1. Changes in CSF cholinergic biomarkers in response to cell therapy with NGF in patients with Alzheimer's disease 
Azadeh Karami, Helga Eyjolfsdottir, Swetha Vijayaraghavan, Göran Lind, Per Almqvist, Ahmadul Kadir, Bengt Linderoth, Niels Andreasen, Kaj Blennow, Anders Wall, Eric Westman, Daniel Ferreira, Maria Kristoffersen Wiberg, Lars-Olof Wahlund, Åke Seiger, Agneta Nordberg, Lars Wahlberg, Taher Darreh-Shori, Maria Eriksdotter 
Alzheimer's & Dementia: The Journal of the Alzheimer's Association 
Volume 11, Issue 11, Pages (November 2015)
DOI: /j.jalz
Copyright © 2015 The Alzheimer's Association Terms and Conditions

2. Fig. 1
Stratification of the patients based on changes in cognitive performances after treatment. (A) The overall changes in MMSE test score during 12 months treatment. (B) MMSE changes in the Single and Double implants subgroups. (C) Subdivision of the patients based on annual changes in the MMSE test scores. Three patients showed ≤2 points annual decline in the MMSE test and were assigned as the Responders, and three patients who showed greater than two points of annual decline in MMSE were the Nonresponders. (D) One-year changes in MMSE test scores in a reference group of patients on stable dosage of the reversible cholinesterase inhibitors (ChEIs). (E) The overall changes in ADAS-Cog test score during 12 months treatment. (F) Changes in ADAS-Cog test in the Single and Double implants subgroups. (G) Changes in the ADAS-Cog is consistent with the Responders and Nonresponders group assignment based on the MMSE. The data is shown for the overall group and for Responders and Nonresponders to ChEI-therapy as were predefined in the section of “Statistical analyses”. *P < .05 and **P < .01 indicate significant difference compared with the baseline levels, ¤P < .05 and ¤¤¤P < .001 indicate significant differences between the groups. MMSE, Mini-Mental State Examination; ADAS-Cog, Alzheimer's Disease Assessment Scale-cognitive subscale.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

3. Fig. 2
Overall CSF biomarkers and the difference based on NGF doses and the performance in cognition. (A) Overall changes of ChAT activity during NGF treatment. (B) No NGF dosage-dependent changes in CSF ChAT activity was observed when the patients were grouped based on Single and Double EC-NGF implants. (C) In contrast, when the changes in CSF ChAT activity were analyzed based on annual cognitive changes, the Responders showed 25% increased ChAT activity compared with Nonresponders or baseline. (D) This finding is in contrast with relative changes in the reference group (n = 17), who were stably on reversible cholinesterase inhibitors for 12 months. Note, also that the nonresponders in this reference group show ∼23% decline in ChAT activity (P < .14) compared with baseline or ∼34% (P < .17) compared with the responders of ChEIs' therapy. (E) CSF AChE activity showed 40% general increases after 12 months of treatment. (F) This increase was most pronounced among AD subjects with Double implants (∼65%), whereas those with Single implant had a nonsignificant 13% increase. (G) Furthermore, no difference in the level of changes in AChE activity was apparent between the Responders and Nonresponders. (H) The changes in the CSF AChE activity among the reference group show in contrast significant decreases in overall group (∼12%, P < .03), and particularly in nonresponders (∼19%, P < .02) but less so among the responders (∼9%, P < .18). *P .05 indicates significant difference compared with the baseline levels, ∗P .05 indicates significant differences between the groups. AChE, acetylcholinesterase; CSF, cerebrospinal fluid; ChAT, choline acetyltransferase; EC-NGF, encapsulated NGF releasing cells.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

4. Fig. 3
Correlation between CSF ChAT activity and clinical and para-clinical in vivo assessments. Correlation between changes in CSF ChAT activity and (A) absolute MMSE test scores, and the 12 months' changes from baseline in (B) the MMSE, and (C) ADAS-Cog tests. (D) Positive correlation between changes in CSF ChAT activity and the in vivo glucose utilization (μmol/100 g/min) in the frontal cortex assessed by FDG-PET at 12 months EC-NGF implant follow-up. Positive correlation between CSF ChAT activity and (E) number of the in vivo nicotinic binding sites (1-k∗2) in frontal association cortex, assessed by 11C-nicotin-PET, and (F) the cortical thickness in right cingulate assessed by MRI after 12 months. Noteworthy, the Responders show relatively the best clinical outcomes in relation to the CSF ChAT activity. The numbers identify individual patient in the graphs. ADAS-Cog, Alzheimer's Disease Assessment Scale-cognitive subscale; ChAT, choline acetyltransferase; PET, positron emission tomography; MMSE, Mini-Mental State Examination; CSF, cerebrospinal fluid; MRI, magnetic resonance imaging; EC-NGF, encapsulated NGF releasing cells' implants; FDG-PET, 18F-fluoro-2-deoxy-D-glucose (F-FDG)-PET.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

