Paraspinal radiation recall myositis after gemcitabine for pancreatic adenocarcinoma

Abstract

Radiation recall (RR) is a chemotherapy-induced reaction that leads to inflammation and necrosis in previously irradiated tissue. Gemcitabine is a cytidine analogue that is often used in conjunction with nab-paclitaxel in the treatment of pancreatic cancer. Herein, we present a case of a 56-year-old woman with stage III pancreatic adenocarcinoma diagnosed with gemcitabine-induced RR when she presented with lower back pain and new rim-enhancing collections within the right and left paraspinal musculature 5 months after radiation therapy to the pancreas. A PubMed search was performed for ‘Radiation Recall Myositis’ and a complete literature review performed. This case and review of the literature of published cases of RR myositis highlight the clinical course and presentation of RR myositis. This review highlights the importance of considering RR in the differential diagnosis when patients who are undergoing chemotherapy and radiation present with inflammatory changes in previously irradiated areas.

Background

Radiation recall (RR) is a rare and poorly understood inflammatory reaction whereby a chemotherapeutic agent, such as gemcitabine, can induce inflammatory changes in any organ or tissue that has been previously exposed to radiation.^{1 2–4} The mechanisms of RR phenomena are not well understood, and a wide number of chemotherapeutic agents have been associated with RR reactions: gemcitabine, cisplatin and actinomycin D.⁵ 2 3 6–8 The most commonly reported type of RR is RR dermatitis; however, fewer cases of gastritis, pneumonitis and myositis have been reported.^{3 9–13}

Herein, we report a case of a 56-year-old woman with stage III locally advanced pancreatic adenocarcinoma, previously treated with localised radiation therapy and chemotherapy. She was found to have multiple paraspinal fluid collections on a restaging CT scan and was admitted due to a concern for multiple infectious abscesses. She lacked clinical signs of an infection and failed to improve on antibiotic therapy. Ultimately, she was diagnosed to have RR myositis while on the fourth cycle of palliative gemcitabine and nab-paclitaxel. She improved after treatment with steroids.

Case presentation

In September 2018, the patient presented to the hospital with a 1-month history of epigastric pain, jaundice, dark urine and light bowel movements. She was found to have a mass in the head of the pancreas. She was subsequently diagnosed with unresectable locally advanced pancreatic adenocarcinoma. She began systemic chemotherapy treatment with fluorouracil, leucovorin, irinotecan and oxaliplatin.

After 2 months of treatment, her Carbohydrate Antigen 19-9 (CA19-9) serum levels continued to rise, and restaging showed interval worsening of pancreatic cancer. In November 2018, the patient decided to proceed with localised, palliative radiation in order to exert local control of the pancreatic tumour (figure 1). She subsequently completed palliative radiation up to 30 Gy using a 3D radiation technique for local control in December 2018 and was found to have improved CA19-9 serum levels and no interval growth on further imaging.

Figure 1.

Computed Tomography (CT) representative imaging demonstrating paraspinal muscle recall reaction with prior radiation fields of the pancreas. (A) CT image showing mild rim-enhancing collections within the right paraspinal musculature consistent with myositis/necrosis. Blue arrows indicate the area of myositis/necrosis in the right paraspinal musculature that corresponds with dose deposition in the radiation field. (B–D) Previous radiation fields of the pancreas. The isodose curves are the coloured lines shown. The red line represents an isodose curve of 100% indicating an area that received 3000 cGy. The bolded green line represents an isodose curve of 94%. The thin orange line (best seen in D) indicates 70% isodose levels that extended to include the anterior and posterior soft tissues.

