethylbenzidine (TMB) Substrate; Stop Solution and Dilution/Transfer Plate (96 well uncoated Plate)

{1}

2

J. Substantial Equivalence Information:

1. Predicate device name(s):
TOSOH BioScience AIA-PACK™ CEA

2. Predicate K number(s):
P910053

3. Comparison with predicate:

| Similarities | | |
| --- | --- | --- |
| Item | Device | Predicate |
| | AMDL-ELISA DR-70®(FDP) | TOSOH AIA-PACK™ CEA |
| Intended Use | Monitoring disease progression in patients previously diagnosed with colorectal cancer | Same |
| Sample | Human serum | Same |
| Differences | | |
| --- | --- | --- |
| Item | New Device | Predicate Devices |
| | AMDL-ELISA DR-70®(FDP) | TOSOH AIA-PACK™ CEA |
| Analytes | Fibrin and Fibrinogen Degradation products | Cancer Embryonic Antigen (CEA) |
| Antibody | Polyclonal (rabbit) | Monoclonal (mouse) |
| Methodology | Manual ELISA | Automated Immunoassay analyzers |
| Solid phase capture | Antibody-coated microwells | Antibody-coated magnetic beads |
| Substrate | TMP | 4-methylumbelliferyl phosphate |
| Detection Method | Chromogenic | Fluorogenic |
| Precision | < 10.6% | 3.2-3.9% |
| Sample Volume | 100 μL | 10 μL |

K. Standard/Guidance Document Referenced (if applicable):
CLSI EP5-A2 for precision study; EP6A for linearity.

L. Test Principle:
The AMDL, Inc. DR-70® (FDP) assay is an ELISA based assay utilizing removable strips in a 96 micro titer plate well format. The wells are coated with affinity purified rabbit anti-DR-70 polyclonal antibodies. The DR-70 antigen in diluted patient serum (1:200) is captured by these antibodies immobilized in the well of a micro titer plate. After a wash step, anti-DR-70 antibodies conjugated to horseradish peroxidase (HRP) are added to the wells. If the DR-70 antigen is present, the HRP-labeled anti-DR-70 Ab will bind to the captured tumor marker to form an immunological sandwich with the immobilized antibodies.

{2}

After a second wash step, the enzyme substrate TMB is added to the well. The end point is read in a micro plate reader at 450 nm once the reaction is stopped with 0.1N HCl. The intensity of the color formed is proportional to the amount of DR-70 in the serum. The amount is quantified by interpolation from a standard curve using the calibrators provided with the kit.

## M. Performance Characteristics (if/when applicable):

### 1. Analytical performance:

#### a. Precision/Reproducibility:

Five specimens were tested by 3 sites, 20 days, 3 lots, 2 runs (per day), and replicate (2 per run) for 720 lines of data per specimen. The five specimens were Pool 1 (low concentration), Pool 2 (medium concentration), Pool 3 (high concentration), QC1 (low concentration), and QC2 (high concentration). The study day is nested within study site and run is nested within day. To obtain variance components for all variables, each was considered random in the mixed model.

The highest mean concentration of serum pool for the precision study of 2.8 ng/mL is accepted based on the rationales that majority (&gt;90%) of observed FDP of colorectal cancer are below 5 ng/mL, and the assay imprecision is usually lower in the upper assay range.

Table 1: Intra-assay, Between-run, Day-to-day, and Total Assay Imprecision

| Sample | Mean Concentration (μg/mL) | Within-Site Imprecision (%CV) | Between-Day Imprecision (%CV) | Between-Lot Imprecision (%CV) | Within-Run Imprecision (%CV) | Total Variability (%CV) |
| --- | --- | --- | --- | --- | --- | --- |
| Pool 1 | 0.315 | 11.60 | 6.48 | 6.64 | 5.53 | 22.53 |
| Pool 2 | 1.389 | 3.51 | 4.31 | 0.83 | 3.06 | 10.60 |
| Pool 3 | 2.739 | 4.45 | 3.43 | 0.76 | 2.74 | 9.91 |
| QC 1 | 0.240 | 13.74 | 8.58 | 11.03 | 4.84 | 28.21 |
| QC 2 | 2.994 | 5.01 | 3.74 | 3.05 | 1.27 | 11.8 |

