Author Contributions
P.K.: conceptualization, methodology, validation, formal analysis, investigation, resources, data curation, writing—original draft, writing—review and editing, visualization, supervision, project administration, funding acquisition; P.G.: conceptualization, methodology, validation, formal analysis, investigation, resources, data curation, writing—review and editing, supervision, project administration; M.C. (Maja Cieślewicz): conceptualization, formal analysis, investigation, resources, data curation, writing—original draft, writing—review and editing, visualization; J.M.: conceptualization, formal analysis, investigation, resources, data curation, writing—original draft, writing—review and editing, visualization; A.M.: conceptualization, formal analysis, investigation, resources, data curation, writing—review and editing; M.C. (Michalina Czupińska): conceptualization, formal analysis, investigation, resources, data curation, writing—original draft, writing—review and editing, visualization; G.M.: conceptualization, validation, formal analysis, investigation, writing—review and editing, visualization, supervision, project administration; M.R.: conceptualization, validation, formal analysis, investigation, writing—review and editing, visualization, supervision, project administration, funding acquisition. All authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
P.K.: honoraria from Novartis, Ipsen, clinical trial remuneration from Neurocrine Biosciences, Crinetics Pharmaceuticas, Ascendis Pharma; P.G.: honoraria from Novartis, Ipsen; M.C.: no conflict of interest; J.M.: honoraria from Novartis; A.M.: no conflict of interest; M.C.: no conflict of interest; G.M.: no conflict of interest; M.R.: honoraria from Novartis, Pfizer, Ipsen, Berlin-Chemie Menarini, Genzyme, Merck, IBSA. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
Figure 1.
The changes in β-hCG concentration over the 48-month study period depending on the treatment response according to RECIST 1.1. Mean β-hCG concentrations at a given time point determined by a sample number are presented as squares in patients with stable disease (SD) and circles in patients with progressive disease (PD). Vertical lines display 95% confidence intervals in each of the time points.
Figure 1.
The changes in β-hCG concentration over the 48-month study period depending on the treatment response according to RECIST 1.1. Mean β-hCG concentrations at a given time point determined by a sample number are presented as squares in patients with stable disease (SD) and circles in patients with progressive disease (PD). Vertical lines display 95% confidence intervals in each of the time points.
Figure 2.
The changes in β-hCG concentration over the study period depending on tumor grading based on 2022 WHO criteria. Mean β-hCG concentrations at a given time point determined by a sample number are presented as squares in patients with G1 tumors and circles in patients with G2 tumors. Vertical lines display 95% confidence intervals in each of the time points.
Figure 2.
The changes in β-hCG concentration over the study period depending on tumor grading based on 2022 WHO criteria. Mean β-hCG concentrations at a given time point determined by a sample number are presented as squares in patients with G1 tumors and circles in patients with G2 tumors. Vertical lines display 95% confidence intervals in each of the time points.
Figure 3.
The changes in β-hCG concentration over the 48-month study period in subgroups depending on liver tumor burden (LTB). Mean β-hCG concentrations at a given time point determined by a sample number are presented as squares in patients with LTB up to 10% and circles in patients with LTB between 10 and 25%. Vertical lines display 95% confidence intervals in each of the time points.
Figure 3.
The changes in β-hCG concentration over the 48-month study period in subgroups depending on liver tumor burden (LTB). Mean β-hCG concentrations at a given time point determined by a sample number are presented as squares in patients with LTB up to 10% and circles in patients with LTB between 10 and 25%. Vertical lines display 95% confidence intervals in each of the time points.
Figure 4.
Delta β-hCG concentration in patients with stable disease (SD) and progressive disease (PD) according to RECIST 1.1. Delta β-hCG represents the individual variation between final and initial β-hCG concentrations (measured in mIU/mL). Box plots depict the interquartile range (Q1 to Q3), with the median highlighted by the internal line. The whiskers extend to the minimum and maximum values of the dataset.
