Long-term second primary cancer risk in adolescent and young adult (15-39 years) cancer survivors: a population-based study in the Netherlands between 1989 and 2018

Background Few studies have comprehensively investigated the long-term second cancer risk among adolescent and young adult (AYA, aged 15-39 years) cancer survivors. This study investigated the long-term second cancer risk by including the full range of first and second cancer combinations with at least 10 observations in the Netherlands between 1989 and 2018. Materials and methods First and second primary cancer data of all 6-month AYA cancer survivors were obtained from the nationwide population-based Netherlands Cancer Registry. Excess cancer risk compared to the general population was assessed with standardized incidence ratio (SIR) and absolute excess risk (AER) statistics up to 25 years after diagnosis. Cumulative incidences were estimated, using death as a competing risk factor. Analyses were carried out with and without applying multiple cancer rules. Results The cohort included 99 502 AYA cancer survivors. Male survivors had a 2-fold higher risk of developing any cancer compared to the general population, whereas this was around 1.3-fold in females. AERs were 17.5 and 10.1 per 10 000 person-years for males and females. The long-term excess risk of cancer was significantly higher for most first and second primary cancer combinations, but comparable and lower risk estimates were also observed. Application of the multiple cancer rules resulted in a noticeable risk underestimation in melanoma, testicular, and breast cancer survivors. Risk outcomes remained similar in most cases otherwise. The cumulative incidence of second cancer overall increased over time up to 8.9% in males and 10.3% in females at 25 years’ follow-up. Highest long-term cumulative incidences were observed among lymphoma survivors (13.3% males and 18.9% females). Conclusions AYA cancer survivors have a higher cancer risk compared to the general population for most cancers up to 25 years after their initial cancer diagnosis. Additional studies that investigate risk factors for the specific cancer type combinations are needed to develop personalized follow-up strategies.


INTRODUCTION
Adolescents and young adults (AYAs, defined as individuals aged 15-39 years at first cancer diagnosis) form a distinct population that is often overlooked within the oncology care setting. 1,2AYAs present themselves with a distinct spectrum of cancers, varying from cancers frequently found among children (e.g.acute lymphatic leukaemia, Ewing's sarcoma), to cancers usually found in older adults (e.g.colorectal, lung, and breast carcinomas), but also cancers with the highest incidence at AYA age (e.g.Hodgkin's lymphoma, melanoma, germ cell tumours, and thyroid carcinomas). 35][6][7][8][9] Current estimates show that w85% of all AYAs are still alive 5 years after their primary cancer diagnosis. 10As such, there is a growing population of AYA cancer survivors that is at risk of developing survivorship-related adverse health outcomes later in life, creating new challenges for AYA survivors and their loved ones. 119][20][21][22][23][24] AYA survivors in high-income countries are shown to have a 1.2-to 2-fold higher risk of developing any second cancer compared to the general population, resulting on average in 11-23 excess cancers per 10 000 person-years. 18,20,23,24][22][23][24] These initial findings provide important information.6][27][28][29][30] As such, there is a general lack of detailed knowledge about the second cancer risk pertaining to a large number of first and second primary cancer types among the vulnerable AYA survivor population.Identification of distinct patient groups that are at a particular high risk of developing a second cancer can further help to guide the development of better survivorship care guidelines aimed at prevention and early detection of second cancers among AYAs, which has been shown to improve relative survival outcomes up to 47%. 31 Therefore, more detailed second cancer risk studies that describe long-term risk patterns in AYAs are warranted.This study investigated the long-term risk of developing a second malignant cancer overall and by including the full range of first and second cancer combinations with at least 10 observations up to 25 years after diagnosis in the Netherlands between 1989 and 2018.

