You are translating the reversible AI feedback on aromatase protein levels to be present after withdrawal of the AI, thus leading to more enzyme leading to more E2. That idea is only accepted in the bro lore - the scientific evidence has NOT shown it occurs.
[Note: For the irreversible AI's, there appears to be NO positive feedback on enzyme levels.]
Below is a table from a study done to determine the effect AI's have on bone mineral density, however, estrogen levels were deemed necessary because of the influence estrogen has on bone density. Changes in the concentrations of estrogens were assessed after daily administration of exemestane (25 mg), letrozole (2.5 mg), anastrozole (1 mg), or placebo for 24 weeks It shows NO "estrogen rebound" after AI withdrawal.
I appologize for the confusing nature of the table. I cut & pasted from a pdf. I only highlighted the results for exemestane in red (for E1 and E2) to make it easier to read but you'll get the idea.
Effects of steroidal and nonsteroidal arom... [Breast Cancer Res. 2007] - PubMed - NCBI
Table 4
Median values and median percentage change from baseline for plasma concentrations of estrogens (mITT population)
Estrogen
Exemestane (n = 20) Letrozole (n = 20) Anastrozole (n = 18) Placebo (n = 20)
E1
Baseline, median (min., max.)
31 (11, 72) 28 (13, 51) 29 (14, 61) 26 (9.9, 61)
Week 12, median (min., max.)
2.0 (1.8, 10) 1.8 (1.8, 20) 1.8 (1.8, 84) 31 (8.4, 58)
% baseline, median (95% CI*)
7.8 (6.9, 9.3) 6.6 (5.5, 8.1) 7.6 (6.4, 11) 101 (89, 110)
Week 24, median (min., max.)
1.8 (1.8, 63) 1.8 (1.8, 33) 1.8 (1.8, 44) 26 (7.2, 56)
% baseline, median (95% CI)
6.7 (5.3, 8.6) 6.6 (5.1, 8.1) 7.9 (7.1, 11) 94 (76, 108)
Week 36, median (min., max.)
23 (10, 62) 19 (10, 44) 25 (8.4, 158) 22 (13, 46)
% baseline, median (95% CI)
72 (63, 86) 76 (61, 93) 70 (66, 121) 89 (72, 95)
E2
Baseline, median (min., max.)
4.4 (1.1, 13) 4.2 (1.8, 8.6) 3.8 (1.9, 32) 3.9 (1.9, 8.8)
Week 12, median (min., max.)
0.70 (0.70, 2.5) 0.70 (0.70, 2.7) 0.70 (0.70, 104) 3.8 (1.6, 15)
% baseline, median (95% CI)
18 (14, 22) 19 (14, 24) 20 (18, 31) 102 (87, 115)
Week 24, median (min., max.)
0.70 (0.70, 29) 0.70 (0.70, 7.3) 0.72 (0.70, 84) 4.4 (1.9, 12)
% baseline, median (95% CI)
19 (13, 22) 18 (13, 24) 22 (19, 31) 128 (95, 139)
Week 36, median (min., max.)
4.5 (1.6, 36) 4.2 (1.9, 15) 4.2 (1.4, 136) 4.0 (1.6, 11)
% baseline, median (95% CI)
103 (89, 134) 109 (81, 119) 106 (92, 130) 105 (90, 130)
E1S
Baseline, median (min., max.) 192 (58, 757) 193 (79, 1075) 202 (64, 896) 200 (62, 1156)
Week 12, median (min., max.) 13 (8.7, 246) 8.7 (6.0, 62) 15 (6.6, 3260) 209 (58, 876)
% baseline, median (95% CI) 9.1 (6.1, 11) 4.5 (3.3, 6.0) 11 (6.7, 13) 110 (91, 146)
Week 24, median (min., max.) 16 (6.1, 839) 10 (6.0, 254) 18 (8.8, 1440) 195 (64, 802)
% baseline, median (95% CI) 11 (7.2, 13) 4.7 (3.0, 7.2) 10 (8.6, 20) 105 (69, 123)
Week 36, median (min., max.) 223 (7.4, 2100) 222 (115, 932) 296 (58, 4890) 205 (68, 864)
% baseline, median (95% CI) 114 (90, 152) 121 (105, 135) 131 (109, 147) 125 (91, 143)
*95% CI, bias-corrected and adjusted 95% bootstrap CI for median.
CI, confidence interval; E1, estrone; E1S, estrone sulfate; E2, estradiol; max; min; mITT, modified intent to treat.