Table 2. Total and Components of Assay Variance and Percentage by Source

| Specimen | Mean Concentration (μg/mL) | Site Variance (%) | Day Variance (%) | Lot Variance (%) | Run Variance (%) | Residual Variance (%) | Total Variance |
| --- | --- | --- | --- | --- | --- | --- | --- |
| Pool 1 | 0.315 | 0.001336 (23.53) | 0.000417 (8.28) | 0.000437 (8.68) | 0.000303 (6.02) | 0.002543 (50.50) | 0.005036 |
| Pool 2 | 1.389 | 0.002385 (11.00) | 0.003585 (16.53) | 0.000132 (0.61) | 0.001810 (8.34) | 0.01378 (63.53) | 0.021692 |
| Pool 3 | 2.739 | 0.01483 (20.11) | 0.008803 (11.94) | 0.000431 (0.58) | 0.005631 (7.64) | 0.04405 (59.73) | 0.073745 |
| QC1 | 0.240 | 0.001088 (23.74) | 0.000424 (9.25) | 0.000701 (15.29) | 0.000135 (2.95) | 0.002236 (48.78) | 0.004584 |
| QC2 | 2.994 | 0.02246 (20.03) | 0.01255 (11.19) | 0.008363 (7.46) | 0.001438 (1.28) | 0.06731 (60.03) | 0.112121 |

## Accuracy/Spiked Recovery

Sera from three normal subjects having DR-70 values ranging from 0.3 μg/mL to

{3}

0.6 µg/mL and a control diluent buffer were spiked with a DR-70 antigen solution to obtain expected levels ranging from 1.5 µg/mL to 10 µg/mL to represent the range of the DR-70 calibrators. The values of DR-70 in the spiked serum samples were measured and compared to the theoretical values and to values obtained for the control diluent buffer. The experiment was designed to compare responses of the analyte in a biological sample versus the standard diluent to assess for any difference in assay response.

Table 3. Spike Recovery

| Sample | Non-spiked | Spiked in DR-70 concentration value (µg/mL) | | | | |
| --- | --- | --- | --- | --- | --- | --- |
| | | 1.5 | 2.5 | 5.0 | 7.0 | 10 |
| Diluent buffer(5x) | 0 | 1.517 | 2.649 | 4.586 | 6.983 | 10.94 |
| Patient 1 | 0.428 | 1.743 | 2.908 | 4.839 | 7.057 | 13.11 |
| Patient 2 | 0.576 | 1.520 | 2.680 | 4.848 | 7.050 | 11.95 |
| Patient 3 | 0.464 | 1.598 | 2.967 | 5.193 | 6.701 | 10.88 |
| Patient mean value | 0.489 | 1.620 | 2.852 | 4.960 | 6.936 | 11.98 |
| % Mean Recovery | --- | 107% | 108% | 108% | 99% | 110% |

b. Linearity/assay reportable range

Linearity

The linearity of a marker over the range of evaluation was analyzed by a method consistent with NCCLS EP6-A. Two-fold serial dilutions were made on serum samples from 5 colorectal cancer patients with DR-70 values in the range of 19.7 to 22.2 µg/mL with assay diluent buffer. For each CRC patient serum sample, a total of 9 diluted samples were tested. The following table lists the average % difference between the actual and the estimated DR-70 concentrations for each dilution as well as the average % recovery. The DR-70® FDP ELISA assay was linear from 1.5- 10 µg/mL with recoveries from 93 to 108%.

Table 4. DR70 Linearity

| Estimated FDP Conc. ng/mL | Dilution Ratio | Highest Positive % Difference (n=5) | Lowest (Negative) % Difference (n=5) | Average % Difference | Average % Recovery |
| --- | --- | --- | --- | --- | --- |
| 20 | 1 | 2.6 | (14.5) | (4.4) | 96 |
| 10 | 1/2 | 7.5 | (7.4) | (1.0) | 99 |
| 5 | 1/4 | 7.0 | (2.6) | 0.1 | 100 |
| 2.5 | 1/8 | 11.8 | (21.2) | (7.0) | 93 |
| 1.25 | 1/16 | 3.6 | (14.5) | (5.7) | 94 |
| 1.125 | 1/32 | 20.1 | (5.8) | 7.6 | 108 |
| 0.625 | 1/64 | 16.7 | 11.8 | 13.6 | 114 |

High dose hook effect (assessment of antigen excess)

No evidence of a hook effect was found up to a concentration of 200 µg/mL

Assay reportable range

{4}

Assay reportable range is from 0.063 to 10 µg/mL.

c. Traceability, Stability, Expected values (controls, calibrators, or methods): There is no recognized reference standard for DR 70. Value assignment of the controls and calibrators was based on gravimetric method.

