Study design

In this retrospective study, we searched our database of allo‑HCT recipients from 2014 to 2024. The only inclusion criterion was the occurrence of at least 1 M protein peak verified by immunofixation after the date of allo‑HCT. The exclusion criterion was the presence of gammopathy before allo‑HCT. Next, we extracted clinical data from the hospital records of patients included in the analysis.

Setting and study size

The analysis concerned relevant clinical data of allo‑HCT recipients, undergoing a transplant from January 2014 to December 2024, who met the inclusion criterion. The data were collected from November 2024 to April 2025. The inclusion period was chosen to maximize the chance of identifying individuals with post–allo‑HCT gammopathy. Data availability before the year 2014 was limited.

Participants

Among 241 allo‑HCT recipients, we identified 36 patients with gammopathy who were divided into 2 subgroups: 20 patients with MG and 16 with OG. In another 43 patients, serum protein electrophoresis with subsequent immunofixation were performed at least once after allo‑HCT. However, in these cases, no M protein was identified, and hence, these patients were not included in the study. Subsequently, we analyzed and gathered relevant data for all 36 participants. Then, in the same initial cohort, we identified 36 individuals without gammopathy, case‑by‑case matched for sex, age at transplantation, donor type, and diagnosis being an indication for allo‑HCT. As our study involved retrospective analysis of already existing data from past hospital records and no intervention requiring patient informed consent was performed, the study did not require an approval of the ethics committee.

Variables

The analyzed data included age at transplantation, sex, disease being an indication for transplantation, donor type, M protein type(s), conditioning regimen, date of transplantation, date of immunofixation, and date of the last follow‑up for all participants. Furthermore, if applicable, we also accounted for date of death, date of relapse / progression, date of cGvHD diagnosis, organ and global severity of the disease according to the 2014 National Institute of Health (NIH) criteria,¹² date of acute GvHD (aGvHD) diagnosis, organ and global severity of the disease according to the MAGIC criteria,¹³ and other variables potentially associated with post–allo‑HCT gammopathy occurrence.

Statistical analysis

Quantitative variables were analyzed and compared in terms of the number of cases, percentage, incidence (in %), incidence rate (in cases per person‑years), mean with corresponding 95% CIs and SD, median with corresponding interquartile ranges, hazard ratio (HR) with corresponding 95% CIs, overall survival (OS), and progression‑free survival (PFS). Normal distribution of the data in the analyzed groups was evaluated with the Shapiro–Wilk test. The equality of variances was assessed with the Levene test. Based on these results, differences between the groups were evaluated with the Student t test, Welch t test, or Mann–Whitney test for continuous data. The χ² test and the method of logistic regression of univariable outcomes were used for categorical data. Survival analysis was performed for the gammopathy group vs the matched control group and for the MG group vs the OG group. Survival curves were evaluated with the Kaplan–Meier method. OS was calculated from the date of allo‑HCT to the date of death from any cause, with living patients censored on the last follow‑up date. PFS was calculated from the date of allo‑HCT to the date of relapse / progression or death from any cause, with living patients without relapse / progression censored on the last follow‑up date. HRs were assessed with the Cox proportional hazards model. Comparisons of survival distributions were performed using the log‑rank test. For all tests, a P value below 0.05 was chosen as a default threshold necessary to reject the null hypothesis. In the case of multiple comparisons for variables with more than 2 categories, the thresholds of significance were adjusted using the Bonferroni correction of P values. All necessary data analyses were performed using Statistica version 13.3 (TIBCO Software Inc., Palo Alto, California, United States).

Gammopathy vs no gammopathy

Among the 241 allo‑HCT recipients, we identified 36 cases of post–allo‑HCT gammopathy (incidence, 14.9%; incidence rate, 0.142 cases per person‑years). Then, their characteristics were compared with corresponding data for the matched control group. Comprehensive data on patient characteristics in both groups are showed in Table 1.

