Analysis of Cause-Specific Mortality in Patients with Retinoblastoma

Background Retinoblastoma (RB) is a rare pediatric tumor with a relatively favorable prognosis. However, RB is associated with cause-specific mortality, some of that should be of great importance to clinicians. In this study, we summarize the characteristics of cause-specific mortality from nontumor disease in patients with RB. Methods This retrospective case series study identified and analyzed cause-specific mortality in patients with RB. The information of cause-specific mortality of RB patients, including detailed clinical characteristics, diagnosis, treatment process, cause-specific mortality classification, and lag time, was assessed. Results A total of 12 eligible patients were selected from 264 patients who died among 3780 patients diagnosed with RB. The cause-specific mortality rate was 4.5% for all patients with RB who died and 0.3% for all patients with RB. The main nontumor cause-specific mortalities were diseases of the nervous, circulatory, and respiratory systems, which specifically included intracranial infection, cerebral hemorrhage, paraplegia, and respiratory failure. The longest lag time was 42 days from the last chemoradiotherapy or surgery. Conclusion Nontumor cause-specific mortality is an essential outcome of RB. Thus, intensive care and differentiation during management need to be taken seriously.


Introduction
Retinoblastoma (RB) is a rare ocular tumor. However, it is the most common intraocular tumor in childhood. It is initially caused by a tumor-suppressor mutation of RB1 [1]. Early diagnosis and multifarious surgery or chemoradiotherapy sharply decrease the mortality rate of patients with RB. At present, comprehensive treatment of RB involves administration of intravenous chemotherapy (IVC), intraophthalmic artery chemotherapy (IAC), and periocular and intravitreal techniques [2]. IVC and IAC in particular have been found to be safe and efficient for the treatment of RB. Both therapies have been reported to gradually preserve the eye and increase the five-year survival rate of patients with RB [3,4]. Meanwhile, a combination of IAC and intravitreal chemotherapy has higher success rates and lower complication rates [5]. However, patients with RB still have a high risk of developing subsequent primary malignancies or other systemic diseases, which could be life-threatening [6,7].
Previous studies have shown that the cause of mortality of patients with RB is related to subsequent malignancies and other cause-specific mortality, including infections, endocrine and metabolic diseases, mental disorders, neurological diseases, circulatory diseases, respiratory diseases, and digestive system diseases (according to the International Classification of Disease (ICD) codes) [6,8]. Subsequent malignancy has been reported to be a greater risk factor among hereditary retinoblastoma survivors and is associated with radiotherapy [9,10]. However, the nontumor causespecific mortality still presents a potential risk for patients with RB [1,11]. e aim of this study was to analyze causespecific mortality in patients with RB, especially those with other systemic nontumor diseases and summarize their clinical characteristics. We emphasized nontumor causespecific mortality in patients with RB to raise awareness of clinical treatment and survival of RB and help decrease the mortality rate of RB.

Methods
In this retrospective case series study, we identified and analyzed cause-specific mortality in patients with RB that was initially diagnosed and treated between 2006 and 2019 at the Beijing Children's Hospital and Beijing Tongren Hospital. All procedures were performed, and written informed consent was obtained in accordance with the relevant guidelines and regulations of the Declaration of Helsinki.
We collected the detailed information of the patients with RB from electronic medical records with standardized forms of basic clinical information, detailed diagnostic and treatment processes, details of the last therapy, and detailed time, circumstances, and classification of cause of mortality. e cause mortality of patients was classified according to the ICD, and the cause-specific mortality was defined as the mortality from other systemic nontumor diseases, excluding mortality due to tumors or metastases.
All patients with RB were clinically diagnosed based on the presence of abnormal ocular symptoms, including leukocoria, vision loss, or hypopyon, and findings of detailed fundus examination, computed tomography, or magnetic resonance imaging. Clinical disease stages at the time of diagnosis and treatment period were categorized according to the standards of the International Intraocular Retinoblastoma Classification (IIRC). All patients were treated using standard treatments, including IVC or IAC, radiotherapy, and surgical enucleation.
Lag time, which is a relatively life-threatening period for RB patients, was defined as the duration from the end of the last therapy to the time of death. Diagnostic delay time was defined as the duration from the detection of the first sign to the time of diagnosis. Treatment delay time was defined as the duration from the time of diagnosis to the initiation of primary treatment. Survival time was defined as the duration from the time of diagnosis to the time of death. All causes of death were considered classified variables, whereas the time of diagnosis, therapy, and death were considered numeric variables. e average diagnostic delay time, death time, treatment delay time, survival time, and lag time were calculated and summarized.

