A total of 74 patients from eleven healthcare centers was included in this cohort study. Thirty-nine (52.7%) of the participants were male, 35 (47.3%) were female. Half of the patients were born until 2002 (IQR: 1995–2009, range: 1975–2019). Health records were available for the years 1986 to 2020 (median: 2016, IQR: 2013–2016). Median survey duration was 2 years (IQR: 1–5, range: 1–17). The median age patients had at the recorded examinations was 11 years (IQR: 6–15, range: 0–42). The study cohort mostly represented underage patients at a total of 240 points in time; only 36 observations were made for patients at the age of 18 years and older.

Median age at diagnosis of INC (n = 67) was 15 months (IQR: 10–29, range: 0–110). In one case the date of initial diagnosis was unknown, while in six cases only the year was given. Spearman's correlation of birth year and age at diagnosis, as well as a scatterplot, showed no correlation (ρ = −0.17, p = 0.16).

Systemic cystine-depleting therapy was initiated at a median age of 16.5 months (IQR: 11–30, range: 0–111). Data was missing on 12 patients born between 1975 and 2003. In 53 patients, cysteamine therapy started within 1 month after initial diagnosis, in seven cases within 4 months, whilst two patients (both born in 1981) started treatment 3 and 7 years after initial diagnosis.

For 72 patients, the treatment with systemic cysteamine was documented, data was missing for two patients. The use of IRC was recorded in 54 cases, with 17 patients changing to ERC at a median age of 11 years (IQR: 6–14, range: 2–16). Two patients changed back to IRC after 1 and 2 years, one of them due to increased vomiting. The use of ERC alone was reported for two patients, whilst two more patients took IRC twice during the daytime and ERC once at night-time.

The exact cysteamine dosage was available for 67 patients, with the median of the 235 documented values being 1.26 g/m² per day (IQR: 1.03–1.48, range: 0.22–1.99). 84 values of 35 patients were >1.35 g/m² per day with a median of 1.57 (IQR: 1.46–1.67, range: 1.36–1.99) at a median age of 12 years (IQR: 7–16, range: 1–35). Two patients received a dose slightly above the recommended maximum of 1.95 g/m² per day, with 1.99 each at the age of 12 and 14 years. There was no documentation on side effects available.

The median dose of 171 values of the patients on IRC was 1.31 (IQR: 1.16–1.51, range: 0.22–1.99). The median dose of 71 values of the patients on ERC was 1.03 (IQR: 0.95–1.21, range: 0.27–1.86).

The median time-averaged LCL values based on 65 patients amounted to 0.57 (IQR: 0.33–0.98, range: 0.07–3.13). Here, only measurements at least 1 year after initiation of therapy were considered. 69.2% of all 198 LCL measurements of 67 patients were within the desired target range ≤ 1 nmol cystine/mg protein. An overview of all documented LCLs is given in Table 1.

There was no association between cysteamine dosage and age (ρ = 0.06, p = 0.38) or LCLs (ρ = (−0.06), p = 0.44).

For 52 patients, 195 values of eGFR could be calculated before KRT and assigned to CKD stages (Figure 1).

Thirty-one patients (41.9%), 17 female and 14 male, underwent KRT (either dialysis or pre-emptive kidney transplantation), data was missing in four cases. The median age at initiation of KRT was 10 years (IQR: 9–15, range: 7–25).

Thirty patients underwent kidney transplantation at least once. 18 of those patients received their first transplant after a period of dialysis therapy, 4 kidneys were transplanted pre-emptively whereas in 8 cases data on preceding dialysis was missing. One patient on chronic dialysis therapy had not yet been listed for transplantation.

Kaplan–Meier analysis was used to evaluate time from birth to KRT (Figure 2). Data was stratified for age at initial diagnosis in months (Figure 3). Thirty-four patients (45.9%) were each assigned to one of two groups (initial diagnosis before 18 months of life vs. at 18 months or older), for which a log-rank test showed a significant difference in renal survival.

When stratified for year of birth (1975 to 2001 vs. 2002 to 2019) or sex, with crossing hazards and therefore failing log-rank test, no difference in renal survival was detected.

Percentiles for height were calculated for 68 patients with a total of 242 documented values available. The overall median lay at the 7^th percentile (IQR: 1–25, range: 1–99). Detailed boxplots are shown in Figure 4, displaying all age groups from 1 year to adulthood and differentiating two subgroups (before and after initiation of KRT).