Data collection
Data of all primary malignancies were obtained from the population-based Netherlands Cancer Registry (NCR), with a national coverage since 1989 and an estimated completeness of at least 95%. 32The NCR is based on notification from the nationwide network and registry of histopathology and cytopathology (PALGA) and the national registry of hospital discharges.NCR records contain information about patient, tumour, and primary treatment characteristics collected from medical records by trained registration clerks of the NCR.Malignancies within the NCR are coded based on the International Classification of Diseases for Oncology (ICD-O), using the first edition between 1989 and 1993, the second between 1993 and 2000, and the third edition since 2001. 33Information on vital status (i.e.dead, alive, emigrated) is obtained through annual linkage between the NCR and the Nationwide Personal Records Database (BRP, last linkage on 1 February 2020).In the current analyses, the first two primary malignant cancers (ICD-O behaviour /3) per AYA were included.Subsequent primary malignancies were omitted, including 372 third, 34 fourth, and 8 fifth malignancies.AYAs were eligible for inclusion when diagnosed with first cancer at ages 15-39 years in the Netherlands between 1989 and 2018. 2 The NCR did not provide data about basal-cell skin and lip carcinomas and second primary squamous cell tumours of the skin.Myelodysplastic and myeloproliferative syndromes were available since 2002.Moreover, cancers were included only when they satisfied the international rules for multiple cancers published by the International Agency for Research on Cancer (IARC), which exclude all extensions, recurrences, and metastasis. 33To minimize detection and information bias, AYA survivors with a followup <6 months after first primary cancer diagnosis were excluded [n ¼ 7002 (6.6%) total, including n ¼ 318 (0.3%) second cancers]. 34

Diagnostic classification
Cancers were grouped into 11 diagnostic main groups and further categorized into more detailed diagnostic subgroups according to the topography (anatomic location), morphology, and behaviour based AYA-specific classification scheme developed by Barr and colleagues. 35,36etailed definitions for each of the diagnostic main groups and subgroups are provided elsewhere. 35,36

Statistical analysis
Expected numbers of second cancer in the general population were calculated by multiplying the accumulated person-time at risk among AYA cancer survivors by the corresponding age-(5-year bands), sex-, calendar year-, and cancer type-specific incidence rates in the general population and then summing the results.Cancer incidence rates for the general background population were provided by the NCR and included annual crude rates per 100 000 person-years for all malignancies diagnosed in the Netherlands between 1989 and 2018.Person-time at risk for each of the strata was obtained by accumulating individual follow-up times (in days) from 6-month survival until the date of second cancer diagnosis, death, loss to follow-up, or 31 December 2018, whichever came first.Standardized incidence ratios (SIRs) and absolute excess risks (AERs) per 10 000 person-years with corresponding exact Poisson distribution-derived 95% confidence intervals (CIs) were calculated from the observed and expected numbers of second cancers, using standard statistical methods. 37Second cancer cumulative incidences with 95% CIs were estimated while using death as a competing risk. 38,39Patients (n ¼ 7) with identical follow-up times for developing a second cancer and death were included within the competing risk group.All statistical analyses over the entire study period (1989-2018)  were carried out separately for males and females.Analyses were further stratified by age at first cancer diagnosis (5-year age bands), follow-up period, and cancer type (first and second cancer).To allow meaningful interpretation, analyses were truncated at 25 years and outcomes were listed only if at least 10 second cancers were observed (all outcomes were listed for analyses by follow-up period).Statistical analyses were carried out with Stata/SE 17.0 (StataCorp LP, College Station, TX).Two-sided P values <0.05 were considered statistically significant.Study approval was granted by the Privacy Review Board of the NCR.Data used in this study can be requested from the NCR (study number K20.066, www.iknl.nl).

Sensitivity analysis
An underestimation of second cancer risk may arise when applying the international rules for multiple cancers in cases with a high frequency of consecutive malignancies that are considered topographically and morphologically identical.Due to this inherent limitation, we repeated all analyses by including all first two malignancies regardless of whether they satisfied the International Association of Cancer Registries (IACR)/IARC rules for multiple cancers.Other research parameters were kept the same.

