General information at baseline
At initial visit, all 41 newly-diagnosed patients had polyneuropathy, endocrinopathy, M-protein band and skin changes. Organomegaly manifested mostly as lymphadenopathy (38/41, 92.7%). Besides, splenomegaly (27/41, 65.9%) and hepatomegaly (24/41, 58.5%) were not uncommon. Endocrinopathy included 71.8% (28/41) of increased ACTH, 41.5% (17/41) of hypothyroidism, 14.6% (6/41) of diabetes mellitus, and 75% (21/28) of male gynecomastia. Thirty-six patients completed 12 cycles of LDex treatment, while 3 patients died due to disease progression and 2 quitted the clinical trial midway. In this study, three patients who did not meet the basic requirement of ocular examination were excluded: one patient with incomplete initial ocular examination data, one with missed pre-treatment ODE area data, and another one with delayed first posttreatment measurement of optic disc at 12 months later. At diagnosis, average age of the enrolled 38 patients with complete data was 50.0 ± 11.5 years (range 21-70). Among them, 26 were male and 12 were female. Mean serum VEGF level was 5655 ± 3290 pg/mL (range 534-14328 ).
Ocular manifestations and ODE grading before and after treatment
At the first ocular examination before treatment, 25 (65.8%) patients were diagnosed with ODE, of which 24 (96%) were bilateral and only 1 (4%) was unilateral. ODE stratification was determined in these 25 patients: 1 eye (2%) was grade 0, 12 eyes (24%) were grade 1, 10 eyes (20%) were grade 2, 11 eyes (22%) were grade 3, 14 eyes (28%) were grade 4 and 2 eyes (4%) were grade 5. Among the 25 patients with ODE, except 8 (32%) patients were unsymptomatic, the other 17 patients (68%) had various symptoms including blurred vision (7/17, 41.2%), decreased vision (4/17, 23.5%) and palpebral edema (3/17, 17.6%). Red eye, photophobia and photopsia also appeared but relatively rare. In contrast, all the other 13 patients (34.2%) without ODE did not show any ocular symptoms at baseline or during follow-up.
All enrolled patients received treatment with LDex after initial examination. Except for 3 patients who either died during treatment or showed very slight ODE and refused to receive follow-up ocular examination, all the other 22 patients had ocular examination follow-up, with a mean duration of 8.6 ± 3.3 months (range 3-13). Twenty-one patients received ocular examination 3 months after treatment. The remission rate was 83.3% (35/42) and the complete remission rate was 14.3% (6/42), with only 7 eyes remained stable ODE grade. At last visit, the remission rate reached 90.9% (40/44) and the complete remission rate reached 54.5% (24/44). ODE grade was 0 in 24 eyes (54.2%), 1 in 14 eyes (33.3%) and 2 in 6 eyes (12.5%) (Figure 2, Figure 3, Figure 4). Ocular manifestations disappeared in almost all patients at last visit except for one patient who had remnant blurred vision even with complete remission of ODE.
ODE area measurement before and after treatment
ODE area was measured in 21 patients with ODE at initial examination. Except one patient died during treatment, all the other 20 patients had their ODE areas measured at follow-up ocular examinations.
The ODE area of right eye and left eye at initial examination was 2.0098 ± 0.8050 mm2 (range 0.8064-3.2659) and 1.9310 ± 0.7747 mm2 (range 0.8124-3.8495) respectively. There was no significant difference between ODE area of right and left eyes (p = 0.595), and there was a relatively good consistency between them, with an intraclass correlation coefficient (ICC) of 0.652 (p < 0.001). Binocular mean ODE area was 1.9704 ± 0.7158 mm2 (range 0.8094-3.5577). There was a significant correlation between ODE area and ODE grading before treatment (right eye, r=0.632, p = 0.002; left eye, r = 0.638, p = 0.002; binoculus, r = 0.620, p = 0.003).
At last visit, ODE area of right and left eye of the 20 alive patients was 1.0501 ± 0.2760 mm2 (range 0.6395-1.7315) and 1.1176 ± 0.3265 mm2 (range 0.6690-1.9767) respectively, without significant difference (p = 0.230). Binocular mean ODE area was 1.0839 ± 0.2768 mm2 (range 0.6646-1.7118), which was remarkably lower than baseline (p < 0.001). The declination in ODE area was noticeable through treatment (Figure 4, Figure 5).
Serum VEGF levels and ODE area
As stated above, mean serum VEGF level of all 38 POEMS patients with or without ODE at initial visit was 5655 ± 3290 pg/mL (range 534-14328). And 3 months post-treatment, serum VEGF level of 37 alive patients declined to 1979 ± 2132 pg/mL (range 181-11179), significantly lower than baseline VEGF (p < 0.001).
At initial visit, baseline serum VEGF level in patients with and without ODE were significantly different, with a mean level of 6564 ± 3257 pg/mL (range 1553-14328) in patients with ODE and 3907 ± 740 pg/mL (range 534-8594) in patients without ODE, respectively (p = 0.016).
Besides, binocular mean ODE area was positively correlated with serum VEGF among patients with ODE at diagnosis (r = 0.460, p = 0.036, Figure 6A). However, the correlation after treatment was not significant (r = 0.015, p = 0.937).
ODE area and peripapillary retinal thickness (pRT)
Since pRT was also supposed to be capable of indicating ODE severity, we did correlation analysis between ODE area and pRT among 20 patients with both two parameters measured at baseline, and the result confirmed the consistency between these two indicators of ODE severity (r = 0.760, p < 0.001, Figure 6B). The measurement of pRT in these patients was described in one of our former published articles.[6]