Results
A total of 12 eligible nontumor cause-specific mortality patients were selected from 264 patients who died among 3780 patients diagnosed with RB and included in this study. e cause-specific mortality rate was 4.5% for all patients who died and 0.3% for all patients with RB. e eligibility evaluation flow diagram ( Figure 1) clearly showed the 12 cause-specific mortality RB patients selected for participating in this study.
e clinical information of the selected 12 patients with RB is displayed in Table 1. e 12 patients were equally distributed in terms of sex and laterality of diagnoses. Leukocoria was the main initial symptom, followed by vision loss and hypopyon. Most patients visited our hospital for the first time, two patients were referred to Beijing Children's Hospital, and one patient was treated in another hospital after consultation in our hospital. Table 2 shows the specific clinical information of 12 cause-specific mortality patients with RB. Regarding RB stages at the time of diagnosis, six cases of unilateral RB were in stages D, E, and E+ at the time of diagnosis, whereas six cases of bilateral RB were in stages B/E, C/D, D/D, ?(unclear)/E, and D/E+. Regarding RB stages at the last evaluation, the six cases of unilateral RB were in stable stage E, chemotherapy duration stages D and E, and enucleation stages E, E, and E+, whereas the six cases of bilateral RB were in the stable/enucleation and chemotherapy duration stages.
Regarding the therapy processes of the patients, nine patients in group D or E initially underwent IVC therapy, carboplatin, etoposide, vincristine (CEV), or carboplatin, teniposide, vincristine (CTV) on an average of 2.7 (1-7) times and subsequently accepted enucleation, radiotherapy, or intraocular laser therapy; one patient in group D with bilateral low vision was initially treated with IAC two times; and two patients in group E initially underwent enucleation and subsequently accepted several sessions of IVC. However, although five of these patients (lines 1, 2, 6, 7, and 11 in Table 2) who underwent enucleation or systematic IVC were evaluated as stable and were considered to should survival, the cause-specific mortality still happened after their last regular treatment.
We summarized the nontumor cause-specific mortality classification, last therapy, and lag time of the 12 patients with RB (Table 3). e cause-specific mortality of the included patients was divided into mortality from diseases of the nervous, circulatory, and respiratory systems, specifically intracranial infection (ICD-G06.003), paraplegia (ICD-G82.205), cerebral hemorrhage (ICD-I61.902), and respiratory failure (ICD-I96.051). Seven patients died of nervous system disease: five of the patients had intracranial infections, whereas two had paraplegia an average of 7.8 and 20.5 days after