## Stability

The DR-70® (FDP) Kit has an expiration date assignment of up to 18 months which is based on the component with the shortest dated stability data for the kit components. Real time stability studies have been conducted on each production lot of kits consisting from the initial date of QC up to the amount of months to the assigned kit expiration.

d. Detection limit

### Analytical Sensitivity

The minimal detectable concentration (MDC) of DR-70® (FDP) is estimated to be 0.06 µg/mL. The MDC is defined as that concentration of DR-70® (FDP) corresponding to the absorbance that is two standard deviations from the mean rate of absorbance of 20 replicate determinations of a zero calibrator.

### Functional Sensitivity

The functional sensitivity was determined by diluting the lowest non-zero calibrator serially, measuring the DR-70® (FDP) concentration and extrapolating to the point where the CV% = 20%. Functional sensitivity for the AMDL-ELISA DR-70® (FDP) was calculated as being between 0.052 and 0.063 µg/mL.

e. Analytical specificity and interference

Interference was defined, for purpose of this study, to be recovery &gt; 10% of the known specimen mean concentration. Interferents (hemoglobin, bilirubin, triglyceride and heparin) were spiked into two patient serum pools with background man FDP concentrations at 1.07 and 2.2 µg/mL. Results showed that hemoglobin (up to 500 mg/dl), bilirubin (up to 30 mg/dl), triglyceride (up to 1000 mg/dl) and heparin (at concentrations of 500 U/ml) do not interfere with the assay.

In addition, the following pharmaceutical agents were tested at levels indicated and found not to cause analyte recovery &gt; 10%: 5'-fluorouracil (Adrucil) 1.0 mg/mL; Acetaminophen 0.2 mg/mL; Adriamycin (Doxorubicin HCl) 0.10 mg/dL; Coumarin 1.4 mg/mL; Cyclophosphamide (Cytoxan) 0.25 mg/mL; Paclitaxel, 3.5 × 10⁻⁶ g/m²; Amethopterin hydrate (Methotrexate) 4.5 mg/mL; Mitoxantrone (Novatrone) 0.5 mg/mL; Folinic acid (Leucovorin) 1.10 mg/mL; Mitomycin C, 0.06 mg/mL and Cisplatin, 0.10 mg/mL.

f. Assay cut-off

A 15% increase from the previous visit was chosen as the threshold for significant % change for the determination of disease progression in the DR-70® (FDP) immunoassay based on the % total CV from the precision study. The total %CV was computed over all runs, days, and intra-assay for each

5

{5}

specimen analyzed and the highest CV values were observed for specimens with low DR 70 concentrations (0.21-0.42 µg/mL). Since over 80% of the DR 70 measurements in the cancer monitoring samples had concentrations ≥ 0.6 µg/mL, the sponsor used the average of the highest %CV obtained for samples with 1.31 (CV=7.85) and 4.11 (CV=7.14) µg/mL which was equal to 7.495%.

The δ (significant change in marker value) is equal to 1.96 times the %total CV. Using the value of 7.495%, the threshold for significant change was estimated to be 14.69% (7.495*1.96) rounded up to 15%. Thus if a later visit has a value that is greater than 15% higher than the previous value, it will be considered evidence of disease progression.

2. Comparison studies:
a. Method comparison with predicate device:
Not applicable since there is no predicate device.
b. Matrix comparison:
Not applicable, use of serum as specimen only.

3. Clinical studies:
a. Clinical Sensitivity and Clinical Specificity:
A clinical evaluation was performed to assess the DR 70 (FDP) assay for monitoring patients with colorectal cancer. Serial samples were taken from 112 colorectal cancer patients resulting in 445 paired observations (335 paired observations were post baseline samples). The sequential draws covered an average longitudinal period of at least nine months. The samples were retrospective banked samples that were collected blindly and without bias to include all patients with diagnosed colorectal cancer in the bank at the time of the collection. Inclusion and exclusion criteria for the samples were provided.

The breakdown of the patient series is presented in Table 5. The average number of observations per patient is 4.0.