Figure 4.
Delta β-hCG concentration in patients with stable disease (SD) and progressive disease (PD) according to RECIST 1.1. Delta β-hCG represents the individual variation between final and initial β-hCG concentrations (measured in mIU/mL). Box plots depict the interquartile range (Q1 to Q3), with the median highlighted by the internal line. The whiskers extend to the minimum and maximum values of the dataset.
Figure 5.
Delta β-hCG concentration in patients with G1 and G2 tumors defined by 2022 WHO grading. Delta β-hCG represents the individual variation between final and initial β-hCG concentrations (measured in mIU/mL). Box plots depict the interquartile range (Q1 to Q3), with the median highlighted by the internal line. The whiskers extend to the minimum and maximum values of the dataset.
Figure 5.
Delta β-hCG concentration in patients with G1 and G2 tumors defined by 2022 WHO grading. Delta β-hCG represents the individual variation between final and initial β-hCG concentrations (measured in mIU/mL). Box plots depict the interquartile range (Q1 to Q3), with the median highlighted by the internal line. The whiskers extend to the minimum and maximum values of the dataset.
Figure 6.
Delta β-hCG concentration in patients with up to 10% and 10–25% liver tumor burden (LTB). Delta β-hCG represents the individual variation between final and initial β-hCG concentrations (measured in mIU/mL). Box plots depict the interquartile range (Q1 to Q3), with the median highlighted by the internal line. The whiskers extend to the minimum and maximum values of the dataset.
Figure 6.
Delta β-hCG concentration in patients with up to 10% and 10–25% liver tumor burden (LTB). Delta β-hCG represents the individual variation between final and initial β-hCG concentrations (measured in mIU/mL). Box plots depict the interquartile range (Q1 to Q3), with the median highlighted by the internal line. The whiskers extend to the minimum and maximum values of the dataset.
Figure 7.
ROC curve analysis illustrating the prognostic accuracy of initial β-hCG concentration on the occurrence of progressive disease. The dotted line represents the ROC curve if the classification is randomly estimated. The red dot indicates the optimal operating point for sensitivity and specificity.
Figure 7.
ROC curve analysis illustrating the prognostic accuracy of initial β-hCG concentration on the occurrence of progressive disease. The dotted line represents the ROC curve if the classification is randomly estimated. The red dot indicates the optimal operating point for sensitivity and specificity.
Figure 8.
Gantt diagrams illustrating study group characteristics and the overlap between variables in the study subjects.
Figure 8.
Gantt diagrams illustrating study group characteristics and the overlap between variables in the study subjects.
Table 1.
Baseline characteristics of the study group.
Table 1.
Baseline characteristics of the study group.
Variable | All Patients (n = 40) |
---|
Demographics | |
Age, years | 63 (8) * |
Sex | |
Male | 12 (30%) |
Female | 28 (70%) |
Clinical characteristics | |
Primary tumor location | |
Pancreas | 40 (100%) |
Liver Tumor Burden | |
Up to 10% | 22 (55%) |
10–25% | 18 (45%) |
Grading by WHO | |
G1 | 18 (45%) |
G2 | 22 (55%) |
Treatment response by RECIST 1.1. | |
Stable Disease (SD) | 19 (47.5%) |
Progressive Disease (PD) | 21 (52.5%) |
Ki 67 | |
1% | 2 (5%) |
2% | 16 (40%) |
4% | 2 (5%) |
5% | 8 (20%) |
10% | 12 (30%) |
Table 2.
Relationship between the serum concentration of β-hCG and treatment response according to RECIST 1.1 across 16 evenly distanced measurements over 48 months.
Table 2.
Relationship between the serum concentration of β-hCG and treatment response according to RECIST 1.1 across 16 evenly distanced measurements over 48 months.