Ninety-eight values (40.5%) from 43 patients were within the defined pathological range of ≤ 3^rd percentile at a median age of 8 years (IQR: 5–13, range: 1–18 and older). Pathological heights were mostly documented for younger children and adults born between 1975 and 2001. Only 3 patients reached the 50^th percentile or higher at least once.

The subgroup not yet in need of KRT (41 patients, 164 measurements) had a median percentile of 8 (IQR: 1–26, range: 1–99), the median of 27 patients after initiation of KRT lay at the 6^th percentile (IQR: 1–23, range: 1–49, including 66 values).

Percentiles for weight were calculated for 71 patients with a total of 248 documented values available. The overall median lay at percentile 15.50 (IQR: 5.00–34.25, range: 1.00–97.00). Fifty-three values (21.4%) from 28 patients were ≤ 3^rd percentile at a median age of 8 years (IQR: 4–13, range: 1–18 and older). Only 16 patients reached the 50^th percentile or higher at least once.

Data on tube feeding was available for 25 patients (33.8%). Median age when beginning tube feeding was 1 year (IQR: 1–3, range: 1–10). For 11 patients, the median duration was 6 years (IQR: 2–13, range: 0.04–23), whilst for 13 patients the median duration was at least 4 years (IQR: 2–5, range:1–11), and was still ongoing through the period of observation.

Thirty-nine (52.7%) of the patients received recombinant growth hormone therapy at some point. In 10 cases, the exact duration was known (mean: 7.70 years, SD: 3.53, range: 1.00–13.00); in those cases the therapy started at a mean age of 8.00 years (SD: 3.62, range: 1–12). In 26 cases, the documented duration was a median of at least 4.00 years (IQR: 1.25–7.50, range: 1.00–15.00), but the exact starting and/or ending points remained unclear due to insufficient documentation.

Observed extrarenal manifestations are listed in Table 2. The available documents did not distinguish the absence of those symptoms from missing information.

The most reported skeletal deformities were genu valgum (33 cases), pes planovalgus (10 cases), and scoliosis (7 cases). Where further details on ophthalmological symptoms were given, photophobia and corneal discomfort were reported to be the main concern. Regarding neurological symptoms, 7 patients with seizures were reported, and there was one case each of movement disorder, dyslalia, motor retardation and stroke with speech impediments. There were no entries on diabetes, but one patient was reported as suffering from exocrine pancreatic insufficiency. Detailed information on the context of these observations was not available.

Gastrointestinal symptoms are presented in Table 3.

Despite the tremendous progress treatment has made within the last decades, INC continues to be a severe chronic condition (16, 28, 29).

Frequently, the first sign of INC noticed by parents and clinicians is failure to thrive (38). We observed that children often needed the support of tube feeding at a young age, continuing for several years. Growth has been an important outcome parameter in previous studies as it reflects the multi-factorial influences of different aspects of the disease as well as therapeutic success (29, 38). We were able to show that even though medication in Germany and Austria started at an early age, patients remained very short in stature. Our results are in concordance with another study focusing on patient's body growth before initiation of KRT (23). However, our data was not sufficient to explain the great interindividual variety, with several patients having fewer difficulties than others. These results might indicate an improvement around length of therapy and medical developments in recent years, as pathological percentiles were mostly concentrated at younger ages and adults born before 2001. While thorough analyses of its impact on growth in INC patients are still necessary, we recommend an early start of GH treatment in infancy for children with INC based on the suggestion directing the general population of infants with CKD (39).

Our study also demonstrated that there was subsequent kidney failure with advancing age. Overall median renal survival in German and Austrian patients corresponded to the large European-Turkish cohort analyzed recently by Emma et al. (29). We also showed that in German-Austrian patients, time of initial diagnosis accounted for a great difference in median renal survival with our cut-off point at the age of 1.5 years (arbitrarily chosen at a younger age than in previous studies (15, 30), oriented to age at initial diagnosis in our cohort).

Information on extrarenal symptoms and their onset could not be determined as thoroughly as planned, but our results still provided an insight into the greatest challenges regarding this disease. Corresponding to international studies (30, 40), hypothyroidism, anemia, ophthalmological, skeletal and gastrointestinal complications were very common manifestations of INC. Considering the incomplete health records on this topic, it is likely that more patients were affected than was actually documented in the present study.

We aimed to characterize the clinical course for German-Austrian patients with INC, with a main focus on assessing the quality of their healthcare using data on initial diagnosis and treatment as well as undertaking an analysis of the dosage and monitoring of specific therapies.

