Nodal peripheral T‐cell lymphoma in Finland between 2007 and 2019: Incidence, late morbidity and survival

Summary

Data on late morbidity and survival in nodal peripheral T‐cell lymphoma (PTCL) are scarce. This study investigated the incidence, subtype distribution, late morbidity and survival outcomes of nodal PTCL in Finland. The study compared 998 patients with nodal PTCL diagnosed between 2007 and 2019 with matched controls. Data from four nationwide Finnish healthcare registries were used. PTCL not otherwise specified was the most common subtype, with a mean annual incidence of 0.62 (0.45–0.87) per 100 000. The 1‐year survival was 34% for patients aged <70 years and 49% for those aged ≥70 years. In the younger group, an increased risk of chronic heart failure (hazard ratio [HR], 4.17 [2.16–8.05]; p < 0.001) was observed after PTCL diagnosis. The risks of pneumonia (HR, 6.47 [4.40–9.53]; p < 0.001 and HR, 2.78 [2.010–3.859]; p < 0.001) and thrombosis (HR, 16.09 [4.22–61.39]; p < 0.001 and HR, 5.34 [2.33 to NA]; p < 0.001) were significantly higher in the younger and older age groups than that in controls. The prognosis of nodal PTCL is poor, and patients have a significantly higher risk of comorbidities than controls, irrespective of age. Therefore, targeted and less toxic treatment modalities are required.

INTRODUCTION

Nodal peripheral T‐cell lymphoma (PTCL) is a rare and aggressive type of malignancy. The annual incidence in Europe is 0.8 per 100 000 population, with a male predominance, and the typical age at diagnosis is >60 years. ¹ According to the WHO classification, PTCL has more than 30 subtypes with different biological and genetic backgrounds. ² , ³ , ⁴ The most common subtypes include PTCL not otherwise specified (NOS); nodal T‐follicular helper cell lymphoma (nTFHL), including nTFHL‐NOS, nTFHL angioimmunoblastic‐type (nTFHL‐AI; previously known as angioimmunoblastic T‐cell lymphoma [AITL]) and follicular‐type nTFHL (nTFHL‐F; previously known as follicular TCL); anaplastic large cell lymphoma (ALCL), including anaplastic lymphoma kinase (ALK)‐positive and ‐negative types; and enteropathy‐associated TCL. ³ , ⁵

Effective standard therapy for PTCL is lacking. Standard treatment consists of anthracycline‐based protocols, including CHOP (cyclophosphamide, doxorubicin, vincristine and prednisone)‐like regimens with or without etoposide and dose‐adjusted EPOCH (etoposide, prednisone, vincristine, cyclophosphamide and doxorubicin). ⁶ , ⁷ The recent ECHELON‐2 trial established brentuximab vedotin plus CHP (BV‐cyclophosphamide, doxorubicin, prednisolone) as a first‐line therapy option for patients with CD30‐positive subtypes, especially those with ALCL. ⁸ In patients with stage I–II ALK‐positive ALCL, radiotherapy combined with limited cycles of chemotherapy is an alternative. ⁷ The COMPLETE study, ⁹ in concordance with a recent prospective study, ¹⁰ found autologous haematopoietic stem cell transplantation (ASCT) to be beneficial in patients with TFHL; however, contradictory results were observed in a retrospective study conducted by the LYSA centres. ¹¹ According to the NCCN guidelines, ASCT consolidation is a treatment option for chemosensitive transplant‐eligible patients with PTCL, but not for those with ALK‐positive ALCL with low‐ or intermediate‐risk International Prognostic Index. ⁷ Regardless of therapy choice, a vast majority of patients with PTCL have refractory disease or inevitable relapse, resulting in a limited 5‐year progression‐free survival (PFS) rate of 20%–30%. ¹²

Another challenge in the treatment of PTCL is chemotherapy‐associated late morbidity, as doxorubicin, an RNA‐ and a DNA‐synthesis inhibitor, increases cardiovascular disease (CVD) morbidity risk. ¹³ Additionally, etoposide, a topoisomerase II inhibitor, may be associated with the risk of secondary leukaemia, ¹⁴ and vincristine, a mitotic spindle inhibitor, has been shown to be responsible for neurotoxicity. ¹⁵ Moreover, cyclophosphamide, an alkylating agent, is associated with a risk of haematological malignancies. ¹⁶ Notably, somatic mutations associated with PTCL may predispose patients to late haematological morbidity. ⁵

This real‐world study investigated the incidence, causes of death, late morbidity and survival outcomes of PTCL in Finland using data derived from nationwide healthcare registries.