Table 5. Patient Observation Series
| Number of Samples in Series | Number of Observation Pairs in Series | Total Number of Series with that Number of Pairs | Percent of the Total Samples | Cumulative Percent of Samples |
| --- | --- | --- | --- | --- |
| 2 | 1 | 1 | 0.9 | 0.9 |
| 3 | 2 | 38 | 33.9 | 34.8 |
| 4 | 3 | 48 | 42.9 | 77.7 |
| 5 | 4 | 18 | 16.1 | 93.8 |
| 6 | 5 | 3 | 2.7 | 96.5 |
| 7 | 6 | 3 | 2.7 | 99.2 |
| 8 | 7 | 1 | 0.9 | 100.0 |

Table 6 presents the stage of the disease at time of diagnosis for 111 of the 112 evaluable serial patients. One patient chart did not contain information related to the stage at time of diagnosis.

6

{6}

Table 6. Stage of Cancer at Time of Diagnosis

| Stage at Diagnosis | Frequency | Percentage | Cumulative Percentage |
| --- | --- | --- | --- |
| 0 | 1 | 0.9 | 0.9 |
| I | 4 | 3.6 | 4.5 |
| II | 18 | 16.2 | 20.7 |
| III | 39 | 35.1 | 55.9 |
| IV | 49 | 44.1 | 100.0 |
| Total | 111 | 100.0 | |

Table 7 demonstrates the relationship between stage at diagnosis and the presence of metastases. As the stage of the disease progressed, the percentage of patients with metastases increased.

Table 7. Distribution of Metastases by Stage at Diagnosis

| | Known Metastases at time of Diagnosis: | | |
| --- | --- | --- | --- |
| Stage | Yes | No | Total |
| 0-I | 0 | 5 | 5 |
| | 0.0% | 100.0% | 100.0% |
| II | 3 | 15 | 18 |
| | 16.7% | 83.3% | 100.0% |
| III | 29 | 10 | 39 |
| | 74.4% | 25.6% | 100.0% |
| IV | 49 | 0 | 49 |
| | 100.0% | 0.0% | 100.0% |
| Total | 81 | 30 | 111 |
| | 73.0% | 27.0% | 100.0% |

## Data analysis

Clinical disease progression was determined by the Subject’s physician based on office procedures and clinical laboratory based analyses that were the standard of care during the time of the monitoring period. Progression of the DR-70 value was determined as a significant percentage change (15%) between the current and previous readings

The 335 paired observations from the post baseline sampling were evaluated in two ways. The initial analysis presents estimates directly from the data. This analysis is followed by a bootstrap sample for each patient by randomly sampling one visit from each sample and recording the sensitivity or specificity for that visit. Note that if there was a progression the sensitivity would be 1 if the DR-70 increased from the previous visit by 15% or more and 0 if it did not. If there were no progression at that visit, then there would be no sensitivity reported at that visit, but the specificity would be reported as 1 if the DR-70® (FDP) value was below a 15% increase for that visit and 0 otherwise. For the per-visit analysis, there were 135 visits for sensitivity and 198 visits for specificity.

{7}

A second analysis was done on a per-patient basis in which the number of progressions across all visits for a given patient was used to compute a patient level sensitivity by taking the number visits that DR-70 value increased by at least 15% among the number of visits in which there was a progression. Similarly, the number of visits at which DR-70 had a lower than 15% increase divided by the number of visits in which there was a non-progression allowed the computation of a per-patient specificity. If a patient had all progressions there would be no specificity for that patient and if a patient had all non-progressions, there would be no sensitivity for that patient. This resulted in a sample of 112 patients with at least one sensitivity, specificity, or both. This resulted in 70 estimates of per-patient sensitivity and 86 estimates of per-patient specificity.

## Results of per-visit analysis

There were 10,000 bootstrap samples of 112 observations taken with replacement from the 334 paired observations. After each sample was taken, the following table was formed.

Table 8. Distribution of Progression by DR-70® (FDP) Value Increase from Previous Visit

| | Disease Progression | | Total | |
| --- | --- | --- | --- | --- |
| | | No | | Yes |
| DR-70®
(FDP) | < 15% | 134 | 47 | 181 |
| | ≥15% | 65 | 88 | 153 |
| Total | | 199 | 135 | 334 |

The computed per-visit sensitivity from the 334 per-visit evaluations was 100*88/135 = 65.19 with standard deviation (SD) 2.58, the specificity was 100*134/199 = 67.34 with SD = 2.94, the sum of sensitivity and specificity was 132.53 with SD = 3.91, the PPV was 100*88/153 = 57.52 with SD = 1.63, and the NPV was 100*134/181 = 74.03 with SD = 2.44.

