Association of Splenic Rupture and Infectious Mononucleosis: A Retrospective Analysis and Review of Return-to-Play Recommendations

Abstract

Background:

Infectious mononucleosis is typically a self-limited disease commonly affecting young adults. Splenic rupture is a rare but serious complication affecting 0.1% to 0.5% of patients with mononucleosis. Current guidelines (based on published case reports) recommend complete activity restriction for 3 weeks after onset of mononucleosis symptoms to reduce rupture risk. We examined actual timing of mononucleosis-associated splenic injury using a large repository of unpublished patient data.

Hypothesis:

The risk of splenic injury after infectious mononucleosis will remain elevated longer than previously estimated.

Study Design:

Retrospective case series.

Level of Evidence:

Level 4.

Methods:

The Military Health System Management Analysis and Reporting Tool (M2) was used to conduct a retrospective chart review. Coding records of TRICARE beneficiaries aged 5 to 65 years between 2006 and 2016 were screened. Patients diagnosed with both splenic injury and mononucleosis-like symptoms were identified, and their medical records were reviewed for laboratory confirmation of infection and radiographically evident splenic injury.

Results:

A total of 826 records of splenic injury were found in M2. Of these, 42 cases met the study criteria. Mean time to splenic injury was 15.4 (±13.5) days. Only 73.8% (n = 31) of injuries occurred within 21 days, and 90.5% (n = 38) of splenic injuries occurred within 31 days of symptom onset.

Conclusion:

A substantial number of splenic injuries occur between 21 and 31 days after symptom onset. While most splenic injuries were atraumatic, consideration should be given to extending return-to-play guidelines to 31 days after symptom onset to minimize risk. Risk of chronic pain after splenic injury may be higher than previously believed.

Clinical Relevance:

The risk for postmononucleosis splenic injuries remains elevated longer than current guidelines suggest. Restricting activity for 31 days after mononucleosis symptom onset may reduce the risk of splenic injury.

Infectious mononucleosis (IM) is a typically self-limited disease, commonly treated in outpatient clinics, and preferentially affecting young, otherwise healthy adolescents and adults. Caused by contraction of the Epstein-Barr virus (EBV) or cytomegalovirus (CMV), IM is characterized by the classic triad of fever, pharyngitis, and cervical lymphadenopathy.^6,10,13 Other common symptoms include fatigue, malaise, and headache.^9,12,13 The symptoms of IM typically resolve within 3 to 8 weeks, though some athletes report symptom duration (mostly fatigue) of 6 months.^3,8 Splenomegaly is seen in nearly all cases, with splenic size increasing 3 to 4 times that of baseline.^7,11

Although rare, splenic rupture is a medical emergency and a justifiably feared complication of IM, even in healthy, young athletes. Splenic rupture occurs in an estimated 0.1% to 0.5% of cases, with as many as 86% occurring via atraumatic mechanisms.^1,2,12 To minimize the risk of splenic rupture, consensus guidelines recommend affected individuals avoid any form of exertion for the first 3 weeks of illness, resuming sport after this time only when asymptomatic. Return to play at 3 weeks is further limited to light activity without risk of chest or abdominal trauma and avoidance of Valsalva maneuvers. However, as noted by the American Medical Society for Sports Medicine 2008 consensus guidelines, it is unclear how long such light activity should continue and whether 3 weeks is an appropriate duration for activity restriction.¹² Current recommendations are based on case reports of splenic rupture as reported in the medical literature. While most cases of rupture occur within the first 21 days of illness, a recent review of published case reports found splenic rupture occurring as far as 8 weeks after onset of disease.² Current guidelines are hampered by the absence of a comprehensive, retrospective review of a large patient database.

We performed a retrospective review of a large military patient database to analyze and describe the temporality of splenic rupture with regard to infectious mononucleosis symptom onset. The goal of our investigation was to provide objective data to further guide safe return-to-play recommendations in young athletes affected by IM.

Methods

This study was approved by the institutional review board at the Walter Reed National Naval Medical Center in Bethesda, Maryland. Investigators used the Military Health System Management Analysis and Reporting Tool (M2) database to conduct a retrospective chart review of medical coding data of individuals enrolled in the military health care insurance program, TRICARE, between January 1, 2006, and December 31, 2016. All coding records in this 10-year period were evaluated to determine the incidence of splenic injury occurring within 12 months of a mononucleosis-like illness. We first identified TRICARE enrollees aged 5 to 65 years, then queried for those diagnosed with both a splenic injury and mononucleosis within a 12-month period via relevant International Classification of Diseases–Ninth Revision (ICD-9) and ICD-10 diagnosis codes; these codes were selected by 3 members of the study team and are listed in Appendix Table A1 (available in the online version of this article). The resulting possible cases were then manually reviewed in the electronic medical records. Initial cases with clinical presentation consistent with mononucleosis and radiographic proof of splenic injury were identified for further screening to determine whether all inclusion criteria were met. Confirmed inclusion cases were defined as those with a recorded time of onset of mononucleosis-like symptoms, laboratory confirmation of EBV or CMV infection (via positive EBV or CMV IgM [immunoglobulin M], EBV or CMV polymerase chain reaction, or heterophile antibody laboratory testing), and radiographic proof of splenic injury.