PATIENTS AND METHODS

All patients diagnosed with PTCL between 1 January 2007 and 31 December 2019, in Finland, were included in this study. Based on care practices, including eligibility for ASCT, ¹⁷ patients were divided into younger (aged <70 years) and older (aged ≥70 years) groups. The cut‐off of 70 years was based on divergent care practices without ASCT in older patients. ¹⁸

Data collection

The study data were obtained from four nationwide Finnish registry controllers by using unique Finnish personal identification numbers: the Finnish Cancer Registry (FCR), the National Institute of Health and Welfare (THL), Statistics Finland and the Digital and Population Data Services Agency (DPA). The data covered the entire population of Finland, encompassing approximately 5.5 million people in 2019. This study included only national registry data; thus, hospital‐administered medication data stored in individual data lakes in Finland were unavailable.

The study population was formed based on the first nodal PTCL diagnoses between 2007 and 2019 collected from the FCR, based on ICD‐O‐3 morphological codes (PTCL‐NOS: 9702; AITL: 9705; ALCL: 9714 and 9715; hepatosplenic TCL [HSTCL]: 9716; enteropathy‐associated TCL [EATL]: 9717; natural killer/TCL [NKTL]: 9719). Every patient was matched with a control according to age, sex and region in a 1:1 ratio using DPA records. Members of the control group were eligible if they were alive at the date of diagnosis (index date) of the corresponding patient and had neither a prior nor a concomitant diagnosis of lymphoma (C81–C85) during follow‐up.

Ethics approval statement

This study was approved by the Finnish Social and Health Data Permit Authority (Findata number, THL/1541/14.02.00/2021) and Statistics Finland (number, TK/3616/07.03.00/2021). The ethical principles of the Declaration of Helsinki and Good Clinical Practice statements were followed while handling and analysing pseudonymized patient data in a secure environment. Owing to the retrospective nature of the study, the patients' treatments were not affected.

Statistics

Statistical analyses were performed using R version 4.0.3. Patient characteristics and demographics were assessed at diagnosis, with follow‐up continuing until 31 December 2019, or death. The Charlson comorbidity index (CCI) was assessed from specialty care diagnoses (THL), excluding lymphoma (C81–C85) from 3 years to 3 months before PTCL diagnosis, and a modified scoring system was used. ¹⁹ Statistical differences were tested using the chi‐squared test for categorical variables and the Kruskal–Wallis test for continuous variables. Paired testing was not performed because of the small sample sizes of some subgroups. The annual incidence per 100 000 people was calculated for each subgroup. The incidence by age group was stratified according to sex.

Specialty care registry diagnoses alone were used to identify the comorbidities, as primary care diagnoses were available no earlier than 2011. Diagnostic records containing data from 3 years up to 3 months before PTCL diagnosis were used to define baseline comorbidities, and late morbidities were defined as those observed 3 months after PTCL diagnosis. The accumulation of comorbidities during the follow‐up was estimated using Aalen–Johansen estimators for diseases (Table S1) and with the mean cumulative function for recurring comorbidities (Table S1). Comorbidity recurrence was defined as a new record of an event after 2 months since the last record of the event in case of pneumonia, septicaemia and bronchitis and later than 1 year in case of venous thrombosis (VT).

Differences in comorbidity risk between patients and controls and factors associating with comorbidity occurrence were studied using Cox regression models including age, sex and case–control pairing as a clustering variable, with death considered as a censoring event. Recurring comorbidities were modelled as recurring events; however, only the risks for the first comorbidity occurrence were presented. The results and p‐values were descriptive, and multiple testing corrections were not applied. Kaplan–Meier analysis was used for overall survival (OS), with follow‐up until the end of the study or death. Differences in OS between patients and controls were tested using the stratified log‐rank test. A Cox regression model was used to study the factors associated with OS. Radiation status and stem cell transplantation (SCT) were included as time‐dependent variables to avoid an immortal time bias. Proportional hazard assumptions were tested with all Cox models. Aalen–Johansen estimators were used to study causes of death. All statistical tests were two‐tailed, and p‐values <0.05 were considered statistically significant. No missing values were imputed.