The results from the tabulations and per-visit bootstrap are shown in Table 9.

{8}

Table 9. Results of Tabulated and Five Repetitions of 2,000 Samples of 112 per-visit Observations of the Sensitivity, Specificity, the Sum of Sensitivity and Specificity, PPV, and NPV

| Run | Measure | Median | Lower 5% | Lower 2.5% |
| --- | --- | --- | --- | --- |
| From Data | Sensitivity | 65.19 | 60.91 | 60.13 |
| | Specificity | 67.34 | 62.50 | 61.58 |
| | Sensitivity + Specificity | 132.53 | 126.10 | 124.87 |
| | PPV | 57.52 | 54.84 | 54.33 |
| | NPV | 74.03 | 70.02 | 69.25 |
| Bootstrap 1 | Sensitivity | 65.85 | 55.81 | 54.17 |
| | Specificity | 66.20 | 59.09 | 57.63 |
| | Sensitivity + Specificity | 132.66 | 120.31 | 118.23 |
| | PPV | 58.33 | 50.91 | 49.23 |
| | NPV | 73.13 | 66.18 | 65.45 |
| Bootstrap 2 | Sensitivity | 65.85 | 56.25 | 54.17 |
| | Specificity | 67.19 | 58.46 | 56.92 |
| | Sensitivity + Specificity | 132.92 | 120.75 | 118.72 |
| | PPV | 58.82 | 50.91 | 48.98 |
| | NPV | 73.21 | 66.18 | 64.91 |
| Bootstrap 3 | Sensitivity | 65.91 | 56.10 | 54.00 |
| | Specificity | 67.16 | 58.73 | 56.92 |
| | Sensitivity + Specificity | 133.20 | 120.29 | 117.81 |
| | PPV | 58.93 | 49.15 | 48.94 |
| | NPV | 73.33 | 66.18 | 65.08 |
| Bootstrap 4 | Sensitivity | 65.85 | 56.25 | 53.66 |
| | Specificity | 66.67 | 58.57 | 56.72 |
| | Sensitivity + Specificity | 132.96 | 119.73 | 116.99 |
| | PPV | 58.73 | 49.18 | 48.84 |
| | NPV | 73.13 | 66.00 | 64.52 |
| Bootstrap 5 | Sensitivity | 65.96 | 56.25 | 54.90 |
| | Specificity | 67.16 | 58.73 | 57.38 |
| | Sensitivity + Specificity | 133.29 | 120.62 | 118.23 |
| | PPV | 59.02 | 50.88 | 49.09 |
| | NPV | 73.24 | 66.20 | 64.91 |

Using alpha = 0.05, the sum of sensitivity and specificity from this analysis clearly is statistically significantly above 100. The median sum is likely to be about 133 with the median sensitivity about 65 and the median specificity about 67. The lower two-sided $5\%$ confidence bound is about 118. The median PPV and NPV are about 59 and 73, respectively, across the five samples. Note that these values are consistent with those computed from the per-visit values given above Table 9. The five repetitions of the sample of 2,000 demonstrate that the result is robust and consistent.

# Results of per-patient analysis

For the per-patient analysis, the computed per-visit sensitivity from the 112 per-patient evaluations was $100^{*}45.68 / 69 = 66.21$ , the specificity was $100^{*}58.63 / 86 = 68.18$ , the sum of sensitivity and specificity was 134.39, the

{9}

PPV was $100*51.83 / 97 = 53.44$ , and the NPV was $100*71.67 / 103 = 69.58$ .

# The per-patient bootstrap per-patient results

The confidence intervals are obtained from the bootstrap evaluations below.

Table 10. Results of Five Repetitions of 2,000 Samples of 112 per-patient Observations of the Sensitivity, Specificity, the Sum of Sensitivity and Specificity, PPV, and NPV