Discussion
In this paper, we focused on the cause-specific mortality of 12 patients with RB and summarized their details of basic clinical information, diagnoses, treatments, and mortality. We specifically analyzed cause-specific mortality from other systemic nontumor diseases, including diseases of the nervous, circulatory, and respiratory systems, especially intracranial infection, paraplegia, cerebral hemorrhage, and respiratory failure.
In this study, a total of 12 cause-specific mortality patients with RB from other systematic nontumor diseases    [12] reported cause-specific mortality due to nontumor disease in 2 out of 43 patients who died in a cohort of 633 patients with RB in Germany. Yu et al. [6] reported cause-specific mortality from nontumor disease in 39 out of 211 who died during a one-year follow-up study in the United States (US). Waddell et al. [13] revealed that among 108 deaths recorded in a cohort of 270 patients with RB in Uganda, one case of leucopenia after chemotherapy may be due to infection. Broaddus et al. [3] identified three cases of septicemia mortality among 63 patients in the SEER database who died. Kleinerman et al. [8] found 98 cases of cause-specific mortality from nontumor disease among 690 deaths recorded in a long-term follow-up from 1914 to 2016 in the US. Marees et al. [14] summarized 99 cases of cause-specific mortality from nontumor disease, mostly circulatory or cardiovascular disease, among 332 patients during a longterm follow-up from 1862 to 2005 in the Netherlands. e infectious mortality rate for patients with RB recorded in these previous studies ranged from 0.7% to 4.8% [3,6,8,12,13], whereas the cerebral hemorrhage mortality rate was 1.5% [14]. e infection mortality rate in our present study was 1.9% and the cerebral hemorrhage mortality rate was 1.1%, whereas the total nontumor causespecific mortality rate was 4.5% and 0.3% for all patients who died and all patients with RB, respectively.
Cause-specific mortality due to nontumor disease occurs during the course of RB. erefore, increased attention should be paid to the occurrence of nontumor diseases in cases of RB. In this study, we determined the lag time from the end of the last therapy to death, which is a vital and risky period for patients with RB. During this period, doctors need to pay attention to patient care from the aspects of chemotherapy, radiotherapy, and surgical management. Regarding the perioperative period of regular chemotherapy or enucleation, close monitoring of patients with RB for at least 19 days is essential to get through the lag time. For radiotherapy, we suggest strengthening intensive care for more than 40 days after the therapy to expedite the effects of the treatment and reduce possible damage due to radiation [15,16]. e most common systemic complications of IVC and IAC are neutropenia, transient fever, and nausea/vomiting, whereas the most common ocular complication is retinal detachment [17][18][19][20]. Researchers have proved that increased frequency of chemotherapy, drug dose accumulation, and previously failed treatment are independent risk factors for ocular motility complications during the treatment of RB [11,21], which may be hazardous to patients. ere are still some accidents and complications that occur during the common treatment of RB that alert clinicians to differentiate the preaccident mortality symptoms. In the present study, all cases of intracranial infection and cerebral hemorrhage mortality occurred after specific IVC or IAC sessions performed for approximately two weeks; other symptoms such as paraplegia and respiratory failure were also observed. ese specific symptoms indicate that more attention should be paid to the condition of patients with RB during or after treatment. In addition, physicians should be alert to the occurrence of cause-specific mortality and consider specific perioperative management, especially for patients with stage D/E disease or those who underwent multiple chemoradiotherapy sessions.
Tumor therapy has advanced over the years; however, the safety and efficiency of novel and traditional therapies still need to be evaluated. Although the new method of IAC is as efficient as traditional IVC or enucleation in some specific cases of RB [20], vascular toxicity and local complications are curtailed with IAC. However, the proper dose of IAC is unpredictable for complex and variable conditions [22]. An increasing number of novel therapies, such as focal lasers [23] and intravitreal chemotherapy [24] have emerged as therapeutic options for RB. Furthermore, researchers have found that a white blood cell count lower than 1 × 10 9 strongly increases the risk of infection. Some clinicians administer recombinant human granulocyte colony-  stimulating factor (rhG-CSF) on the fourth day after regular chemotherapy, which may provide a protective benefit against infection-related accidents. However, it has been suggested that rhG-CSF be used only when anticipating long-term agranulocytosis. Preventing cause-specific mortality is essential for patients with RB who are judged to have chances of survival. In the present study, 5 of the 12 included patients were in the stable stage at the last evaluation and were considered to should survival after appropriate treatment (lines 1, 2, 6, 7, and 11 in Table 2). As the multifarious treatment advanced, we should still take the original and intensive care for RB therapy period and differentiate the severe symptom avoiding the cause-specific mortality from nontumor disease. In this study, the average ages of the 12 patients at the time of diagnosis and death were relatively young. In addition, delays in diagnosis and treatment were observed in this study (Table 4). e average survival time (6 months) of the included patients was much shorter than that of the overall patients with RB. Proper diagnosis and appropriate timing and planning of treatment may also play essential roles in the prognosis of RB and the prevention of nontumor cause-specific mortality [25,26].
One limitation of the present study was information bias, which selected cases based on the hospital's electronic medical record system in the regional center in China and limited its generalization. Besides, the present study does not assess causal relationship for cause-specific mortality in RB patients, and thus, more prospective collaboration in multicenter studies with longer follow-up is needed to generate higher-quality evidence in the future.

Conclusion
RB could have a good prognosis after comprehensive treatment; however, cause-specific mortality from nontumor diseases still needs to be considered. e most common cause-specific mortalities in present cases of RB are nervous, circulatory, and respiratory system diseases. Intensive care and differentiation are essential during the lag time from the last chemoradiotherapy or surgery.

Data Availability
e clinical information data used to support the findings of this study are available from the corresponding author upon request.

Consent
All methods and the written informed consent were performed based on the Declaration of Helsinki.

Disclosure
Lei Yang and Xiaolian Fang are the co-first authors.

Conflicts of Interest
e authors declare that they have no conflicts of interest.