Measuring Time Point | Median β-hCG—SD [mIU/mL] | Median β-hCG—PD [mIU/mL] | Sum of Ranks SD | Sum of Ranks PD | U | Z Corrected | p |
---|
1. | 67.80 | 356.40 | 223 | 597 | 33 | −4.4964 | <0.001 |
2. | 76.50 | 376.30 | 220 | 600 | 30 | −4.5783 | <0.001 |
3. | 86.30 | 389.50 | 219 | 601 | 29 | −4.6047 | <0.001 |
4. | 93.40 | 402.60 | 220 | 600 | 30 | −4.5776 | <0.001 |
5. | 102.30 | 420.40 | 219 | 601 | 29 | −4.6047 | <0.001 |
6. | 113.40 | 438.40 | 218 | 602 | 28 | −4.6314 | <0.001 |
7. | 123.40 | 451.20 | 219 | 601 | 29 | −4.6045 | <0.001 |
8. | 134.20 | 462.40 | 217 | 603 | 27 | −4.6587 | <0.001 |
9. | 143.50 | 476.50 | 218 | 602 | 28 | −4.6314 | <0.001 |
10. | 153.40 | 487.60 | 219 | 601 | 29 | −4.6043 | <0.001 |
11. | 159.40 | 496.70 | 219 | 601 | 29 | −4.6043 | <0.001 |
12. | 167.80 | 515.80 | 220 | 600 | 30 | −4.5772 | <0.001 |
13. | 178.50 | 527.80 | 222 | 598 | 32 | −4.5230 | <0.001 |
14. | 189.70 | 546.70 | 223 | 597 | 33 | −4.4964 | <0.001 |
15. | 199.60 | 576.30 | 223 | 597 | 33 | −4.4962 | <0.001 |
16. | 212.80 | 598.70 | 223 | 597 | 33 | −4.4959 | <0.001 |
Table 3.
Relationship between the serum concentration of β-hCG and grading according to RECIST 1.1 across 16 evenly distanced measurements over 48 months.
Table 3.
Relationship between the serum concentration of β-hCG and grading according to RECIST 1.1 across 16 evenly distanced measurements over 48 months.
Measuring Time Point | Median β-hCG—G1 [mIU/mL] | Median β-hCG—G2 [mIU/mL] | Sum of Ranks G1 | Sum of Ranks G2 | U | Z Corrected | p |
---|
1. | 67.70 | 354.35 | 211 | 609 | 40 | −4.2822 | <0.001 |
2. | 76.50 | 368.75 | 211 | 609 | 40 | −4.2829 | <0.001 |
3. | 85.40 | 375.70 | 210 | 610 | 39 | −4.3094 | <0.001 |
4. | 95.60 | 390.60 | 213.5 | 606.5 | 42 | −4.2143 | <0.001 |
5. | 102.40 | 409.55 | 214 | 606 | 43 | −4.2007 | <0.001 |
6. | 114.40 | 427.60 | 213 | 607 | 42 | −4.2275 | <0.001 |
7. | 123.45 | 442.00 | 216 | 604 | 45 | −4.1461 | <0.001 |
8. | 135.45 | 457.80 | 215 | 605 | 44 | −4.1733 | <0.001 |
9. | 143.60 | 473.15 | 217 | 603 | 46 | −4.1187 | <0.001 |
10. | 154.90 | 480.40 | 216 | 604 | 45 | −4.1459 | <0.001 |
11. | 162.55 | 488.95 | 216 | 604 | 45 | −4.1459 | <0.001 |
12. | 171.05 | 507.25 | 215 | 605 | 44 | −4.1731 | <0.001 |
13. | 182.95 | 522.15 | 216 | 604 | 45 | −4.1459 | <0.001 |
14. | 192.55 | 545.20 | 217 | 603 | 46 | −4.1191 | <0.001 |
15. | 200.95 | 570.35 | 218 | 602 | 47 | −4.0917 | <0.001 |
16. | 213.10 | 593.15 | 216 | 604 | 45 | −4.1459 | <0.001 |
Table 4.