Population and tumour characteristics
A flow chart of the study population selection procedure is presented in Supplementary Figure S1A

Overall SIRs and AERs
The overall risk of developing any second cancer after 25 years of follow-up was 2-fold higher in male survivors and 1.3-fold higher in female survivors compared to the general population.AERs of any second cancer were 17.5 and 10.1 per 10 000 person-years for male and female survivors, respectively (Figure 1, Supplementary Figure S3, available at https://doi.org/10.1016/j.esmoop.2023.102203,and Table 2).The overall higher cancer risk among AYA survivors maintained throughout the entire follow-up duration for all cancers combined (Supplementary Figure S4, available at https://doi.org/10.1016/j.esmoop.2023.102203).A closer look at survivors of specific cancer types showed that only male lymphoma and gastrointestinal tract carcinoma survivors had a higher second cancer risk throughout the entire follow-up period, whereas no such trend was observed among female survivors (Supplementary Table S2A, available at https://doi.org/10.1016/j.esmoop.2023.102203).A higher cancer risk compared to the general population was observed regardless of age at first cancer diagnosis, with SIRs peaking among those aged 15-19 years (3-fold risk for both sexes).AERs were highest among males aged 35-39 years at first cancer diagnosis, whereas this was the case for females aged 15-19 years.Cancer type-specific outcomes showed higher cancer risks among survivors of most of the 11 first primary cancer diagnostic main groups.SIRs in males ranged from 1.2 for melanoma survivors to 4.7 for blood and lymphatic vessel tumour survivors, whereas for the same cancers AERs ranged from 3.6 to 69.3 per 10 000 personyears.In females, SIRs ranged from 1.1 to 2.5 and AERs from 3.9 to 37.4 per 10 000 person-years for melanoma and lymphoma survivors, respectively.SIRs for the carcinoma diagnostic subgroups showed higher cancer risks after most first primary carcinomas among males, ranging from 1.7 to 4.0 for thyroid and other invasive carcinoma survivors, respectively.Meanwhile, female breast carcinoma survivors displayed a cancer risk comparable to the general population [SIR ¼ 0.9 (95% CI 0.9-1.0)],whereas higher cancer risks among female carcinoma survivors were observed otherwise (Table 2).

Second cancer type-specific SIRs and AERs
A higher cancer risk compared to the general population was observed for most second cancer types when grouping all first cancer survivors together.In general, male survivors had a lower second testicular cancer risk [SIR ¼ 0.6 (95% CI 0.5-0.8)],whereas a lower risk for second melanomas [SIR ¼ 0.9 (95% CI 0.8-1.0)]and breast carcinomas [SIR ¼ 0.8 (95% CI 0.7-0.9)]was found among female survivors (Supplementary Table S3A, available at https://doi.org/10.1016/j.esmoop.2023.102203).SIRs and AERs for a distinct range of first and second cancer combinations are displayed in Figure 1 (significant outcomes only) and Supplementary Tables S4A-S20A, available at https://doi.org/10.1016/j.esmoop.2023.102203.Second primary carcinomas were observed after nearly all first primary cancer types and most often included the gastrointestinal and lung carcinoma diagnostic subgroups.Second skin (males) and breast carcinomas (females) occurred after five of eight first primary cancer types (Figure 1 and Supplementary Tables S4A-S20A, available at https://doi.org/10.1016/j.esmoop.2023.102203).Male gonadal and related tumour survivors had a lower second primary testicular cancer risk [SIR ¼ 0.4 (95% CI 0.2-0.8)],whereas female breast carcinoma survivors had a lower risk of developing another breast carcinoma compared to the general population [SIR ¼ 0.0 (95% CI 0.0-0.1)].Comparable cancer risks between AYA cancer survivors and the general population were observed for several first and second primary cancer combinations, including (non-Hodgkin's) lymphoma risk among (breast) carcinoma and melanoma survivors (both sexes) and male gonadal and related tumour survivors.In females, comparable breast cancer risk was observed in first primary leukaemia [SIR ¼ 0.9 (95% CI 0.5-1.4)],gonadal and related tumour [SIR ¼ 0.9 (95% CI 0.6-1.2)],and central nervous system and other intracranial and intraspinal neoplasm survivors [SIR ¼ 1.0 (95% CI 0.5-1.7)].Likewise, breast cancer risk was comparable after most carcinoma diagnostic subgroups (Supplementary Tables S4A-S20A, available at https://doi.org/10.1016/j.esmoop.2023.102203).

