Regression of Neonatal Cardiac Rhabdomyoma in Two Months Through Low-Dose Everolimus Therapy: A Report of Three Cases
- 1k Downloads
Cardiac rhabdomyoma (CR) is the most common cardiac tumor in newborns. Approximately 75% of cases are associated with tuberous sclerosis complex. Although these tumors usually spontaneously regress after 2 years of age, they can be life-threatening when they obstruct major cardiac inflow or outflow pathways. Everolimus is an inhibitor of the mammalian target of rapamycin, reducing its production of the proteins harmartin and tuberin. Everolimus has demonstrated a remarkable suppression effect in children with tuberous sclerosis complex at doses of 4.7–5.6 mg/M2/day and serum trough levels of 5–15 ng/mL. Since 2012, five case reports of neonates with CR have also reported the tumor-regressing effect of everolimus. However, the optimal dosage for neonates is still unknown. Over the past 2 years, we have deliberately used a low dose everolimus regimen (0.3–0.67 mg/M2/day) in three neonates with large CRs, in an effort to maintain serum trough levels at 3–7 ng/mL. In all three cases, the tumors regressed smoothly within 2 months. Regarding the drug’s side effect of predisposing patients to infection, we observed that adenovirus pneumonia occurred in one case at 3 months of age, and chicken pox occurred in another case at 9 months of age; both recovered smoothly. Our three cases of neonatal CR demonstrate that a low-dose everolimus regimen is an effective treatment for tumor regression.
KeywordsCardiac rhabdomyoma Everolimus Neonate Side effect
Cardiac tumors occur in only 0.027% of children, and cardiac rhabdomyoma (CR) comprises 45% of these occurrences . Approximately 70–90% of CR patients also present with cerebral tuberous sclerosis, exhibiting a phenotype of tuberous sclerosis complex (TSC) . Pathologically, CR is a type of hamartoma composed of striated muscles that usually follows a benign course of spontaneous regression after 2 years [1, 3]. However, when a CR severely obstructs the cardiac inflow or outflow tract of a neonate, urgent cardiac surgery to remove the tumors must be considered [1, 2, 3, 4, 5]. Since 2012, five case reports have been published in the relevant English literature that present successful cases of using everolimus, an inhibitor of the mammalian target of rapamycin (mTOR), to trigger tumor regression in neonatal patients with CR [6, 7, 8, 9, 10]. These case reports offer hope of nonsurgical survival among high-risk patients. However, the proper dosage and duration of using everolimus in neonates, and knowledge about avoiding its side effects are largely unknown . Over the past 2 years, we have observed three neonates prenatally diagnosed with CR, including one with total left ventricular obstruction. Attempting to reduce the drug’s potential side effects as much as possible, we used a lower-than-recommended-dosage regimen of everolimus on all three patients.
After providing an explanation to the parents, we began an off-label use of oral low-dose everolimus therapy (0.0625 mg/day, 1/4 of a 0.25 mg tablet, equivalent to 0.334 mg/M2/day), aiming to achieve a trough serum everolimus level of 3–7 ng/mL. Her cardiac function was also supported with digoxin and furosemide. We monitored the complete blood counts, renal and liver functions, and serum everolimus levels once a week. After the first week, the everolimus level was 3.50 ng/mL. We escalated the dosage to 1/3 tablet a day (0.427 mg/M2/day) for the second week, and 1/2 tablet a day (0.558 mg/M2/day) for the third week. Echocardiography revealed a remarkable reduction of tumor size. With the same dose, she was discharged on the 17th day of therapy. By the 50th day of therapy, the LV tumor had reduced to 0.29 cm2 surface area, approximately one-eighth of the original size (Fig. 1b). However, the everolimus serum level unexpectedly rose to 20 ng/mL (Fig. 4) and we tapered the dosage back to 1/4 tablet a day (0.241 mg/M2/day). The serum level remained high (14.18 ng/mL) on the 80th day of therapy, so we further decreased the dosage to 1/8 tablet a day (0.12 mg/M2/day). Two weeks later, she suffered fever, cough and respiratory distress. Everolimus was withheld by the parents, and she was readmitted to our ward. Her weight (3.9 kg), height (55 cm) and head girth (39 cm) were all below the third percentile for an average 3-month-old female infant. The chest roentgenogram showed an interstitial type of pneumonia, and adenovirus was isolated from her throat. Because the everolimus serum level had become immeasurable (<1.5 ng/mL), we suspended the therapy. She was discharged after 1 week. Later, follow-up echocardiograms noticed a slow but steady increase in the tumor size, whose surface areas were 0.69 cm2 at 4 months, 0.86 cm2 at 5 months, 1.05 cm2 at 9 months, and 1.13 cm2 at 1-year-and-4 months of age (Fig. 1c). However, with growing of the LV chamber and its improving compliance, the cardiac output remained adequate. By the age of 1-year-11-months, her bodily statistics had improved to the 50–70th percentile for her age: she weighed 10 kg, was 93.8 cm tall, and had a head circumference of 47.5 cm. Finally, the genetic studies for TSC-1 and TSC-2 anomalies were both negative.