In January 2019, she was started on second-line gemcitabine/nab-paclitaxel and subsequently underwent four, 21-day cycles of treatment with Abraxane (125 mg/m²) given on days 1, 8 and 15 and followed immediately by gemcitabine (1000 mg/m²) given on days 1, 8 and 15. In early April 2019, she began to experience dull lower back pain, primarily on the right side. She described the pain as a dull ache accompanied by a sensation of pressure localised to the area of discomfort that she described as a sensation that ‘something is there’. On physical exam, she had mild tenderness and fullness in the lower back that was initially thought to be due to muscle strain. A CT of the abdomen and pelvis revealed new rim-enhancing paraspinal muscle fluid collections that extended from the T12 to L3 vertebral levels measuring 4.2×3.9×7.6 cm and 1.2×1.3×6.4 cm on the right and left sides, respectively. Imaging also revealed a similar rim-enhancing collection in the right rectus that measured 1×4×6 cm.

The patient was admitted to the hospital for intravenous antibiotics as this was thought to be multiple paraspinal abscesses. On admission, she had leucopenia but no associated fever, chills or signs of disseminated infection. Ultrasound-guided biopsy of the paraspinal muscle and aspiration of fluid collections were performed on the second hospital day. It was noted during the procedure that the large, oedematous muscle was consistent with a diagnosis of myositis. The paraspinal fluid aspirate was non-purulent, and cytology revealed sparse lymphocytes with no malignant cells. Paraspinal aspirate cultures, gram stain, acid-fast bacilli smear, fungal cultures and smear body fluid results were all negative for infection. Further, the patient had negative blood cultures and no signs of disseminated infection.

We found that the paraspinal fluid collections specifically overlapped with the patient’s previously irradiated areas, and surgical pathology confirmed that the muscle biopsy consisted primarily of necrotic cells and sparse inflammatory cells, consistent with myositis. Based on clinical presentation, lack of infectious signs and symptoms and rapid improvement with steroid therapy, we determined that the paraspinal fluid collections were likely due to RR myositis. This is further supported by the timing and anatomical overlap of the patient’s fluid collections with their radiation. One week after steroid therapy, the palpable paraspinal mass, lower back pain and tenderness to palpation of the lower back resolved.

Literature review

While RR reactions have been reported in numerous case studies and even assessed in retrospective studies, most reported cases of RR reactions noted are for RR dermatitis or visceral organ involvement, but myositis remains a rare entity.^{3 12 14} We conducted a literature review on cases specific for recall myositis. A PubMed search for ‘Radiation Recall Myositis’ was performed. Twenty-two cases of RR myositis were identified (table 1).

Table 1.