| Run | Measure | Median | Lower 5% | Lower 2.5% |
| --- | --- | --- | --- | --- |
| Bootstrap 1 | Sensitivity | 66.12 | 58.78 | 57.11 |
| | Specificity | 68.20 | 62.58 | 61.46 |
| | Sensitivity + Specificity | 134.38 | 123.82 | 121.46 |
| | PPV | 53.30 | 45.88 | 44.55 |
| | NPV | 69.74 | 62.71 | 60.90 |
| Bootstrap 2 | Sensitivity | 66.17 | 58.58 | 57.13 |
| | Specificity | 68.16 | 62.39 | 61.41 |
| | Sensitivity + Specificity | 134.26 | 123.85 | 122.17 |
| | PPV | 53.41 | 45.34 | 43.96 |
| | NPV | 69.50 | 62.75 | 61.37 |
| Bootstrap 3 | Sensitivity | 66.13 | 58.59 | 57.27 |
| | Specificity | 68.20 | 62.29 | 61.23 |
| | Sensitivity + Specificity | 134.32 | 123.79 | 121.36 |
| | PPV | 53.37 | 45.71 | 43.89 |
| | NPV | 69.50 | 62.06 | 60.89 |
| Bootstrap 4 | Sensitivity | 66.43 | 58.53 | 56.99 |
| | Specificity | 68.28 | 62.64 | 61.55 |
| | Sensitivity + Specificity | 134.69 | 123.87 | 121.17 |
| | PPV | 53.70 | 45.92 | 44.39 |
| | NPV | 69.55 | 62.71 | 61.33 |
| Bootstrap 5 | Sensitivity | 66.42 | 59.00 | 57.92 |
| | Specificity | 68.29 | 62.82 | 61.64 |
| | Sensitivity + Specificity | 134.62 | 124.38 | 122.92 |
| | PPV | 53.41 | 45.75 | 44.15 |
| | NPV | 69.64 | 62.66 | 61.38 |

Using alpha $= 0.05$ , the sum of sensitivity and specificity from this analysis is statistically significantly above 100. The median sum is likely to be about 134 with the median sensitivity about 66 and the median specificity about 68. The lower two-sided $5\%$ confidence bound for the sum is about 121. The median PPV and NPV are about 53 and 69, respectively, across the five samples. Note that these values are consistent with those computed from the per-patient values given above Table 10. The five repetitions of the sample of 2,000 demonstrate that the result is robust and consistent.

# The per-patient bootstrap results by cancer stage

The data are limited with respect to the stage of the cancer, but the analysis below presents the results of the estimates of sensitivity, specificity, sum, PPV, and NPV with confidence intervals computed with variances from the Emir et al (1998) method. The number of patients at each stage was 1 in

{10}

Stage 0, 4 in Stage 1, 18 in Stage 2, 39 in Stage 3, and 50 in Stage 4. For the purposes of this analysis, Stage 0-2 are combined in one analysis, and Stages 3 and 4 are analyzed separately.

The table of sensitivity, specificity, sum, PPV and NPV by cancer stage is given below. It should be noted that there were only 6 patients with 8 progressions in the Stage 0-2 group so the estimate of sensitivity will have large variability.

The per-patient bootstrap results by cancer stage are in the following table:

Table 11. Results of Sensitivity, Specificity, the Sum of Sensitivity and Specificity, PPV, and NPV by Cancer Stage

| Stage | Measure | Median | Lower 5% | Lower 2.5% |
| --- | --- | --- | --- | --- |
| 0-2 | Sensitivity | 62.50 | 44.36 | 40.88 |
| | Specificity | 60.94 | 51.19 | 49.32 |
| | Sensitivity + Specificity | 123.44 | 102.84 | 98.90 |
| | PPV | 16.67 | 12.27 | 11.43 |
| | NPV | 90.70 | 78.58 | 76.26 |
| 3 | Sensitivity | 58.70 | 47.90 | 45.83 |
| | Specificity | 73.08 | 58.35 | 55.53 |
| | Sensitivity + Specificity | 131.77 | 113.51 | 110.02 |
| | PPV | 56.25 | 49.22 | 47.87 |
| | NPV | 75.00 | 63.60 | 61.40 |
| 4 | Sensitivity | 69.14 | 62.68 | 61.44 |
| | Specificity | 66.67 | 59.72 | 58.40 |
| | Sensitivity + Specificity | 135.80 | 126.32 | 124.51 |
| | PPV | 74.32 | 68.73 | 67.65 |
| | NPV | 60.32 | 54.75 | 53.69 |

Note that even among the stages with the lowest numbers of patients, the two-sided $5\%$ lower confidence limit excludes 1 and the test retains its informative nature for Stage 3 and Stage 4. Predictably, as the prevalence increases across the stages, PPV increases and NPV decreases.

b. Other clinical supportive data (when a. is not applicable):

# 4. Clinical cut-off:

For monitoring of disease status, fifteen percent (15%) or greater change from the previous visit for monitoring of disease progression.