Cumulative incidence
From 6 months after the date of first primary cancer diagnosis, the overall 5-year cumulative incidence of second cancer was 0.9% for male and 1.2% for female survivors and  Cancer types were grouped according to the AYA-specific classification scheme developed by Barr and colleagues (2020). 35Only cancer combinations with at least n ¼ 10 observed second cancers and significant risk estimates are presented.CNS, central nervous system.increased to 8.9% for male and 10.3% for female survivors at 25 years of follow-up (Figure 2 and Table 3).Age-specific cumulative incidences at 25 years of follow-up ranged from 3.4% among male and 7.1% among female survivors aged 15-19 years to 12.8% and 11.7% among male and female survivors aged 35-39 years, respectively (Figure 2 and Table 3).Five-year cumulative incidences of any second cancer remained below 8% for males and below 12% for females.At 25 years of follow-up, the cumulative incidence of second cancer was overall the highest for male lymphoma survivors (13.3%), followed by carcinoma survivors (11.5%), with the highest point estimates observed after first primary skin and other head and neck carcinomas.Among females, the highest 25-year cumulative incidence was observed among first primary lymphoma survivors (18.9%).Among female carcinoma survivors, highest cumulative incidences were observed after skin and other head and neck carcinomas (Table 3, and Supplementary Figures S5 and S6, available at https://doi.org/10.1016/j.esmoop.2023.102203).

Sensitivity analysis
A flow chart of the study population selection procedure for the sensitivity analysis is presented in Supplementary Figure S1B, available at https://doi.org/10.1016/j.esmoop.2023.102203.Outcomes of the sensitivity analyses are presented in Supplementary Figure S7 and Tables S1B-S22, available at https://doi.org/10.1016/j.esmoop.2023.102203.SIRs, AERs, and cumulative incidences from the sensitivity analyses were generally higher compared to those that were obtained after applying the IACR/IARC rules for multiple cancers, with overall SIRs now indicating a 3fold higher second cancer risk among male and a 2-fold higher risk among female survivors.AERs were, respectively, 32.5 and 39.5 per 10 000 person-years for male and female survivors, indicating a two times higher number of excess cancers in the sensitivity analyses in males and four times in females.Albeit still higher, outcomes by diagnostic group were more consistent between both analyses settings.However, SIR estimates obtained from the sensitivity analyses were noticeably two times higher among male testicular cancer, melanoma, and urinary tract carcinoma survivors.Testicular cancer survivors displayed a 10-fold higher risk of developing a second primary testicular cancer compared to the general population, whereas second melanoma risk was 13 times higher in male melanoma survivors.A noticeable 27 times higher risk of developing second kidney cancer was observed among male urinary tract carcinoma survivors (mostly kidney).SIRs also doubled for female melanoma survivors.Among female breast carcinoma survivors, a three times higher risk compared to the general population was observed in the sensitivity analyses, whereas risk was comparable in the main analyses.The risk of developing another breast cancer was four times higher in female breast cancer survivors compared to the general population, whereas female melanoma survivors had an 8fold higher second melanoma risk.Higher risk throughout the entire follow-up duration was now also observed for male melanoma and gonadal and related tumour survivors.In female survivors this was the case for head and neck, genital site, and breast carcinomas (Supplementary Tables S1B-S22, available at https://doi.org/10.1016/j.esmoop.2023.102203).
) 8.1 Superficial spreading/low cumulative sun damage melanoma 0.8 (0.5-1.Cancer types were grouped according to the AYA-specific classification scheme developed by Barr and colleagues (2020). 35