Reported Cases of Radiation Recall Myositis

Cancer type	Chemotherapeutic agent	Chemotherapeutic dose during reaction (mg/m2)	Time to recall reaction (months)*	Radiation technique	Total dose in Gy(Dose per fraction in Gy/fraction)	Reaction	Treatment; Resolution	BED (Gy)	EQD2 (Gy)	Reference
Nasopharyngeal	Gemcitabine	1000	6	isocentrically-mounted linear accelerator with 18 MV photon energy	70.2 (1.80)	Bilateral Cervical neck oedema	No treatment, Spontaneous Resolution	112.32	67.39	(Patel et al., 2017)
Breast	Gemcitabine	1000	3	PMRT 5000 cGy in 25 fractions Breast; SBRT 2400 cGy one fraction T12; Left Femur RT 3000 cGy in 10 fractions	50 (3.0)	Left Thigh and possible Paraspinal myositis of the thoracic spine	D/C Gemcitabine and Dexamethasone; Resolved with slow taper	100.00	60.00	(Ravishankar et al., 2018)
Breast	Irinotecan	NR	7	palliative radiation of R acetabulum- NR type	20 (4.0)	R buttock and pelvic myositis	D/C irinotecan, ibuprofen; Resolved	46.67	28.00	(Prabhakar, Goyal, & Gonzalez, 2017)
Rectal Adenocarcinoma	5-fluoruracil, leucovorin, oxaliplatin	NR	5	Six-field technique (anterior oblique, oblique and posterior oblique fields bilaterally)	50.4 (2.0)	Hip Flexor myositis	Prednisone	84.00	50.40	(Florczynski et al., 2016)
Cervical	Carboplatin, Paclitaxil	NR	5	NR	60 (1.43)	L Gluteus maximus and medius	Symptom management with analgesics; no recurrence	88.60	53.16	(Maeng et al., 2014)
Metastatic adenocarcinoma of undetermined origin (Lung vs Pancreas)	Gemcitabine- carboplatin	NR	1.5	anteroposterior or posteroanterior field arrangements (palliative intent rad treatment)	30 (3.0)	Left Hip and Shoulder	Gemcitabine continued, narcotics for pain management; Spontaneous resolution in months	60.00	36.00	(Grover, Jones, Teitelbaum, & Apisarnthanarax, 2015)
Breast	Gemcitabine	NR	3.5	single fraction	39 (3.0)	Left posterior thigh with overling dermatitis	Dexamethasone; 1 week symptoms resolved	78.00	46.80	(Delavan et al., 2015)
Lung adenocarcinoma and Anal Squamous Cell Carcinoma	Gemcitabine/carboplatin	NR	2	NR		Hip, buttocks, and groin	Prednisolone			(Graf et al., 2014)
Synovial Sarcoma	Gemcitabine and docetaxel	900 gemcitabine; 100 docetaxel	1.2	NR		Forearm	Dexamethasone			(Eckardt, Bean, Selch, & Federman, 2013)
two cases: Ewing Sarcoma	Vincristine and cyclophosphamide	NR NR	4.5 4	proton radiotherapy proton radiotherapy	45 with 57.6 boost (1.8) 45 with 55.8 boost (1.8)	Left iliac, left psoas, left gluteus Left iliac, left psoas, left oblique	Rofecoxib, ibuprofen and narcotics Corticosteroids and NSAIDs	243.0 236.3	145.8 141.8	(Hattangadi et al., 2012)
Pancreatic Cancer	Gemcitabine, doclitaxel, and capecitabine (two cases)	1000	0	NR		Bilateral lower extremities with overlying dermatitis	D/C gemcitabine and Dexamethasone			(Pentsova et al., 2012)
Adenocarcinoma	Gemcitabine	1000	2	NR	44.1 (2.94)	Abdominal muscles	Continued gemcitabine and 200–400 mg Ibuprofen	87.32	52.39	(Lock, Sinclair, Welch, Younus, & Salim, 2011)
Ewing Sarcoma	vincristine, ifosfamide, doxorubicin, etoposide	NR	6	NR	64 (1.6)	Multifidus and erector spinae	Spontaneous resolution in 16 weeks	98.13	58.88	(Heirwegh, Bruyeer, Renard, Uyttebroeck, & Demaerel, 2010)
Lung adenocarcinoma metastases	Gemcitabine	1000	3	NR	30 (3.0)	Left hip and buttocks	Prednisone continued gemcitabine	60.00	36.00	(Squire et al., 2006)
Lung adenocarcinoma	Carboplatin and gemcitabine	1000	3.5	NR	30 (3.0)	Right pectoralis	NR	60.00	36.00	(Pinson, Griep, Sanders, & Lelie, 2006)
Pancreatic	Gemcitabine	2000	4 5	NR NR	50.4 (1.8) 36 (3.0)	Rectus abdominus with overlying dermatitis Rectus abdominus with overlying dermatitis	D/C Gemcitabine D/C Gemcitabine	80.64 72.00	48.38 43.20	(Fakih, 2006)
Multiple Myeloma	Cyclophosphamide	NR	6	Photons (15 MeV) from linear accelarator; using anteroposterior and posteroanterior fields; 6 MeV to lumbar region in supine position	36 coxofemoral; 35 lumbar	Left coxofemoral and lumbar region	NR	75.83	45.50	(Borroni et al., 2004)
Pancreatic adenocarcinoma	Gemcitabine	1000	3	NR	50.4 (1.8)	Rectus abdominus	Corticosteroids	80.64	48.38	(Friedlander et al., 2004)
Two cases: Lung adenocarcinoma	Gemcitabine	800	35.5	NR NR	50 (2.0) 60 (2.0)	Right thigh Pectoralis	Analgesics Analgesics	83.33	50.00	(Miura et al., 2003; Ravishankar et al., 2018))
Papillary thryoid and anal canal carcinoma	Unknown	NR	36 and 252	NR	NR	unknown	NR	100.0	60.00	(Indinnimeo et al., 2003)
NSCLC	Gemcitabine	1000	4.5	NR	36 (3.0)	Chest wall muscles	NSAIDs and oral steroids	72.00	43.20	(Fogarty, Ball, & Rischin, 2001)
Two cases:Pancreatic Cancer	Gemcitabine with tegafur/gimeracil/oteracil	NR NR	4 4	NR NR	25 50.4	Right Obturator externus and pectineal muscles Rectus abdominus	NSAIDs, opioids, and betamethasone D/C chemotherapy NSAIDs and Continued Chemotherapy			(Nishibuchi & Wadasaki, 2018)