RESULTS

Patient characteristics

Altogether, 998 patients were included in the study; of those, 550 were aged <70 years and 438 aged ≥70 years. Most patients were male (younger group, 62.2%; older group, 53.9%; Table 1). The proportions of chosen baseline comorbidities were comparable with those of controls, whereas the CCI at diagnosis was significantly higher in patients than in controls in both younger and older age groups (p = 0.029 and p < 0.001 respectively; Table 1). A minority of the patients received radiotherapy: 12.7% and 9.1% in the younger and older groups respectively. Less than one‐fourth of the patients in the younger group proceeded to transplantation, and the vast majority of those received ASCT alone (n = 124; 88%).

TABLE 1.

Characteristics and demographics of 998 patients with peripheral T‐cell lymphoma and their 998 controls according to two age groups. Testing was not performed for matching variables (age & sex).

Variable	Younger (<70 years) group			Older (≥70 years) group
	Patients (n = 550), n (%)	Controls (n = 550), n (%)	Sig. p‐value	Patients (n = 438), n (%)	Controls (n = 438), n (%)	Sig. p‐value
Age, years (median, IQR)	59.14 (49.11, 64.82)	59.14 (49.11, 64.82)		77.97 (73.8, 82.72)	77.97 (73.8, 82.72)
Gender
Female	208 (37.8)	208 (37.8)		202 (46.1)	202 (46.1)
Male	342 (62.2)	342 (62.2)		236 (53.9)	236 (53.9)
Baseline comorbidities
HA	35 (6.4)	35 (6.0)	0.899	80 (18.3)	66 (15.1)	0.246
CVD	13 (2.4)	19 (3.5)	0.361	54 (12.3)	59 (13.5)	0.684
Dysrhythmia	11 (2.0)	23 (4.2)	0.518	57 (13.0)	57 (13.0)	1.000
CHF	<5	6 (1.1)	0.505	39 (8.9)	27 (6.2)	0.169
Diabetes	20 (3.6)	20 (3.6)	1.000	30 (6.8)	30 (6.8)	1.000
Cerebral vascular disease	9 (1.6)	<5	0.267	24 (5.5)	16 (3.7)	0.256
Depression	12 (2.2)	6 (1.1)	0.238	7 (1.6)	<5	0.547
CCI			0.029			<0.001
0	466 (84.7)	495 (90.0)		309 (70.5)	321 (74.0)
1–2	73 (13.3)	49 (8.9)		102 (23.3)	89 (20.3)
3+	11 (2.0)	6 (1.1)		27 (6.2)	25 (5.7)
Subtype of nodal PTCL
PTCL‐NOS	239 (43.5)			205 (46.8)
AITL	109 (19.8)			122 (27.9)
ALCL	133 (24.2)			67 (15.3)
EATL	47 (8.5)			38 (8.7)
NK/T‐cell nasal	17 (3.1)			6 (1.4)
Hepatosplenic	5 (0.9)
Radiotherapy used	70 (12.7)			40 (9.1)
Transplantation
Autologous	124 (22.5)			9 (2.1)
Allogeneic	9 (1.6)
Autologous + allogeneic	8 (1.5)
Follow‐up time, years (median, IQR)	1.77 [0.46, 5.27]	6.46 [2.87, 9.69]	<0.001	0.53 [0.13, 1.90]	3.93 [0.82, 6.96]	<0.001

Abbreviations: AITL, angioimmunoblastic T‐cell lymphoma; ALCL, anaplastic large T‐cell lymphoma; ALK, anaplastic lymphoma kinase; CCI, Charlson Comorbidity Index; CHF, chronic heart failure; CVD, cardiovascular disease; EATL, enteropathy‐associated T‐cell lymphoma; HA, hypertension arterialis; IQR, interquartile range; NK, natural killer; NOS, not otherwise specified; PTCL, peripheral T‐cell lymphoma.

Incidence

The mean annual incidence per 100 000—with wide yearly variations—was 0.62 for PTCL‐NOS, 0.32 for AITL, 0.28 for ALCL and 0.13 for EATL (Figure 1A). HSTCL (0.02 per 100 000) and NKTL (0.04 per 100 000) had the lowest mean annual incidence. The annual incidence per 100 000 population by age group was highest in men aged 80–85 years (Figure 1B).

FIGURE 1.

(A) Annual incidence by subtypes of PTCL and (B) incidence by age groups of PTCL per 100 000 population in Finland between 2007 and 2019. PTCL, peripheral T‐cell lymphoma.