Considering its difficulty and high risk of mortality, our cardiac surgeon elected to not perform tumor resection. After a consultation with the parents, they agreed to off-label everolimus therapy. We began the therapy at 0.0625 mg/day (1/4 tablet, 0.316 mg/M2/day), in addition to a combination of intravenous vasoactive drugs (namely, dopamine, bosmin, and prostaglandin E1). Although cerebral ultrasound did not find a significant tubercle in the brain, a generalized seizure occurred the next day, so the anticonvulsant therapy of dilantin and phenobarbital was begun. The everolimus serum level was 5.53 ng/mL on the 7th day of therapy, but decreased to 1.68 ng/mL on the 14th day, which was assumed to be caused by a drug–drug interaction with anticonvulsants (Fig. 4). Nevertheless, echocardiography observed an initiation of tumor regression, and both mitral flow and subaortic flow along the shrunken tumor appeared. We then increased the everolimus dosage to 0.125 mg/day (0.613 mg/M2/day), but the serum level remained at 1.92 ng/mL on the 21st day, and even dropped to less than 1.5 ng/mL on the 28th day, when the phenobarbital serum level rose to 80 mg/L. Thereafter, we reduced the phenobarbital dosage and increased everolimus to 0.25 mg/day (1.2 mg/M2/day). On the 35th day, the serum level rose to 5.59 ng/mL (Fig. 4). Another echocardiogram displayed a remarkable regression of the tumor, and blood flow through both the mitral and aortic valves became satisfactory. We ended all the courses of anticonvulsants and reduced everolimus to 0.125 mg/day (0.6 mg/M2/day). Finally, the endotracheal tube was removed on the 37th day of therapy. The patient was discharged on the 74th day of therapy when his weight was 4.4 kg (3rd percentile) and his height was 56 cm (5–15th percentile).
After discharge, the tumor continued to reduce in size. At 6 months of age, under an everolimus dose of 0.35 mg/M2/day (serum level: 1.81 ng/mL), the remaining tumor appeared as only a small nodule attached at the lateral papillary muscle of the LV (Fig. 2c, d). We thus suspended everolimus therapy. However, 16 days later, the shrunken tumor rebounded to a surface area of 1.35 cm2 producing a mild degree of mitral regurgitation. Everolimus therapy was reinstituted at 0.125 mg/day (1/4 tablet bid, 0.35 mg/M2/day). Although all serum levels afterwards were less than 1.5 ng/mL, the tumor remained as a small nodule throughout follow-up. At 1 year of age, his weight (8.8 kg), height (75 cm) and head girth (45.5 cm) had all improved to the 15–50th percentile for his age, his development was normal, and the gene studies of TSC-1 and TSC-2 were both negative.
In this cases series of three neonates with CR, we deliberately used a low dose regimen of everolimus therapy, beginning with 0.3–0.67 mg/M2/day to facilitate a serum level between 3 and 7 ng/mL. All cardiac tumors regressed smoothly within 2 months. This dosage is considerably lower than the recommended dose of 4.5 mg/M2/day for managing cerebral subependymal giant cell astrocytomas (SEGA) in TSC children , which aims for a serum level of 5–15 ng/mL and has been demonstrated to be effective at reducing the tumor size of SEGAs by half in 35% of patients (p < 0.0001).