For the groups with and without post–allo‑HCT gammopathy, 5‑year OS rates were 79.6% and 50.9%, respectively (HR, 3.05; 95% CI, 1.18–7.86; P = 0.02). After 10 years from transplantation, survival was highly comparable, with OS rate of 51.7% and 50.9%, respectively (HR, 2.19; 95% CI, 0.95–5.03; P = 0.06). Similarly, 5- and 10‑year PFS rates were 70.1% and 51.2% (HR, 1.91; 95% CI, 0.86–4.21; P = 0.1) and 45.5% and 51.2% (HR, 1.54; 95% CI, 0.74–3.22; P = 0.25), respectively. The Kaplan–Meier curves are showed in Figure 1.

The frequency of aGvHD in the gammopathy and control groups was also similar (22.2% vs 19.4%; P = 0.77), with no major differences in aGvHD severity according to the MAGIC criteria (mild, P = 0.64; moderate, P = 0.45; and severe, P = 0.15).

On the other hand, cGvHD occurred more frequently in the gammopathy group (63.9% vs 30.6%; OR, 4.02; 95% CI, 1.5–10.74; P = 0.005). There was no difference in terms of the severity of cGvHD according to the 2014 NIH criteria (mild, P = 0.5; moderate, P = 0.06; and severe, P = 0.21). The organs involved more frequently in the gammopathy group included the skin (P = 0.004), oral cavity (P = 0.009), gastrointestinal tract (P = 0.006), and liver (P = 0.01). No differences were observed for the eyes (P = 0.55), lungs (P = 0.45), joints (P = 0.08), and the hematopoietic system (P = 0.09).

Monoclonal vs oligoclonal gammopathy

Among the 36 individuals with post–allo‑HCT gammopathy, we identified 20 MG and 16 OG cases. There were no major differences between the MG and OG groups in terms of sex (P = 0.42) and age at transplantation (P = 0.22). Mean (SD) age at transplantation was 47.1 (17) years (95% CI, 38.9–55.3) for the MG group and 42.2 (13) years (95% CI, 35–49.3) for the OG group. Moreover, the groups did not differ in terms of donor type, type of conditioning, and the most common underlying diseases.

The most commonly identified M protein types were IgGλ and IgGκ, which occurred in 9 and 7 patients with MG, respectively, and in all but 1 patient as part of protein combination in the OG group. Moreover, 9 in 16 patients in the OG group had both IgGλ and IgGκ identified with immunofixation. Comprehensive data on patient characteristics are showed in Table 2.

For patients with MG and OG secondary to allo‑HCT, 5‑year OS rates were 81.2% and 75.4%, respectively, and 10‑year OS rates were 46.4% and 50.3%, respectively. The groups did not differ much in terms of either 5‑year OS (HR, 0.81; 95% CI, 0.16–4.04; P = 0.8) or 10‑year OS (HR, 1.02; 95% CI, 0.27–3.83; P = 0.98). Similarly, 5‑year PFS rates were 82% and 55.9%, respectively, and 10‑year PFS rates were 46.8% and 37.2%, respectively. The groups did not differ markedly in terms of either 5‑year PFS (HR, 0.3; 95% CI, 0.08–1.17; P = 0.07) or 10‑year PFS (HR, 0.43; 95% CI, 0.14–1.31; P = 0.13). The Kaplan–Meier curves are showed in Figure 2.

There was no significant difference in terms of the frequency of aGvHD between the MG and OG groups (30% vs 12.5%; P = 0.21). No difference was also observed regarding the severity of aGvHD according to the MAGIC criteria.

The incidence of cGvHD was very high in both groups. The patients with MG and OG did not differ in terms of the frequency of cGvHD (55% vs 75%, P = 0.22). There were also no differences in terms of particular organ involvement. Considering global severity, the incidence of moderate (P = 0.44) and severe cGvHD (P = 0.72) was similar in both groups. Mild cGvHD occurred more frequently in the OG group (P = 0.04).