Relationship between the serum concentration of β-hCG and liver tumor burden according to RECIST 1.1 across 16 evenly distanced measurements over 48 months.
Table 4.
Relationship between the serum concentration of β-hCG and liver tumor burden according to RECIST 1.1 across 16 evenly distanced measurements over 48 months.
Measuring Time Point | Median β-hCG—up to 10% [mIU/mL] | Median β-hCG—10–25% [mIU/mL] | Sum of Ranks up to 10% | Sum of Ranks 10–25% | U | Z Corrected | p |
---|
1. | 72.15 | 382.20 | 275 | 545 | 22 | −4.7717 | <0.001 |
2. | 79.40 | 397.15 | 272 | 548 | 19 | −4.8539 | <0.001 |
3. | 87.40 | 413.05 | 271 | 549 | 18 | −4.8804 | <0.001 |
4. | 95.60 | 429.95 | 271 | 549 | 18 | −4.8804 | <0.001 |
5. | 102.40 | 442.80 | 271 | 549 | 18 | −4.8804 | <0.001 |
6. | 114.40 | 457.90 | 270 | 550 | 17 | −4.9071 | <0.001 |
7. | 123.45 | 472.40 | 271 | 549 | 18 | −4.8802 | <0.001 |
8. | 135.45 | 490.95 | 268 | 552 | 15 | −4.9617 | <0.001 |
9. | 143.60 | 510.25 | 269 | 551 | 16 | −4.9343 | <0.001 |
10. | 154.90 | 523.00 | 269 | 551 | 16 | −4.9343 | <0.001 |
11. | 162.55 | 535.05 | 269 | 551 | 16 | −4.9343 | <0.001 |
12. | 171.05 | 550.30 | 269 | 551 | 16 | −4.9343 | <0.001 |
13. | 182.95 | 562.30 | 271 | 549 | 18 | −4.8800 | <0.001 |
14. | 192.55 | 578.30 | 271 | 549 | 18 | −4.8804 | <0.001 |
15. | 200.95 | 594.70 | 272 | 548 | 19 | −4.8530 | <0.001 |
16. | 213.10 | 612.15 | 272 | 548 | 19 | −4.8530 | <0.001 |
Table 5.
Relationship between delta β-hCG and study variables: treatment response, grading, liver tumor burden, and patient age. Mann–Whitney U tests were employed for treatment response, grading, and liver tumor burden, while the Spearman rank correlation test was used for patient age.
Table 5.
Relationship between delta β-hCG and study variables: treatment response, grading, liver tumor burden, and patient age. Mann–Whitney U tests were employed for treatment response, grading, and liver tumor burden, while the Spearman rank correlation test was used for patient age.
Variable | Median β-hCG [mIU/mL] | Sum of Ranks | U | Z Corrected | p |
---|
Treatment Response |
| SD | 147.50 | 263 | 73 | −3.4127 | <0.001 |
PD | 260.70 | 557 |
Grading |
| G1 | 143.23 | 261 | 90 | −2.9227 | 0.003 |
G2 | 256.25 | 559 |
Liver Tumor Burden |
| up to 10% | 146.65 | 307 | 54 | −3.9014 | <0.001 |
10–25% | 267.00 | 513 |
Patient Age | R Spearman | T (N-2) | 0.97 |
−0.0069 | −0.0423 |
Table 6.
Logistic regression analysis of initial β-hCG concentration for predicting progressive disease.
Table 6.
Logistic regression analysis of initial β-hCG concentration for predicting progressive disease.
Variable | Beta Coefficient | 95% CI for Beta | p-Value | Odds Ratio | 95% CI for OR |
---|
Intercept | −3.2841 | (−5.2093, −1.359) | 0.0008 | 0.0375 | (0.0055, 0.2569) |
Initial β-hCG | 0.0171 | (0.0079, 0.0264) | 0.0003 | 1.0173 | (1.0079, 1.0268) |