Overall survival

The 1‐year survival rates in the younger and older groups were 66% and 41% respectively. The median OS in the younger group was 28.7 months (14.2–46.9) for PTCL‐NOS, 35.7 months (22.6–52.2) for AITL and not reached (NR) for ALCL subtypes, whereas the median OS was NR in any of the control groups. Moreover, the shortest median OS was 8.3 months (5.0–14.1) in patients with EATL. In the older group, the median OS was only 4.9 months (2.7–7.1) in patients with PTCL‐NOS versus 106.3 months (86.6–136.4) in controls (p < 0.0001), 11.9 months (8.8–19.4) in patients with AITL versus 111.3 months (83.6‐NR) in controls (p < 0.0001) and 10.5 months (3.0–22.3) in patients with ALCL versus 84.6 months (54.0–114.3) in controls (p < 0.0001; Figure 2). Additionally, in older patients with EATL, the median OS was extremely short (5.9 months [2.6–12.1]).

FIGURE 2.

Overall survival of patients with the most common subtypes of nodal PTCLs according to two age groups at diagnosis: (A, B) PTCL‐NOS, (C, D) AITL and (E, F) ALCL. ALCL, anaplastic large cell lymphoma; NOS, not otherwise specified; PTCL, peripheral T‐cell lymphoma.

Higher age at diagnosis was associated with poorer survival (HR per year, 1.04 [1.03, 1.04]; p < 0.001) and later year of diagnosis with better survival (HR per year, 0.97 [0.95, 1.08]; p < 0.022). Furthermore, patients who received radiotherapy (HR, 0.65 [0.50, 0.86]; p < 0.002) and those who proceed to SCT (HR, 0.55 [0.41, 0.73]; p < 0.001) had better outcomes. AITL (HR, 0.71 [0.58, 0.869]; p < 0.001) and ALCL (HR, 0.58 [0.46, 0.74]; p < 0.001) were associated with better prognosis than PTCL‐NOS (Figure 3). No significant increase in OS between the 2014–2019 and 2007–2013 time periods was observed in both the younger (OS, 49.0 months [23.7 to NA] vs. 37.5 months [22.6–55.4]; p = 0.219) and older (6.3 months [4.7–9.3] vs. 8.8 months [5.6–12.1]; p = 0.488) groups (Figure S1).

FIGURE 3.

Cox proportional hazards model for overall survival of patients with peripheral T‐cell lymphoma in Finland during 2007–2019.

During the first year after diagnosis, altogether 25% of younger patients died from nodal PTCL. During the 5‐year follow‐up, 40% of younger patients died owing to lymphoma (Table S2a, Figure 4). B‐cell lymphoma was the cause of death in 15% of older patients, comprising mainly those with a primary diagnosis of ALCL or PTCL‐NOS (Table S2b, Figure 5).

FIGURE 4.

The risk of the first (A) late morbidity or (B) thrombosis or infection in the patients aged <70 years and (C) late morbidity or (D) thrombosis or infection in the patients aged ≥70 years.

FIGURE 5.

Causes of death in the patients with PTCL (A) aged <70 years and (B) aged ≥70 years and their matched controls. PTCL, peripheral T‐cell lymphoma.

Morbidity after PTCL diagnosis

Younger patients had an elevated probability of heart diseases, manifested by a more than fourfold risk of CHF versus controls (HR, 4.17 [2.16–8.05]; p < 0.001). Also, younger patients were more likely to be affected by depression than controls (HR, 2.86 [1.25–6.52]; p = 0.013). Most remarkably, younger patients had a 16‐fold risk of VT (HR, 16.09 [4.22–61.39]; p < 0.001) and also a higher risk of septicaemia (HR, 12.90 [6.59–25.24]; p < 0.001) after PTCL diagnosis than controls, whereas the most common late morbidity was pneumonia (HR, 6.47 [4.40–9.53]; p < 0.001; Figure 4A,B, Table S3, Figure S2).

In the older group, the manifestations of late morbidity were infections such as pneumonia (HR, 2.78 [2.01–3.859]; p < 0.001) and septicaemia (HR, 3.54 [2.13–5.89]; p < 0.001) and a more than fivefold risk of VT (HR, 5.34 [2.33–12.26], p < 0.001; Figure 4C,D, Table S3). Notably, the risk of subsequent thrombosis or infection was not significantly different in either group compared to controls.