5. Fig. 4
Correlations between CSF AChE activity and different clinical and paraclinical measures. Positive correlation between AChE activity and (A) average in vivo glucose utilization (μmol/100 g/min; FDG-PET), and (B) number of the in vivo nicotinic binding sites (1-k∗2), assessed by 11C-nicotine-PET, in the whole brain at 12 months EC-NGF implant follow-up. (C) Positive correlation between CSF AChE activity and cortical thickness of right cingulate assessed by MRI after 12 months. The numbers identify individual patient in the graphs. ADAS-Cog, Alzheimer's disease Assessment Scale-cognitive subscale; AChE, acetylcholinesterase; CSF, cerebrospinal fluid; PET, positron emission tomography; MMSE, Mini-Mental State Examination; MRI, magnetic resonance imaging; EC-NGF, encapsulated NGF releasing cells' implants; FDG-PET, 18F-fluoro-2-deoxy-D-glucose PET.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

6. Fig. 5
Patterns of correlation between ChAT activity, cognition and the CSF biomarkers of AD pathology at 12 months of EC-NGF treatment. Correlations between absolute levels or changes of the CSF AD biomarker, Aβ42, and (A and B) ChAT activity in CSF, and (C) the changes in MMSE test result after 1 year of EC-NGF implant treatment. The corresponding pattern of correlations between absolute levels and the changes in total tau (another AD biomarker) and (D and E) the CSF ChAT activity, and (F) the changes in MMSE test results after 12 months of EC-NGF therapy. The numbers identify individual patient in the graphs. Aβ, amyloid-β; CSF, cerebrospinal fluid; ChAT, choline acetyltransferase; MMSE, Mini-Mental State Examination; EC-NGF, encapsulated NGF releasing cells' implants.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

7. Fig. 6
Patterns of changes in MRI measures after 1 year of EC-NGF treatment. (A) The cortical thickness statistically shows a significant cortical atrophy changes in the Nonresponders when compared with the baseline. No statistically significant atrophy was observed in the Responders. (B) The pattern of the volumetric changes, and the CSF expansions are also in agreement with the pattern of cortical atrophy changes between the Responders and the Nonresponders at 12 months. *P < .05 and **P < .005 indicate significant difference compared with the baseline levels. #P < .05 significant differences between the groups. MRI, magnetic resonance imaging; CSF, cerebrospinal fluid; GM, gray matter; WM, white matter.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

8. Fig. 7
Longitudinal follow-ups of the patients' cognitive performances before, during, and after EC-NGF implant study. (A) MMSE test scores in average, 20 months prior (Pre-NGF) to the start of EC-NGF therapy and up to 15 months after the removal of the implants (post-NGF). No difference between MMSE test scores exists between the Responders and Nonresponders at in-average 20 months before and at the baseline of EC-NGF study, indicating that the observed longitudinal cognitive stabilization seen among the Responders is most likely related to response to EC-NGF therapy rather than difference in the stage of the disease that could be caused by a selection bias. The arrow and asterisk above the horizontal line indicates statistically significant changes among the Nonresponders compared with the pre-NGF MMSE assessment. The arrow and asterisk below the horizontal line indicates statistically significant changes among the Nonresponders compared with the NGF-baseline assessment. (B) ADAS-Cog test scores from baseline until up to 7 months after removal of EC-NGF implants. These analyses are consistent with observations on MMSE assessment and suggest that the Responders have a preserved cognition for at least 3 months after removal of the implants even in this clinical test. In contrast, the Nonresponders show cognitive deterioration compared with both baseline and the Responders. *P < .05 compared with baseline, **P < .01 compared with the pre-NGF assessment; ¤P < .05 between the groups. MMSE, Mini-Mental State Examination; CSF, cerebrospinal fluid; ADAS-Cog, Alzheimer's Disease Assessment Scale-cognitive subscale; EC-NGF, encapsulated NGF releasing cells' implants.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

9. Supplementary Figure 1
Pattern of correlation between phosphorylated tau, ChAT, and cognition after EC-NGF therapy. (A–C) The pattern of correlation between absolute levels or the changes in phosphorylated tau (p-tau) and ChAT activity in CSF. (D and E) The corresponding pattern of correlation between percent changes in p-tau in relation to cognition after 1 year, supporting the above notion. Nonetheless, the overall pattern supports the observed pattern in the Fig. 5. The numbers identify individual patient in the graphs. ChAT, choline acetyltransferase; MMSE, Mini-Mental State Examination; EC-NGF, encapsulated NGF releasing cells' implants.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions

10. Supplementary Figure 2
Trajectory of changes in MMSE and ratio of ChAT to AChE activity in responders and nonresponders of EC-NGF and a MMSE-matched subgroup of the reference ChEI group. A subgroup of the ChEI reference group was MMSE-matched with the NGF group (A), and then the MMSE trajectories was compared for 12-month treatment. Similar 12 months analyses were then done with regards to changes in the ratio of ChAT to AChE activities (B). This ratio reflects relative changes in activities of two enzymes, one that synthesizes Ach, i.e. ChAT, and one that breaks down Ach, i.e. AChE. Thereby this ratio may be a cholinergic index of extracellular ACh homeostasis. The ChAT/AChE ratio trajectory in the NGF responder group is different from both the responder ChEI reference group and the nonresponder NGF group, indicating that balanced changes in ChAT/AChE ratio (i.e. a balanced ACh homeostasis) as an important indicator of responsiveness to NGF therapy (B).
Alzheimer's & Dementia: The Journal of the Alzheimer's Association  , DOI: ( /j.jalz )
Copyright © 2015 The Alzheimer's Association Terms and Conditions