Although both SEGA and CR are hamartomas and should be responsive to everolimus therapy, we were unsure whether the recommended dosage and range of serum levels for 1–23.5 year-old children with SEGAs would be safe for neonatal patients with CR [1, 2, 10]. Therefore, we deliberately used a low-dose regimen to treat these patients. The distribution of the everolimus serum levels in all three patients indicated that most levels were under 7 ng/mL. Nevertheless, this low level was still effective at reducing the tumors in all three neonatal patients within 2 months (Fig. 4).
Since 2012, five case reports have described the effectiveness of using everolimus to treat neonatal patients with CR and substantial intracardiac obstructions [6, 7, 8, 9, 11]. The idea of using everolimus to treat CR was derived from the successful experiences of using everolimus on patients with SEGA [2, 10]. Accordingly, the high-dose regimens for achieving serum trough levels between 5 and 15 ng/mL that had been applied to patients with SEGA, were used in the initial case reports [6, 7, 11]. The case report by Demir et al. (2012) was the pioneer instance of using everolimus to treat neonatal patients with CR. Initially, they used a high dose regimen of 0.25 mg every 6 h for four doses (equivalent to 5 mg/M2/day), which resulted in a high serum level of 83 ng/mL. That dosage was immediately reduced to 0.25 mg every 12 h, 2 days per week, and the tumor uneventfully regressed to a small size after 75 days of therapy . Hoshal et al. used a high dose of everolimus (0.5 mg/day, equivalent to 2.5 mg/M2/day) on a 2-day-old neonate after a successful cardiopulmonary resuscitation. Although the tumor had obstructed the outflow tract of the RV and infiltrated into the LV myocardium, it was reduced to a small size within 2 months of therapy .
By contrast, we targeted a lower serum trough level (3–7 ng/mL) from the very beginning in all three of the cases presented herein by using a lower dosage (0.3–0.67 mg/M2/day). Notably, the tumors in all three of our patients successfully regressed in 2 months. Despite our efforts to maintain a low-dose everolimus regimen throughout, two occasions of unexpectedly high serum levels still occurred. One high serum level (16.1 ng/mL) occurred in a 4-day-old neonate who was receiving a dose of 0.67 mg/M2/day. Another high level (over 20 ng/mL) occurred in a 49-day-old infant who was receiving a dose of 0.56 mg/M2/day at home; she also suffered adenovirus pneumonia immediately following this high-serum-level episode. These two occasions of high serum levels reflect the unpredictable nature of drug metabolism during early infancy, and suggest the impairment of hepatic drug metabolism following a viral infection .
Maintaining an adequate serum level during acute CR is difficult if the patients are also receiving drugs that interact with everolimus. For example, we used luminal and dilantin to control seizure attacks in our Case 2 patient. These two anticonvulsants are CYP3A4/P- glycoprotein(PgP) Inducers and possess a drug–drug interaction with everolimus . Accordingly, the serum levels of everolimus dropped from 5.53 ng/mL to an undetectable level. We subsequently had to escalate the patient’s everolimus dosage from 0.3 to 0.6 to 1.2 mg/M2/day, until his serum levels returned to 5.59 ng/mL.
Two cases of using low-dose everolimus to manage neonatal CR from its outset have recently been reported. First, Goyer used a dose of 0.65 mg/M2/day in two premature neonatal patients. Their cardiac tumors, which exhibited significant hemodynamic derangement, regressed smoothly, and the drugs were discontinued on day 36 for one patient and on day 81 for the other . Bornaun also used a low-dose everolimus course of 0.25 mg two times a day, and twice a week (equivalent to 0.7 mg/M2/day), and aimed for a serum trough level of 2.6–6.1 ng/mL. Although the actual serum levels of that patient were only between 0.4 and 2.6 ng/mL, Bornaun did not try to increase the dose, and the tumor regressed in 4 weeks . Therefore, including our case series, a total of six case reports have demonstrated that a dose of everolimus of less than 0.7 mg/M2/day is effective for producing regression of neonatal CR.
Notably, everolimus is an immunosuppressive drug that may predispose recipients to a higher risk of infection, particularly by opportunistic pathogens . In our case series, for example, two episodes of unexpected infection occurred. Case 1 contracted adenovirus bronchopneumonia when she was 3 months of age, which occurred immediately following unexpectedly high drug levels (20 ng/mL). Case 3 contracted chicken pox when he was 9 months of age, though the drug dosage had already been tapered to 0.33 mg/M2/day and the serum levels were less than 1.5 ng/mL. Although both patients smoothly recovered from these infection episodes, they demonstrated the threatening infections that can be triggered in everolimus-treated infants.