DISCUSSION

This large real‐world study reports the incidence, subtype distribution, late toxicities and survival outcomes of nodal PTCL in Finland. The highest incidence was observed among patients aged 80–84 years, with PTCL‐NOS being the most common subtype. In particular, older patients had a poor prognosis and only 41% of those were alive 1 year after diagnosis. Regarding later morbidity, younger patients had a more than fourfold higher risk of CHF and a 16‐fold higher risk of VT than controls. Most remarkably, infection risk—manifested by pneumonia and septicaemia—was significantly higher in both younger and older patients than in controls. More effective and less toxic novel treatment modalities are warranted to improve outcomes in these patients.

In Western countries, PTCL accounts for 5%–15% of lymphomas, ²⁰ and the age‐adjusted incidence of PTCL is 2.1 per 100 000. ²¹ In contrast to the increased occurrence of B‐cell lymphomas, the prevalence of nodal PTCL remains quite stable. ²² Consistent with previous data, we found no major changes in the incidence of PTCL. Women aged between 75 and 79 years and men aged between 80 and 84 years had the highest incidence. Consistent with data from other Western countries, ⁵ we found that PTCL‐NOS was the most common. However, we did not perform a pathology review; the disease was classified according to electronic medical records, and the classification was valid at each time point. In the WHO HAEM V classification, PTCL‐NOS with the TFH phenotype was moved under the heading nTFHs; thus, we cannot rule out the possibility that this might be as large group as PTCL‐NOS. ³

Peripheral T‐cell lymphoma incidence is higher in Asia than in Western countries; for example, in the Beijing Cancer Registry, the age‐adjusted incidence was 0.35 per 100 000. ²³ Additionally, Epstein–Barr virus–associated PTCL is more common in East Asia and South America, ²⁴ probably leading to a higher PTCL incidence in Asia. ⁶ We found a higher EATL incidence (0.13 per 100 000) than in the US (0.014 per 100 000); however, in the US, EATL incidence has increased by 2.58% annually. ²⁵ Biopsy‐associated coeliac disease, a well‐known risk factor for lymphoma, has increased in the US and Europe ²⁶ owing to the high prevalence of HLA DQA1*0501 and DQB1*0201, especially in the Nordic population. ²⁷ The early detection of coeliac disease and a healthy lifestyle may protect against PTCL.

In line with previous data, we observed poor OS in both age groups, except patients with ALK‐positive ALCL. Unfortunately, outcomes have not significantly improved during the last decades, which was also observed in our analyses, irrespective of age. However, possibly improved diagnostics was causative for better OS in later years. Radiotherapy was associated with better OS in our analyses, possibly because of the limited stage of the disease in patients treated with radiotherapy. A Nordic research group reported a 5‐year OS rate of 58% in patients with limited‐stage PTCL treated with CHOP therapy with or without radiotherapy. Age ≥60 years and B‐symptoms were predictive for inferior OS in those patients ²⁸ in concordance with our findings. A plausible explanation may be the suboptimal treatment used in older patients owing to comorbidities at baseline, which limit the assortment of available treatments. Additionally, SCT resulted in improved OS in our population; however, in clinical practice, only younger patients responding to induction therapy are transplant‐eligible; thus, these patients are highly selected.

PTCL can be split into 30 subgroups with distinct immunophenotypic and molecular characteristics, and subtypes have variable clinical pictures of the disease, including nodal, extranodal or leukaemic manifestations. ⁷ , ²⁹ , ³⁰ Increasing biological data regarding various subtypes have questioned the ‘one‐fits‐all’ approach in PTCL and enabled the dawn of new, more precisely targeted treatments. However, owing to the rarity of this disease, only a few prospective trials have been conducted to determine optimal treatment strategies. The phase 2 EATL‐001 study with a limited number of patients with EATL suggested that BV‐CHP plus ASCT resulted in 2‐year PFS and OS rates of 63% and 68% respectively. ³¹ The same combination in the ECHELON‐2 study produced 5‐year PFS and OS rates of 51.4% and 70.1%, respectively, in patients with ALCL in a front‐line setting. ⁸ In the relapsed setting, EZH1/2 and JAK/STAT pathway inhibitors have shown promising results. ³² In the future, the therapeutic landscape of nodal PTCL may include epigenetic modifying agents, such as romidepsin, belinostat and azacitidine, alone or in combination with immuno‐oncologic agents. ³²

Other malignancies were a more frequent cause of death among patients with PTCL than among controls. B‐cell lymphomas were unexpectedly overrepresented in both age groups as a cause of death. That may be partly explained by the fact that nTFHLs may relapse with a B‐cell phenotype; however, we cannot fully exclude the possibility of misclassification. Myeloid and solid tumours were more common causes of death among younger patients than among controls. Myeloid malignancies may be therapy‐related or arise from haematopoietic stem cells carrying the same driver mutations as those in treated T‐cell disease. ³³ However, a great majority of patients with nodal PTCL die early after diagnosis, before the appearance of secondary malignancies.