According to the findings of Bertholet-Thomas et al. (15), our cohort was representative for management of INC in developed nations. We showed that systemic application of cysteamine was offered to patients very soon after initial diagnosis, addressing the importance of cystine-depleting therapy. Prescribed doses of systemic cysteamine were at an optimal level according to international standards (12), regardless of whether IRC or ERC was used. No relevant overdoses were documented.

As the measurement of LCLs provides an indirect insight into patients' adherence and overall success of cysteamine treatment, analyzing this aspect thoroughly was of importance in this study. The aim of the therapy is to remain ≤ 1 nmol cystine/mg protein, with an optimal outcome being a level ≤ 0.5 (31). We found our cohort to have overall satisfying LCLs when calculating the time-averaged mean for each patient as well as when considering all measurements individually. The therapy in our cohort seemed to be particularly efficient after a few years of treatment which is indicated by the observation that those values above 1 nmol cystine/mg protein are found at younger ages than the sufficient LCLs. The comparison of our findings with international data (15, 29) is remarkable as it shows that, though there may have been room for improvement on an individual level, this cohort had exceptional overall therapy control based upon LCL. However, the assay to detect LCL, the current gold standard to monitor cysteamine therapy, is technically demanding and not always available. Biomarkers of macrophage activation are promising monitoring candidates, reflecting the macrophage-mediated inflammation in cystinosis. Among them, chitotriosidase enzyme activity showed to be a good predictor of LCL and was significantly correlated to extrarenal complications (41). Concentration of ceroid in macrophages (a product of lysosomal oxidation of LDL that plays an important role in the development of atherosclerosis) might become a future marker to monitor the antioxidative properties of cysteamine (42). Monitoring of therapy adherence should not be restricted to the analysis of direct cysteamine effects, but should also include biomarker tracking of disease complications, such as those of bone status (43), thyroid function and glucose metabolism. The most relevant clinical biomarker correlated to cysteamine effect doubtlessly remains preservation of glomerular filtration rate (44) and might be the best motivation for therapy adherence.

Age at initial diagnosis in our patient cohort was comparable to recent experiences in other countries (15, 29). Interestingly, in our cohort spanning four decades, a more recent date of birth was not associated with earlier diagnosis or relevantly improved renal survival, leading to the hypothesis that the awareness for diagnosis of cystinosis in the first year of live has not increased and an optimal therapeutic approach might have already been reached. Thus, new initiatives, such as new-born screening to enable earlier diagnosis, need to be implemented to achieve even more favorable outcomes. It has to be noted, though, that it is currently unclear whether cysteamine treatment bears more advantages than risks when initiated before children experience symptoms of INC.

Limitations of our study were closely linked to its retrospective nature, and mainly consist of incomplete records on some patients as well as heterogeneous distribution of duration of individual follow-up.

There was no information available on the underlying genetic defect.

Particularly in cases where there was no reporting on extrarenal manifestations, it remained unclear whether the information was missing or whether the symptoms did not occur, so more patients might have been affected. Our numbers on those manifestations are therefore to be understood as minimal counts. Also, patient ages at onset remained uncertain. Concerning all extrarenal manifestations and especially our data on growth, important information could not be evaluated: Data on supporting medication, nutritional intervention, acidosis, bone disease and other disease-related as well as therapeutic factors was not available.

For Kaplan–Meier-analyses, data on eight patients was included although it was not known whether these persons had received dialysis before their documented transplantation date. In our opinion, however, it was deemed important to include data on these patients nevertheless to avoid loss of valuable information.

Additionally, despite being a large cohort for national studies on this orphan disease, statistical analysis was still flawed because of its relatively small sample size.

A certain level of bias might have also played a role as patients' dates of birth varied greatly and health care has changed fundamentally for patients born more recently. Changes in awareness and profound knowledge of INC, pharmaceutical developments such as the availability of cysteamine and growth hormones, advances in KRT but also individual patients' characteristics such as adherence and psychosocial aspects in general may have had an impact on our observations that we cannot differentiate nor quantify.

Furthermore, there may have been sampling bias as there are rarely coordinated programs for transitioning care to specialized centers in adulthood and it was therefore very difficult to include adult patients in this study. Adult nephrologists were very widely contacted but did not report back. The majority of all pediatric nephrology centers in Germany that treat patients with cystinosis have included their data in the registry of the German Society for Pediatric Nephrology. In our view, establishing national and international registries will be the key to understanding and managing this orphan disease in the future.