* Time to recall, indicates the time interval, in months, between the last dose of radiation and the recall myositis reaction

D/C, discontinue; Gy, Gray; NR, Not reported.

Numerous chemotherapeutic agents have been associated with RR myositis reactions including gemcitabine, most frequently, and vincristine. RR myositis occurred in the girdle, back and upper and lower extremity musculature with radiation doses ranging from 25 to 70.2 Gy.

To assess the effects of radiation dose normalised across different fractionation regimes, we retrospectively assessed biologically effective dose (BED) and the equivalent dose in 2 Gy fractions (EQD₂). We calculated the BED using the established formula derived from the linear–quadratic model: BED=total dose×(1+d/(α/β)), with d=dose per fraction and (alpha/beta)=3 Gy given that recall phenomena are late effects.^{13 15} We calculated the EQD₂ with the following formula: total dose×((d+(α/β))/(2 Gy+(α/β))).¹⁵ The mean calculated BED across reported cases of RR myositis was 95.9±11.6 Gy (mean±SEM) with a median of 82 Gy. The mean EQD₂ was 57.6±6.93 Gy (mean±SEM) with a median of 49.2 Gy. Radiation techniques were also probed, and of the reported cases, only six reported on the radiation technique used (table 1).

The time between the last dose of radiation and the associated RR myositis ranged from ‘concurrent’ radiation to a reaction that reportedly took place 21 years after the last radiation treatment with approximately 5.3±1.5 months being the average (mean±SEM). Treatment for RR myositis varied, though all cases reported eventual resolution of symptoms. In most cases, the suspected chemotherapeutic agent was suspended; however, in three cases, the patient was kept on the chemotherapeutic agent and concurrently managed with corticosteroids or Non-steroidal anti-inflammatory drugs (NSAIDs). Oral steroids were reported as used for treatment in 10 cases.Five cases of myositis resolved spontaneously in weeks with no treatment changes.

In this patient’s case, the time interval between radiation and the recall myositis, the chemotherapeutic agent and the symptoms are consistent with previous reports. The current case is the only reported case to our knowledge of recall myositis in the paraspinal muscles as a result of pancreatic cancer treatment. In the previously reported cases of recall myositis in the setting of pancreatic cancer, muscles that became involved included the rectus abdominis and lower extremities. Two of the three reported cases of pancreatic cancer were associated with overlying dermatitis. In this case, there was no sign or previous report from the patient of overlying dermatitis. In this case, as in a few reported cases, the myositis progressed to necrosis and resulted in fluid-filled collections that raised concern for infectious abscess formation.

Discussion

RR is a rare, acute inflammatory reaction in a previously irradiated region of the body associated with certain systemic agents that has been associated with various medications including chemotherapeutic agents and antibiotics.^{3 16–18} Most commonly occurring as inflammation of the dermis, RR has been reported in numerous organs and is believed to be possible in any tissue.³ RR myositis appears to be a particularly rare subset of RR.