DISCUSSION
This study comprehensively investigated the long-term second primary cancer risk in 6-month AYA (15-39 years at first diagnosis) cancer survivors for both sexes in the Netherlands between 1989 and 2018.Male survivors overall had a 2-fold higher risk of developing any cancer at 25-year follow-up compared to the general population, whereas this was around 1.3-fold in females, resulting in 17.5 (males) and 10.1 (females) excess cancers per 10 000 person-years.Higher long-term cancer risks compared to the general population were observed for most first and second primary cancer combinations.The cumulative incidence of second cancer steadily increased during the 25-year follow-up period to 8.9% in male and 10.3% in female survivors.Nevertheless, there were some noticeable risk differences depending on whether the IACR/IARC rules for multiple cancers were applied.1][22][23][24] Nevertheless, direct comparison of outcomes between available studies is difficult due to the use of different age ranges (e.g.12-24 and 15-39 years), classification systems (e.g.Birch 40 and Barr 35 ), and applied primary cancer, multiple tumour, and latency period definitions.This is compounded by the use of different diagnostic periods, follow-up times, and general population definitions, which influences SIR and AER estimates by affecting the expected number calculation.
1][22] As in other studies, the highest long-term cumulative incidences were found among both male and female (Hodgkin's) lymphoma survivors. 229][20][21][22][23][24] Still, contradicting observations have been described and likely relate to the previously described methodological and cohort differences.For some specific combinations, this likely also relates to smaller sample sizes, which can inflate the resulting excess risk estimates and it is therefore generally advised to carefully interpret outcomes that are based on small sample sizes.
Despite falling outside the scope of most AYA papers, the literature suggests several intrinsic and extrinsic factors that (in composite) might explain the higher cancer risk among AYA cancer survivors. 42,43Depending on first cancer type, Chao and colleagues observed different second cancer risk factors. 21For instance, higher risk of solid malignancies among AYA survivors was independently associated with white race/ethnicity, receipt of radiotherapy, female sex, and older age and advanced stage at first cancer diagnosis. 21Female sex was only associated with increased risk when second breast cancers were included, whereas an inverted association was observed otherwise. 21Elsewhere, female sex was found to be a risk factor among teenage and young adult Hodgkin's lymphoma survivors, 44 and a protective factor among AYA melanoma 45 and thyroid cancer survivors. 46As stated by Chao et al., the observed risk factor variation between first cancer types might hint at different pathogenic mechanisms, but additional research is needed to ascertain this. 21Nevertheless, current findings do emphasize the importance of risk stratification by first primary cancer type.
9][50] Higher cancer risk after radiotherapy and chemotherapy exposure in Hodgkin's lymphoma survivors is also well-documented. 26,27,51Radiotherapy volumes and doses have become smaller over time, 52,53 but the introduction of more modern therapies does not appear to have resulted in a lower risk of second malignancies. 26,27,5424]55 Considering the disease prevalence, most previous studies included a young survivor population (aged 50 years), making it likely that the increased risk of second cancer among AYA lymphoma survivors is to some extent also treatment-induced.Nevertheless, investigation of risk estimates by first cancer treatment fell outside the scope of this and most other AYA-centric studies, warranting the need for more tumour-centric studies incorporating treatment exposure among AYA survivors.
Changes in radiotherapy and chemotherapy dose over time have resulted in lower breast cancer rates among childhood cancer survivors. 56Our main analyses showed female AYA survivors to have a noticeable lower second breast cancer risk, especially among first breast cancer survivors.However, this was not supported by the results from the sensitivity analyses, showing a 3-fold higher breast cancer risk among female AYA survivors compared to the general population.In the sensitivity analyses, a higher risk for developing another breast cancer was also found among AYA breast cancer survivors, which is in line with the substantial higher same site breast cancer risk that has been reported by previous studies. 21,24he differences in breast cancer risk between our main and sensitivity analyses most likely relate to the IACR/IARC rules for multiple cancers, which were applied in the main analyses and exclude all consecutive malignancies with identical topography and morphology.This might also explain why male gonadal and related tumour survivors in the main analyses had a lower second testicular cancer risk compared to the general population, whereas outcomes from the sensitivity analyses indicated a marked 10-fold higher risk.Similar high melanoma risk outcomes were observed among melanoma survivors (both sexes), whereas this risk combination was not available when the IACR/IARC rules were applied.Still, the outcomes of the sensitivity analyses are likely biased and it should be noted that other studies have reported a 5-38 times higher risk of developing second testicular cancer (different laterality) among AYA testicular cancer survivors whilst applying multiple cancer rules, including the IACR/IARC rules. 21,240][61] BRCA1 and BRCA2 mutation carriers typically have increased breast, ovarian, uterine, and pancreatic cancer risk. 59,62Breast cancer survivors in this study also had a higher risk of developing second ovarian and pancreatic cancers, which might hint at a shared genetic predisposition pattern.