We found a higher proportion of CVD‐related deaths among younger patients than among controls, whereas the proportion was lower in older patients. This is probably explained by the healthy survivor effect, as older patients with the highest comorbidity load also have the highest risk of dying of lymphoma. This insists that CVD risk factors of asymptomatic patients should be treated.

Well‐known risk factors for VT in patients with lymphoma early after diagnosis include advanced stages of the disease and exposure to anthracyclines. ³⁴ Although a remarkable 16‐fold risk of VT was detected in patients aged <70 years in the present study, the incidence rate was lower than that in a previous publication, ³⁴ probably owing to the different study populations.

A notable finding in our study was the high number of pneumonia and sepsis cases among patients with lymphoma aged <70 years (up to 25%) versus controls. The impaired immune system, treatment‐related neutropenia and possibly a slightly higher CCI at diagnosis in patients play a causative role in infections.

This study had some limitations. First, owing to its retrospective nature, there may be missing or inaccurate data caused by various recording practices. Second, detailed data on the treatments used were unavailable; however, owing to uniform treatment practices, the effect of this on the results is probably minimal. Additionally, data on precise medical history, including way of life, were unavailable. We did not have data on patients included in clinical trials. The follow‐up was also limited owing to the early deaths of patients, especially in the older age group. Additionally, the healthy survivor effect in older patients may have hampered the results concerning late comorbidities. The pathological anatomical classification of nodal PTCLs is challenging, and there is some interobserver variance. Our study did not include a pathological review, which may have caused some variance. Moreover, the diagnosis was reported to FCR according to the existing lymphoma classification; thus, the subclassification does not fully match the novel WHO HAEM V classification ³ and will likely overestimate the proportion of the PTCL‐NOS subgroup. However, the strengths of this study include the large study population and well‐matched controls, encompassing Finnish national registries and real‐life data on the incidence and outcomes of nodal PTCL.

In conclusion, patients with nodal PTCL have a significantly higher risk of pneumonia and VT than controls, irrespective of age, as well as an excess risk of CHF in the younger group. Older age at diagnosis was associated with inferior outcomes, and nodal PTCL remained the most common cause of early death. New, innovative, more effective and less toxic treatment approaches are warranted to improve outcomes, especially in older patients with PTCL.

AUTHOR CONTRIBUTIONS

OK, AR, AA, LU‐V, IT, AP and TM designed the study. AA, LU‐V and IT coordinated data collection. AA, LU‐V and IT performed the analyses. AA, LU‐V and IT provided expert insight into the statistical analyses. AP, OK, AA and AR wrote the manuscript. OK and HK coordinated research funding. All authors reviewed the manuscript critically and read and approved the final version of the manuscript.

CONFLICT OF INTEREST STATEMENT

AP reports consultancy fees from Behring and Abbvie and has participated in Scientific Advisory Board meetings organized by Abbvie, AstraZeneca, Janssen‐Cilag, Novartis, Pfizer and Takeda. AR reports personal fees for speaking at symposia and financial support for attending conferences from Amgen, Bristol‐Myers Squibb, Merck, Novartis, Pfizer, Daichii Sankyo and Roche. TM was a contractor for Takeda Oy during the study course. LU‐V, IT and AA are employed by Medaffcon Oy. All other authors declare no competing interests.

Supporting information

Data S1.

Partanen A, Rönkä A, Anttalainen A, Ukkola‐Vuoti L, Toppila I, Kuitunen H, et al. Nodal peripheral T‐cell lymphoma in Finland between 2007 and 2019: Incidence, late morbidity and survival. Br J Haematol. 2025;207(3):851–859. 10.1111/bjh.20253

DATA AVAILABILITY STATEMENT

This study was based on healthcare registry data. Data can be acquired with permission by following the guidelines and application processes of the registries.