After identification of confirmed cases, we recorded the following demographic variables for each study case at the time of splenic injury: age, sex, TRICARE eligibility status (active duty servicemember, family member, or military retiree), days from onset of mononucleosis symptoms to time of splenic injury, days from onset of symptoms to evaluation by health care provider, events surrounding splenic injury (traumatic vs atraumatic), baseline activity level, time to return to play, and presence of other chronic medical conditions. Return-to-play time was defined as the time needed to return to athletic endeavors completed prior to splenic injury, either in sports or in military duties.

The above data were then analyzed using Excel (Microsoft) to evaluate for possible trends. STATA (StataCorp LP) was employed to create a Kaplan-Meier survival analysis curve to evaluate time to splenic rupture.

Results

Examination of the M2 database involved a review of medical coding data of approximately 9.5 million TRICARE enrollees annually over the 10-year study period (amassing approximately 43.5 million inpatient and outpatient encounters per year). Analysis of these data identified 826 possible cases over the study period; of these, 775 were not included as case studies since they did not meet the radiographic and clinical symptom criteria above. There were 51 likely cases of splenic rupture with co-occurring mononucleosis-like illness discovered. Four cases were excluded as the onset of mononucleosis symptoms could not be determined. Five additional cases were excluded as no laboratory confirmation of mononucleosis was evident. In total, there were 42 confirmed cases with documented onset of illness symptoms as well as laboratory and radiographic confirmation of mononucleosis and splenic rupture, respectively (Figure 1).

Figure 1.

Participant flow study diagram. EBV, Epstein-Barr virus; CMV, cytomegalovirus.

Patient Demographics

Demographic data were available from all confirmed cases and are displayed in Table 1. Most patients were active duty (76.2%), male sex (83.3%), and younger than 25 years (71.4%). All but 2 were younger than 34 years. Fourteen (33.3%) were active tobacco users. Most were otherwise healthy, though 8 (19.2%) had systemic comorbid conditions. These conditions and time of splenic rupture on day of illness (in parentheses) included hypertension (1-12 days), ulcerative colitis (16 days), common variable immunodeficiency (5 days), and Von Hippel-Lindau syndrome (12 days). None of these individuals with comorbid conditions experienced splenic injury after day of illness 16.

Table 1.

Patient demographics

Total cases	42
Sex, n (%)
Men	35 (83.3)
Women	7 (16.7)
Age, y
Range	14-51
Mean (SD)	23.98 (6.92)
Median	23
Military status, n (%)
Active duty	32 (76.2)
Dependent, pediatric	4 (9.5)
Dependent, adult	6 (14.3)
Tobacco use, n (%)
Yes	14 (33.3)
No	28 (66.7)

Time to Splenic Injury

The mean time of splenic rupture is presented in Table 2 and Figure 2. Overall, the mean time from symptom onset to splenic injury was 15.4 days, with a standard deviation of 13.5 days and median of 11.5 days (range, 1-68 days). In all, 73.8% of splenic injuries occurred by day 21 of symptoms (95% CI, 0.600-0.859); by 28 days, 85.7% of observed ruptures had occurred (95% CI, 0.736-0.942). A total of 90.5% of observed ruptures occurred by 31 days after IM symptom onset (95% CI, 0.794-0.970).

Table 2.

Time to splenic injury ^a

	Time to Rupture, d
All cases
Mean	15.4
SD	13.5
Median	11.5
Atraumatic
Mean	13.8
SD	13.9
Median	10.0
Traumatic
Mean	22.9
SD	11.9
Median	23.0

^aTime to rupture for all cases as well as subgroup analyses based on atraumatic versus traumatic nature of splenic injury.

Figure 2.

Kaplan-Meier survival curve of time to event of splenic rupture, demonstrated as a proportion of the whole. All studied patients experienced a splenic rupture event. Vertical lines signify the 25th, 50th, 75th, and 90th percentile of time to rupture.