Modifiable lifestyle factors likely also play an important role in second cancer development.Smoking is a wellknown risk factor for developing cancer in general and has been shown to increase the risk of second lung and other tobacco-related cancers (e.g.4][65][66] Previous findings show that AYA survivors (especially females) more often smoke compared to a similar population without cancer (33% versus 22%) and that ever (those who had regularly smoked before or at diagnosis) and current smokers and drinkers at the time of first primary cancer diagnosis in general had a higher second cancer risk compared to those who never smoked or consumed alcohol regularly during their lifetime. 63,67,68here is also evidence that second cancers are biologically different compared to first primaries of the same type. 29,46,69As such, better survivorship care guidelines and targeted interventions aimed at prevention and early detection of second cancers among AYAs are desperately needed, especially among radiotherapy recipients.Less invasive radiotherapy and chemotherapy regimens have resulted in lower cancer rates among childhood cancer survivors. 56Safer treatments that minimize toxic effects may also lower the long-term subsequent cancer risk among AYA survivors and should be explored considering that AYAs have many life years remaining.Results from this and related studies could help conceptualise such guidelines for AYA cancer survivors, which was also suggested elsewhere. 21Nevertheless, more studies investigating (modifiable) risk factors for the specific cancer type combinations among AYA survivors are required to best inform and further guide the development of age-appropriate survivorship care guidelines, as detailed information about second cancer risk factors among AYAs remains scarce, especially about treatment dose/exposure and modifiable factors like obesity, smoking, and physical activity. 43his study has several strengths, including the generalizability of outcomes to the entire Netherlands population by using data from the population-based NCR, which contains complete records of all malignant cancers in the Netherlands since 1989 and has a near 100% coverage. 32Use of these data also minimizes selection bias, which is another major strength.Furthermore, we included all first and second cancer types based on observed numbers rather than making a prior selection, as was done in most previous studies.Outcomes in this study were also presented for males and females separately, resulting in a detailed risk overview in general and for distinct diagnostic subgroups.Adding to this, we reported both SIRs and AERs, whereas previous studies mostly reported SIRs.Self-contained interpretation of relative measures like the SIR is difficult, as they can easily inflate in a low number setting, providing no guidance to health services.
This study also has several limitations.Firstly, cancers with a non-malignant behaviour were not taken into consideration, but their existence and treatment may have affected the second cancer risk profile within our cohort.Furthermore, we included AYA 6-month survivors, whereas other studies usually adopted a 5-year latency period to not misclassify recurrences and metastases as new primaries.Albeit considerably shorter, a recent study into the most appropriate period to define synchronous cancers showed 4 months to be sufficient. 34As confirmed by our sensitivity analyses, application of the IACR/IARC rules resulted in a noticeable second cancer risk underestimation in melanoma, testicular, and breast cancer survivors, all of which had a high frequency of developing subsequent malignancies with identical topography and morphology [n ¼ 125 (20.2%) testicular, n ¼ 155 (25.1%) melanoma, and n ¼ 278 (45.0%) breast cancers].Our general background population likely amplified this effect by including all malignancies, including those that were excluded from the survivor population based on the IACR/IARC rules.Although risk outcomes remained similar in most cases otherwise, researchers should carefully consider how to best define their survivor and background populations and in all cases provide a clear description of this process when investigating second cancer risk.Lastly, despite being available in the NCR, treatment fell outside the scope of this study due to the large number of detailed combinations of diagnostic subgroups.Considering that treatment can vary greatly per cancer type, we believe that it is better to address this topic in more tumour-centric studies.
In conclusion, this study demonstrated that AYA cancer survivors in the Netherlands have a higher cancer risk compared to the general population for most first and second primary malignant cancer combinations up to 25 years after their initial diagnosis between 1989 and 2018.Considerable risk variation was observed for both sexes and between cancer types, but interpretation requires some caution due to low numbers and underestimation of risks caused by the IACR/IARC rules.Nevertheless, these findings highlight the need for personalized follow-up strategies.Additional studies that investigate risk factors for the specific cancer type combinations are needed to best inform and develop such tailored survivorship care guidelines for AYA cancer survivors.