Everolimus can also inhibit the formation of the human protein mTOR, which suggests that the drug may inhibit children’s growth . In all three of our cases, we found that growth was stunted during their early infancy. For example, at 3 months of age, the patients’ body weights dropped to the 3rd, 3rd, and 15th percentile for their ages, respectively. Later, as we tapered the doses, their body weights all improved to 15–50th percentile. Ideally, everolimus would be administered to neonatal patients using the minimum possible dosage and duration. We tried to suspend the drugs when our patients were 3–6 months old, but found that the tumors rebounded in 2–4 weeks (Fig. 1c). We readministered everolimus to our Case 2 patient because of the tumor-related mitral regurgitation, and to our Case 3 patient because of the compression on the left ventricle outflow tract.
Although surgical resection of the obstructive CR through open heart surgery is a considerably more invasive procedure than our using tumor suppression medication [5, 6, 7, 13], surgery might still play a critical role in some cases. For example, we treated one neonatal patient with CR three years ago who had a tumor growing from the free wall of the RV outflow tract, which was severely obstructing outflow in the RV. We successfully applied tumor resection (not reported in this paper). Therefore, because of the tumor-specific characteristics of different neonates, treatment plans should always be individualized.
Everolimus therapy must be administered for at least 2 weeks before a significant reduction in tumor size can occur. Therefore, during that period of hemodynamic instability, the presence of a highly integrated cardiac intensive care team is crucial. In Hoshal’s case, they had to perform cardiopulmonary resuscitation before any medication could begin . In Demir’s and Bornaun’s cases, and our Case 2 patient, continuous prostaglandin infusions had to be used to maintain the patient’s ductal arteriosus sufficiently open. Moreover, multiple cardiovascular drugs including dopamine, milrinone, and bosmin were provided [6, 11]. Therapeutic cardiac catheterization procedures, including PDA stent and balloon atrioseptostomy, were undertaken in Mlczoch’s case . Similarly, because the CR had totally obstructed the mitral and aortic areas in our case 2 patient, he was dependent on ventilator support through tracheal intubation until the tumor resolved (Fig. 2a–d).
Some proposed natural timings of spontaneous CR regression showed a wide range of variation (range 0–105 months, median 4.3 months) . In our 3-cases series, we have observed that significant tumor rebounds occurred in two patients, two to four weeks after suspension of everolimus therapy at 3 months and 6 months of age, respectively. Because most CR may regress to only small tumors within the first year of life, it is never necessary to treat an asymptomatic CR patient. However, for the symptomatic cases of neonatal CR, including heart failure, arrhythmia or cyanosis, the patients will need specific medications and close observation in hospital until their symptoms end. When CR causes severe obstructions on the heart of a neonate, as that of our case 2, a life-threatening situation emerged. Before the use of everolimus therapy was advocated, most reported cases of obstructive CR had received tumor resections [1, 2, 3, 4, 13]. Only one case report with severe obstruction found spontaneous tumor regression in 3 weeks without surgery .
In summary, a low-dose everolimus treatment course, in the range of 0.3–0.67 mg/M2/day that aimed for serum levels between 3 and 7 ng/mL, appeared effective at triggering tumor regression in neonatal patients with CR within 2 months. However, its potential predisposition of patients to infections and its inhibiting growth should be cautiously attended to.
No financial assistance was received to support this study.
Compliance with Ethical Standards
Conflict of interest
All authors declare that he/she has no conflict of interest.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
- 1.Sciaca P, Giacchi V, Mattia C et al (2014) Rhabdomyomas and tuberous sclerosis complex: our experience in 33 cases. BMC Cardiovasc Disord 14:66Google Scholar
- 2.Robert B, Hinton RB, Prakash A, Robb RL et al (2014) Cardiovascular manifestations of tuberous sclerosis complex and summary of the revised diagnostic criteria and surveillance and management recommendations from the International Tuberous Sclerosis Concensus Group. J Am Heart Assoc 3(6):e001493CrossRefGoogle Scholar
- 12.Afinitor (Everolimus) tablets for oral administration (2016) Highlights of prescribing information. Initial U.S. Approval: 2009Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.