The diagnosis of RR myositis is challenging, in part, because it can appear through patient presentation and imaging as infection, thromboses, or traumatic injury.^{14 16 19} As previously reported, in order to make the diagnosis of RR myositis, the area of acute inflammation on imaging must correspond with and be confined to an area of prior radiation exposure.¹⁶

Previous studies have suggested that important factors affecting the development of RR myositis include radiation dose, chemotherapeutic agents involved and the time interval between radiation and the inflammatory reaction.^{14 16 20} In our study, the dose of radiation exposure ranged from 25 to 70.2 Gy, which appears to be within the standard range of radiation dosing for each specific cancer treatment suggesting that RR myositis may not be associated with radiation doses that differ significantly from the extreme.²¹ We also aimed to assess the agents associated with RR myositis, and from the existing literature, our case included, it appears that RR myositis has only been reported with certain chemotherapeutic agents, gemcitabine being the most common.

The time between radiation and a subsequent RR myositis reaction ranged from concurrent radiation to 21 years after radiation in our review, with most reactions occurring between 1 and 6 months. Studies have suggested that a minimum timing after radiation should be used to distinguish RR reactions and radiosensitisation.^{3 17 18} Radiosensitisation is a relatively common phenomenon during which, days after radiation, irradiated sites are exquisitely sensitive to systemic agents.³ Researchers have suggested that acute inflammatory reactions in previously irradiated sites after more than 7 days of radiation exposure are RR reactions, while those occurring sooner are radiosensitisation changes. While the mechanisms of both phenomena are not understood, our case and review of the literature suggest that RR myositis, with the exception of one reported case, occurs several months after the last exposure to radiation.

Limited data exist on features of radiation technique and exposure involved in RR, and we found that most cases of RR myositis did not report radiation techniques used. We found that the average BED among cases for recall myositis phenomena was consistent with previously reported data on BED for other RR pathology including dermatitis. This literature review does not show that RR myositis may be associated with particularly high BED from radiation exposure. Much more data are needed to examine whether certain radiation techniques play a role in RR myositis.

Given the rarity of the procedure, no clear guidelines exist on proper treatment of RR myositis. In our case study, the patient was started on oral steroids, achieving symptom resolution within weeks. In the literature, corticosteroids were used successfully to treat myositis, but some cases were treated successfully without steroids table 1(table 1). Similarly, most groups discontinued the inciting chemotherapeutic agent, but some clinicians continued the medication. In all reported cases, it appeared that RR myositis symptoms resolved, suggesting that neither steroids nor stopping the possible inciting agent is strictly required. NSAIDs were widely reported for pain management regardless of other treatment plans.

RR remains a rare, poorly understood phenomenon. Some hypothesise that radiation alters stem cell function or vascular growth factors that predispose these areas to acute inflammation with certain systemic agents.^{3 13 17} Another prominent theory suggests that RR is a type of drug-induced hypersensitivity reaction, which could explain its rarity, unpredictability and variability in inciting agents and individual responses.^{3 17} The current limited understanding of RR contributes to the clinical challenge of prevention and treatment. Less, still, is known about how radiation plans, chemotherapeutic agent choices and other factors can be leveraged to prevent and/or reduce the severity of RR reactions.

Conclusions

RR myositis is an important consideration when patients on chemotherapy present with acute inflammation of any muscle with a history of prior radiation therapy. RR myositis appears to be associated with any therapeutic dose of radiation to the body and specific chemotherapeutic agents. RR myositis appears to occur primarily between 1 and 6 months after radiation. RR myositis appears to be transient and may be treated with oral steroids and withdrawal of the offending agent. NSAIDs may be useful in managing myositis-associated pain. A pathophysiological mechanism for RR myositis remains unknown.

Learning points.

• Radiation recall (RR) myositis is a rare but important clinical consideration in patients receiving chemotherapy and radiation treatment.

• A thorough review of patient radiation treatment plans in cancer patients receiving radiation is important for assessing for and diagnosing RR myositis.

• RR myositis has been treated by removing offending chemotherapeutic agents, steroid use and NSAID use, but there is no standard approach or apparent differences in treatment efficacy from reported cases.