# 5. Expected values/Reference range:

The distribution of DR 70 concentrations determined in 1185 serum samples from normal subjects and patients with nonmalignant or malignant disease was evaluated. In this study, $84.5\%$ of the normal subjects had DR-70 levels less than $1.5~\mu \mathrm{g / mL}$ (Table 12).

{11}

Table 12. Distribution in Percent of Serum DR-70® (FDP) Values **

| | | Percent (%)
95% CI (lower-upper %)* | | | |
| --- | --- | --- | --- | --- | --- |
| Disease | n | 0-1.4 μg/ml | 1.5-2.4 μg/ml | 2.5-4.9μg/ml | ≥ 5.0 μg/ml |
| Normal | 420 | 94.5 (91.9, 96.5) | 5.0 (3.1, 7.5) | 0.5 (0.1, 1.7) | 0.0 (0.0, 0.9) |
| < 65 years | 337 | 96.4 (93.9, 98.2) | 3.3 (1.6, 5.8) | 0.3 (0.0, 1.6) | 0.0 (0.0, 1.1) |
| ≥ 65 years | 83 | 86.8 (77.5, 93.2) | 12.1 (5.9, 21.0) | 1.2 (0.0, 6.5) | 0.0 (0.0, 4.4) |
| | | | | | |
| Benign | 326 | 75.5 (70.4, 80.0) | 6.8 (4.3, 10.0) | 0.6 (0.1, 2.2) | 17.2 (13.2, 21.7) |
| GU Disease | 94 | 94.7 (88.0, 98.3) | 4.3 (1.2, 10.5) | 0.0 (0.0, 3.9) | 1.1 (0.0, 5.8) |
| GI Disease | 61 | 90.2 (79.8, 96.3) | 3.3 (0.4, 11.4) | 0.0 (0.0, 5.9) | 6.6 (1.8, 16.0) |
| Pancreas | 84 | 60.7 (49.5, 71.2) | 15.5 (8.5, 25.0) | 2.4 (0.3, 8.3) | 21.4 (13.2, 31.7) |
| Heart Disease | 87 | 58.6 (47.6, 69.1) | 3.5 (0.7, 9.8) | 0.0 (0.0, 4.2) | 37.9 (27.7, 49.0) |
| | | | | | |
| Malignant | 439 | 44.0 (39.3, 48.8) | 24.2 (20.2, 28.4) | 19.6 (16.0, 23.6) | 12.3 (9.4, 15.7) |
| Colon | 187 | 55.6 (48.2, 62.9) | 21.4 (15.7, 28.0) | 15.0 (10.2, 20.9) | 8.0 (4.6, 12.9) |
| Lung | 44 | 34.1 (20.5, 49.9) | 38.6 (24.4, 54.5) | 18.2 (8.2, 32.7) | 9.1 (2.5, 21.7) |
| Liver | 44 | 31.8 (18.6, 47.6) | 27.3 (15.0, 42.8) | 22.7 (11.5, 37.8) | 18.2 (8.2, 32.7) |
| Breast | 31 | 54.8 (36.0, 72.7) | 25.8 (11.9, 44.6) | 12.9 (3.6, 29.8) | 6.5 (0.8, 21.4) |
| Ovarian | 31 | 25.8 (11.9, 44.6) | 6.5 (0.8, 21.4) | 32.3 (16.7, 51.4) | 35.5 (19.2, 54.6) |
| Cervical | 28 | 50.0 (30.7, 69.4) | 28.6 (13.2, 48.7) | 7.1 (0.9, 23.5) | 14.3 (4.0, 32.7) |
| Gall Bladder | 19 | 42.1 (20.3, 66.5) | 26.3 (9.2, 51.2) | 31.6 (12.6, 56.6) | 0.0 (0.0, 17.7) |
| Pancreas | 28 | 25.0 (10.7, 44.9) | 17.9 (6.1, 36.9) | 35.7 (18.6, 55.9) | 21.4 (8.3, 41.0) |
| Gastric/ Other | 27 | 22.2 (8.6, 42.3) | 33.3 (16.5, 54.0) | 29.6 (13.8, 50.2) | 14.8 (4.2, 33.7) |

*Exact binomial confidence limits.

## N. Proposed Labeling:

The labeling is sufficient and it satisfies the requirements of 21 CFR Part 809.10.

## O. Conclusion:

The submitted information in this premarket notification is complete and supports a substantial equivalence decision