Figure 1 .
Figure 1.Standardised incidence ratios (SIRs) and absolute excess risks (AERs) of first and second primary malignant cancer combinations diagnosed among 6month adolescent and young adult (AYA, aged 15-39 years) cancer survivors compared to the general population in the Netherlands between 1989 and 2018.Cancer types were grouped according to the AYA-specific classification scheme developed by Barr and colleagues (2020).35Only cancer combinations with at least n ¼ 10 observed second cancers and significant risk estimates are presented.CNS, central nervous system.

Figure 2 .
Figure 2. Cumulative incidence of any second primary cancer diagnosed up to 25 years after first primary cancer among 6-month adolescent and young adult (AYA, aged 15-39 years) cancer survivors in the Netherlands between 1989 and 2018.Outcomes are presented for males and females overall and by age group at first primary cancer diagnosis.Death of any cause was included as a competing risk event.

Table 1 .
Population, tumour, and treatment characteristics of 6-month AYA (aged 15-39 years) cancer survivors diagnosed with first and second primary malignant cancer in the Netherlands between 1989 and 2018 35ncer types are grouped according to the AYA-specific classification scheme developed by Barr and colleagues (2020).35AYA,adolescent and young adult; CNS, central nervous system; FIGO, Fédération Internationale de Gynécologie et d'Obstétrique; IQR, interquartile range; NA, not applicable; TNM, tumourenodeemetastasis. a Percentages might not add up to 100% due to rounding.b Age at diagnosis of first and second primary cancer.c Years of follow-up from 6-month survival until the date of second cancer diagnosis, death, loss to follow-up, or 31 December 2018, whichever came first.D. J. van der Meer et al.

Table 2 .
SIRs and AERs of any second primary malignant cancer diagnosis after first primary malignant cancer among 6-month AYA (aged 15-39 years) cancer survivors compared to the general population in the Netherlands 35ncer types are grouped according to the AYA-specific classification scheme developed byBarr and colleagues (2020).35Cancercombinationswithless than n ¼ 10 observed second cancers were excluded from the analyses.AERs, absolute excess risks; AYA, adolescent and young adult; CNS, central nervous system; Exp, expected number of second cancers; FIGO, Fédération Internationale de Gynécologie et d'Obstétrique; NA, not applicable; NET, neuroendocrine tumour; NK, natural killer; NLP, nodular lymphocyte predominant; NOS, not otherwise specified; Obs, observed number of second cancers; TNM, tumourenodeemetastasis. a Age at diagnosis of first primary cancer.D. J. van der Meer et al.ESMO OpenVolume 9 -Issue 1 -2024 https://doi.org/10.1016/j.esmoop.2023.102203

Table 3 .
Cumulative incidence of any second primary cancer up to 25 years after first primary cancer among 6-month AYA (aged 15-39 years) cancer survivors in the Netherlands between 1989 and 2018 Death of any cause was included as a competing risk event.AYA, adolescent and young adult; CNS, central nervous system; FIGO, Fédération Internationale de Gynécologie et d'Obstétrique; NA, not applicable; NET, neuroendocrine tumour; NK, natural killer; NLP, nodular lymphocyte predominant; NOS, not otherwise specified; TNM, tumourenodeemetastasis. .J. van der Meer et al.
a Age at diagnosis of first primary cancer.D