Acute leukemias in pregnant women: Results of a retrospective study at a local tertiary‐care hospital in Japan

Abstract Leukemia may rarely develop in a woman during pregnancy, posing clinical challenges to the patient, fetus, family, and medical staff managing malignancy and pregnancy. We retrospectively analyzed cases of pregnancy‐associated leukemia consecutively diagnosed and treated at a local tertiary‐care hospital in Nagano, Japan, over the past 20 years. Five cases were identified among 377,000 pregnancies in the area (one in every 75,000 pregnancies), all involving acute leukemia (three acute myelogenous leukemia [AML] and two acute lymphoblastic leukemia [ALL]). The cases were diagnosed in the first trimester (n = 1), second trimester (n = 3), or third trimester (n = 1). There were no apparent pregnancy‐associated delays in diagnosing and treating the cases. Three patients underwent induction chemotherapy during pregnancy, two of whom eventually delivered healthy babies. One of the five patients chose abortion before chemotherapy initiation. Two cases showing high‐risk features at the diagnosis (AML with an FLT3‐ITD mutation [n = 1] and relapsed ALL [n = 1]) eventually died despite consolidative allogeneic hematopoietic stem cell transplantation. Our results suggested that patients with pregnancy‐associated acute leukemia can be treated similarly to nonpregnant patients, although pregnancy imposes particular clinical challenges that should be resolved with multidisciplinary care.

Furthermore, such clinical dilemmas are exacerbated when a patient presents with acute rather than chronic leukemia, as it will likely result in both maternal and fetal mortality in a relatively short period of time if left untreated [4,13,14]. In addition, while delaying the induction chemotherapy may negatively impact the likelihood of remission, patients and their families still need ample time and sufficient support to make their treatment decisions [4,11].
Despite the above-mentioned issues, a standard approach to addressing this clinical dilemma has not yet been established, partly due to the wide range of personal beliefs among patients [15]. Furthermore, the rarity of such cases also precludes large-scale prospective controlled trials, and many recommendations in clinical guidelines regarding the management of pregnancy-associated leukemia are derived from expert opinions [11,16]. Although it has been suggested that standard chemotherapeutic intervention may help achieve a similar maternal survival rate to that in nonpregnant leukemic patients, a tailor-made approach is mandatory to resolve potential and actual problems a pregnant patient may have to face, such as adverse effects on the pregnancy outcome, long-term effects on the fetus, and impairment of the patient's future fertility [4].
We herein report the fetal and maternal outcomes as well as the dilemmas and their resolution in consecutive cases of pregnancyassociated leukemias diagnosed and treated at a local tertiary-care hospital the past 20 years. Our descriptions depict our perspective concerning rare but very challenging situations hemato-oncologists may have to confront.

Patients
Patients with pregnancy-associated leukemia who were referred to and treated at the Department of Hematology and/or Department of Gynecology and Obstetrics of Shinshu University Hospital in Nagano, Japan, between 2001 and 2022 were enrolled. There were 44 medical offices and hospitals offering obstetrics services in Nagano Prefecture as of 2014 [17]. When a pregnant woman presents with a severe hematological abnormality, such as leukemia, in Nagano Prefecture, the patient is generally referred to Shinshu University Hospital, the only tertiary-care hospital in the prefecture offering both obstetric service for high-risk pregnancy, neonatology services, and adult hematologyoncology services. The cohort of patients in the present study may therefore closely reflect the total incidence of pregnancy-associated leukemia in this prefecture. Relevant clinical data on leukemia as well as the pregnancy and fetus were retrieved from patients' medical records.

The diagnosis and treatment
The indication and procedures for bone marrow biopsies were the same in pregnant women as in patients without pregnancy when a leukemia was suspected. The diagnosis and classification of leukemia were confirmed in accordance with the WHO Classification Tumours of Haematopoietic and Lymphoid Tissues, 4th edition 2017 [18].
Radiographic evaluations, such as X-ray and computed tomography (CT), were performed if necessary for differential diagnoses and evaluating complications, but all efforts were made to minimize potential radiologic exposure to the fetus. Cumulative anthracycline exposure dose was calculated by adding the doxorubicin-equivalent dose of each anthracycline received. The conversion factors used to calculate doxorubicin equivalent were 1.0 for daunorubicin and 4.0 for mitoxantrone [19].
All patients were treated at the Department of Hematology, and their pregnancies were managed at the Department of Gynecology and Obstetrics of Shinshu University Hospital.

Statistical analyses
Vital statistics were obtained from a governmental database freely available online (e-Stat), a portal site of official statistics of Japan managed by the Statistics Bureau [20]. The data included the annual local population as well as the total number of pregnant women in Nagano Prefecture. The incidence of leukemia was referenced from the Cancer Statistics issued by the National Cancer Center, Japan, a registry dataset also freely available online [21].

Ethical considerations
Leukemia patients who did not agree to be included in the present study were excluded from the cohort using an opt-out method. This study was approved by the institutional review board of Shinshu University School of Medicine (approval number: 5756, February 14, 2023) and conducted in accordance with the Declaration of Helsinki.

Patients
We identified and enrolled five patients with pregnancy-associated acute leukemia during the study period: three AMLs and two ALLs ( F I G U R E 1 Clinical courses of pregnancy in patients with pregnancy-associated leukemia are depicted. Each horizontal black bar represents a patient, with four time-points, each indicating the day of diagnosis of acute leukemia (red circle), the day of initial presentation (white triangle), the day of initiation of induction chemotherapy (yellow rectangular), and the day of delivery or abortion (blue triangle). The colored bar at the top represents three trimesters of pregnancy.
According to the e-Stat, there were 377,000 pregnancies in Nagano Prefecture, Japan, between 2001 and 2022, so we estimated the incidence of pregnancy-associated acute leukemia to be 1.33/100,000 pregnancies, or 1 in 75,000 pregnancies in Nagano (Supplemental Table). Nagano Prefecture had a population of two million, including 244,000 women between 20 and 44 years old as of 2019 [20]. Registry data retrieved online showed that the incidence of leukemia, including chronic and acute cases, among the general female population between 20 and 44 years old varied with age between 2.5/100,000 and 4.5/100,000 in 2019 [21].