Relationship between allogeneic hematopoietic cell transplantation, gammopathy, and chronic graft‑versus‑host disease

The mean (SD) time from allo‑HCT to immunofixation that confirmed the occurrence of secondary gammopathy was 16 (14.5) months for the MG group and 7.9 (6.3) months for the OG group. This difference (P = 0.06) implied a high probability of MG to be identified later than OG. The mean (SD) difference in time from allo‑HCT to the diagnosis of cGvHD was 9.1 (4.8) months for the MG group and 14.9 (9.3) months for the OG group (P = 0.09). Considering the diagnosis of cGvHD as a fixed point in time and relating it to the occurrence of gammopathy, the P value reached 0.009. Interestingly, OG was identified mostly before cGvHD diagnosis (median [IQR] of 6.3 [14.2–2.5] months following allo‑HCT), and MG mostly following cGvHD diagnosis (median [IQR] of 4.5 [0.9–20] months following allo‑HCT). Moreover, most MG and OG cases were identified in the interval from 12 months before to 12 months following cGvHD diagnosis. Comprehensive data are presented in Figure 3.

Gammopathy risk factors

Almost every patient in our study was diagnosed with infectious complications following allo‑HCT during immunosuppressive therapy. Such complications included viral reactivations (EBV, CMV, hepatitis B virus, human papillomavirus, varicella zoster virus, human herpesvirus 6, human polyomavirus 1), bacterial sepsis, and fungal pulmonary infections. Moreover, 2 patients were diagnosed with post‑transplant lymphoproliferative disease as a result of EBV reactivation. There were also 3 other secondary malignancies, namely renal cell carcinoma, lung squamous cell carcinoma, and head and neck carcinoma. All 3 cases of secondary malignancies occurred late following transplantation (about 3.5, 8, and 8.5 years following the procedure, respectively) in patients who remained in complete remission of the disease that was the original indication for allo‑HCT. One patient was diagnosed with Wiskott–Aldrich syndrome in childhood. Some individuals suffered from vasculitis. However, in these cases, it was difficult to differentiate between microangiopathy following allo‑HCT, drug‑related complications, or atypical cGvHD. None of the patients received alemtuzumab‑based conditioning. The gammopathy‑inducing potential of the mentioned factors has already been documented.¹⁴ Furthermore, 6 patients were preemptively treated with donor lymphocyte infusions. In 2 patients, natural killer cell lymphocytosis of uncertain significance was identified. One patient was diagnosed with hemophagocytic lymphohistiocytosis. One patient was treated with chimeric antigen receptor T‑cells following allo‑HCT. The gammopathy‑inducing potential of these factors has not been confirmed yet.

Limitations

The study was based on the hospital records of historical patient groups. As it was impossible to find more participants, the group sizes remained small. Although we did our best to include matching patients without gammopathy in the control group, there is always a possibility of selection bias. Moreover, serum protein electrophoresis and subsequent immunofixation were performed only in the patients suspected of monoclonal gammopathy by a clinician. It was not performed routinely in all patients after allo‑HCT or in all patients diagnosed with cGvHD due to a lack of such recommendations. As a result, post–allo‑HCT gammopathy is probably underdiagnosed. The fact that cGvHD occurs more frequently in the patients with post–allo‑HCT gammopathy might be confounded with superior OS of these patients. As they survive longer following transplantation, they have an increased chance of cGvHD diagnosis.

Conclusions

Post–allo‑HCT gammopathy is a laboratory finding frequently identified following allogeneic transplantation, most probably as a transient result of the immune reconstitution. It is not associated with an increased risk of plasma cell dyscrasias, but is most probably a reactive phenomenon. The occurrence of post–allo‑HCT gammopathy improves long‑term survival following transplantation and correlates with frequent incidence of cGvHD. The number of monoclonal proteins identified in patients’ serum via immunofixation probably does not have a clinical significance. As the differences between MG and OG groups are minor, both situations are probably different laboratory presentations of a single phenomenon, rather than distinct conditions.