All but 1 patient (n = 41) had documentation reporting events surrounding their splenic injury, which we coded as atraumatic or traumatic. There were 80.5% of cases (n = 33) reported as atraumatic. Of the 8 reported traumatic splenic injuries, 2 noted onset of pain with running, while 6 experienced a preceding fall or blow to the abdomen. Analysis of traumatic splenic rupture demonstrates a mean time to rupture of 22.9 days from onset of IM symptoms, with a standard deviation of 11.9 days (range, 1-40 days). Thirty-seven of 42 patients with splenic rupture were diagnosed with IM after the occurrence of splenic rupture. None of those with traumatic splenic injuries had been diagnosed with IM previous to their rupture.

Methods of Diagnosis

All confirmed cases in this study had laboratory testing that confirmed EBV (n = 38) or CMV (n = 4) infection. There were 5 (11.9%) patients diagnosed with mononucleosis prior to developing a symptomatic splenic injury. Of those diagnosed prior to splenic injury, each had a positive heterophile antibody (Monospot) test. Of those diagnosed after splenic injury, 25 (59.5%) patients had IM confirmation using the heterophile antibody test, 6 (14.3%) via EBV IgM or CMV IgM titers, and 6 (14.3%) via complete blood count with atypical lymphocytes on peripheral smear that were later confirmed via positive EBV IgM titers. Three cases had initial negative heterophile antibody tests but subsequent positive EBV IgM titers, confirming active infection.

All confirmed cases underwent computed tomography (CT) of the abdomen and pelvis over the course of their diagnosis. Two were preceded by abdominal radiography that demonstrated splenomegaly, prompting advanced imaging.

Treatment and Outcome

None of our confirmed study cases of splenic rupture were fatal. Approximately 90% of patients were managed nonoperatively (n = 38). Of patients managed nonoperatively, 84.2% (n = 32) were admitted to inpatient floor units for observation, while 14.3% (n = 6) were observed in intensive care settings. Four patients (9.5%) were managed operatively with open splenectomy (n = 2) or splenic artery embolization (n = 2). One case was managed as an outpatient.

Of the 42 confirmed cases, 7 (16.7%) did not have any information regarding long-term outcome. Return-to-play data are presented in Table 3 for the remaining cases. Of the 35 individuals where pre- and post-illness activity level is documented, 71.4% (n = 25) returned to preinjury athletic activities. Sixteen (45.7%) returned to their pre-illness activity level within 90 days, while an additional 9 (25.7%) returned to activity within 180 days. Average time to return to play or military duties was 94 days. All observed cases that returned to their pre-illness baseline did so within 180 days.

Table 3.

Return to play ^a

Return-to-Play Status	Number of Cases (N = 35)	%
Returned	25	71.4
≤90 days	16	45.7
91-180 days	9	25.7
≥181 days	0	0.0
Did not return	10	28.6
Chronic abdominal pain	8	22.8
Panic attacks after hospitalization	1	2.9
Unrelated knee pain	1	2.9

^aOf 42 cases, 35 included information about return to previous activity levels.

While 1 patient retired from sport due to unrelated joint pains, 9 (25.7%) did not return to sports or military duty as a result of their splenic injury event. Eight patients (22.8%) reported chronic abdominal pain that was exacerbated by physical activity, and 1 patient reported new anxiety and panic attacks after hospitalization. Two patients subsequently underwent splenectomy 1 to 2 years after their splenic rupture, without improvement in pain symptoms. Eight of the patients who did not return to activity were on active duty, and all were subsequently medically retired from the military due to complications from their disease course.

Approximately half (n = 20) of all identified cases underwent at least 1 follow-up imaging study: 6 with CT alone (30%), 8 with ultrasound alone (40%), and 2 with radiography (10%). Four underwent more than 1 follow-up imaging study: 2 with serial CT scans and 2 with 1 further CT followed by serial ultrasounds. Only 1 of these individuals demonstrated interval worsening of the splenic injury from the initial imaging study.

Discussion

The incidence data from a large patient medical record database show that 1 in 4 splenic ruptures occurs after the current return-to-play recommendation of 21 days after symptom onset. Of these later ruptures, 17% of splenic ruptures occur between 3 and 4 weeks. While these data are similar to previous case reports showing that the majority of splenic injuries occur within 3 weeks of IM symptom onset, the incidence data demonstrate the prolonged risk of atraumatic ruptures. Not until day 31 of illness had 90% of the cohort’s ruptures occurred. These data suggest that 31 days after onset of IM symptoms may be a more evidence-based recommendation for safe return to activity.

These data also show a longer risk period for traumatic post-IM splenic rupture. Whereas prior analyses of case reports suggested that most traumatic ruptures occurred within 2 weeks of symptom onset,² our data demonstrate an average time to traumatic rupture of 22.9 days. This again suggests that individuals with IM are at an elevated risk of splenic injury for a longer period of time than previously thought.