Management of pregnancies and neonatological care
Leukemia was diagnosed in the first trimester in one patient, second trimester in three patients, and third trimester in one patient ( Figure 1).
The patient whose gestation was in the first trimester (UPN#4) selected abortion before the initiation of anti-leukemic treatment as opposed to continuing the pregnancy throughout the treatment. The three patients in the second trimester proceeded with the induction chemotherapies with the intention to continue their pregnancy until the gestational week of 32 or longer, in order to avoid high neonatal morbidity and mortality associated with respiratory immaturity, which would be expected in cases of very preterm birth earlier than 32 gestational weeks [22]. Intrauterine chemotherapeutic exposure of the fetus was eventually observed in 3 patients ( Table 2)

Management of leukemias
The median (range) duration from the initial presentation to the diagnosis of leukemia was 7 (0-28) days, and the median (range) duration from the diagnosis to the start of cytotoxic chemotherapy

Management of perinatal dilemmas
Multidisciplinary discussions and consultations were frequently held as needed for each case to identify and resolve medical problems as well as ethical dilemmas, such as conflicts of interest between the mother and fetus [24]. Supportive talk sessions with the patients and their families were repeated as requested, even after critical informed consent was obtained. Any disagreement on treatment plans among physicians and other medical staff were resolved through discussions.
Detailed descriptions of representative perinatal dilemmas identified in our cohort of pregnancy-associated leukemia and a result of multidisciplinary discussions to resolve the dilemmas are summarized in the Appendix.

DISCUSSION
We identified a total of five cases of pregnancy-associated acute leukemia occurring in Nagano Prefecture in the past 20 years, suggesting an estimated incidence of 1.33/100,000 pregnancies or 1 in 75,000 pregnancies in this region. We did not find any cases of adult T-cell leukemia (ATLL) with pregnancy in our cohort, largely because the prefecture is located in less-endemic area for HTLV-1 [25,26].
It is generally accepted that, unlike ATLL, geographical distribution of other acute leukemias is not significantly different among prefectures in Japan, so our results may accurately reflect the incidence of pregnancy-associated acute leukemia in Japan as a whole. This frequency, as well as the ratio of myeloid and lymphoid leukemia, was comparable to the findings in previous studies conducted overseas [2].
Our results also suggest that pregnancy-associated acute leukemia is rare in Japan and that pregnancy was not associated with an increased frequency of leukemia.
The small size and short observation period of our study hamper the drawing of any hard conclusions concerning the prognostic impact of pregnancy on the clinical course of leukemia. However, it may be worth mentioning that we did not note any significant delays in the diagnosis or initiation of treatment of leukemia due to pregnancy itself.
In a previous study of malignant lymphoma during pregnancy, the start of chemotherapy was deferred in approximately 40% of cases [27]. Our findings may reflect the fact that any delay in the start of induction therapy for acute leukemia may negatively impact on both the patient and their fetus [4,14] and may also reflect the urgent nature of this clinical situation. Our results may also confirm that a standard anti-leukemia treatment strategy may be carefully sought even during pregnancy [4,11,13,15].

FUNDING INFORMATION
The authors received no specific funding for this work.

CONFLICT OF INTEREST STATEMENT
The authors have no conflict of interest to be disclosed.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

ETHICS STATEMENT
Informed consent of participating patients was obtained using an optout method. This study, including this opt-out method, was approved by an institutional review board of Shinshu University School of Medicine, approval number 5756, approval date February 14th, 2023. And this study was conducted in accordance with the Declaration of Helsinki.
The statement of IRB approval is described in the Method section. Discussions were also frequently held with obstetricians in different occasions. Informal discussions among these staff, as a curbsideconsultation manner, were also held countlessly throughout the clinical course of the patient.
All team members were expected to share the pertinent clinical information of the patient, regarding the diagnosis, treatment options, expected results, and adverse effects in each option, expected prognosis in each option, patient's character and preferences, social background of the patient and available support from her family members. Social workers also input necessary information to the team members at their request. Any concerns or dilemmas were encouraged to be addressed by team members, either formally during a conference or informally during a curbside, and we tried to identify medical, ethical or emotional issues behind the concerns in straightforward terms.
Our multidisciplinary discussions concluded that one of the medical challenges as ethical dilemmas we encountered in this case was a conflict of interest between the expectant mother and the fetus [24].
It is generally accepted that suboptimal treatment of the mother's condition may harm the fetus as well as the mother. The treating physicians acknowledged the concerns about the risks the high-dose consolidation therapy posed to the fetus but eventually concluded that such concern for the fetus should not lead to inappropriately withholding more-effective treatment from the mother. Furthermore, the patient and her family understood and consented to the advice.
There was, however, an equally compelling question from some staff members about whether or not the benefit of pursuing a better DFS rate of 57%, instead of 39%, without a significant OS benefit was worth risking the fetus, although we all agreed that the very-high-dose chemotherapy was not the sole cause of the fetal distress. All staff members also agreed that there would not be a single perfect answer to solve this clinical dilemma, and that repeated multidisciplinary discussions and shared-decision-making process with patient and family is crucial.
This case may be representative of the clinical challenges facing patients with pregnancy-associated acute leukemia and the multidisciplinary teams treating such patients in similar situations. Our results suggested that repeated discussions among staff members as well as the patient can help resolve such lingering misgivings.