Presented at ASMS 2018

Troubleshooting LC-MS/MS Biomarker Assays Transfer for Sphingolipids with Multiple Isomers in Human Dried Blood SpotsAllena J. Ji1, Nan Zhao2, Yi Zhu1, Mona Hdeib1, Troy Voelker*2 and Scott Reuschel21Biomarkers and Clinical Bioanalyses-Boston, Sanofi, Framingham, MA; 2Covance, Salt Lake City, UT; *Poster Presenter and corresponding author

IntroductionTotal ceramide (11 isomers, CER) and total sphingomyelin (17 isomers, SPM) are surrogate biomarkers in plasma and blood for monitoring patients with Niemann-Pick diseases when they have enzyme replacement therapy (Figure 1). The two bioanalytical methods used for several clinical trials in the sponsor’s lab were transferred to Covance. Both methods were cross-validated at Covance and met precision and accuracy acceptance criteria. However, the normal reference ranges from the same 100 individual blood samples showed significant difference between the two labs (Figures 2 and 3). A series of systematic troubleshooting with innovative approaches were conducted, and results showed that modification of instrument collision energy to match isomer patterns of patient samples and standards (porcine CER and SPM) between labs and elimination of matrix difference using the same pre-spotted DBS cards for STDs preparation are necessary to obtain comparable reference ranges between labs.

Novel Aspect The results of total sphingolipids in human blood can be significantly different between two labs due to isoform pattern mismatch between animals (for standards) and humans (unknown). Adjusting instrument parameters could lead to isoform pattern similarity and comparable reference ranges between labs.

MethodsDried blood spot (DBS) samples were punched into a 96-well plate and soaked in the solvent with IS. The supernatant was transferred to a clean plate for LC-MS/MS analysis. Sample analyses were performed on Sciex API 4000 with Aquity UPLC, by using isocratic mobile phase and a reverse phase HPLC column. MRM mode was used for CER (11 isomers and its IS for 12 transitions), and SPM (17 isomers and its IS for 18 transitions), respectively. All method procedures from the sponsor were followed. Stock solutions for CER and SPM were shared by the two labs. Mass spectrometer’s parameters, including collision energy and decluttering potential, were adjusted through all isomers to match the isoform patterns of standards from the sponsor’s lab. The chromatogram of total ceramide and total SPM in human dried blood spot containing endogenous are shown in Figures 4 and 5, respectively.

Figure 1. Structure of ceramide and sphingomyelin

0

5

10

15

20

25

# of

lots

DBS SPM Conc., µ/mL

CovanceSanofi

Before Modification (SPM)

Figure 3. Different reference ranges from the same 100 donors for SPM in human dried blood spots before SPM method parameter modification using identical instrument parameters in both labs.

DataPrecision and accuracy of both assays met the validation acceptance criteria, when standards and QCs shared the same porcine reference standard for each assay. In contrast, total isomer results for endogenous CER or SPM from human unknown samples were different when evaluated on LC/MS/MS systems from the two labs even when the same stock solutions and same instrument parameters were shared. The values for CER and SPM using the same 100 human blood lots showed a significant differences (Figures 2 and 3). For DBS CER, the sponsor’s lab result was 12.1 ±3.43 µg/mL (Mean ± SD), while Covance’s original result was 17.3 ± 4.23 µg/mL (Mean ± SD). For DBS SPM, the sponsor’s lab result was 1240 ± 161 µg/mL (Mean ±SD), while Covance’s original result was 2314 ± 237µg/mL (Mean ± SD).

Based on our investigation, the different values between labs were not caused by sample preparation, extraction method or instrument condition. Instead the shifts in values between labs appeared to result from differences in isoform patterns between the porcine reference standard and endogenous human samples which required further instrument adjustments to align (Figures 8 & 9). For SPM, adjustment of collision energy was sufficient to match isoform patterns between labs due to the relatively small difference in isoform patterns between porcine and humans isoform. The final SPM results became 1096 ± 109 µg/mL (Mean ± SD), which is similar to the sponsor’s results (Figure 7). For CER, because the reference standard’s isoform pattern has significantly different in C18:H2O and C18:0 than endogenous CER from human blood (Figure 10), adjustment of collision energy and decluttering potential to match sponsor standard’s isoform pattern was not enough. The same lot of pre-spotted human DBS cards with low endogenous CER was also required for standard curve preparation. The final CER result was 12.1 ± 3.40 µg/mL (Mean ± SD), which matched results from the sponsor’s lab (Figure 6).

Figure 10. Comparison of CER isomer profile pattern in STD (Porcine brain CER reference standard spiked on a pre-spotted human DBS with a typical human healthy subject (sponsor’s isoform data), 5 major isoforms are different in abundance, which caused total CER different in two labs although both API-4000s had same parameters.

0.05.0

10.015.020.025.030.035.0

DBS CER_Human Sample Isoform Pattern

% Is

omer

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0 DBS CER_STD (Porcine) Isoform pattern

% Is

omer

smaller

Before Modification (CER)

Figure 2. Different reference ranges from the same 100 donors for CER in human dried blood spots before CER method parameter modification using identical instrument parameters in both labs.

0

2

4

6

8

10

12

14

16

18

20

# of

lots

DBS CER Conc., µg/mL

Covance

Sanofi

Figure 5. Chromatogram of a 500 µg/mL calibration standard [LLOQ (containing endogenous sphingomyelin)]: sphingomyelin in human dried blood spot.

SPM SPM-C12:0

Figure 4. Chromatogram of a 4.00 µg/mL calibration standard [LLOQ (containing endogenous ceramide)]: ceramide in human dried blood spot.

CER CER-C-19:0

Figure 6. Overlapped ceramide reference ranges from the same 93 donor DBS samples after CER instrument parameters’ modification to match sponsor’s isoform pattern in porcine CER standards.

After Modification (CER)

0

2

4

6

8

10

12

14

16

18

20

# of

lots

DBS CER Conc., µg/mL

Covance

Sanofi

0

5

10

15

20

25

30

# of

lots

DBS SPM Conc., µg/mL

CovanceSanofi

After Modification (SPM)

Figure 7. Overlapped sphingomyelin reference ranges from the same 100 donor DBS samples, after SPM Instrument parameters’ modification to match sponsor’s isoform patterns in porcine SPM standards.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

Before Adjust CE&DP (Porcine CER STD4, 32 µg/mL)

Covance

Sanofi

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

After Adjust CE&DP(Porcine CER STD4, 32 µg/mL)

Covance

Sanofi

Figure 8. Comparison of % isomer profile patterns before and after adjusting CE and DP: CER in human dried blood spot.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

C14:

0C1

5:10

C16:

1C1

6:0

C17:

0C1

8:1

C18:

0C2

0:1

C20:

0C2

1:0

C22:

1C2

2:0

C23:

1C2

3:0

C24:

2C2

4:1

C24:

0

Before Adjust CE&DP(Porcine SPM STD4, 2000 µg/mL)

CovanceSanofi

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

C14:

0

C15:

10

C16:

1

C16:

0

C17:

0

C18:

1

C18:

0

C20:

1

C20:

0

C21:

0

C22:

1

C22:

0

C23:

1

C23:

0

C24:

2

C24:

1

C24:

0

After adjust CE&DP(Porcine SPM STD4, 2000 µg/mL)

CovanceSanofi

Figure 9. Comparison of % isomer profile pattern before and after adjusting CE and DP: SPM in human dried blood spot.