These data and the author’s clinical experience indicate the need for revisited return-to-play guidelines. Athletes with a “mild” case of IM may prompt difficult return-to-play decisions for team physicians, as these athletes and their coaches often push to return to activity well before the current recommended 21-day window. However, these data clearly show that mild symptoms do not equate to low risk of splenic rupture. Of patients with documented rupture, 88% were diagnosed with IM after their splenic injury had already occurred. Presumably, these patients were not previously diagnosed with mononucleosis because their symptoms had been “mild.” Yet they still experienced a serious complication. These actual registry data will be helpful in guiding better informed choices for at-risk athletes returning to play after IM.

The data from the current study agree with previously reported case series analyses in that most IM-associated splenic ruptures occur in males (83.3%) and individuals under age 35 (95.2%) and are atraumatic (78.6%). These data may be skewed toward male and young patients given the predominance of active duty military; yet, other authors have reported similar outcomes with regard to sex, age, and injury patterns in civilian case report analyses.² Dishearteningly, similar to prior analyses, diagnosis of IM often occurs after discovery of splenic injury, limiting the medical provider’s ability to engage in shared return-to-play decision making with the patient. While it is unreasonable to restrict athletes from play after each upper respiratory infection, clinicians should maintain a high index of suspicion of IM in patients with typical symptoms, particularly in male patients younger than 35 years.

Treatment and Long-Term Outcomes

Treatment in this series was predominantly nonoperative, which differs from previous studies. This pattern correlates with recent trends in surgical literature supporting nonoperative management of splenic rupture.¹⁴ It is also possible that the military’s “closed population” of medical care, with medical follow-up readily available and free of charge, could have affected this difference.

Previous studies cite that up to 91% of those who underwent splenic injury resulting from mononucleosis returned to their previous activity levels²; however, data demonstrated that nearly 1 in 4 were unable to return to sport or military duty due to persistent pain after splenic injury. One of these individuals with chronic pain underwent splenectomy at the time of rupture, while 2 underwent elective splenectomy 1 to 2 years later. None of these 3 cases experienced relief in their pain outcomes after splenectomy.

No case of splenic rupture proved fatal in our series. This contrasts with the generally accepted mortality rate of splenic rupture, reported as 9% to 13%.^2,5 As these data were derived from diagnostic codes, it is conceivable that some patients presented de novo to the hospital with splenic rupture and subsequently died without the diagnosis of preceding IM. Our review was constructed to analyze the temporal relationship between splenic rupture and IM symptom onset, hence our design is inherently limited in describing true mortality rate after splenic rupture.

Notably, tobacco users (33%) accounted for a higher proportion of those experiencing splenic rupture than would be expected from smoking rates in the general (15.5%) or military (24%) population.⁴ The significance of this finding remains unclear. However, given the known pleiotropic, adverse vascular effects of nicotine and tobacco use, these data suggest that tobacco use may indeed be a significant risk factor for splenic injury after IM.

Limitations

There are several limitations to this study. As a retrospective, chart review of level 4 evidence, this research was limited by nature. Potential injuries were identified by medical coding data, which may have resulted in the exclusion of potential patients. Furthermore, confirmation of both disease, splenic injury, and time to splenic injury relied on documentation within the medical record, which resulted in exclusion of cases from the data set as previously discussed. Additionally, it is unknown how many patients may have been treated emergently at civilian health care centers (for reasons of convenience/expediency) or expired prior to evaluation, and thus without evidence of treatment in the M2 data pool. However, M2 data do contain ICD codes for any civilian care billed to TRICARE insurance, so most civilian care should be captured within our data set. Though this study population is the largest to date involving direct patient data, the sample size remains small at 42 cases, limiting the conclusions gleaned from the data. Prospective studies would likely be limited by the large size and high costs needed to capture the rare occurrence of splenic injury after IM. A national registry of splenic rupture in athletes after IM could provide a more comprehensive data set for future studies.

Conclusion

Splenic injury is an uncommon but serious potential complication seen in the setting of IM. Post-IM splenic injury occurs on average 15 days after symptom onset, though with a wide standard deviation. A significant portion (17%) of splenic ruptures occur between 21 and 31 days after symptom onset, outside the window of current return-to-play guidelines. Nonsurgical management is successful in most cases of post-IM splenic injury, but a significant risk of chronic pain and failure to return to play/duty exists regardless of operative versus nonoperative management. Additional studies, including a national registry of post-IM splenic injuries, would be valuable in further elucidating risk factors and outcomes.